Shakespeare, or Bottom?

The Myth of "Memorial Reconstruction"

Encounter, January 1989; © the estate of eric sams 



In June 1987, I proposed in these pages that the 1986 Oxford Shakespeare Complete Works (CW for short) already needed complete revision. It has now been joined by its long-delayed Textual Companion (TC), which takes the same wrong turning into oblivion via such outmoded myths as "memorial reconstruction" by "actors who, like Bottom, attempted to play the parts of all the characters" (p. 26).1  Like Bottom indeed; a fundamental yet asinine fantasy from a literary dream-world. If you believe, clap your hands! You will be joining in the frenzied applause of a few academics, amid silence or dissent from everyone else over the last 400 years. But Bottom has become the basis for sky-high piles of editorial assumptions. Overturn it, and these towering but tottering edifices will collapse, leaving all our multi-volume editions exposed as radically wrong about the dates, the life, the works, the lot.2 This article aims to apply some vigorous leverage.

     I should again declare an interest. My edition of the Tudor drama Edmund Ironsideattributed it to the young Shakespeare, with due acknowledgment to the Renaissance specialist E. B. Everitt who in 1954 had identified its British Library manuscript as a Shakespeare autograph. In 1986, that same identification was independently announced, on quite different grounds, by the professional palaeographer Charles Hamilton, who specialises in studies of Shakespeare's handwriting.' Hamilton has publicly testified that he would stake his lifetime reputation as a documents expert on the correctness of his conclusion. The Oxford editors, who pro­fess no such expertise, had already rejected Ironsideon stylistic grounds, together with most of Henry VI, Part I, much of Henry VIII, Macbeth,Pericles and Timon of Athens, and perhaps some of Henry VI, Parts II and III, The Taming of the Shrew, and Titus Andronicus. They have thus pledged their own credit that Hamilton (like the 1623 First Folio editors) must have been much mistaken. His attribution ofIronside. the editors of the Textual Companion admonish us (p. 138), "depends upon acceptance of his controversial views on a number of other alleged autographs (1985) views which are too recent to have been subjected to thorough scholarly scrutiny".


No such scruples restrained these same Oxford editors from announcing their own controversial Shakespeare attribution of the poem "Shall I die?" as a new discovery, long before it had been subjected to independent scrutiny of any kind. Further, their present affirmations are false. As they must have seen in the source they cite, Hamilton's Ironsideidentification in fact follows from his 1985 demonstration that Shakespeare's will is an autograph document  an assessment first advanced by a Lord Chancellor of England in 1859.4 If that is really too recent for scholarly scrutiny, the reason must be a shortage of scholarship, not time.

     Meanwhile, how has editorial opinion contrived to out­weigh palaeographic expertise? Let the Companion itself (p. 7) be our guide:


“Editorial controversy, like all other forms of discourse, is an instrument of power...Editors are the pimps of discourse.”


So they can certainly go astray, and mislead others. They and their associates should therefore temper their power with re­sponsibility. Their own conclusion is that "we look forward to our future obsolescence". Unfortunately for them and their readers, that was built in; it has overtaken them far sooner than they foresaw.

    Their fate was sealed some 60 years ago, when the Oxford faith in "memorial reconstruction by actors" (MRA) was first founded.5 Before then, no one had ever heard of it. Now it is freely feigned to be the cause of the so-called "Bad Quartos", once well known to Shakespeareans as his own early versions which he later revised, not MRA at all. They are now thrice lost  from the canon; to the public; and by academic intransigence. If they were readily available for study, the general reader could soon decide whether they emanated from Shakespeare or from various Bottoms. These are the exclu­sive choices; heads or tails. Take for example the 1594 Taming of A Shrew, the 1594-95 versions ofHenry VI, Parts II and III, the 1603 Hamlet, and the 1608 King Lear. All five have been condemned as "Bad Quartos" and blamed on the bad memories and bad behaviour of imaginary Tudor actors. The consequent chaos in Shakespeare studies should instead be blamed on the bad logic and bad example of actual modern editors. Two earlier culprits are Peter Alexander (1894-1969) and George Duthie (1915-67), whose records may now be examined.


