13b. Richard III: by 1597


previously unpublished; © the estate of eric sams


(Numbers in bold refer to the Documents section and bibliographical citations refer to the Bibliography of The Real Shakespeare II, on-line edition)




     The play that Henslowe's Diary calls 'Buckingham', perhaps an early version of Richard III, was performed by Sussex's Men at the end of 1593 and the beginning of 1594. That latter year also saw the anonymous publication of Titus Andronicus, performed by Sussex's Men at the same period. But no Richard III was registered until October 1597 (233), two months after Richard II (232). By the end of that year, both those plays had appeared in print, again anonymously but with due title-page acknowledgement to the Lord Chamberlain's Men.

     So everyone knew who their playwright was. By this time, he would be increasingly aware of the power exerted by his name and fame. He may well have been personally responsible for the splendidly orotund sub-title of Richard III (233), so reminiscent of Contention and True Tragedy. Such lively descriptions would have attracted readers, as the plays themselves had attracted spectators. And their author would feel it was high time that his works were duly acknowledged on their title-pages, above the publisher and printer, especially when in 1598 he was praised by Francis Meres for his narrative poems and sonnets and identified as the most excellent for stage comedy and tragedy (242). So when the two Richards were duly re-issued in that same year (248,249) they both proudly announced 'By William Shakespeare'. A third Quarto edition of Richard III (306) was published as 'Newly augmented' in 1602, though no substantive additions can be identified. After transfer of the rights (327) a fourth, fifth and sixth edition appeared in 1605 (354), 1612 (422) and 1622 (497) respectively. All six represented the same basic text, with occasional variants and amendments, and cumulative errors; so did the seventh and eighth editions of 1629 and 1634 (517540). But the very different version printed in the History section of the 1623 First Folio (502) and reprinted in the further Folios of 1632, 1663 and 1685 (534,595 and 615) contains thousands of differences in spelling, punctuation, stage-directions and speech-prefixes as well as hundreds of textual variants. Specifically, the 3,485-line Quarto version (Q) includes twenty-seven passages, totalling some forty speech-lines, not found in the 3,871-line Folio version (F), which has some 50 passages, totalling some 200 speech-lines, not found in Q (data in Hammond 1981, 2).

     'The respective origin and authority of the first Quarto and first Folio texts of Richard III is perhaps the most difficult question which presents itself to an editor of Shakespeare' (Clark and Wright 1864, v. xvi). But only to an editor; everyone else will sensibly infer from the facts that Q 1597 was turned into F 1623 by detailed revision, not vice versa by the 'bad memories' of 'actors' for their own profit and that of equally venal administrators who registered a stolen fake and publishers and printers who sold it to a gullible public as a genuine Shakespeare play over a period of thirty-five years – and had the nerve to announce their own names1 and addresses into the bargain. For a start, 1597 is earlier than 1623. Nor would the F text have been preserved unpublished for some thirty years. Nor should five-act plays and theatre troupes and performances and eccentric behaviour and bad memories and worse morals and purported explanations and reversible datings ever have been conjured into existence just to gratify a private whim. Besides, the Tudor theatre relied on memory and its book trade on probity; a conspiracy theory that equates both professions with knavery and their public with gullibility borders on paranoia.

      But it has been believed and asserted for some seventy years. 'I regard [F as the earlier]'...some of [Q1's] textual features...are such as appear in "reported" plays' (Chambers 1930 i, 298). Q1 is an abridgement of F reconstructed from memory by the entire company that had acted F (Patrick 1936). 'No doubt I am prejudiced in Dr. Patrick's favour, since his conclusions are very much what I had arrived at myself...the prompt-copy [of F] had been temporarily mislaid or the actors were on tour in the country and out of reach of their stock of "books"' (Greg 1938, 118, 120). [Q1 was] 'almost certainly printed from a "report" and not from a transcript of the author's manuscript' (McKerrow 1939, 111). [Q1 was] 'corrupted by some form of memorial transmission' (F. Wilson 1942, 1970, 90). [Q1 was] 'perhaps an abridged memorial reconstruction by [Shakespeare's] company when on tour' (Halliday 1952, 414). [Q1 was] 'as Patrick has now shown ... reconstructed by actors from memory' (Wilson 1954, viii). 'D. Patrick [showed] that Q is indeed none other than a report...made for stage use, very likely in the country...an occasion is suggested by a known provincial tour by the Chamberlain's Men in the summer of 1597, supposing them to have needed to produce a play the prompt-book of which had remained in London' (Greg 1955, 191-2). 'The actors of Shakespeare's company... reconstructed [Q1] from memory' (Eccles 1964, 177). 'Recent editors agree that [Q1] was reconstructed from memory by Shakespeare's company, presumably to replace a lost prompt-book' (Honigmann 1968, 242). Q1 'is some kind of memorially reconstructed version', as 'recent scholarship is now generally agreed' (Evans 1971, 231). 'There are strong reasons for believing [Q1] to have been produced by an act of collective memorial reconstruction involving virtually the entire company' (Hammond 1981, 19). '[Q1 was] an unusually full memorial reconstruction (probably the work of the complete cast)' (Taylor 1981, 34). '[Q is] widely believed to have been compiled by an entire acting company, who delivered their lines from memory to a scribe, in a corporate attempt to reconstruct a missing prompt-book while on a provincial tour' (Jackson 1986, 175). 'The hypothesis of communal reconstruction satisfactorily accounts for [Q1]' (Taylor and Wells 1988, 229). 'Close study of the texts has revealed that Q1 is a "memorial" version of the play ... it has been concluded that this version was prepared by an entire acting company...probably because [Shakespeare's] company lost its copy of the play...the playwright may have had a hand in reconstructing the play' (Boyce 1990, 548). '[Q1 was] apparently put together as a memorial reconstruction by actors who no longer had their prompt-book to consult' (Baker 1996, 748).

