Compositors’ Spelling Preferences and the Integrity of 2 Henry VI

Abstract

The recent New Oxford Shakespeare Authorship Companion states that 2 Henry VI was co-authored by Shakespeare and Marlowe, citing an essay by Paul Vincent arguing that a change in the spelling of the interjection O/Oh showed a change in authorship. Vincent affirmed that ‘playwrights usually favoured one alternative far more than the other, and that the compositors of the First Folio regularly followed their copy text.’ He compiled a list of the variant spellings, arranged by Act and scene, identifying which compositors had set each section. Vincent argued that a passage corresponding to Act 3 in modern editions (the Folio is undivided) containing few instances of either O or Oh signalled an authorial change. When the evidence is arranged according to the sequence in which the two compositors set the pages, no consistent preference can be seen, thus disproving the case for Marlowe’s co-authorship.

In the recent New Oxford Shakespeare Authorship Companion, Gary Taylor and Rory Loughnane state that 2 Henry VI was written by ‘Shakespeare, Christopher Marlowe, and another (?); revised by Shakespeare’. They believe that ‘in recent years, the hypothesis for the play’s co-authorship has gained greater support and is approaching acceptance’.¹ One of their chief witnesses is an essay by Paul Vincent claiming that ‘some inconsistencies in 2 Henry VI’ challenge ‘the prevailing complacency in modern scholarship that accepts the play as having been written solely by Shakespeare’.² Of course, ‘complacency’ is a loaded term, as if those scholars who accept the play’s integrity were guilty of ignoring an established fact; this is not the case here. Vincent supported his challenge with three types of evidence, one of which was bibliographical, the other two linguistic and literary.³ He based his bibliographical argument on two assumptions, that Elizabethan dramatists had clear preferences in spelling the interjection O/Oh, and that compositors faithfully preserved their preference.⁴ Taylor and Loughnane cite this argument to support their claim that Marlowe wrote parts of the play.

Vincent’s reference-point for his assumptions was an essay by Gary Taylor and John Jowett that surveyed the occurrence of these variant spellings in both Shakespeare quarto and Folio editions, together with plays by other dramatists derived from Ralph Crane’s transcripts.⁵ Taylor and Jowett tabulated the occurrence of each form in the Shakespeare ‘good’ quartos (Table II. 1) and in twenty-five First Folio texts ‘apparently set from manuscript copy’ (Table II. 2). These comprised five plays alleged to have been set from ‘foul papers’, eleven from ‘scribal copy’ (three of which are of uncertain status), three from ‘fair copy’, and six from Crane transcripts.⁶ They claimed that ‘Shakespeare’s good quartos show an overwhelming preference for the spelling O’ (250), although they recorded that Titus Andronicus (1594) has 19 (+1) instances of O and 39 (+5) of Oh (the figures in brackets are for justified lines), while Richard II (1597) has 8 instances of O and 26 (+2) of Oh. Taylor and Jowett dub the Oh spellings ‘anomalies’, a term reused for their list of the twenty-five Folio plays ‘apparently set from manuscript copy’. According to their counts, ten of these plays were ‘anomalies’—although it might be stretching the term ‘anomaly’ a shade too far when referring to forty per cent of a sample. Two of the five plays copied by Crane favoured the Oh spelling: Two Gentlemen 12:33, Measure for Measure 7:56 (including the totals for justified lines). Of the remainder, several did so, and by considerable margins: the O/Oh ratio for Coriolanus is 12:40, for Hamlet 39:69, for Othello 52:117, for Antony 13:83, and for Cymbeline 22:60. Among these ‘anomalies’, Othello, Antony, and Cymbeline were ‘all set by Compositors B and E’, thus: ‘compositorial interference cannot be ruled out; but it would require us to assume that both compositors simultaneously developed a very sudden preference for Oh, which one of them abandoned when he set Troilus. Such a scenario is implausible’.⁷ Since ‘the manuscripts consulted for Folio Othello and Cymbeline (and Hamlet) were scribal’, Taylor and Jowett concluded that these ‘anomalies’ can be put down to ‘an unknown scribe’s spelling preference’. They noted another ‘anomaly’ in King John, where spellings change at line 1893. However, since ‘a single compositor spelt the same word differently in the two parts of the text’, they deduced that the play was set from ‘a scribal transcript, in which a second scribe took over towards the end of

