Variability as a Criterion of Shakespeare's Authorship: the First (1600) edition of King Henry V
previously unpublished; © the estate of eric sams
Shakespeare's mind and hand went together. The hand was equally multifarious; but its clear fingerprints have been left largely unexamined. Even its six signatures deploy more than forty distinct forms of the fifteen letters required. This helps to validate the Thomas More manuscript scene, which contains some seventy separate symbols, although thirty would have sufficed. It bristles with minim errors, flourishes, arbitrary capitals and abbreviations. Its lineation and punctuation are equally unpredictable. On its third page, the same hand writes in a markedly different style of penmanship. Above all, its speech-prefixes, stage directions and text are insistently inconsistent. They include some 130 variants, such as Countrey/Countrie/Country, moo/moor/more/moore [More] seriant/Seriant/seriaunt [sergeant], shreef/Shreiff/shreeve/ Shreiue/Shreue [sheriff]. Prima facie, such multiformity was habitual with Shakespeare. So it should also prove identificatory.
The 130 More variants all occur within 147 lines. A complete playscript penned in the same free style could easily have contained changes and obscurities by the thousand; a compositor's nightmare. Perhaps Shakespeare himself, or a skilled copyist, or the compositors, strove to regularise the copy for printing purposes. Certainly someone did; the early Quarto editions of Shakespeare published between 1594 and 1603 contain far fewer unfamiliar forms than the three More pages, which though undated Such More forms as 'yt' and 'yf' for example were already old-fashioned in the 1590s, and might well be normalised. But many other equally Shakespearean spellings with y instead of i survive, even in the 1623 Folio. On the evidence, he was anything but consistent. Compositors, however, were not critics or editors, as the first editions' unnoticed false starts and redrafts  clearly confirm. Nor were printers editors, while copyists are at best conjectural and were anyhow employed to reproduce what they saw before them. Nobody would have wittingly added to the Shakespearean Niagara of variability. But that would surely spray, or at least sprinkle, the printed texts. As Walter Greg says,  'If we find any considerable number of eccentric or archaic spellings in a print, the likelihood is great that it was set up from the author's own manuscript'.
No expert will be surprised to note that this applies to the 1623 Folio version (F) ofHenry V, which exhibits old spellings (such as 'besmyrcht') by the score, and variant readings by the hundred. Indeed, the professional consensus  and the present thesis are thus alike confirmed; F was set up from the author's own papers, which show his hand. But most editors will be astonished, and should be dismayed, to learn that the 1600 Quarto (Q1) shows the selfsame features. Both Q1 and F share, letter for letter, the assorted speech-prefixes Con/Const/Constable, Dol/Dolphin, Exe/Exeter, Gour/Gower, Her/Herald/Herauld, Kin/King, Pist/Pistoll, Sal/Salisbury and Warwick/Warwicke, as well as unfamiliar spellings (such as 'lugyge') in plenty. Both texts also bestow different names on the same characters (three for Mrs. Quickly, two for Fluellen) and the same name on different characters ('King', whether of England or France). Further, each text contains such variants as answer/answere, apparance/appeard, backe/backward, be/bee, black/blacke, blow/blowes, boy/boyes, camp/campe, certain/certaine, cheare/cheerd, claime/clayme, come/comming, day/dayes, do/doo, doost/dost, dog/dogge, enemie/enemy, fellow/fellowes, follow/followes, foot/foote, France/Fraunce, friend/friendes, go/goe, grow/growes, hangd/hanged, he/hee/heele, hear/heare, heeres/here, herald/herauld, honor/honour, host/hoste, I/ay, keep/keepe, kild/killd, kind/kinde, know/knowes, law/lawe, litle/little, loose/lose, maides/maydens, maiestie/maiesty, maisters/masters, me/mee, mercie/mercy, mightie/mighty, mind/minde, mock/mocke, mouth/mouthes, naught/nought, neck/necke, new/newes, O/oh, oath/oathes, onely/only, pasport/passe, pay/payes, Pistol/Pistoll, quarrel/quarrell, remember/remembred, say/sayes, shalbe/shall be, she/shee, sicke/sicknesse, speak/speake, straight/strait, sun/sunne, tel/tell, thankes/thanks, think/thinke, throat/throate, throte/throtes, to/too, vnckle/vncle, warlike/warres, Warwick/Warwicke, way/wayes, winde/windpipe.
