How do you spell Shakespeare?

When Oxford decided to do Shakespeare they clearly made up their minds that the scale of the operation must be very grand, and a team of scholars has been working hard for eight years to get it done quickly, done right, and done with the greatest possible display and novelty. One has to admire not only the industry of Professor Wells and his associates, but their flair for publicity, as evidenced by the enormous solemn fuss about the poem ‘Shall I die?’, now accorded an honoured place in their canon, and also by the proclaimed scope and originality of their enterprise, which, though not essentially different from other such enterprises, is different in many eye-catching ways, and must have set the Press some unique problems.

The work is not yet fully finished. What we have at present is two vast volumes, the Complete Works in modernised and again in original spelling, and a fair number of the single-play volumes edited by a lot of able people who are presumably in agreement about the general editorial line and who can provide the detailed introductions and notes missing from the big collections. Of these more portable and more useful books I can’t say much; they will be judged by comparison with the New Arden editions, some of which are now showing signs of age (some of them always did), and possibly with the New Penguin. There is a Cambridge set now well under way; the Oxford Hamlet is the third considerable edition of that play in five years, following hard upon Harold Jenkins’s magisterial Arden and Philip Edwards’s serious Cambridge version.1

It may well be asked by non-Shakespearians and non-publishers whether all this editorial activity is needed, and by whom, and the Oxford team anticipates the question by asserting the boldness as well as the unparalleled scope of its enterprise. It is worth asking how much there is in this claim.

The Original-Spelling volume greatly increases the size of the undertaking and is probably the greatest novelty. In some respects it is a very odd compilation. Back in 1960 John Russell Brown wrote an article, celebrated in the trade, in which he argued against the value of old-spelling Shakespeare, saying among other things that the amount of ‘silent alteration’ an editor would have to introduce would make such a text almost useless for close study; that it would be naive to suppose the spelling reproduced was that of the author; and that to believe the old spelling imparted ‘an Elizabethan flavour’ to the words was a pseudo-historical nonsense, for the strangeness of the spelling isn’t something the Elizabethan reader could have been expected to notice. Brown’s article caused a stir because, as he remarked, it ran counter to assumptions rarely questioned. W.W. Greg had been firm about it in the Prolegomena to his authoritative book, The Editorial Problem in Shakespeare (2nd ed., 1951), where he argued that modernisation seriously misrepresents Elizabethan English when it changes ‘murther’ to ‘murder’, ‘mushrump’ to ‘mushroom’, ‘vild’ to ‘vile’, ‘wrack’ to ‘wreck’, and so forth. Less controversially, he indicated that modernisation can in some cases destroy rhymes, and hinder the business of emendation (but surely nobody would be ass enough to try that without recourse to the original). Greg’s case really depends upon a conviction that the needs of critics and of ordinary readers are different, which is true, and another conviction, that there is always a chance that the original spelling has some trace of the author’s hand and should therefore be preserved. Of course he knew better than anybody how slim this chance was.

Brown saw that Greg’s general case, though somewhat dogmatically stated, was less strong than it appeared. Why not work from photographs instead of type facsimiles or old-spelling editions? His opponents minimised the treachery of type facsimiles and stressed the fallibility of photographs: but that was before Hinman’s Norton Facsimile of the First Folio (1968) – a remarkable achievement. By using only the best leaves of a great many copies of the Folio, Hinman produced photographs of an ‘ideal’ version, so that his facsimile is in practice a better copy than any genuine exemplar, and anybody who wants to see what Jaggard’s compositors actually printed, without modern editorial interferences except for line numbers, should look there.

In 1965 Fredson Bowers, who more or less inherited Greg’s authority, took a look at the problem in the light of the enormous recent expansion of bibliographical techniques, in part at least brought on by himself. Though he did not deny that an original-spelling text was desirable, he maintained sadly but strongly that it couldn’t yet be achieved: there was still so much to be done in the way of studying printing-house practices, compositors’ vagaries and so forth, along the lines of Hinman’s extraordinary investigation of the printing of the First Folio, published in 1963. A great deal of recondite research has gone on since then, but work of this kind, as the present Oxford editors admit, usually opens up rather than closes the prospect of more work, and one would have expected Bowers’s prediction, that an old-spelling edition of Shakespeare was a matter for the 21st century, to have held at least into the Eighties.

