The Lexicons of Early Modern English

Early Modern English (EME) was spoken and written in Britain and British America from the late fifteenth to the eighteenth centuries. The Lexicons of Early Modern English (LEME), a project funded by the Social Sciences and Humanities Research Council of Canada, builds on the Early Modern English Dictionaries Database (EMEDD; 1966-99), an online resource of sixteen lexicons from 1530 to 1656 searchable by registered researchers. LEME consists of a bio-bibliography of over 800 primary sources from 1480 to 1700 (a quarter of them manuscripts, surveyed for the first time), a bibliography of 1450 secondary books and articles, searchable lexicons (about one hundred at present), and a broader searchable corpus of period literature.

EMEDD and LEME are steps towards making an EME period dictionary. Word-entries in lexical texts from that period are an important source of historical quotations for modern lexicographers. LEME draws on four kinds of historical record: (a) bilingual dictionaries (which give English-language equivalents for other European languages), (b) monolingual glossaries (which usually translate `hard words' imported from other languages into easier English words), (c) lexical encyclopedias such as herbals, medical compendia, and law lexicons, and (d) miscellaneous treatises on language, grammars, thesauri, and so forth. They offer complementary perspectives on how contemporaries of Sir Thomas More, Shakespeare, Thomas Hobbes, and John Dryden understood the English language at a time when the English dictionary, as we know it, did not yet exist.

Nearly blinded from close reading, Randle Cotgrave published his huge French-English dictionary in 1611. The years had taken their toll of him. Like Samuel Johnson, who in 1755 defined `lexicographer' as “a harmless drudge”, Cotgrave had few illusions about his craft. He pointedly explains the French phrase “Pont aux asnes” (the ass's bridge) as “Any shift, euasion, helpe at a pinch for th'ignorant; any ease, or direction vnto dull, or vnlearned people, for the resolution of difficulties which otherwise they cannot conceiue; as, a Dictionarie; and, in Logicke, the conuersion of Propositions” (Cotgrave, “Oreille d'asne”).[1] He presents lexicography and logic as bridges for lesser minds. The phrase, “ass's bridge”, originates in the geometric figure that conventionally illustrates the fifth proposition of Euclid's Elements, that the base angles of an isosceles triangle are equal. This figure looks like a bridge truss that is too steep for a horse to cross, but manageable for a donkey. Cotgrave's allusion to the “ass's bridge” predates the earliest OED example (“ass”, n.1, ca. 1780) by 150 years.

Ass's bridge or not, an Early Modern English period dictionary is, in the humanities, equivalent to a scientific “grand challenge”. Twenty years ago, Richard Bailey and his colleagues at Ann Arbor published a collection of materials for one that documented 4,400 words which antedate their first recorded occurrence in the Oxford English Dictionary (OED), or which are not there at all. These antedatings came from the Early Modern English Dictionary project (EMED) initiated at Ann Arbor in the 1920s by the visionary Charles C. Fries. Huge projected costs and a task that proved to be much larger than the OED itself contributed to the lapsing of the project. Other projects have since revived this dream. Bror Danielsson in 1974 projected DEMP: A Dictionary of Early Modern English. Robin Alston, followed by Greg Waite at the University of Otago, proposed a Tudor glossary of Literary English. The late Jürgen Schäfer proved the inadequacy of Early Modern English citations in the first OED. The early Oxford lexicographers had used major authors like Shakespeare too much and excerpted too little from neologistic writers like Thomas Nashe. In his Early Modern English Lexicography (1989), Schäfer went on to survey 133 English glossaries from 1480 to 1640 and developed from them some 5,000 entries that add to or correct the OED. Schäfer still believed that the OED would serve all our needs because he did not computerize the texts of these 133 glossaries.

The two dozen diverse lexical texts in the LEME primary bibliography that date from 1595 to 1599 illustrate the typically rich mix of language information in works of the period. Besides John Florio's A Worlde of Wordes (1598) and John Minsheu's Spanish-English dictionary (1599; both already in the EMEDD), bilingual lexicons include Andrew Duncan's Latin grammar with a Latin-English dictionary (1595), Simon Sturtevant's English-Latin glossary of Greek words (1597), and Cornelius Gerritszoon's vocabularies for the island of St. Lawrence, the Bay of Tangil, and East India. The first general hard-word glossary appears, Edmund Coote's The English Schoole-maister (1596). John Gerard contributes two botanical works, a catalogue of plants in his garden (1596) and his famous herbal (1597). Four hard-word glossaries of medical simples turn up, three by John Banister, Christoph Wirsung, and Oswaldus Gaebelkhover, and one in British Library Sloane MS 1032. Other hard-word glossaries document Biblical words, Turkish terms used in Giovanni Minadoi's Warres betweene the Turks and the Persians (1595), and the vocabularies of hawking, hunting, fowling, and fishing (W. G. Faulkner), alchemy (Roger Bachon), warfare (Robert Barret), classical goddesses (Thomas Rogers), logic (Thomas Blundeville), and rhetoric (Angel Day). A glossary for Chaucer's works and John Marston's commentary on hard words in “The Scourge of Villanie” are printed in 1598, and The Theatre of the Earth, a pocket place-name dictionary by John Thorius, in 1599.

