University of Virginia Library

Principles of Annotation: Some Suggestions for Editors of Shakespeare
by
Alice Walker

I HAVE CHOSEN PRINCIPLES OF ANNOTATION AS MY topic[1] for two main reasons: the first is because new principles would greatly reduce the formidable weight of explanatory matter with which editions of Shakespeare are often overloaded; the second is because a more critical standpoint would appreciably strengthen the hand of the literary critic in what will always be his related responsibility—the editing of the text.

It should cause no surprise that I place the task of editing Shakespeare on the literary expert. During the present century there has been a great deal of confused thinking about this matter. Briefly, what the position amounts to is this. An editor's first duty is to arrive at some conclusion about the transmission of the text. This is the field in which the bibliographer operates, but only in the most fortunate circumstances can the bibliographer recover a definitive text. That the most fortunate circumstances did not attend the transmission of Shakespeare's plays is now common knowledge: one adverse factor to be reckoned with (perhaps the most important) is that most of Shakespeare's printers made no serious effort to guard against compositors' errors. When, therefore, all is known that we are likely to know about the transmission of Shakepearian texts, there will always be a residuum

of compositors' errors which can only be detected by literary means. The two great fallacies of twentieth-century editorial theory have been, (1) the assumption that fuller knowledge about transmission would establish which readings were right and which were wrong, and (2) that palaeography would serve as the main tool in emendation. Errors due to carelessness (omission, interpolation, repetition, anticipation, and so on) are, in fact, commoner than misreadings in Folio plays set from autograph.

The text of Antony and Cleopatra, which I shall use for demonstration in this paper, illustrates the position. We do not yet know precisely what Jaggard's proof-reading of this play amounted to, but the known proof-corrections make it fairly certain that it was perfunctory and that proof was not read with copy. The text was set by the untrustworthy Compositor B and it would not be remarkable if it contained well over 200 errors, mostly due to carelessness. Only the most intimate understanding of Shakespeare's idiom will enable an editor to determine which readings are errors. It is not enough to conclude that sense can be made of the matter as it stands in the Folio. The question is whether the sense is what Shakespeare intended and whether the words an editor is interpreting are Shakespeare's words or compositor's errors. About 80 readings are usually emended in this text, leaving probably 2 out of every 3 errors unaccounted for. Of the 80 errors now usually emended, only 1 out of every 3 is explicable as a misreading.

I take a very simple example of a questionable reading. 'With news the time's with labour' is the Folio reading at III.vii.80[2]. Rowe substituted 'in labour'. Is the Folio reading, which is contrary to Elizabethan idiom, what Shakespeare wrote or is the repeated 'with' merely a compositor's error? Mr. Ridley (in the New Arden edition) accepted Rowe's emendation (rightly, I think), remarking that 'the compositor would easily pick up and repeat the with from five words before' and that 'with labour is an almost impossible phrase'. I have no quarrel with this, except that the last observation sounds too much like a personal opinion on a matter that should be approached from an historical standpoint. There are here two questions: (1) what was the idiom of Shakespeare's day, and (2) how far Shakespearian usage was in accordance with it. On the evidence of the Oxford English Dictionary, the facts appear to be that 'with labour' in the sense of 'in travail' is without parallel and that the idiom was either 'in travail' (the older expression) or 'in labour'. That Shakespeare knew this is evident from his use of both expressions elsewhere.

If I seem to be making a mountain out of a molehill over a reading which most editors would probably have emended had they realised that nothing effective was done to guard against compositors' slips of this kind in the Folio, it is because what I want to stress is the need for a new attitude towards Shakespeare's language in the light of the material now available in the Oxford Dictionary. The Dictionary was completed in 1928 and ought to have revolutionised the annotation of Shakespeare. The response to this wealth of new material has been sluggish, and editors now seem more interested in Elizabethan phonology and orthography (matters about which we as yet know very little) than in the critical appreciation of the words Shakespeare used. How many, I wonder, could say off-hand how far Shakespeare's language was the current coin of his day and how far it was his own minting? Editors leave the question obscure in spite of the fact that an historical approach to a writer's language is fundamental to both textual and literary criticism.