It is also a reasonable inference that such an edition would take a lot of people a long time: yet this has apparently not proved to be the case. An old-spelling Oxford edition was mooted long ago by R.B. McKerrow, but he produced only the Prolegomena (1939) before he died, and Alice Walker, who took over the project, did not issue a single play. With this discouraging history behind him, Professor Wells seems not to have been thinking of old spelling when he mounted the present assault on the Shakespearian summit. ‘The newly proposed Oxford editions,’ he wrote in 1979,

will be in modern spelling. This procedure, traditional in editions of Shakespeare, removes unnecessary barriers to understanding, making it possible for the reader to concentrate on the text itself, undistracted by obsolete and archaic accidentals of presentation. Thus, his reading experience is closer to that of Shakespeare’s contemporaries, who also read the plays in what was, for them, a modern form. We plan both a new single-volume edition of Shakespeare’s works for the Oxford Standard Authors (OSA) series, and a detailed scholarly edition, devoting a volume to each play, for the Oxford English Texts (OET).

Wells went on to give by far the best account of the problems of modernising (Modernising Shakespeare’s Spelling, 1979), but said nothing whatever about an old-spelling edition, and it is reasonable to infer that eight years ago the very idea of one must have seemed terrifying to the Press if not to the general editor, who was in any case ready with all manner of good reasons for doing without such a thing. So it is at least a little surprising that the decision was reversed, and that the work, once undertaken, occupied not the decades of editorial drudgery envisaged by McKerrow and Bowers, but five or six years at most of doubtless frenzied activity, years during which work continued on the modernised version, on the single-play volumes, and on the Textual Companion, still missing but promised for later this year.

Apparently the editors decided that the objections to old spelling were less cogent than they had thought. Does the present volume show their second thoughts to have been right? To answer with confidence one would need to have worked with the edition for a long time, and I have only sampled it. The most obvious difficulty is ‘silent alteration’, and so I thought I might look at a few passages celebrated for their difficulty, and see how the ‘original spelling’ of this Oxford edition compared with the original spelling.

1. There is a well-known crux in Antony and Cleopatra, I. iv. 47, where the original (Folio) text speaks of an ‘Arme-gaunt Steede’. In Oxford OS we read ‘Arme-iaunct’. The modernised text has ‘arm-jaunced’. Presumably these readings will be defended and explained in the volume we have not yet got, but the point is clear: an emendation of ‘Arme-gaunt’ has been read back into Elizabethan spelling, and the word ‘Arme-iaunct’ is a pretty modern guess at what the author, or a compositor, or a proof-reader, wrote or rewrote. It is, in fact, a fake antique.

2. There is another crux in the same play, solved by the despised 18th-century editor Warburton in one of the most brilliant and certain of all Shakespearian emendations. Antony is complaining that all is lost by Cleopatra’s treachery. His followers have deserted him:

             All come to this? The hearts
That pannelled me at heeles, to whom I gaue
Their wishes, do dis-Candie, melt their sweets
On blossoming Caesar: And this Pine is barkt,
That ouer-top’d them all.

Warburton read ‘spanieled’ in the second line. Antony has picked up the idea of ‘discandying’ (melting) from Cleopatra a little earlier.

                        Ah (Deere) if I be so,
From my cold heart let Heaven engender haile,
And poyson it in the sourse, and the first stone
Drop in my necke: as it determines so
Dissolue my life, the next Caesarian smile [smite]

Till by degrees the memory of my wombe,
Together with my brave Egyptians all,
By the discandering of this pelleted storme,
Lie grauelesse ...

The idea is of the hailstones melting (‘discandering’ = ‘discandying’) and so killing her, her son and her other children, and then all the Egyptians. But the word for melting stays with Antony and he remembers its other association with candy, a sweet resembling ice: so he launches into a strange conceit, supposing the ‘hearts’ of those who followed him, dog-like, to have let his gifts (their ‘wishes’) melt in their mouths (the candy discandies), and then go and drop the sticky remnant fawningly before Caesar, represented as a tree in blossom, while Antony is a tall barked pine. In this extraordinary sequence the verb ‘pannelled’ must refer to the fawning action of a dog, which is why Warburton, conscious no doubt that Shakespeare had elsewhere linked fawning and slavering dogs with candy, emended to ‘spanieled’. In the Oxford modern-spelling edition this emendation is accepted. In the original-spelling version we get ‘spannell’d’, ‘spanel’ or ‘spannel’ being an old form of ‘spaniel’ presumed, no doubt, to have been familiar to the compositor, though not current in Shakespeare’s time so far as the OED is aware.