Bilingual lexicons give, for words or phrases, their translations in one or more other languages or registers. Usually neither headword nor equivalent is explained, as the entry, “Agognanti cani, howlyng dogges”, from William Thomas's Italian-English lexicon (1550) shows. We may know the sense of the Italian or the English phrases, of course, but the lexicographer does not explain it with a phrase like “curs continuously barking”. However, these post-lemmatic explanatory segments are richer, word for word, with information about our language than are monolingual hard-word glossaries. Bilingual works not in EMEDD, but added to LEME, include Sir Thomas Elyot's Latin-English dictionary (1538), Welsh-English lexicons by William Salesbury (1548) and in British Library Add. MS 15048 (ca. 1580), English-Latin and Latin-English dictionaries by John Withals (1553) and William Lily (1567), Spanish-English dictionaries by Antonio del Corro (1590) and William Stepney (1591), and the Algonquin-English glossary by William Strachey (1612). As well, thanks to the collaboration of the Perseus Project, LEME will have Thomas Cooper's Latin-English dictionary (1584) and John Florio's second Italian-English dictionary (1611). These, too, tend to be neglected. The OED extracts only about 6,500 quotations from the over-200,000 entries in the sixteen dictionaries in the EMEDD. Almost half, 3,000, are from Thomas Blount's Glossographia (1656). None appears from Thomas Thomas (1587), Robert Cawdrey (1604-17), and John Garfield (1659); only one each from John Palsgrave (1530) and Edmund Coote (1596). There are 516 quotations from 26,000 entries in Sir Thomas Elyot's Latin-English dictionary of 1538 (the first English book to be called a dictionary), 898 from about 50,000 in Cooper's Thesaurus (1565), and 1,054 from 70,000 in John Florio's second Italian-English dictionary (1611).

These bilingual lexicons have linguistically fascinating English explanations, full of synonyms and idioms, but the word-entries are alphabetized by foreign-language headwords so that the only way to look up English words in paper editions is by scanning the text manually. Converted to a searchable database, these texts open up their wealth of information about the English language. Consider word-entries in John Florio's first Italian-English dictionary (1598; see appendix) that use the word “English”. Nine, a third of them, add information to the OED. Florio thought “terwort”, “set-wort”, “prick-pride”, “lust-pride”, “ship-like bone”, and “verdacchi” were English words, but I cannot find them in the OED. There are also two antedatings (“bear's foot” and “Abraham's balm”), two post-datings (“enpossess” and “sun-stead”), occasional novel denotations (such as “ghing” as galley slaves), and mention of something for which, it seems, no English word existed (a shrinking of sinews in a horse). Many entries are gossipy and tell us things the OED might not, such as that the French league at sea has three English miles. Properly encoded, any uni-directional bilingual dictionary can, in effect, be inverted and so made to serve English instead.

During Shakespeare's lifetime, the term “dictionary” meant only a bilingual lexicon. If Renaissance authors ever considered Richard Mulcaster's plea, in 1582, that English needed its own dictionary, and Sir Philip Sidney and Ben Jonson did so briefly, they quickly dismissed the idea: that native speakers needed to be taught their own tongue was thought absurd and demeaning. Yet the flood of words from foreign languages into English caused many readers trouble. For them, publishers put out hard-word glossaries, including Coote's English Schoole-maister (1596), Cawdrey's Table (1604), John Bullokar's Expositor (1616), and Henry Cockeram's Dictionarie (1623): they prefigure later full English dictionaries. They all give easy equivalents for difficult words, that is, for words either archaic or adopted from other languages. Coote's word-entry, “define shew what it is”, is an example: “shew what it is” can be used, Coote indicates, instead of “define”. The earliest monolingual lexicons imitate the structure of bilingual dictionaries.

