"If you have tears...": Oxford and Onions

© Connotations 1991 (Vol. 1.2), with kind permission 



The Oxford editors [1] Gary Taylor and Stanley Wells have spent seven years disintegrating Shakespeare and distributing the pieces among "pirates" [2] and "collaborators" [3]. Not even his vocabulary has escaped attack. From the latest edition [4] of the Oxford Shakespeare Glossary, all its eight hundred specific Shakespeare references have been silently excised.

     The Glossary was originally conceived as the brain-child of the distinguished grammarian and lexicographer C. T. Onions, who served for fifteen years as an editor of the Oxford English Dictionary. His declared intentions were to show how far Shakespeare's use of vocabulary was idiosyncratic, what special senses it exemplified, and what new usages it introduced into the language.

     For these purposes, Onions adapted the OED system of illustrative quotations, which avowedly aimed to show the age as well as the source of each usage by citing its first known occurrence. In this exacting task the OED had been aided by teams of specialist researchers. Of course their results were neither exhaustive nor infallible, and several antedatings have since been discovered. An Oxford monograph [5] has been devoted to counselling caution about the validity of OED first citations, especially in such disputable categories as hyphenated compounds and participial adjectives. Nevertheless, there is no doubt that Shakespeare was a linguistic innovator of the highest and most prolific order, whose immense contribution to the growth and development of English included thousands of new-minted words and expressions, most of which will have been duly documented in the OED in accordance with its explicit intention. Its recent second edition [6] continues to record Shakespeare's idiosyncrasies, most first citations of which are preserved unchanged from their earliest printing, up to a century ago.

     Such historical data, objectively analysed, should therefore prove an invaluable adjunct to Shakespeare studies, including the attribution of authorship among the accredited apocrypha and indeed all other candidates for inclusion in the canon. Thus if an uncollated play of known or factually inferable date is found to contain words or expressions which the 1989 OED's illustrative quotations ascribe to Shakespeare in later contexts, then the initial hypothesis for testing would be that this play too was written by Shakespeare in a prior phase of his development, during the so-called lost years between 1582 when (teste John Aubrey) he first came to London at about eighteen and began to write for the stage, and 1592, when (teste Robert Greene) he had become a known playwright.

     That strong argument for Shakespearean authorship would apply a fortiori to those selected usages which Onions singled out as "not pre-Shakespeare" or even "peculiar to Shakespeare," because the criteria for such categories would by definition be stricter and more positive than those for the earliest known usage. Indeed, the presence of even one word which was truly peculiar to Shakespeare would eo facto identify his hand throughout any work of single authorship in which it appeared. Conversely, if a previously unacknowledged play were now to be authenticated as Shakespeare's, then its use of vocabulary definded by Onions in his original Glossary as essentially Shakespearean would tend to confirm that Onions was not only on the right track but hot on the scent.

     This latter proposition can now perhaps be tested. There is a current academic consensus in favour of one particular candidate for the canon, namely Edward III, which was registered for publication in December 1595 and printed in 1596. For the last fifty years, every specialist [7] who has objectively analysed this play has found good (and widely varied) reasons for assigning it to Shakespeare, in its entirety. Even the Oxford editors, [8] who steadfastly rejected it for seven years because of its failure to conform with their statistical tests and other preconceived criteria, [9] have now expressed regret [10] at having omitted it from their so-called Complete Works.

     No parallel apology has yet been offered for the unfortunate consequence that all such a priori assumptions must have been fundamentally mistaken, thus vitiating the entire Oxford edition. This inference has now been further confirmed by the professional mathematician Dr. M. W. A. Smith of Ulster University, who has recently shown that all the Oxford statistical stylometric tests of authorship are misconceived and invalid. [11] But Onions, conversely, is corroborated; for his original Glossary identifies, as ostensively Shakespearean, usages which are in fact still first cited from Edward III in the 1989 OED. So it was the Oxford Shakespeare Complete Works and Textual Companion which stood in manifest need of comprehensive and drastic revision, not the Glossary at all.