The point is that the Oxford editors, having accepted the emendation and printed ‘spaniel’d’ in the modern version, have to invent a likely-looking older form as what the compositor misread as ‘pannelled’. This is a guess, and a ‘silent alteration’ – and it is not easy to see that a modern reader is better-off with it than he would be with Hinman’s photographs and a modern edition to tell him or her how the original ‘pannelled’ has been altered to ‘spaniel’d’.

3. The most famous crux of all is in Henry V, when it is said of the dying Falstaff that ‘his Nose was as sharpe as a Pen, and a Table of greene fields’. Theobald’s emendation is widely though by no means universally accepted, and occurs in the Oxford modern version as ‘a babbled’, in the ‘original’ as ‘a babeld’ – again an attempt to reproduce the manuscript reading a compositor might have misread as ‘a Table’. I don’t know that every expert would think this a probable misreading, but it is in any case pure conjecture, and argues back from Theobald’s bright guess – which took little account of the peculiarities of secretary hand – to a version which looks more plausible and scientific. No doubt the whole matter will be argued out in the missing Textual Companion, which, as we see from the bits and pieces of it sent out in advance, devotes so much grave argument to the text of ‘Shall I die’ that it is sure to be very expansive on rather larger issues.

4. At the beginning of the third act of The Tempest Ferdinand remarks that ‘these sweet thoughts, do euen refresh my labours,/Most busie lest, when I doe it.’ Much ink has been expended on this difficulty, which Oxford-modern resolves by printing ‘busil’est’ – an emendation duly defended in Stephen Orgel’s excellent single-play Oxford version2 (‘my thoughts of Miranda are most active when I am busiest at my work’). The Oxford-original gives ‘busielest’ as what the compositor or possibly the scribe misread or wrongly divided in an attempt to make sense where there seemed to be none. There is of course no guarantee, and some would say ‘lest’ means ‘least’, while others might give up and say there is here, as elsewhere, an indeterminable corruption: that, as anybody who has ever read proof knows perfectly well, minds wander, and it is sometimes impossible to explain how things came to be as they are. But the original-spelling editor is obliged to produce a rational fake.

5. The Oxford editors have decided that the differences between the Quarto and Folio versions of King Lear are so considerable that they must treat them as two separate plays: they therefore print them as two separate versions, The History of King Lear (Quarto) and The Tragedy of King Lear (Folio). This is one of their bolder innovations, but at present I am interested in what happens when the two versions differ only slightly and not, on the face of it, as the result of revision. Here is what Kent says about Oswald in the Quarto:

     ... such smiling roges
As these, like Rats oft bite those cords in twaine,
Which are to intrencht, to vnloose ...

In the Folio this is:

    such smiling rogues as these,
Like Rats oft bite the holly cords a twaine,
Which are to intrince, t’vnloose ...

Since they give both versions, the editors naturally have both ‘intrencht’ and ‘intrince’ in the original-spelling text; in the modernised version they have ‘entrenched’ (Quarto) and ‘intrinse’ (Folio). But it is quite difficult to believe that these are independent versions, or that one is an improvement made in revision. The holy cords must always have been ‘too intrinse’ and never ‘too entrenched’, since the rats are biting at knots that cannot be untied, not at knots that have dug themselves into something. The Quarto must be wrong, perhaps because the compositor or proofreader changed an unfamiliar to a familiar word; and the editors, pursuing their idea of two independent texts, have given canonical status to a sophistication. ‘Silent alteration’ would here have made a small point against the independence theory, or at any rate complicated it. Once again we shall have to see how the Textual Companion argues the matter, or consult the single-volume edition, not yet published. It will also, no doubt, offer explanations for pastiche stage directions. It certainly shouldn’t be supposed that this old-spelling edition remains very close to the original, and in the absence of annotation in the volume itself the reader who hasn’t got the Companion will always be uncertain, even when the spelling of everything looks reassuringly Elizabethan, as to what is editorial and what is not.