Although usually English-only works too, lexical encyclopedias explain things, not supply verbal equivalents. This class of texts in LEME includes herbals by Richard Banckes (1525) and Nicholas Culpeper (1652), law lexicons by John Rastell and John Cowell (1523, 1579, 1607), a glossary of schemes and tropes by Richard Sherry (1550), a rhyming dictionary by Peter Levins (1570), medical lexicons such as John Hall's Chirirgia parua Lanfranci (1565), and Nicholas Culpeper's English physitian (1652), and dictionaries of sea-terms like Henry Mainwaring's “Nomenclator Naualis” (1620-23). Their explanations are referential, discussing the thing named by the headword, often as if the headword itself (inflection, spelling, part of speech) were irrelevant. The contemporary name for referential explanation is “logical definition”, that is, the least wordy complete description of a thing that differentiates it from anything else. Because some in the Renaissance, notably John Florio, thought of words as things, and thus as things, words themselves were susceptible to logical definition, but this type of non-lexical definition seldom classified senses. A word's logical definition gave information like spelling, etymology, equivalencies in other languages, usage restrictions, sound correspondences in English, grammatical part of speech, and things denoted. The word itself remained empty of content, a sign-thing that pointed to or labeled other things outside it, but the word as sign could nonetheless be described uniquely so as to distinguish it for native speakers.

The most important classes of miscellaneous works are grammars, etymologies, and thesauri, also neglected. The OED also took only one quotation from ten principal English grammars from that of William Bullokar (1586) to that of Jeremiah Wharton (1654). Proper-name glossaries, including the manuscript by Sir John Dodderidge (British Library Sloane MS 3479), show the interest in etymology as a tool for characterizing the world. By the mid-seventeenth century, concept-mediated lexicons, that is, thesauri, appear. John Wilkins' An essay towards a real character, and a philosophical language (1668) pioneers the modern lexicon for which philosophers and scientists were then pressing. Wilkins uses a definition, not to explain some thing, but to define a word precisely as denoting a mental concept that mediates between the word and the thing. Here the rule-based logical definition has shifted from the thing to the concept for the thing. Wilkins creates a universal conceptual framework and then embeds it in semantics, as in his entry for “utensils”:

Such kind of Vtensils as serve to contain other things, are usually called VESSELS, Cask, Receptacle, Pan, Plate, &c. These are distinguishable by their Matter, Shapes and Uses, into such as serve for the Keeping and carriage of things; being either Pliable to the things they contain; whether more loose: or more close ....

This shift away from thing-based explanation to onomasiology follows from skepticism (for which Sir Francis Bacon's idols of the tribe gave ample grounds) that words can be trusted as labels for things. The OED only recognizes, quite late, an English sense of the word “definition” that is lexical. Readers interpret the explanations or definitions for headwords in concept-based lexicons as modifying the headword, the lexical unit. Wilkins marked the beginning of what might be termed the Great Definition Shift, which transferred the logical definition of the thing, pointed to by a word, to that word itself as a lexical definition. This shift was hardly sudden. Even Samuel Johnson, almost one hundred years later, did not believe in lexical definitions.

The Web-based LEME is being prototyped in two forms. It is a textbase, currently searchable by lexeme (normalized headword) and in full text by XTeXT. There are 600,000 word-entries or about six million words of lexical text. LEME is also growing into an Access database that will give readers lemmatized, annotated word-entries. LEME is planned for distribution and licensing by the University of Toronto Press, and for prioduction by the University of Toronto Library. To judge from the technologies being prototyped now, LEME will be a considerable improvement on the EMEDD.

Because LEME uses original language materials in order to show what the Renaissance thought its own language was, it takes seriously Charles Fries' interest in “contemporary comments” on words, whether they are right or wrong in our eyes. Post-1700 theories of word-meaning -- what we think -- play no part in LEME word-entries. An historical principle, for this reason, informs LEME encoding practice. Although every LEME-encoded text must be sufficiently well-formed for conversion into an SQL database, the LEME encoding language must be simple enough to allow for the many varied structures of EME lexicons. Principal LEME elements are the word-group (for example, alphabetical or topical headings), the word-entry, and its two nested subelements, the “form” and the “explanation”. The encoding suggests a bilingual dictionary. LEME form and explanation are not headword and definition, as they would be today, but two equivalent units. Entries like Robert Cawdrey's “fuluide, yealow” (1617) receive two lemma attributes: the first time for the “hard” word by which he alphabetizes the entry, the second for the “easy” word with which he explains it. LEME does not encode for either lexical definition or sense because, except in certain sciences, English speakers were familiar instead with “logical definition”, the sufficient description of a thing in the world.