4.2 … In any case, no one will want to attribute the disparity in Oh spellings here to Shakespeare’.⁸

My aim in tracing Taylor and Jowett’s line of argument is not to dispute their data for individual plays, but to highlight how their discussion exonerated the compositors of having caused these so-called ‘anomalies’—which they could then ascribe either to scribal preferences or to the nature of the copy. This assumption is of long standing. For instance, in the 1987 Textual Companion Wells and Taylor, discussing Cymbeline, noted a theory of Honigmann, ‘confirmed’ by Jowett and Taylor, ‘that at some stage in its transmission the manuscript … was affected by a change of hands between

2.4.152 and 2.5.0.1/1154 and 1154.1. The most striking evidence for this is the abrupt change in the spelling of ‘O(h)’, which cannot be compositorial in origin’.⁹ I wish to challenge their assumption that compositor spellings regularly followed their copy, a thesis that they use to attribute authorship generally, not just in the case of 2 Henry VI. Taylor and Jowett argued that the ‘simplest explanation for … all but one of these discrepancies, in quartos and Folio alike, is chronological’.¹⁰ It is often the case that ‘the simplest’ explanation is usually the one that first occurs to the researcher, but that further enquiry can discredit it. In the Folio, seven of Shakespeare’s earliest plays (the three parts of Henry VI, Titus, Errors, Taming, and Two Gentlemen) prefer Oh (162 instances) to O (85) an ‘agreement’, they suggested, that can only be explained by ‘their date of composition’.¹¹ Surprisingly, they then shifted their focus from Jaggard’s shop, where the Folio was edited and printed in 1622–23, to that of Richard Field, Shakespeare’s fellow Stratfordian, who set up as a London stationer and printed the two narrative poems in 1593–94.

Venus and Adonis … tolerates Oh (10 of 25 occurrences); but The Rape of Lucrece firmly comes down on the side of O (43 of 45 occurrences). We find it difficult to avoid the conclusion that Shakespeare’s habitual spelling of this word changed, perhaps consciously, c. 1593.¹²

The first problem with this ‘conclusion’ concerns Taylor and Jowett’s data. In Venus they count 15 uses of O and 10 of Oh, but Henry Woudhuysen has given the properly differentiated figures: ‘both O and Oh are used at the beginnings of lines (11 and 8 times respectively’ while oh occurs twice within a line, and ô occurs 4 times.¹³ For Lucrece they count 41 (+2) instances of O

and 2 instances of Oh, but Woudhuysen gives the correct figures: ‘Neither Oh nor oh appears in Lucrece, instead O occurs 36 times in the poem, usually to begin a line.’ On one occasion O follows an opening bracket, and ô occurs 6 times within a line.¹⁴

The more serious issue is that Taylor and Jowett grounded their ‘conclusion’ of Shakespeare’s having changed his spelling preference in 1594 on the evidence of a quarto printed and edited by Richard Field. Field has long been recognized for his high standards in most aspects of printing. R.

B. McKerrow described him as ‘a more than usually careful printer’,¹⁵ and

A. E. M. Kirkwood, in his survey of Field’s career, judged him to have been ‘painstaking and thorough in all his work’.¹⁶ Apprenticed to Thomas Vautrollier for six years, ‘the habits formed in Vautrollier’s workroom persisted’, but Field developed an individual ‘sense of fitness and proportion’ in his ‘choice and arrangement of type for title-pages’ and his use of ornaments and initials.¹⁷ ‘Field printed with ease in several languages other than English … the accuracy of Field’s compositors in setting texts in Latin called forth tributes of praise’ from the authors. Able to set passages in Greek and Hebrew type in varying quantities and sizes, ‘Field printed a few books in French, half a dozen in Italian, five in Spanish, and one in Welsh’. Field also stood out among London printers ‘for his skill in setting up mathematical treatises in which the regular processes of composition and type-arrangement were complicated by the necessity for diagrams’.¹⁸

It is well known that Field’s compositors followed house style when it came to the accidentals of the texts they were setting. We can see how they exercised control over the spelling and punctuation of Sir John Harington’s translation of Ariosto’s Orlando Furioso when he printed it in 1591. In 1923