These examples are confined, for simplicity's sake, to clear alternative readings in spoken lines of ordinary English text, excluding all the many additional variants of these same words in each source, and of other words in each source separately, and all italics, capitals, dialect, French, stage-directions and speech-prefixes. Otherwise the total tallies would run into thousands, and call for computerisation. Meanwhile, these few typical examples have selected themselves; they were not chosen to corroborate the thesis that Q1 and F show the same hand. So any such indications can occur only by chance. How is it, then, that so many of them correspond so closely with both sources? Thus, to take only the first three doublets listed above, 'answer' appears at 'arrest them to the answer of the law' and 'answere' at 'let his neck answere', in Q1 as well as F. Similarly with 'chased your blood out of apparance' and 'you appeard to me as a common man', or 'a strait backe' and 'perish the man whose mind is backward now', again in both Q1 and F.
No one aware of such facts could ever have denounced Q1 as a 'Bad Quarto', a mere 'memorial reconstruction'  made by forgetful 'traitor-actors' from their supposed pre-1600 performances of F1623. For why should, or how could, the corrupting activities of corrupt people resemble the known authorial attributes of Shakespeare at all, let alone so very often and in such minute particulars?
The phantoms conjured up for much of this century by 'Bad Quarto' necromancy, though as the air invulnerable (and for much the same reason) may now at last fade in this new light, or at least be dispelled by counter-spells. Sadly, however, the profession has become inured to inventions, which have come to include 'compositors' who not only consult 'Bad Quartos'  but also insert their own special spellings. Thus all such variants as shalbe/shall be  are offloaded on to 'Compositors A and B' or doe/doo  on to 'Compositors X and Y'. Similar variants in King John and Cymbeline are also freely attributed to imaginary 'compositors'.  But as the More manuscript shows, and the facts verifiably confirm, Shakespeare's own hand could write alternatives, by the hundred. There is no need to invent other Tudor artists or artisans to account for such proclivities. The simple explanation of the same vast variability in the same vocabulary found in both Q1 and F is that they were the products of the same hand, in every sense; and if the latter was set from authorial holograph then so was the former.
Their shared authorship is a view inadvertently arrived at via the route of 'memorial reconstruction' bent backwards.  In Q1 1600 it is the Duke of Bourbon, not the Dauphin, who is present at Agincourt. This striking (and historically correct) difference is anticipated in earlier scenes, where 'Q but not F brings this same Duke of Bourbon on to the stage and gives him something to say'. This is 'impossible to account for as an error of memory'. So the 'simplest explanation' is that 'Shakespeare reverted to his original intention'. But this is a complex non-explanation, requiring yet more invention. The evident reason why Q1 has an admitted non-memorial aspect is that it is not memorial.
Taylor also discovered 'creative profanity' in Q1 which is 'probably authorial'.  But what of the self-evident corollary that Q1 actually is authorial? That would also explain how it can 'furnish some apparently authentic readings lost in the Folio, including two whole lines'. 
All such exceptions disprove the rule. The remedy is to treat Q1 as an independent script, not a mere corruption of F. Then their close correspondence becomes manifest; phrase after phrase verifiably stands letter for letter the same in both texts, a result which once again replaces a dozen hypotheses by one actual author. Similarly the shared spellings require the 'traitor-actors' to forget their correct words but magically reconstruct idiosyncratic orthography and vocabulary they had never seen, such as 'owze' [ooze], and 'crasing', which still perplexes experts after four hundred years. There are many other examples of such shared but surprising spellings.