The LEME research group now consists of research assistants Jennifer Roberts Smith and Anna Guy, Access/Cold Fusion programmer Dr. Marc Plamondon, digital librarian Sian Meikle, and myself as director. Transcription, encoding, and bibliography are shared activities that usually go hand-in-hand with personal research. Jennifer Roberts Smith draws information from LEME that contributes to her doctoral research on Shakespearean prosody. I have reported findings on research topics such as the herbal source for the name of Shakespeare's first villain, Aron in Titus Andronicus, the concepts of definition in Early Modern English, the first English dictionary, and manuscript sources of lexical data. That a database of word entries from old dictionaries can be useful has surprised some researchers accustomed to using the OED and, very occasionally, paper facsimile reprints of early dictionaries.

One of the principal applications of LEME will likely be in editorial annotation. For example, Sir John Dodderidge's manuscript lexicon of proper names covers Early Modern English words seldom explained in period dictionaries. Shakespeare's first villain and Moor,Aron, in Titus Andronicus, for example, is named after a common English weed, a black-spotted member of the dragon family. Iago and his wife Emily in Shakespeare's Othello also have significant names. Dodderidge says Iago is “in frenche Iaques”: this leads us to the melancholic Jacques in As You Like It. Dodderidge resolves Emily's name to “Emme or Amy Goodnurse”. Thus Emily, Desdemona's lady in waiting, resembles the Nurse to Juliet in Romeo and Juliet. We also learn something about minor characters. Desdemona's father Brabantio takes his name from the region around Antwerp and is associated with Randle Cotgrave's Meurte de Brabant, “The sweet shrub Gaule, or sweet Willow; the Dutch Mirtle tree”. This may explain why the “song of `willow'” is in Desdemona's mind on the night when Othello murders her (IV.iii.28). Her mother's maid had sung the song to her, and her father, whose name recalls its subject, had objected to her marriage to Othello.

These examples show how early lexicons can improve our reading of Shakespeare, by far the most widely read EME author. They illuminate the now-obscure things in the world which his words denote. A speech by the Queen to Cloten in Shakespeare's Cymbeline III.i.15-34, a passage that recalls John of Gaunt's patriotic description of England in Richard II, illustrates how editorial annotation will benefit from LEME.

15    That opportunity
16    Which then they had to take from's, to resume
17    We haue againe. Remember Sir, my Liege,
18    The Kings your Ancestors, together with
19    The naturall brauery of your Isle, which stands
20    As Neptunes Parke, ribb'd, and pal'd in
21    With Oakes vnskaleable, and roaring Waters,
22    With Sands that will not beare your Enemies Boates,
23    But sucke them vp to'th'Top-mast. A kinde of Conquest
24    Cæsar made heere, but made not heere his bragge
25    Of Came, and Saw, and Ouer-came: with shame
26    (The first that euer touch'd him) he was carried
27    From off our Coast, twice beaten: and his Shipping
28    (Poore ignorant Baubles) on our terrible Seas
29    Like Egge-shels mou'd vpon their Surges, crack'd
30    As easily 'gainst our Rockes. For ioy whereof,
31    The fam'd Cassibulan, who was once at point
32    (Oh giglet Fortune) to master Cæsars Sword,
33    Made Luds-Towne with reioycing-Fires bright,
34    And Britaines strut with Courage.

A non-exhaustive survey of two hundred years of representative editorial glosses for expressions in this passage shows that, besides describing Neptune, Caesar, and Cassibelan, Shakespeare editors comment on about half-a-dozen words: “resume” (16), “brauery” (19), “Parke” and “ribb'd” (20), “Oakes” (21), “Baubles” (28), and “giglet” (32). LEME improves on their commentaries by giving richer detail about the concrete things conventionally labelled by three of these words.[2]

A few editors correctly cite the legal meaning of the first, “park”, as OED, sense 1a (“An enclosed tract of land”), but most are uncomfortable with Shakespeare's transference of this sense to a water-world. No one cites OED, sense 6, “An enclosed area in which oysters are bred ...”, which is attested from 1603 to 1883. Presumably James I would not appreciate having his ancestors compared to oysters. However, word-entries in lexicons by Thomas Thomas (1587) and John Minsheu (1599) are more applicable. A park can be “a lock in a riuer, a solewse, a floud gate or water stop”, “a fish-poole”. Shakespeare would have seen these things from time to time and, unlike ourselves, would have referred to enclosed bodies of water that protect living things as parks.