W. W. Greg drew attention to a manuscript in the British Museum, partly scribal, partly autograph, which had been used by Field in preparing his edition, containing his mark-up for the signatures and pagination.¹⁹ Greg described Harington’s spelling as ‘archaic and irregular’, while Field’s was ‘regular and modern’. Harington’s punctuation was ‘curiously mechanical’, each stanza containing ‘a colon at the end of the fourth and a full stop at the end of the eighth line, quite regardless of the sense … Field evidently had no intention of following his copy in this respect’, making wholesale changes.²⁰ Greg noted that in Field’s printing house, one of the best in London, ‘the

compositors had a recognized standard of their own in the matter of spelling and to a lesser extent in punctuation’.²¹ His conclusions were endorsed in 1978, when Philip Gaskell compared Harington’s manuscript with Field’s printed text, using a different section.²² He found 21 changes of punctuation in these 48 lines of verse and far more numerous changes to the spelling, ‘for here Field’s compositors normalized vigorously’, altering the spelling of 149 of the 390 words in the extract. ‘Most of these changes were from an old-fashioned form used by Harington to one that is still in use’ (such as ‘yow’ to ‘you’, ‘yf’ to ‘if’), but since no one in 1591 could predict the course that English orthography would take, Gaskell noted that Field’s compositors also made ‘changes from a form which has survived to one which has not’.²³ From these two well-known studies Taylor and Jowett might have expected that the text of Shakespeare’s narrative poems would have been affected by Field’s practices as an ‘interventionist editor’. That possibility would have become unavoidable if they had known the work of A. C. Partridge, who published an essay in 1964 discussing ‘Shakespeare’s Apparent Orthography in Venus and Adonis and some Early Quartos’.²⁴ By ‘apparent’ Partridge implied that many of the spelling choices in that text were made by Field, an argument that he developed in two full-length chapters in a book that—despite its having been published for the Bibliographical Society of the University of Virginia—seems not to have attracted much attention from textual critics, A Substantive Grammar of Shakespeare’s Nondramatic Texts (1976).²⁵ This is partly Partridge’s fault for having chosen such a forbidding title.²⁶ The book discusses orthography (spelling, punctuation, elision), syntax, prosody, vocabulary, typography, and compositorial practices in the two narrative poems, ‘The Phoenix and the Turtle’, and the Sonnets. From Partridge’s meticulous study of Venus (pp. 23–43) and Lucrece (pp. 44–77), I select his comments on orthography to illustrate Field’s interventions in Shakespeare’s text. Rather than compositors’ faithfully reproducing their copy-spellings, Partridge found in both texts abundant evidence ‘that the same compositor was capable of

alternative spellings’.²⁷ He noted that in both punctuation and spelling Shakespeare’s was old-fashioned, Field in touch with modern practice, a difference that accounts for the ‘ubiquity of alternative spellings; some probably owe their origin to the copy’ (p. 59). Partridge listed over fifty words for which he could differentiate authorial from compositorial spellings (pp. 59–63). The influence of house style on Lucrece can be seen from a copy of the corrected state of the first edition, now in the Bodleian, which demonstrates ‘the overzealous care the text received as it passed through the press’ (p. 48). In line 24, ‘As is the morning silver melting dew’, Field’s proof-reader ‘unwisely added an s to morning, converting it to a possessive singular. Shakespeare used the noun adjectivally’. In line 30, ‘What needeth then Appologie be made’, the proof-reader made the ‘unnecessary correction Apologies’, adding further erroneous changes at lines 50 and 125–26.²⁸ We may now pose the leading question: which is more likely, that Field’s control of every aspect of printing produced the remarkable unanimity by which neither Oh nor oh occurs in Lucrece, while O occurs 36 times, usually to begin a line; or that, as Taylor and Jowett argued, ‘Shakespeare’s habitual spelling of this word changed, perhaps consciously, c. 1593’? In any case, if it had changed, we would expect O to be the dominant form for the remaining twenty years of his career, which is demonstrably not the case.

Returning to 2 Henry VI,²⁹ which Taylor and Jowett classified as having been set from ‘fair copy’, their figures were O, 17 (+9); Oh, 16 (+2),³⁰ with O slightly dominant.