On this front, reinforcements have recently arrived from unexpected quarters. Firstly, Emeritus Professor Ernst Honigmann's recent work The Texts of 'Othello' and Shakespearian Revision, with a supportive foreword from Professor Richard Proudfoot, identifies  many variant spellings found in Othello Q 1622 as typically Shakespearean. No one will deny that some of them may have been used by other playwrights, at a time when orthography was in a state of flux. But, as Professor Honigmann points out, this large stock of spellings 'shared by Q Othello with other Shakespeare texts' cannot be ascribed to coincidence. So no one will be surprised at their copious occurrence throughout the F 1623 text of Henry V in such words as aboord, atchieue, atchieued, atchieuement, Battaile, Battell, beleeue, comming, coppy, Countrey, dam'd [damned], deuest [divest], ey'd [eyed], Iland, Island [Iceland], lowd, Lyon, mary [marry, exclam.], Musique, peece [piece], peeces, pitty, Q. [cue], rellish, sayd [said], sence, shew, Souldier, Souldiers, stroake, Syens [scions], tearmes, vertue.
But once again the same applies to the 1600 Quarto version of King Henry V, which has aboord, atchieue, atchieued, battell, beleeue, boord, caytiffe, comming, countrey, deuest, eyde, grones, grose [gross], Iseland, Lyons, merrited, peece, roules [rolls], sed [said], sence, shew, souldier, souldiers. There is often no reason why such words should appear at all, in either version, let alone in the same odd spellings; Shakespeare was not short of vocabulary or variety. And in fact the very short 1600 Q1 text actually contains a far higher proportion than F of spellings independently identified as Shakespearean. So again the inference is that Q1 1600 was just as authorial in origin as F 1623; and the former is the earlier, as their dates demand. Furthermore, as future researchers can confirm, Q1 also contains scores of other Shakespearean spellings identified by Professor Honigmann, amplified by additional categories described by John Dover Wilson,  Walter Greg,  A.C. Partridge  and other specialists. It even contains variants freely attributed by Gary Taylor and John Jowett  to their own supposed 'compositors', such as alreadie, controlles, diuel [devil], eyther, flye, lynes, readie, souldier, turne, from the first (1623) edition of King John and charitie, dogges, doo, intreat, iniury from the first (1623) edition of Cymbeline. For 'compositors' read Shakespeare; and for 'memorial corruption by traitor-actors' read Henry V Q1, set up from his own early manuscript.
So the establishment of an old-spelling database should now empower computer-literate analysts, including stylometrists, to offer improved assessments of Shakespearean style and attribution. Meanwhile, manual computation and preliminary inspection indicate that the approach adumbrated in this essay will show that not only Q Henry V but all the other so-called 'Bad Quartos' represent (as many researchers have already argued, in vain ) other Shakespearean first versions, not 'memorial reconstructions'. Analogous analyses confirm that Edward III and other unacknowledged plays were also set up from a Shakespeare manuscript. Finally, some of his spellings strongly suggest that such schooling as he received left him - exactly as his contemporaries insisted - anything but literate, let alone Latinate. As a bonus, therefore, he cannot have been Francis Bacon, Lord Verulam or Edward de Vere, Earl of Oxford, or any other well-educated contender. Many current schools of thought, both orthodox and heterodox, are overdue for closure.
 e.g. in Love's Labour's Lost Q 1598 (ed. Kerrigan, New Penguin Shakespeare, 1982, p. 243) or A Midsummer Night's Dream Q 1600 (ed. Wells, ibid., 1967, pp. 166-7).