Second, editors widely gloss the word “Baubles” as “toys” or “trifles”, following OED, sense 3, “A showy trinket or ornament such as would please a child, a piece of finery of little worth, a pretty trifle, a gewgaw”, an explanation confirmed by Florio's synonym-list, “toyes, vanities, bables, iests, nifles, fooleries, fopperies”. OED, sense 1, goes further, identifying the concrete thing back of this term, the fool's bauble, but Randle Cotgrave's French-English dictionary (1611) tells us more about Shakespeare's usage in translating “Vitette” as “A verie little pricke, bable, member”. If (as seems likely) Shakespeare's term “Shipping” in this speech means a fleet of men-of-war, we should remember that sailors described such ships as if they were women being ridden by men. A section entitled “How to chuse a Ship fit to make a man of warre” in John Smith's Sea Grammar (1627) makes this association.

IN Land seruice we call a man of warre a Souldier either on foot or horse, and at Sea a Ship, which if she be not as well built, conditioned, and prouided, as neere fitting such an imploiment as may be, she may prove (either) as a horseman that knoweth not how to hold his raines, keepe his seat in his saddle and stirrops, carry his body, nor how to helpe his horse with leg and spur in a curuet, gallop, or stop; or as an excellent horseman that knoweth all this, mounted vpon a Iade that will doe nothing, which were he mounted according to his experience, hee would doe more with that one, than halfe a dozen of the other though as well prouided as himselfe.

Among concrete nautical things, then, the term “Baubles” may have a sexual connotation as the man-of-war's weapon, or, as the OED says with a straight face, his “stick with a mass of lead fixed or suspended at one end, used for weighing, and apparently for other purposes”. That Shakespeare intended a sexual innuendo becomes plausible when we reread his words about the “kinde of Conquest / Caesar here made”, his failed “bragge / Of Came, and Saw, and Ouer-came”, his “shame”, and Cassibulan's mastering of “Caesars sword”.

The third word is “giglet”, for which mildly nonplussed editors supply, time after time, the antiquated OED word, “strumpet”. Only Furness tells us that the OED admits the term is “Of obscure origin”, that is, that he and it do not know why this word describes prostitutes and the faithless lady Fortune. The word “giglet” turns up, unexplained, in 1582 among the 8,000 English headwords that Richard Mulcaster believed should form the core of a wished-for monolingual dictionary of English, but earlier Italian-English dictionaries by William Thomas (1550) and John Florio (1598) give us a more satisfying context. A giglet or giglot is a Florentine coin “stamped with the flower-de-luce”, that is, with the lily, or (as Cotgrave says) “Th'Imperiall Lillie, or Crowne Imperiall; a great, beautifull (but stinking) flower”. Thus Shakespeare means that Fortune is money, and what does money buy but things that are lovely but stinky? When Shakespeare uses the word “giglot” to mean “strumpet”, he conceivably has that lily-stamped Florentine coin in his mind. LEME can also highlight which of the various word-senses that the OED recognizes for a word applies in a given speech. For example, most editors explain “brauery” as aggressive courage in the field. Thomas Thomas and Cotgrave, however, stress what the OED, in sense 3d, defines as “A thing of beauty” (attested from 1583 to 1657). Parks, oaks, roaring waters, and sands belong to the beautiful garb in which England's isle is clothed.

Language data will also help editors of Shakespeare and other EME authors decide what readings merit emendation. Some popular editions today emend “Oakes” into “banks”, but clearly “Oakes” -- the 1623 folio reading -- does nicely. Shakespeare imagines Neptune's park “ribb'd” like a ship's sides (as Minsheu's example shows), and ships, like other “strong” places (as Thomas Thomas says), were built from good English oak.

The primary language resources of the British Renaissance are delightful in their richness. Only once they are searchable in well-appointed databases and textbases will researchers be able to appreciate them. In time, in LEME and other, similar projects, these texts will go online, where they will be available for the hoped-for future Early Modern English period dictionary. And why should that great work be done? It will aid everyone in understanding the still murky language that Shakespeare spoke and that underlies our own speech today.


[1] Quotations from the lexical works by Randle Cotgrave, John Minsheu, Thomas Thomas, John Florio, William Thomas, Robert Cawdrey, and Richard Mulcaster are from EMEDD.

[2] I do not name the editors responsible for these annotations because my purpose is to illustrate the usefulness of historical dictionaries rather than to re-edit Shakespeare's text.

Works Consulted

Appendix: Entries with “english” in John Florio's 1598 Italian-English dictionary