Crystal, David. Oxford Dictionary of Original Shakespearean Pronunciation. Oxford UP, 2016. li + 648 pp. ISBN: 978-0199668427. $32.00 cloth.
The Oxford Dictionary of Original Shakespearean Pronunciation (hereafter DOSP) is undoubtedly one of the most significant works of scholarship to have appeared in recent decades. It is not the first work to have attempted to reconstruct the original pronunciation (OP) of literary works written in earlier periods. OP, as Crystal notes in the introduction to the volume, can be traced back to the mid-19th century. More than anything else, it seems to have come about as a significant sideline of the pioneers of British Phonetics—Alexander Ellis, Henry Sweet (upon whom G.B. Shaw supposedly modeled the linguist in Pygmalion), Daniel Jones and A.C. Gimson. David Crystal himself is part of this tradition: he was taught by Gimson. He is perhaps the best-known English language scholar of today, having published a vast number of works, including several on aspects of the history of English, and made his writings accessible to the general public. Within the linguistics community, he made his first notable mark through a contribution to the study of spoken language, specifically, his work on prosody and intonation (e.g. Crystal 1969). This broad background makes him the ideal author of DOSP. Expectations run high.
The flyleaf of DOSP proclaims that it will be an “invaluable resource for producers, directors, actors, and others wishing to mount a Shakespeare production or present Shakespeare’s poetry in original pronunciation, as well as for students and academics in the fields of literary criticism and Shakespeare studies more generally.” Interestingly, linguists like me are not included. One might easily imagine that this is a commercial decision: linguists are numerically insignificant compared with literary scholars or theater practitioners. But more than this, Crystal states that “OP aims to meet a need that comes from outside linguistics” (ix). The enthusiasm for OP died out over the 1950s. Its present revival has been stimulated by theater, including, notably, an initiative by Shakespeare’s Globe in London to present early plays in OP; DOSP is designed to meet the needs of this revival. This does not mean, of course, that there is nothing in it for linguists, far from it. But it does mean that DOSP is not simply an exercise in historical phonology; it is an exercise in “applied historical phonology” (ix).
An important question to address early is: why bother? Crystal puts forward the following reasons: directors “should find the experience fresh and illuminating”; “for dramaturges, and also for literary critics, OP should provide solutions to some of the difficulties encountered when speaking a text, and suggest fresh possibilities of character interpretation and interaction”; “for audiences, OP offers a new auditory aesthetic, a contrast with British received pronunciation (RP) or the local modern regional accent in which they have experienced Shakespeare hitherto” (x). These are important points. OP can reveal, for example, wordplay. Words such as, to use Crystal’s examples, hour and whore, voice and vice, and own and one, sounded the same then, allowing Shakespeare possibilities that are closed off now (xxvi-xxvii). And the idea of a “new auditory aesthetic” is not at all trivial. Any modern accent carries with it the baggage of its social associations, a sense of where they come from and where they are on the social ladder. A speaker of OP cannot be “placed” so easily. In my experience, this lends OP performances something of an earthy, visceral quality.
However, there are some downsides to OP. It is not as if people switch off their modern accent antennae: they will perceive OP through the prism of their knowledge about today’s accents. In fact, judging from comments made in the introduction to DOSP, British people seem to take OP as some kind of Irish, West Country or Lancashire accent, probably because these, from a British perspective, are all (at least in their most marked forms) rhotic accents (post-vocalic <r> is pronounced in words like car), and rhoticity is what they are struck by in OP (xxxii, xxxv). Such accents do, of course, have the baggage of social stereotypes. One can understand why some might initially balk at high-ranking characters speaking OP and sounding “rustic.” Crystal comments that OP is “no more difficult for an audience to understand than any modern regional accent” (xxxix). However, there is the issue of familiarity: more familiar regional accents are easier to understand. It might be more accurate to say that OP is not more difficult for an audience to understand than any unfamiliar modern regional accent. Crystal notes that OP requires more rehearsal time (xxxvii). Although many words in a play are repeated, that is still a lot of looking up of individual words in the dictionary, though, of course, one would hope that actors gain a “feel” for it and need to check fewer words as they become more experienced.