Vincent made two assumptions concerning these variant spellings, based on Taylor and Jowett’s essay. He affirmed that ‘playwrights usually favoured one alternative far more than the other, and that the compositors of the First Folio regularly followed their copy text’ (p. 270).³¹ Accordingly, Vincent constructed a table (p. 271) identifying the compositor who set each instance, following Blayney’s revised identifications in the second edition of

the Norton facsimile.³² Vincent listed the O/Oh spellings with Through Line Numbering (consecutively numbering each line of text, excluding stage directions) in a vertical table, side-by-side with Act and scene divisions (as defined by the Oxford Shakespeare). He claimed to find ‘some of the O/Oh variations equating to these superimposed divisions’ (p. 271). But the Folio text of 2 Henry VI is undivided, and it is not clear that any bibliographical or authorial significance can be attached to the uneven distribution of these spellings according to the modern division (not necessarily unanimous) into Acts and scenes. Correlating them in this way simply begs the question.

To avoid prejudging the issue I shall set them out according to their place in the Folio text, where 2 Henry VI occupies sigs. m2^v–o3^v (pages 120–46) of the Histories. The Histories consists of 22 quires, 66 sheets, and 132 formes, each containing two double-column pages of sixty-seven lines. It used to be thought that plays were set seriatim, in reading order. But we now know that printers adopted a method that allowed two compositors to work simultaneously. This involved ‘casting off copy’, dividing the text into the quantity needed for each page and treating the quire as a unit, with the three pages folded inside each other. Composition would begin on the outer sheet, with pages 1 and 12, followed by the reverse side, pages 2 and 11, then 3 and 10, 4 and 9, 5 and 8, finally 6 and 7.³³ Charlton Hinman showed that most of the plays were set by two compositors, and if the compositors worked simultaneously, ‘steadily turning out a sheet a day [they] could have set this section of the book’—that is, 16 quires (sigs. d–s, including ^χgg)—’in about fifty working days’.³⁴ But Hinman discovered that many formes were set by only one man, with output reduced to half speed, suggesting a total of ‘about sixty-five working days’ for the whole section, completed between January and March 1623.³⁵ We have no evidence as to other printing jobs that Jaggard took on between spring 1622 and autumn 1623, but we do know the sequence in which the compositors set this play.³⁶ Where Vincent arranged his data according to the reading order, I have decided to display it according to the order in which the pages were actually set. This arrangement reveals that compositors sometimes changed their preferred spelling from one page to the next, or even within the same page.

In Table 1, I take the compositor identification from Hinman–Blayney; for the distribution of O/Oh spellings, and for the indication whether lines

are justified or not (y = yes, n = no), I am indebted to Pervez Rizvi’s database.³⁷

As it turned out, most instances of this exclamation occurred in the sections of the text set by B (who set fifteen pages), which also display more variation than in those of A (who set twelve). For his first page (m3^v) B preferred O, but on his second page (m3^r) he used both forms. On m2^v he

again chose O, but on n4^v he set all five occurrences as Oh. He chose that form again for his next page (5^r) for the first three occurrences, only to shift to O for the fourth. He stayed with O on n5^v and n6^r, but on n6^v he switched to Oh. His final three pages on which the exclamation appears (o3^r, o2^v, and o2^r) show an indifferent mixture of both forms. Compositor A was less given to switching, preferring O on his first three pages (m5^r, m5^v, and m6^r). On n4^r he changed to Oh, repeating that form three times on n3^r, and returning to his preference on n1^r, with three instances of O. Compositor A changed his spelling preference twice, Compositor B changed his six times. There is no visible rationale for these variants. Nor were they affected by the line spacings. On o3^r Compositor B set four unjustified lines, spelling two of them O and two Oh, an indifference he showed again on n5^r, o2^v, and o2^r. Compositor A was equally indifferent, spelling both forms on unjustified m3^r, keeping to Oh three times on m6^r, although setting both justified and unjustified lines. He preferred O three times on m6^r and O on n1^r, although both pages mixed justified and unjustified lines.

Unfortunately, by correlating the distribution of forms with modern editors’ act and scene divisions, Paul Vincent was unable to see the indifference showed by both compositors to spelling choices, whatever the line spacing. He claimed great significance from his data for the distribution of choices in the Folio text corresponding to modern Act 3, scenes 2 and 3:

[Act III scene i] contains not one instance of O or Oh, despite its being the play’s second longest, running to 383 lines … [III. ii] has nine examples of O in 416 lines, or one every forty-six lines. On average, the appearance rate in the whole play is one every seventy-six lines … the almost complete absence of O in act III, with some 832 lines of verse, is not easy to explain, nor is the fact that Oh appears so sparsely outside act III.³⁸