'Your dismission is come from Caesar' says Cleopatra (I.i.26-7): 'dismission. Similarly for dismissal in Cymbeline, II.iii.57' say the notes of both Arden editions. But would it not be more useful if readers were told that 'dismission' was the normal English word down to the early nineteenth century when 'dismissal', which has now replaced it, first (seemingly) appeared? Similarly, both Arden editions, glossing the adjective 'semblable' (III.iv.3), compare 2 Henry IV and mention, quite irrelevantly, that the word was also used as a noun, citing Day, 1599. The word (both as adjective and noun) was in use from the time of Chaucer, so that here again there was nothing remarkable in Shakespeare's use of the word, since it was current both long before and long after his time. It would have been more remarkable had he used its equivalent 'similar', which apparently first appears in Cotgrave, 1611.

In the case of both these words what a reader needs to know is (1) what the word means, and (2) whether there was anything unusual in the use of it. The Arden notes meet the first requirement but not the second. Comparison with Cymbeline and 2 Henry IV is unenlightening and further parallels from other writers would merely show that Shakespeare's use of the word was not exceptional. The justification of the existence of a word by means of parallels is, of course, a legacy of the past. Before our day, editors were compelled to rely on their own reading, or that of earlier editors, and had no such panoramic view of the history of our vocabulary as the Oxford Dictionary now provides. Despite some shortcomings, it is now the authority for the vocabulary of Shakespeare's day. Would it not therefore be of critical service to

have parallel passages, which so often give a misleading impression of the facts, swept out of our notes and to have a gentleman's agreement between editor and reader that, when a word is merely glossed without further comment, it can be assumed that the word was in ordinary use in Shakespeare's time in the sense given? A lot of matter that is merely distracting, and critically quite useless[3], could then be discarded and enable editors to place the emphasis on what readers of Shakespeare are now most interested in-his artistry in the use of words and craftsmanship as a poet.

As a fuller example of this suggestion, I take the word 'renegue' (I.i.8). Both Arden editions cite the Lear quarto reading, as well as parallels from Stanyhurst and Sylvester, and what is called 'a late instance' of 1684. This blurs a great deal of interesting information. According to the Oxford Dictionary, Udall introduced this Latinised form of the word; by the beginning of the seventeenth century it had supplanted the older 'renay', borrowed from the French, and it was still in use at the end of the nineteenth century[4]. There are here three facts that the ordinary reader needs to know: (1) that the word meant 'renounce'; (2) that it was in common use in Shakespeare's day; and (3) that the pronunciation was 'renegue' [g]. This could be conveyed in three lines by the simple gloss 'renegue' 'renounce', followed by the information that Sylvester's rhyme of 'renegue' with 'league' is evidence for the pronunciation. How much further information an editor gives depends, of course, on the kind of reader he is catering for. For school purposes more would be wanted; but if some of the three inches of space now occupied by parallels in the Arden edition could be used for developing points of interest, would it not be better to use it for relating this word to the general tendency of the period to remodel early French loan-words on their Latin form? My main concern here, however, is not with refinements of this kind but, rather, how to convey the necessary information about Shakespeare's vocabulary with the minimum of pother over what is now merely obsolete.

If we suppose that 'dismission' 'dismissal' might serve as a conventional substitute for the older method of justifying a word and its meaning by parallels, the same convention would serve for indicating what was common usage in Shakespeare's figurative language. 'Stomach', for instance, at III.iv.12 is used figuratively for 'resent', and

the Arden editions compare Danett's translation of Comines (?1596). What is interesting about the use of this word is that from the early sixteenth century, when the verb first appeared, it was invariably used figuratively and was never used for physical digestion until the nineteenth century. It may be a Ciceronianism, since Cicero seems to have been much given to the metaphor and had, perhaps significantly, a poor stomach. But, whatever the history of the verb in this transferred sense, 'stomach' (fig.) 'resent' (O.E.D.1)' would briefly show that the transferred sense was not only pre-Shakespearian but also the primary meaning in English.[5]