In order to appreciate DOSP, it is essential to understand its target. The reader might ponder what accent OP is trying to reconstruct—is it actually Shakespeare’s accent? However, Crystal makes it clear that “OP is a phonology—a sound system—which would have been realized in a variety of accents” (xi). Phonology involves relatively abstract categories of sound that are used by speakers to distinguish meanings, and is often discussed in terms of whole languages (e.g. “English phonology”), though it certainly need not be restricted to this (e.g. there are phonological distinctions between accents in the north and south of England). Those categories of sound are shared by a number of different groups of speakers, each having their own accent. Accents are also distinguished by how different groups realize those abstract sound categories. Thus, in Shakespeare’s time, there would have been a phonological system which a number of groups of speakers with different accents shared, their different accents being distinguished by how they realized particular categories of sound.
Crystal’s choice of OP as a phonology is, in my view, thoroughly practical, for the following reasons: (1) there is less evidence for any one specific accent, (2) it would be near impossible to decide which to choose to reconstruct (e.g. the actors at the Globe came from various regions, and there is no evidence about Shakespeare’s accent), and (3) one specific accent would be of less use to the modern reader, as it would not help them, for example, distinguish characters by means of accent. To illustrate, today, the vowel that most speakers in the south of England would produce when they say the words strut, mud or cup did not exist in Shakespeare’s time. This vowel sound, which in phonetic symbols would be represented as /ʌ/, is not part of OP phonology (no speaker would have said it then, as it has only been in existence for around 200 years). Instead, in Early ModernEngland, such words would have had the vowel sound that is typical of today’s northern speakers. This vowel can be heard in the word foot, whether said by northern or southern speakers, and is represented in phonetic symbols as
Let us extend this illustration a little. Things are somewhat more complex for words spelt <oo> (e.g. book, look, nook). For these, today, most speakers in England would also use /ʊ/ (i.e. the vowel sound of foot is the same as the one used for book, look and nook). However, the original vowel sound for these <oo> words was the one speakers in England today would associate with the word goose. In phonetic symbols, this is represented as /u:/. In fact, this vowel sound is still produced in words such as book, look and nook in some northern accents (e.g. in southern Lancashire and Liverpool). In Early ModernEnglish, the goose (/u:/) vowel was not universal in words spelt <oo>, but appeared in a wider range (for instance, blood would have rhymed with mood). However, things were changing; speakers sometimes had a choice between the older and the newer pronunciation. Thus, according to DOSP, nook was not universally /u:/ (as in goose), but it could also be /ʊ/ (as in foot). Where there is more than one major Early Modern pronunciation, as in nook, DOSP records the alternatives. This allows the actor a choice which they can make to suit a character (the /u:/ pronunciation in some words is perceived as distinctly regional in Britain today, and thus might, depending on one’s perspective, suit some characters better than others). However, sometimes the recording of alternatives in DOSP seems a little less than consistent. For example, today, a speaker of an accent characterised by /u:/ would typically use that vowel across the range of <ook> words, nook, book, look, and so on. DOSP, in contrast, mentions /ʊ/ for all three of the words just mentioned, but only /u:/ for nook. Of course, DOSP is explicitly not attempting to represent an accent, and this variation in the representation of alternatives probably reflects patchy evidence. Nevertheless, if an actor used /u:/ for nook, it might be a little odd if they did not use it also for book and look, despite that not being suggested as a possibility in DOSP.
The dictionary contains pronunciation transcriptions of the individual words in Shakespeare—20,672 headwords in total. A particular strength of DOSP is that it displays most of the evidence upon which the proposed pronunciations are based. A key—probably, the key—source of evidence is the spellings in the First Folio. Spellings in Shakespeare’s time were not fully standardized: they often provide clues about pronunciation. To cite Crystal’s example, murder is spelt thus 84 times, but it is spelt murther 175 times (xxi). Clearly, this gives a clue about the quality of that middle consonant. As one might imagine, not every spelling is equally helpful in this way, and some may even be misleading (e.g. they could be compositor errors). This source of evidence also explains why DOSP was 10 years in the making: “Incorporating frequency information about the use of spellings in the First Folio was one of the reasons the project took so long, as I had to go through each count, initiated using the Find function in Word, to check on such things as word-class, compound words, and lexical status (e.g. proper versus common nouns), and also to eliminate irrelevant strings (such as speech character-identifiers)” (Preface). Crystal notes that one day more sophisticated software will be able to do the job faster. That software is almost up to the job now, but Crystal got there first. He extracted his spellings from a single electronic edition of the First Folio, and records them and their frequencies for every headword in DOSP. He does not say which Folio copy and transcription he used, despite noting that “we know from Elizabethan typesetting practice that no two folios are identical in every respect” (xvii). However, any differences are doubtless small and unlikely to affect the proposed pronunciations.