     The Edward III usages in question are worth dwelling upon in some detail. They are readily verifiable by comparison between Onions and an Edward III concordance. [12] A preliminary total tally of thirty such prima facie Shakespearean usages was reduced by eliminating those with any element of ambiguity in their definition or application. That procedure left no fewer than eighteen clear examples, as follows (with line references cited from Ule [13] and descriptions from Onions [14]):

accent (line 388) = peculiar mode of utterance, "first in S."; bandy (2261) = fight, "first in S."; bury (2302) = consign to oblivion, "not pre-S."; character (674, 2200) = inscribe, "not pre-S."; civil (2065) = having proper order, "not pre-S."; clangor (2654) = loud resonant ringing sound, "not pre-S."; content (1636) = be calm, "recorded only from S."; cope with(1411) = have to do with, "not pre-S."; defiance (92, 93, 2038) = declaration of aversion, "only S."; epithet (388) = term, "S."; fairly (215) = courteously, "recorded only from S.";form (557) = military formation, "not pre-S."; health (2358) = a toast, "recorded first from S."; honourable (1906) = decent, upright, "not pre-S."; lottery (2103) = what falls to one by lot, "S. only"; opposition (988) = resistance, "not pre-S."; profit (1872) = something advantageous, "only S."; reflect (231) = shine, "not pre-S.".


    The original Glossary explains that "S." means "peculiar to Shake­speare," and that "not pre-S." is used with the same implication. So here prima facie are twenty-one separate instances of eighteen special Shakespeare idiosyncrasies being deployed by him in a long lost play, recently rediscovered and acknowledged, which will well repay serious study. Further analysis of Edward III will doubtless disclose other such indicators; and their total tally may gain in evidential value from their juxtaposition in the same passage or line, such as "with epithets and accents of the Scot." [15] Similarly, the identification of such Shakespearisms in other plays, for example Edmund Ironside, [16] could also provide pointers to authorship.

     Meanwhile a further test can be applied. If the eighteen usages listed above are indeed genuine Shakespeare coinages, as Onions plainly says or implies, then they should continue to appear as first known usages in the 1989 OED, which incorporates the carefully-corrected results of a century's further research. If they are absent, conversely, then something is surely amiss with the OED's data or methods. In fact, all eighteen are duly found there. This can hardly be mere coincidence. It speaks volumes for the validity of OED first citations and their evidential relevance. It also tends to corroborate all the other original Onions examples of special Shakespearean coinages. This correlation too can be checked. In 1980, Dr. Jurgen Schafer [17] supplied a list of OED main lemmas first cited from Shakespeare. This basic category alone contains some 2,000 items. Some 300 of them had been specifically defined by Onions in such terms as "peculiar to" or "not before" Shakespeare. All 300 continue to figure as first citations, unchanged and unchallenged, in the 1989 OED second edition.

     Yet the latest edition of the Glossary has blindly deleted all of them, together with another 500 of almost equal interest described by Onions more generally as "first in Shakespeare" and hence also very possibly his own identity cards. All this vital evidence has now been destroyed. Yet this ceaseless chopping of Onions has elicited only crocodile tears, and very few of those. His successor Robert Eagleson [18] blandly claims that modern scholary expertise has "contributed to resolving previous difficulties and clarifying past obscurities." The back-cover blurb is even more self-congratulatory. "Previous interpretations have been altered; earlier problems have been resolved." No evidence is offered for the implication that we now know much more about Shakespeare's personal language than in 1911.

     On the contrary, all the positive factual asseverations then made on that subject have been deleted and replaced by such self-contradictory statements as: "This third edition is based solidly on the earlier editions, and much that Dr. Onions originally prepared remains . . ."; "the original conception of the Glossary has been assiduously preserved"; these are the words used by an Oxford editor to indicate that this third edition is totally different from the first two, that all their specific references to Shakespearean usage have been deleted without mentioning the fact, and that the original conception of the Glossary has thus been assiduously destroyed.

     Knowing one's Onions was once a by-word for competence among Shakespeare scholars and students; and that essence will not be easily dispersed.—On the contrary; it should be collected and concentrated in a new fourth edition, designed to restore every single word and expression first cited from what has come down to us as "Shakepeare." That vital information should be made readily available, not suppressed. Oxford has a duty to give Shakespeare his words back as well as his works.