Alexander's conquest of his colleagues confirms the Oxford power-struggle model of editorship. In 1926 he fantasised that the 1594 Shrew play was an MRA of its allegedly pre-existing 1623 Folio counterpart. The latter was accordingly dragged backwards through time for 30 years and fitted out with imaginary theatre companies, performances, audiences, popularity and piratical plagiarists, all without any objective evidence whatever. This self-refuting hallucination entails absurd and impossible consequences, for example that the memorising actors had forgotten the title, location, characters and almost every line of the Shakespeare play in which they had allegedly been acting. The scholarly consensus of bellwether and flock, misled by Alexander's lifelong advocacy, conferred such respectability on his disgraceful nonsense that various versions of it are still on sale today  for example in the separate Oxford edition of The Taming of the Shrew.6 Yet its own general editor has been publicly rejecting, this particular MRA theory for the past six years. Even more amazingly, the Oxford editorial team accepts and admires the same Alexander's equally preposterous MRA explanations of the 1594-95 Henry VI plays.7

    In 1949, George Duthie fantasised that the 1608 King Lear was an MRA begotten by an entire theatre company upon its allegedly pre-existing 1623 Folio counterpart. The Bottoms had put their heads together. Eleven years later, even Duthie came to realise that this position was ridiculous. Yet he was still so obsessed by MRA that he simply switched Bottoms, and plumped for "two boy actors" instead of the whole company as the source of "memorial error". All such MRA theories of Lear 1608 are now publicly rejected by the editors of the Textual Companion (for instance, on p. 509) — without, however, mentioning either Duthie’s absurd 1949 theories, or their 1960 retraction, or the equally absurd theories with which he replaced them. Yet they amazingly accept and admire the same Duthie's equally preposterous MRA explanation of Hamlet 1603, and even swallow the "entire company" MRA explanation of the 1597 Richard III.8

     Typically, each of the chief Oxford editors offers a quite different justification for the dogma that dictates their canon and chronology. Like Alexander and Duthie in their books, Wells and Taylor treat MRA as true, while also conceding that it is only a theory. Indeed, the Oxford editors in effect confess that MRA is no more than their own personal feeling, as my italics serve to underline: “...the texts which seem to us to have been created by memorial reconstruction...the hypothesis that such texts derive from a process of aural transmission...the agents of transmission seem to have been actors...” (TC, pp. 26-7). In other words, the hypo­thetical agents are inferred from the hypothetical process, and conversely. Equally hypothetical, and circular, although these essential supplementary assumptions remain largely unmentioned, are the plays, companies, performances, audiences, players, plagiarisms, piracies, times, places, methods, reasons, motives, circumstances and consequences concerned.

     General readers will recognise at a glance what general editors have failed to see in 60 years: that all this is a tissue of fabrications, not a material hypothesis. Its seams split apart at the first test. If it is true, then the actual named printers and publishers as well as the imaginary actors must have betrayed Shakespeare for money. The actual paying audiences and readers in Tudor times, as well as editors, administrators, critics and commentators, then and for generations there­after, could not tell their Shakespeare from their Bottoms. Further, the "actors" were, as we shall see, fools as well as knaves. Whole professions and populations have to be de­nounced and derided for the sake of MRA. The only category exempt from such condemnation is that of modern academic editors, although they alone (e.g. George Duthie, on his own confession of folly) are known to have deserved it.


Now for the official proof of MRA, as set forth in the Textual Companion, which like Hamlet proceeds from "seems" to "is" in the same breath. Chief among the so-called anomalies allegedly explained by "memorial reconstruction" are the fluctuations in "the qual­ity of the memorial reconstruction" (TC, p. 27). The ques­tion-begging circularity is breathtakingly blatant. As the next illogical step, the whole complex kaleidoscope of base­less MRA speculation is ecstatically hailed as (of all things) "single" and "simple". On these incredible grounds it is vaunted over its exclusive alternative, the truly single and simple explanation of one revising hand. Ockham's Razor can never have been so suicidally applied.

     Stanley Wells, in response to my public challenges about MRA,9 chose not to go along with his own Companion in any such attempted justification. Instead, he relied solely on a supposed consensus. “A large and varied body of scholars subscribe to” MRA. But only the few editors thus promoted will be impressed by that, especially in view of the proviso: “perhaps they are all wrong, but they are all there”.