These complete editions are alike in their sparing provision of ancillary material, and the passages I’ve been discussing are of course random samples, for hundreds of such cases exist. The modern-spelling version, like many of its rivals in the market, is content to provide bare texts, rather handsomely printed though in double columns, plus a general introduction, a list of contemporary allusions to Shakespeare, and 17 illustrations, many though not all of them perfectly familiar. The general introduction includes a short Life of Shakespeare, a section on the drama and theatre of his time, a brief survey of the early printing of the plays, and a final section called ‘The Modern Editor’s Task’, which usefully indicates the remarkable variety of problems set by the different sorts of copy used when the plays were printed, but also raises the question as to when variation between two Jacobean versions becomes so great that it is best to think of them as distinct and independent.

This is the point at which the Oxford editors have been boldest. In Hamlet there is a Quarto of authority, probably set for the most part from Shakespeare’s own papers, which contains 230 lines not in the Folio version of 1623, and does not contain 80 more that are in the Folio. Since the latter was set mostly from a theatrical manuscript, the editors regard it as a revision of the Quarto which Shakespeare presumably authorised. Not feeling that they could print the longer early version as well as the slightly abbreviated Folio one, they have used the latter. Formerly editors following F have included the Q parts, but here they are printed as an appendix to the play, on the ground that if you put them in you have a composite that was never performed. The editors seem proud of this decision, though it means that their text no longer includes some famous verse from the opening scene and the Closet scene, a lot of Osric, and the whole of the final soliloquy, ‘How all occasions do inform against me ... ’

We may well applaud this move as brave and logical, and it is true that modern editors tend to err on the side of timidity: yet there is something to be said against it. The habit of editors is understandably to climb back as far as possible towards the version they think had the final approval of the poet; they know very well that the best they can hope for is a rough approximation, but, given a choice between an early version and a version used in Shakespeare’s lifetime and possibly under his direction, they will, if forced to choose, choose the latter, even if the reason for the cuts is uncertain (since the Folio version is extremely long without additions from Q, the reason is unlikely to have been simply performance time; a net cut of 150 lines would save only ten or twelve minutes, and by the same arithmetic the remaining 3900 lines would occupy around four and a half hours; to get it done on a winter afternoon at the Globe you would need to cut another two hours of it, roughly half the play).

However, the real objection to leaving out the bits of the first scene and the Fortinbras soliloquy is simply that they belong to our Hamlet. The problem is one that comes up in Biblical studies: it is recognised for instance that two independent and mutually contradictory versions have been combined in the opening passages of Genesis, but the canonical Genesis is the one we have, not the disentangled fragments. They are, we may say, too intrinse to unloose; they are entrenched in tradition and have for millennia been one book and not more. The Oxford editors profess amazement at the timidity of their predecessors, all save one, who was brave enough to put the intrusive Q material into square brackets. G. R. Hibbard, editor of the Oxford single-volume Hamlet, argues strongly that the Folio was printed from a fair copy made by the author himself; naturally he would make changes, some slight, as he worked his way through, and ‘the cuts and additions seem to be parts of a definite policy designed to make the play more accessible to theatregoers in general by giving it a more direct and unimpeded action, pruning away some of its verbal elaborations, and smoothing out its more abrupt transitions.’3 But he admits that many of the changes seem not to further this cause, and it is obvious that the most striking of F’s additions, the comments about the boy actors, didn’t exactly promote direct and unimpeded action. Hibbard also relegates the Q parts of the opening scene, and the great soliloquy, to an appendix. We are being told to get used to a new Hamlet.

The whole question of the text of Hamlet is horribly complex, and the editors must have been tempted to print Q as well as F, but they decided that it would be ‘extravagant’ to do so, and, forced to choose, chose the more ‘theatrical’ version. However, they made an exception for King Lear, because in that case the differences between the two versions affect the story-line, especially in the last two acts. A tendency to think of this play as extant in two versions has been growing of late. The arguments are extremely technical, as several books have recently shown, but the immediate question is whether readers of such plain texts as these will benefit much from the editors’ decision. They may not be very interested in the conduct of the military operations which lead up to the deaths of Cordelia and the King – as Albany says of the death of Edmund, ‘That’s but a trifle here’ (both texts). It is a perpetual irritation to scholars that so many people still think of these plays as they first came on them in such editions as the old Globe, and it is their pardonable professional deformity to exaggerate the differences superior information can impose. This doesn’t mean that editors should simply mark up the texts of other editors, only that the scope for sensible change is less than it pleases them to believe. Bowers was right to warn his colleagues that the incredible refinement of bibliographical technique in our time might not succeed in making many substantial emendations to Shakespeare’s text, and it is worth remembering that Hinman’s work on the printing of the Folio, which made available vast quantities of new information about the way it was put together, the habits of various compositors, etc, didn’t, so far as I know, establish or make certain a single new reading.