Vincent concluded that ‘the O/ Oh distribution’s significance is strengthened by the fact that the alternating pattern cannot be attributed to compositorial preferences: both set both variants in their stints’, hence the alternations were authorial.³⁹ This is the standard explanation of the Oxford editors in 2016 as thirty years earlier. Vincent inferred an ‘alternating pattern’ while denying ‘compositorial preferences’; I argue that there is no pattern, and that the compositors had no preferences. The fact that Act 3 scenes 2 and 3 contain one O and 12 Ohs, is just as random as the fact that 3.1 contains no examples of either, but does have 7 instances of a different interjection, Ah, which also occurs once in 3.2 and 3.3. (Vincent found these variations ‘not easy to explain’, which risks assuming that they can be explained in terms of consistent compositorial reproduction of authorial preferences.)

Vincent supported this claim in a footnote, recording that ‘Separate studies have confirmed both compositors habitually followed the spelling of their copy, with respect to this interjection’.⁴⁰ Of the studies he cited, one was a joint essay by John Jowett and Gary Taylor concerning Richard II, in which they discussed the claim that, although Q3 (1598) was accepted as the main source for F, some passages may derive from Q5 (1615).⁴¹ Jowett and Taylor examined Compositor A’s setting of the abdication episode (4.1.154– 318), listing seventeen spellings ‘for which Folio Compositor A departs from the spellings of Q5, against his own preferences—or his own indifference— elsewhere’ (p. 152; my italics). These ‘anomalies’ include O (Q5) as against Oh (F), but since Jowett and Taylor were dealing with plays set from manuscript they could ‘not know the spelling of the word in his copy’. But we do know that Compositor A worked from printed copy when setting Richard II and Richard III, where Jowett and Taylor recorded that ‘elsewhere Compositor A followed copy O seventeen times, but he did alter to Oh on three occasions (Richard II 1215, 1426; Richard III 710)’.⁴² This is a small instance, but, as their word ‘alter’ indicates, Compositor A evidently expressed a preference here.

The other study cited by Vincent to support his contention that both Folio compositors habitually followed their copy for the spelling variant O/ Oh was by Taylor alone, evaluating the evidence that would distinguish between the stints of Compositors B and E in King Lear.⁴³ Taylor recorded that ‘In the Comedies Compositor B simply reproduces his copy spelling without exception: 41 O/2 Oh’, as again in 1 Henry IV, with ‘26 O/ 2 Oh’, but that his ‘remaining plays from printed copy are slightly more complicated’.⁴⁴ Others might think that setting plays from printed copy should be considered as quite different from setting them from manuscript (of any kind); it is not clear that extrapolating from one to the other can be done without damage. Taylor gave two tables, one for Compositor B’s figures for five early tragedies and histories, the other for three late tragedies.⁴⁵ I simplify his data in Table 2, giving only the totals for each group.

These findings provide much material for discussion, but the evidence of 133 compositor alterations in these eight plays negates Vincent’s claim that these two ‘studies have confirmed both compositors habitually followed their copy, with respect to the spelling of this interjection’.

Indeed, the time has come to query whether Shakespeare’s compositors regarded this variant spelling as—to use Jowett and Taylor’s terms—a matter of ‘preference’, or ‘indifference’. It may have escaped general notice that in 2016, the compositor analysis procedure developed by Charlton Hinman based on individual spelling preferences was turned upside down by Pervez Rizvi in an essay that deserves the epithet seminal.⁴⁶ Rizvi performed the laborious task of manually transferring all of Hinman’s data to computer spreadsheets. This allowed him to make a global evaluation of each compositor’s spelling preferences, with surprising results. He began with Compositor B, who set ‘almost half the Folio pages; he is a workman everyone believes in; and we think he can be instantly recognized by his distinctive do-go-heere preferences’. The Folio pages currently assigned to him indicate that ‘he set 225 pages up to and including q1^r, which occurs in Act 4 of 3 Henry VI, and 224 pages in the rest of the Folio. By drawing a dividing line at the end of q1^r, we can split his work into two approximately equal parts’.⁴⁷ Rizvi listed thirteen words set by Compositor B before and after page q1^r, showing that ‘in every one of these thirteen examples the compositor’s spelling preference reverses as we move from the first to the second group of pages’.⁴⁸ From the spelling prethee he moved to prythee; he changed the -y spelling for beauty to -ie and did the same for eight other words.⁴⁹ Compositor B was not the only one to reverse his spellings, all of Jaggard’s compositors did so in varying amounts, depending on the number of pages they set.⁵⁰ It is as if a compositor decided at some point that he had set a sufficient number of spelling x, and it was time to change to spelling y. Rizvi concluded that the accepted ‘spelling-preference technique … is not safe to use for attribution. It mistakenly sees significance in data that, on closer examination, appears only to show that volatility was normal in compositors’ spellings’. Not only did compositors ‘have no discernible preferences for some words, alternating between spelling variants on successive plays, but that even when they did have preferences, they sometimes reversed them overnight’.⁵¹ That much was evident from my specimen table