     Of course the 'lost prompt-book' and the 'country tour' and the 'abridgement' (absurdly resulting in the longest known Quarto) are all just further inventions. But such phantoms are not only impossible to exorcise from the literary profession; they have come to haunt 'stylometry' as well. Hence the equally insubstantial conclusion, also announced as irrefutable truth (Smith 1993, 202-5), that Ironside c. 1588 cannot be by Shakespeare, because its style, as measured by 'prescribed words', fails to conform with that of 'Richard III' – which (though this fact remains unmentioned) actually means the Folio text first printed in 1623.

     This sample selection could have extended for pages, and years. Yet if no one 'proved' any such notions until 1936, some three and a half centuries later, one obvious reason for that hiatus is that they are all illusory. Indeed, the Oxford editors (Wells 1986, 207) began by ignoring them (as in Wells 1978, 147), and treating Q 1597 as the earlier text. Within two years however the same authorities, without even hinting at their joint conversion, announced that the 1623 Folio text must have come first, because 'Q is, undoubtedly, unusually accurate as memorial texts go; hence the hypothesis of communal reconstruction' (Taylor and Wells 1988, 229). In other words, the question had by then become 'what kind of memorial reconstruction do we have here?'. It just has to be some kind; so the real question (i.e. 'which came first – Q 1597 or F 1623?') is by-passed.

     It might have been confronted by asking how Q could come to contain some 50 lines not found in F (including the famous clock scene, now usually inserted as IV.ii.101-19)? But not even the questioners can see the force of their own objections. Thus the stage directions 'Enter Catesby' in Q and 'Enter Ratcliffe' in F (IV.ii.83, 85-7), with four lines allotted to each respectively 'can hardly have resulted from error in a communal memorial reconstruction (in which Richard, Catesby and Ratcliffe all clearly [sic] participated)' (Taylor and Wells 1988, 244). But the self-evident inference, i.e. that no such 'reconstruction' ever occurred, passes unnoticed. The same applies to Q's attribution to the Earl of Surrey of the line allocated to the latter at F V.iii.3, or the difference of stage business at F IV.iv.514 where Richard himself (not Ratcliffe as in Q) rewards the messenger. Again, 'memorial error' is disproved by the fact that many Q readings are actually closer than F to the Holinshed source (Honigmann 1965, who however remained convinced that 'memorial contamination…played an important part in the transmission of Q1'). Even the skilled analyst who rebutted Patrick's speculations in minute detail (Smidt 1964) later gratuitously supposed some memorial link from F to Q (ibid. 1970).