A second important source of evidence is rhymes. What counts as a rhyme can be tricky; Crystal took a strict, cautious view. Moreover, for rhyme evidence to work, you need to know one side of the rhyme. For this reason, Crystal (xxi) points out that rhyme evidence is most convincing when the target word is paired with a range of clear cases, as in the case of war, which is rhymed with jar, scar, afar and bar—clearly, war was not pronounced as it is today! DOSP records all rhymes under headwords. Illuminating rhymes is a significant merit of DOSP in itself. For instance, Crystal notes that only a third of the rhymes in the sonnets work perfectly in modern English (xxii). A third fairly important source of evidence is that of contemporary commentators. Crystal exemplifies with Richard Mulcaster’s comment on the pronunciation of <t> in the context of words like apparition: it “soundeth like the full s” (xxi; Elementarie 122). However, that evidence could simply reflect the inadequacies or biases of the commentator. There are other, rather more minor, sources of evidence too, such as puns and the wider range (beyond Shakespeare) of spelling variants given in the Oxford English Dictionary. Where evidence is lacking or ambiguous, Crystal says that he simply “takes a view’’ (ix).
Each entry contains the headword and its pronunciation, spellings (with frequencies), a list of its rhymes, and, more rarely, information on puns. Full pronunciation transcriptions are not always given. Where they match a modern pronunciation, specifically RP, an equals sign is given. One might argue that RP is used by only a few percent of the world’s population, and thus is not a good choice. However, as a model of pronunciation it is widely known. Clearly, in a dictionary containing so much information, abbreviations such as the equals sign aid economy. An inevitable downside of this is that it makes entries more difficult to follow. To illustrate, here is an entry for a simple word, hill:
hill / ~s n
ɪl, = / -z
sp hil2, hill37 / hilles6, hills3, hils2
rh still S 7.5, VA 697; until
Mac 4.1.92; will PP 9.5
> dung-, mole-hill; Gad’s Hill
Needless to say, the way the entries work is logical, there is a key to the abbreviations, and everything is fully explained in the introduction. But such entries do suggest that the reader will require a certain tenacity.
Those who have written about the history of English will know that, when it comes to dealing with speech, problems are multiplied. Crystal has the benefit of standard IPA (International Phonetics Association) phonetic symbols. As anybody who has struggled to decode the pronunciations given in early works related to OP will know, this is no trivial point. Nevertheless, “decoding” what is written on the page into a sound in your head is still a real challenge. Where one can refer to a sound that a reader might be familiar with, as indeed I did above in reference to the northern England
/ʊ/ (as in foot) versus the southern England /ʌ/ (as in strut), it is not so bad. But historical stages of English contain sounds that are likely to be unfamiliar to the majority of readers, such as the rather common Early Modern /əɪ/. DOSP contains detailed explanation of all the symbols, but it is still likely to be challenging—to what extent will a reader imagine the result of a description that a certain vowel should be “more open”? In this multimodal world, solutions are at hand, and Crystal has made full use of them. Every copy of DOSP is supplied with an access code to a website for the book. Here we find lists of rhyming pairs in the plays and poems, a table of distinctive features for half-rhymes, a note that the e-text of Folio is available on request, and, most importantly, a sound file of every word in the dictionary. This alone is a tremendous achievement. It opens up the world of sound for those who have no knowledge of IPA. Even just for the vaguely curious it is a treasure box of interest. Of course, minor quibbles are possible. For the vowel in the word noise, I hear /əɪ/, and this is reflected in the transcription. The final vowel sounds of abruptly and noddy are transcribed in the same way. Yet what we hear is somewhat closer to /i:/ (i.e. the sound that many in the south of England would produce, if they said them). The issue here is probably that in the latter examples the vowel is at the end of the word in an unstressed position and thus reduced. Nevertheless, there remains a bit of a mismatch between what we hear and what we see.
As with any work, one might nitpick, as indeed I have been doing in this review. However, one’s final assessment cannot escape the fact that this is a monumental work of fine scholarship, a landmark that will endure for decades. Crystal deserves hearty congratulations for producing it.
 See David Crystal, Prosodic Systems and Intonation in English, Cambridge UP, 1969.