I have so far suggested what might be done to reduce the annotation of the ordinary vocabulary of Shakespeare's day to its right proportions with the idea that this would throw the right emphasis on what is of greater significance—the words that he coined and his strategic use of figurative language for dramatic ends. 'He words me, girls, he words me' (V.ii.190) requires, for instance, more discussion than 'renegue', though neither Arden edition has here a note. The Oxford Dictionary describes this use of 'word' as obsolete and rare, glossing as 'to ply or urge with words' and citing only this instance. Onions, following Schmidt, more plausibly glosses as 'to flatter with words'. This may be a Latinism (dare verbato flatter, deceive) and here comparison with another Shakespearian text is not otiose but apt, since 'words' in Cymbeline, I.iv.14-5 ('words him . . . a great deal from the matter'), possibly means 'flatters' too, though the Oxford Dictionary cites it as a nonce use, meaning 'to represent as in words', and Onions glosses as 'to speak of'.

Similarly, Caesar's use of 'missive' for 'messenger' at II.ii.78 is on a very different plane from the figurative use of 'stomach' already mentioned. The Oxford Dictionary describes the sense as obsolete and rare, citing only the parallel in Macbeth and a verse example of 1649. The description of the sense as obsolete and rare, though probably true enough, is a little misleading, as this is an example of metonymy, frequently employed by Shakespeare in talks at high level, and Shakespeare's contemporaries would consequently have deduced that Caesar was very much on his dignity—as Cleopatra, for instance, was not when she spoke of Antony's 'dismission'. Time has blurred these distinctions, and we can only recover them by the kind of annotation which establishes an historical perspective. I am not forgetting that a poet's art depends largely on his selection of words from the common store; but

if all new words, or rare meanings, were distinguished by even a brief comment (such as 'he words me' requires), we should, of course, recognise to what extent Shakespeare found what he needed in the language of his day.

What he seemingly coined would, for the most part, need only the briefest notice. The old idea that there was something peculiarly licentious in Shakespeare's language unfortunately persists—and hence perhaps the tendency to see the hall-mark of Shakespeare in the oddity rather than in the norm.[6] In fact, most of Shakespeare's neologisms bear the stamp of nature—not eccentricity. Inspired by the same powerful sense of analogy as his imagery, they follow traditional patterns. Thus 'lowering' (I.ii.122) would need no more than the briefest of comment (or even no more than an asterisk) to draw attention to the fact that this is the first recorded use of the adverb as a verb, and it should now be possible to assume that the majority of well-taught school children will recognise the patterns (so admirably sketched by Bradley[7] which have contributed, at all periods, to the making of English.

What I want to stress is the need to get the strands in the fabric of Shakespeare's language disentangled, so that one can see what his inheritance was and by what means he enriched it, both by art and nature. The happy mingle of what he selected and what he invented was Shakespeare's style, and until his language and style are understood textual criticism lacks the technique for discriminating between truth and falsehood in readings. I do not myself believe that the critical faculty of the best editors of the past was as unprincipled as it is sometimes made out to be. We now know that Shakespeare's formative years at the Stratford Grammar School were spent in the analysis and imitation of Latin poets and orators, with text-books like the Epitome of Susenbrotus to explain the poet's tools; and we know that in his early work he consciously applied the lessons he had learnt and that, with maturity, his style grew bolder and technique second nature. Shakespeare's early editors had the great advantage of working in much the same tradition and time may show that where editors between Pope

and Johnson shared common interests with Shakespeare—as in their appreciation of an antithesis or oxymoron—they contributed more to the solution of textual problems than is now thought.