 W. Greg., Editorial Problems in the First Folio1955, p. 148.
 '...[F] was set up from a manuscript in Shakespeare's hand...', ed. J.D. Wilson, Cambridge 1947 etc., p. 111; 'It has certain characteristics which suggest that it was set up from Shakespeare's own draft...', The Arden Shakespeare, ed. J.H.Walter, 1954 etc.', xxxv; 'It was set up from Shakespeare's manuscript...', New Penguin Shakespeare, ed. A.R. Humphreys 1968 etc., p.221; '[F] shows every every sign of having been printed directly from an authorial manuscripts draft',The Oxford Shakespeare ed. G. Taylor, 1982 etc., p.12; ...printed from Shakespeare's own papers...', S. Wells and G. Taylor, Complete Works, Oxford, 1986, p. 637; '[F was] printed, in all likelihood, from the manuscript that the author submitted to the company...', The New Cambridge Shakespeare, ed. A. Gurr, 1992, p. 213; '[F] was printed from Shakespeare's draft', J. Jones, Shakespeare at Work, 1993, p. 43; '[F appears] to have been set from holograph copy...', G. Taylor, in Shakespeare Reshaped, by G.Taylor and J. Jowett, 1993, p. 242; 'There is general agreement that [F] was printed from Shakespeare's "foul papers"', The Arden Shakespeare, ed. T. W. Craik, 1995, p. 24.
 J.D. Wilson, op. cit., 111-3; J.W. Walter, op.cit., xxxiv-v; A.R. Humphreys loc. cit., G. Blakemore Evans, 'Shakespeare's Text: Approaches and Problems', A New Companion to Shakespeare Studies, Cambridge, 1971, pp. 229-232, 238; Taylor, op. cit. 1982, pp. 20-26; Taylor and Wells op. cit. MacD. Jackson, 'The transmission of Shakespeare's text' in The Cambridge Companion to Shakespeare Studies, 1986, pp. 171-5, 179; Taylor, op. cit., pp. 22-5; Wells and Taylor, loc. cit.; G.Taylor and S.Wells in Henry V, Textual Companion to the Oxford Shakespeare, 1988, pp. 375-85; Gurr op. cit., 220-5; Jones, op. cit., pp. 43, 87; Craik, op. cit., pp. 20-32.
 Q2, 1602; Q3, 1608, attributed to Shakespeare.
 'Punctuation and the compositors of Shakespeare's Sonnets, 1609, The Library, fifth series, 30 (1975), pp. 1-24, followed by G. Blakemore Evans, ed. The Sonnets, New Cambridge Shakespeare, 1996, pp. 276-82.
 H.R. Hoppe, The Bad Quarto of Romeo and Juliet, New York, 1948, pp 46-50.
 G.Taylor and J. Jowett, Shakespeare Reshaped, Oxford 1993, pp. 248-259.
 Taylor, op. cit., 1982, pp. 24-5.
 Taylor, op. cit., 1988, p. 382
 Humphreys, op. cit. 1968, p. 222.
 op. cit., pp. 158-161.
 'Bibliographical Links between the Three [More] pages and the Good Quartos', in Shakespeare's Hand in the Play of Sir Thomas More, 1923, pp. 113-4. and The Manuscript of Shakespeare's 'Hamlet' and the Problems of its Transmission, 1934.
 W. Greg., op. cit., 1955, p. 363.
 A. Partridge, Orthography in Shakespeare and Elizabethan Drama, 1964, p. 58.
 e.g. Y.S. Bains, in Making Sense of the First Quartos of Shakespeare's Romeo and Juliet, Henry V, Merry Wives of Windsor and Hamlet, 1995; E. Sams, in The Real Shakespeare 1995; S. Urkowitz: 'Textual generosity and the printing of Shakespeare's multiple-text plays by contemporary editors', Critical Survey vol. 7 No. 3, 1995, pp. 292-8, and 'Two Versions ofRomeo and Juliet 2.6 and Merry Wives of Windsor 5.5.215-45' in Elizabethan Theater: Essays in Honour of S. Schoenbaum, ed. R.Parker and S. Zitner, 1996, 222-37.