The principal effort of the main Oxford editors has gone into the text. Each play, in both original and modern spelling, is given a page of bland appreciative comment (‘In Titus Andronicus, as in his early history plays, Shakespeare is at his most successful in the expression of grief and the portrayal of vigorously energetic evil’; in Romeo and Juliet ‘Shakespeare’s mastery over a wide range of verbal styles combines with his psychological perceptiveness to create a richer gallery of memorable characters than in any of his earlier plays,’ etc). It is claimed that the chronological order of composition has been freshly established, but the evidence isn’t available here, and will, when it appears, chiefly concern the order of the earlier plays. Some may think it a nuisance to have the second and third parts of Henry VI before the first part, with Titus in between: it was sensible of Heminge and Condell to give them in the usual order. Pericles, a special case because it survives only in a quarto which is dependent on memorial reconstruction, and does not appear in the First Folio, is the kind of challenge the editors most enjoy. They offer a ‘reconstructed’ text in the modern-spelling volume and a diplomatic reprint in the original-spelling, which excuses them from the job of inventing Elizabethan-compositor’s spellings.

The ancillary material in the original-spelling version is identical with that in the modern version, except for an impressive essay on ‘The Spelling and Punctuation of Shakespeare’s Time’ by Vivian Salmon. It is mostly what it says it is, but it is also a defence of the original-spelling version, suggesting that the funny appearance of the old spelling is less off-putting than Professor Wells supposed in 1979. As far as can be told, Shakespeare was himself a fanciful speller (‘scilens’ for ‘silence’) and despite some efforts to control them the compositors made almost as free; certainly they added letters to fill lines, substituted letters they had for letters they had temporarily run out of, and so on. Sometimes they can be identified by their orthographical habits. Nor were they the only re-spellers; occasionally they worked from copy prepared by a scribe, notably the well-known Ralph Crane, who also had his own way with a text; and there were proof correctors, who perhaps added their pennyworth. But Ms Salmon is still alert to the evidence for genuine Shakespearian spellings, and hopes that when fully armed with the Textual Companion as well as ‘a critical old-spelling edition’ we shall join in the game of identifying ‘patterns of Shakespearian orthography’. Well, perhaps. At least if we make ourselves aware of spelling conventions we can judge for ourselves ‘the meanings and the forms which Shakespeare and his fellow actors found appropriate to a printed text, unaffected by the cumulative modernisations of generations of editors’. ‘The general reader’ is included in this invitation, though it is not explained why he would be worse-off with Hinman’s photographs.

My overall impression is that despite the undoubted learning and skill of the editors the edition suffers in two main ways. First, it is altogether too anxious to surprise us. There may be an argument for calling Falstaff ‘Oldcastle’ in 1 Henry IV, and there is certainly an argument against doing so, but it is characteristic of this edition that the former argument prevails.4 There is a case for calling 2 and 3 Henry VI The First Part of the Contention and Richard Duke of York: it isn’t a very good case, since by the time 1 Henry VI was written and so entitled it is evident that the second and third parts had acquired the titles by which they have been known ever since, but these editors have committed themselves to the odder choice of title and of order.

There will doubtless be a vast amount of detailed comment and criticism when the whole operation is completed. At the moment it seems that the single-play volumes and the modern-spelling collection, despite the absence of apparatus and the continual self-advertisement, can hold their own against most opposition. The raison d’être of the old-spelling collection is doubtful: for defences of its silent alterations and for readings at present lacking any explanatory support we shall have to wait for the Textual Companion. At £50 it is going to be an expensive supplement, but if you want to know exactly what is going on in the Collected Editions you will have to have it. One can’t help thinking there were some flaws in the strategic thinking of those who planned this large publishing enterprise.