The importance of rhetoric and what Shakespeare owed to Latin writers needs to be constantly borne in mind whenever the Oxford Dictionary describes Shakespeare's meaning as obsolete and rare. Shakespeare's figurative use of language (especially simple instances of metonymy—abstract for concrete, the part for the whole, and so on) was often classified by the Dictionary in this way, giving an appearance of licentiousness or oddity to what were merely the recognised devices of a poet; and what the Dictionary occasionally describes as 'poetic' is sometimes absurdly glossed: 'sea-wings' for 'sails' is explained, for instance, as 'means of flight by sea'. Schmidt's Lexicon, as in the gloss just cited, generally shows, in fact, a readier grasp of poetic devices than the Oxford Dictionary. Poetic language, unless it is conventional poetic diction, is, of course, resistant to semantic treatment: poetry must break loose; semantics must pin down.

Having stressed the need for the kind of historical approach to Shakespeare's language which the Oxford Dictionary now makes possible, this is the first reservation I want to make about its use—that where meaning 'o'erflows the measure' it is intractable material for the lexicographer. The second necessary warning is that reliance cannot be placed on the Dictionary's terminal dates.[8] In a work of this magnitude oversights are inevitable. In the first volume of Jaggard's Treasury of Ancient and Modern Times, 1613,[9] for instance, some of the words escaped notice altogether:

• congruate (congruous), 585.b.38-9[10]
• gripply (covetously), 727.b.44
• hinderous (impeding), 89.b.27
• table pins (styles), 790.a.28

Other words are considerably later than the Dictionary's last entry. Among these are

• metally (metallic), 813.b.43—last entry 1559
• prolocute (speak), 40.b.9—1st and only entry Levins, 1570
• terin (siskin), 805.b.29-30—1st and only entry Chaucer
• wakewort (wild arum), 564.a.42—described by the Dictionary in a single entry for 1530 as 'some plant', though the Treasury identifies it (839.a.23) with the arum or wake-robin about which the Dictionary is more fully informed. Two other names for the same plant in the Treasury (loc. cit.), Calves-foot and Ramp, are similarly later than the Dictionary's last entries (for 1607 and 1611 respectively).

Other words antedate the Dictionary's first entry:

• abra, cadabra (described as two barbarous words and an infallible help against the ague), 694.a.37—1st entry 1696
• ambulatory (nomadic), 176.a.43—1st entry 1622
• maniglions (bracelets), 523.a.14—1st and only entry 1704
• trabilious (for 'atrabilious'), 477.a.15—1st entry 1651
• Wisigoths, 795.b.19—1st entry 1906

These are but a selection[11]

I offer this evidence not in any cavilling spirit, but as a warning against the assumption, sometimes too readily made, that the Dictionary's dates can be relied on. It is only reasonable to suppose that many words were current long before they found their way into print, even if the Dictionary was fortunate enough to spot their first appearance; and any word familiar from Latin authors read at school or popular books, like the Adagia of Erasmus, may have been in circulation decades before the first recorded use. It is very curious, for instance, that, although collision was apparently in use from the fifteenth century,

collide does not appear until Burton in 1621, in spite of the Latin verb's having been used by much-read Latin writers—by Horace, for instance, and Quintilian. Hence, no one should be surprised if Burton's collide proved not to be the first use of the word in English. Similarly, it would be absurd to suppose that, since gripple, meaning 'covetous', had been in use from the Old English period, the adverb gripply was never used before Jaggard's Treasury of 1613. Though it is interesting to know when a word or meaning first seems to have made its début, no useful purpose would be served by encumbering the notes of an ordinary edition with information of this kind. The main aim should be to separate the current coin of Shakespeare's day from his own minting, bearing in mind that what is described by the Oxford Dictionary as 'rare' may be merely a recognised figure of speech and that some of the words which Shakespeare seems to have introduced may already have been current.