     Again, why should three hypothetical actors misremember F 'between' as Q 'betwixt'? 'Buckingham' does so thrice (F I.iii.37 and 38, and III.vii.48), and 'Brackenbury' and 'Queen Elizabeth' once each (I.iv.82 and IV.i.20). But revision from Q to F is not only far more economical but far more conformable with these and many other such textual facts; thus 'betwixt', although much the less frequent form throughout the canon (occurring only 55 times all told, as against 241 for 'between') is the favoured form in early plays such as Richard II (three to one) and Titus Andronicus (one to zero). Again, F, with only one exception, has 'mine' instead of 'my' before vowels, as in 'mine eyes', whereas the Q equivalent often has 'my eyes' etc. This is common in the early quartos, whereas it would be odd if not only 'Anne', twice (I.i.148, IV.i.80) but 'Clarence', twice (I.iv.22-3), and 'Boy' (II.i.13) misremembered 'mine' as 'my'. Similarly F 1623 much prefers 'Ah' to 'O' or 'Oh', by a dozen to one, while the latter vowel is a common quarto form. Yet in their hypothetical communal endeavour of reconstruction, many actors are required to misremember F 'ah' as Q 'o/ohs: 'Queen Elizabeth' six times (I.iii.11, II.ii.34 and 71, IV.i.33, IV.iv. 31 and 34) and 'Queen Margaret' (I.iii.162), 'Clarence' (I.iv.66), 'King Edward' (II.i.134), 'Duchess of York' (II.ii.27), 'Children' (II.ii.72) and 'King Richard' himself (IV.ii.8) once each. Finally the F cast has also communally forgotten that its preferred affirmative form is regularly 'ay' (sometimes spelt 'I') rather than 'yea'; memorial reconstruction theory requires the former to be often misremembered as the latter by 'Richard', five times (I.iii.97 and 120, I.iii.135 and 262, IV.iv.249) and by 'Queen Margaret' (I.iii.125), 'Duchess of York' (II.ii.29) and 'Queen Elizabeth' (IV.iv.282) once each.

     Many F lines, such as 'sons and brothers, haught' instead of Q 'kindred hauty', effectively preserve the metre, which is often rough in Q and smooth in F. Others avoid the inadvertent repetition often found in the Q version; thus F 'adversaries' for Q 'enemies' (III.ii.52) follows 'enemies' three lines earlier in both texts. A third category substitutes a more effective word, such as F 'cast in darkness' at I.iii.326, which improves on Q 'laid in darkness' as a description of the fate of Clarence, up-ended in a butt of malmsey.

     Among the more arcane assumptions of modern theory is that some of the F 1623 text was set up from an annotated copy of the third Quarto edition (Q3, 1602). In the present state of scholarly chaos, no one notices that this cuts clean through 'memorial reconstruction'. A given text can have only one origin; and if any part of F was set up from an annotated copy of Q3, at some unknown time between 1602 and 1622, it cannot also have been written, performed and memorially reconstructed before 1597. But nobody knows exactly how much of F was so assembled. The latest estimate (Taylor and Wells 1988, 229) is about 1,000 lines, which should surely suffice to sink whatever wreckage of 'memorial reconstruction' theory is still left afloat. But it also places a sensible inference on firm ground, namely that the annotator was Shakespeare himself, who (like any other author) submitted his revisions on amended copies of extant editions. That entails original composition by 1597 or earlier and revision after 1602 by one known agent for one known purpose instead of the sky-high heaps of hypothesis generated by modern theory.

     This in turn will explain to the Oxford editors (Wells 1986, 250-1, with Taylor) the true nature of what they call the 'Additional Passages' found only in F. These were surely not early Shakespeare speeches supposedly written and acted before 1597 but later discarded or forgotten by him and his colleagues; they were prima facieadded by him to the revised version prepared for Folio publication.

     Now Academia should at last set its own house in order. But over the door of Richard III can now be written 'NOT MEMORIAL RECONSTRUCTION', on professional authority, for the first time in sixty years (Maguire 1996, 324-5). One clear corollary is that its 1597 Quarto is just as authentically Shakespearean as that of its counterpart Richard II, printed in the same year. So perhaps both texts were set up from holograph copy, with far-reaching consequences? That inference could be tested if an old-spelling concordance of Q were publicly available for comparison with the Oxford concordance of Richard II 1597 (ed. Howard-Hill 1971). Meanwhile a preliminary comparison of spelling habits confirms the proposed inference. For example, modern 'precedent' is spelt 'president' in both texts, which would corroborate the disadvantaged countryman's lack of Latin schooling. The Richard III context (III.vi.7) occurs in a law-clerk's soliloquy, which would have been an especially part for Shakespeare to play (RS1 43). Elsewhere, the dominant imagery is that of plants and growth, with added allusions to animals, butchery and the slaughterhouse (Spurgeon 1935, 219-20, 227-8, 232-3), trades associated with fatherhood (RS1 29). As so often in Shakespeare, the verbal background is biographically as well as aesthetically fertile.


1 which are worth recording, though it is too late for apologies and damages: the administrators Barlow and Mann, the printers Sims, Creede and Norton, the booksellers Wise and Law, from 1597 to 1634.



[For the first draft of this chapter, see The Real Shakespeare II, Appendix 4.4]