Sanderstead, Surrey



[1] S. Wells, Shakespeare: An Illustrated Dictionary (Oxford: OUP, 1978; 2nd ed. 1985); The Complete Works, ed. S. Wells and G. Taylor (Oxford: Clarendon P, 1986); A Textual Companion by G. Taylor and S. Wells, with J. Jowett and W. Montgomery (Oxford: Clarendon P, "1987," recte 1988)

[2] Held responsible, on no factual evidence at all, for The Taming of A Shrew (1594), The First Part of the Contention (1594), The True Tragedy of Richard, Duke of York (1595), Richard III (1597), Romeo and Juliet (1597), Henry V (1600), The Merry Wives of Windsor (1602) and Hamlet (1603)

[3] Thus four-fifths of I Henry VI are attributed to "Nashe," 'IX" or "Y"; one third of Timon of Athens and substantial sections of Macbeth to "Middleton"; two-fifths of Pericles to "Wilkins"; and three-fifths of Henry VIII to "Fletcher.

[4] C. T. Onions, A Shakespeare Glossary (Oxford: Clarendon P, 1911; 2nd ed. 1919; many times reprinted 1922-80 with corrections and additions; 3rd ed. 1986, ed. R. Eagleson)

[5] J. Schiffer, Documentation in the OED (Oxford: Clarendon P, 1980); see also his Early Modern English Lexicography, 2 vols. (Oxford: Clarendon P, 1989).

[6] 0xford English Dictionary, 2nd ed., prepared by J. Simpson and E. Weiner (Oxford: Clarendon P, 1989).

[7] E.g. A. Hart, Shakespeare and the Homilies (Melbourne: Melbourne University P, 1934) 219-41; K. P. Wentersdorf, "The Authorship of Edward III," DAI 21 (1960): 905-06 (U of Cincinnati); F. Lapides, "A Critical Edition of TheRaigne of Edward III," DAI 27 (1966): 1788A (Rutgers U); E. Slater, The Problem of the Reign of King Edward III (Cambridge: CUP, 1989); Edward III, ed. E. Sams (in preparation). See also G. R. Proudfoot, "The Reign of King Edward III and Shakespeare," Proceedings of the British Academy 71 (1985): 159-85, and William Shakespeare, The Sonnets and A Lover's Complaint, ed. J. Kerrigan (Harmondsworth: Penguin, 1986) 293-95.

[8] See note 1.

[9] Notably Gary Taylor's own "function-word" test, on which the entire Oxford canon is based, and his unsubstantiated assumption (typically stated as a fact) that Shakespeare wrote only a fifth of 1 Henry VI, the play to which the rare-word vocabulary of Edward III is most closely related. See A Textual Companion 112-13, 136-37, 217-18.

[10] Shakespeare Newsletter 40 (Summer 1990): 28.

[11] M. W. A. Smith, "Statistical Inference in A Textual Companion to the Oxford Shakespeare", N&Q 236 (1991): 73-78; see also his artide "The Authorship of Timon of Athens," Text 5 (1991): 195-240.

[12] L. Ule, A Concordance to the Shakespeare Apocrypha, vol. 2 (Hildesheim: Georg Olms Verlag, 1987) 145-173.

[13] See note 12.

[14] See note 4.

[15] 0ne of several such anti-Scottish gibes, which would no doubt have precluded publication during the reign (1603-1625) of James I of England and VI of Scotland. But of course there are many other possible explanations, such as copyright difficulties, for the absence of Edward III (as of Pericles) from the 1623 First Folio.

[16] The presence of which in other apocrypha should also provide pointers to Shakespearean origin; thus "bury," "content," "defiance," "health," "honourable," and "profit" occur with the same meaning in Edmund Ironside. For further detailed comparisons see Edmund Ironside, ed. E. Sams (London: Fourth Estate, 1985; 2nd ed., Wildwood House, 1986; Menston: Scolar Press, 1991) and E. Slater, note 7 supra.

[17] See note 5.

[18] See note 4.