We are accustomed to conventions in collation notes and in our day the traditional load has been considerably lightened by cutting out useless matter. Any similar conventions which removed the dead wood from explanatory notes, making it possible to separate the significant features of Shakespeare's diction from words which were ordinary to him but strange to twentieth-century readers, either because they are now obsolete or have changed in meaning, would provide a critical approach to his craftsmanship in the use of language which would be of service to all. It would, in particular, fortify judgment in determining what is right and wrong in substantive texts. As an instance of what is wrong, I cited at the beginning of this paper 'With news the time's with labour'. Is 'my knee shall bow my prayers' (II.iii.3), similarly emended by Rowe (to 'in prayers') another error of the same kind? This is the sort of question which is re-opened by what bibliographers have revealed about the transmission of Shakespeare's plays. The assumption that proof-correction was much more scrupulous than it is now known to have been has led to the disastrous state of affairs when greater confidence is placed in the accuracy of compositors than in the common sense and artistic sense of our greatest poet.

The Oxford Dictionary is not, of course, the only work of editorial importance that has not been advantageously used by annotators. Everything relating to Shakespeare's style is significant—especially Baldwin's Small Latine, which takes us back to the Stratford classroom where Shakespeare served an apprenticeship to Latin poets. This, in turn, should renew interest in the texts of early eighteenth-century editors—especially Pope who, much as he deplored some aspects of Shakespeare's style, recognised the common bond in others. Is there,

for instance, another error of repetition in the following lines (V.i.17-9)? Pope emended to 'in that name lay', presumably to secure the right emphasis for sense and metre. Theobald, Johnson, and Capell accepted the change, but Malone (who had a most mischievous hand with style) restored the Folio reading. Did a similar error occur at IV.xiv.38-9? The special shield should surely be balanced by the particularity which Johnson's conjecture 'This battery' supplies.

The text of Antony and Cleopatra is rife with typical compositor's errors of all kinds. Sometimes the problem for an editor is one of sense and style (as in the examples just cited). Sometimes it is one of imagery. Is, for instance,

(II.v.24) really an image of forcible feeding, as the customary comparison with Julius Caesar and The Tempest implies—or is Hanmer's emendation 'Rain' (an easy misreading) more apt to the context? At other times, the problem is one of dramatic intention. Did Shakespeare write 'made their bends adornings' (II.ii.212), intending by this clutter of consonants no more than a graceful tableau of Cleopatra and her Nereids—or did he write 'adoring',[12] carrying on the love-symbolism of 'in the eyes' and the motive of the love-sick winds and amorous water? I mention these points as reminders that every kind of interest in Shakespeare's art has its part to play in the detection of errors and emendation. The need to get the new material in the Oxford Dictionary into focus is, however, the prime necessity because past editors, however enlightened their appreciation of craftsmanship, can give the twentieth-century critic no help in getting Shakespeare's idiom into historical perspective. This has only recently been made possible.

At the beginning of this paper I mentioned that a more critical use of the Dictionary would greatly reduce the burden of annotation with which texts of Shakespeare are still encumbered. I have also pointed out that we have now cleared collation notes of much useless

matter and that the same policy might advantageously be followed in explanatory notes. Gould's suggestion that Antony's eyes should glow like plated 'stars' rather than plated 'Mars' (I.i.4) is, for instance, the kind of emendation that is best forgotten. Might not the New Arden edition similarly have ignored Staunton's suggestion, discussed in the explanatory notes, that the 'varying shore of the world' was an error for the 'varying star' (IV.xv.11)? Any reader of Latin poets would have recognised what the 'varying shore of the world' meant[13], though he might, with reason, have suspected that the plural 'shores' was more in accordance with poetic idiom. Explanatory notes should, in fact, show the same kind of discrimination as collation notes: what is not worth recording in the one does not merit discussion in the other, unless an editor is rejecting a traditional reading. The left hand should know what the right hand is doing. The disproportionate amount of space now devoted to the discussion of readings should, however, shrink to more manageable proportions when it is realised to what extent and along what lines emendation is wanted.

I have assumed in this paper that all editions of Shakespeare have the common purpose of presenting a text as verbally pure as possible, with explanatory matter to bridge the gap between Shakespeare's language and our own. In the present century, editors have discarded the traditional literary tools. They need them all —and new ones too— for building on the foundations laid by bibliographers; and among the new tools none is of greater importance than the historical study of Shakespeare's language which has only been possible since the completion of the Oxford Dictionary.

Notes