of preferences in the setting of 2 Henry VI. B was capable of changing his spelling within the same page: on three pages (n5^r, o3^r, and o2^v), he starts with Oh and then changes to O; on two pages (m3^r and o2^r), he starts with O and then changes to Oh.

Rizvi returned to these variants later in his essay, giving a more detailed account of the larger picture of O and Oh spelling preferences in F1. He noted that in two Folio plays set from quartos, Much Ado about Nothing and A Midsummer Night’s Dream, the compositors reproduced their copy with a predominance of O spellings (Ado, 46 O and no Oh—another remarkable instance of unanimity; Dream, 55 O and 6 Oh). As Rizvi summed up his data,

The typesetting of both plays has been attributed to Compositors B, C, and D, all of whom set a much higher proportion of Oh spellings elsewhere in the Folio than in these plays … Compositor B set 598 O spellings and 483 Oh spellings; Compositor C set 122 and sixty-eight respectively, and Compositor D set fifty-one and thirty-one. Excluding justified lines reduces the totals to 468 and 405, ninety-five and fifty-five, and thirty-four and twenty.⁵²

This indisputable evidence negates any claim that compositors used the variant O/ Oh spelling consistently, or that changes from one form to another could indicate a change of authorship.

In their attempt to find rhyme or reason in compositors’ spelling preferences, Taylor and Jowett frequently used the term ‘anomaly’ to indicate any variation from the patterns they thought that they could discern, in one instance using the word to describe forty per cent of their data. Discussing Venus and Adonis they used a related term, noting that the poem ‘tolerates’ the Oh spelling, a word that they may have borrowed from Charlton Hinman. Discussing the setting of As You Like It by Compositors B, C, and D, Hinman had identified D as preferring the spelling doe to do, but Rizvi pointed out that ‘in three of Compositor D’s four pages non-preferred spellings predominate’, with 11 instances of do to 5 of doe, ‘an aberration we cannot explain without access to the source copy spellings’.⁵³ On five of the twelve pages in another quire ‘the non-preferred spellings are in the majority’, especially do. Hinman argued that Compositor D ‘was “tolerant” of [do] while truly preferring doe’, to which Rizvi made this trenchant objection:

Such appeals to a compositor’s supposed tolerance are disturbingly numerous, both in Hinman’s book and in papers by subsequent investigators. It is hard to see how any rigor can be maintained if we … cite spellings to support our attributions when they suit us but ignore them when they do not and claim the compositor must have tolerated them. There is an alternative explanation: the compositor is not the man

we thought he was or … did not have the preferences we thought he had. All too often it appears that Folio scholars have succumbed to unintentional confirmation bias when identifying compositors based on spelling preferences.⁵⁴

This is a much larger issue than the supposed co-authorship of 2 Henry VI, yet the misuse of compositors’ spelling preferences by Vincent, Taylor, and Loughnane in ascribing that play to ‘Shakespeare, Christopher Marlowe, and another (?)’ proves the accuracy of Rizvi’s critique. It also adds further scholarly evidence undermining the Authorship Companion’s claim that ‘the hypothesis for the play’s co-authorship … is approaching acceptance’. Since that claim was made in 2016, however, a series of studies in peer-reviewed journals has thoroughly discredited the methods on which the New Oxford Shakespeare’s attributions have been based.⁵⁵ Far from there being any ‘prevailing complacency’ in modern scholarship treating 2 Henry VI as the unassisted work of Shakespeare, that is a position based on the reliable evidence of its appearance in the First Folio and the absence of any evidence to the contrary.

I should like to thank Henry Woudhuysen and the two anonymous readers for this journal for their helpful suggestions and corrections.

Footnotes