    In fact they are anything but all there. They are all over the place, loudly denouncing each other's opinions, including their various MRA theories.10 They are not "varied" in any other respect: they are all university academics. Nor are they in any sense "large": they are a petty minority among Shakespeareans. Such Oxford claims to broad and high consensus are sheer effrontery. They derive solely from the passionate preconviction that what "we believe", or what "all scholars", "most scholars", "most reasonable modern scholars", and so forth, believe (CW, TCpassim) must somehow be true. This express-train of thought can be seen running non­stop from literary opinion to historical truth, feeling to fact, "seems" to "is", "we believe" to "therefore" (TC, p. 398), on editorial power alone. No wonder it has gone off the rails with calamitous consequences.

     What if this much-vaunted consensus may rightly be de­fined as what happens when a caucus swallows a lot of non­sense? Even Stanley Wells concedes that he and his fellow-editors may be "all wrong". He has also confirmed, in the Oxford Magazine, that he is "not ashamed of speaking in terms of belief, opinion, view and hypothesis". But he should be, in so far as his beliefs, opinions, views and hypotheses are or may be wrong because they are offered to and accepted by the paying public, world-wide, as fact and scholarship. Sooner or later, even the most ill-informed and partisan of readers and reviewers will demand some historical evidence for the crucial dogma of MRA. They will then discover that no such evidence has ever existed. On the contrary; all the documented signposts, such as "1623" printed on the Folio, and "William Shakespeare" on the so-called Bad Quartos, point permanently and implacably in the opposite direction. The various MRA speculations are thus not only unnecessary, untestable, incompatible and unevidenced, but also counter-factual. They are, further, flagrantly self-contradictory.

     Thus "memorial reconstruction by actors" cannot really be memorial at all, because that would require the names Po­lonius and Reynaldo (for example) to be remembered as Corambis and Montano. Nor can any process of reconstruc­tion conceivably have created the hundreds of otherwise un­known lines which ex hypothesi were just not there to be re­constructed. Nor can actors be rationally recruited for their opportunities, memories and intelligence and then dismissed as deficient in all three. And how could these amnesiac actors have written plays which were and are widely held to be good enough for Shakespeare? Take for example the lone bit-part player who suddenly decided to reconstruct the whole of Hamlet, which ex hypothesi he had never heard, without any preparation whatsoever. Yet he somehow created a five-act play, duly published as the First Quarto, 1603, by William Shakespeare. This unusual actor then belatedly realised that he could hardly hope to stage Hamlet without the Prince, or indeed without the rest of the cast either. That would be too absurd, even for editors. So he has to be reinforced by an entire new troupe...who chanced to be passing.


Dozens of specialist experts, in scores of sources, have offered hundreds of such objections to MRA and all its works, over the last 50 years. All but eight of those specialists, and every single one of their detailed counter-arguments, remain unmentioned throughout 2,000 pages of Oxford scholarship. For example, the works cited in connection with Henry VI, Part II (TC, p. 178), include Alexander's MRA theories but omit all the articles, reviews, theses and books that refute them.11 In that context, though, we are at least offered two points in favour of them  the only such samples to be found anywhere in two massive tomes that revolve around MRA as the axle on which their editorial methodology explicitly turns (e.g. "our dating presumes .. . memorial reconstruction", TC, pp. 111, 112). Both these items are labelled "linchpins". Ominously, both were manufactured by Alexander in 1929. Close inspection confirms that the bandwagon must soon crash.

     The first linchpin is just nine words long: "the second son [of Edward III] was Edmund Langley, Duke of York". In the 1594 version of Henry VI, Part II (line 899), this phrase is allotted to a Duke of York who is claiming the crown by direct descent from the third son. Therefore, we are instructed, his claim was already made when he declared that his ducal ancestor was the second son, and all the rest of his long speech is superfluous. Let the Oxford editor William Montgomery explain, in his own words (TC, p. 175). "No one who understood what he was writing  that is, no author  could have made this error, but someone parroting someone else's work, of which he himself had but a dim understanding, easily could. Once having established that part of [the 1594 version] is clearly a report it is natural to suppose that the rest of the text is also a report." The argument typically runs, or rather pole-vaults, from what possibly may, to what undoubtedly must, have happened. So, from nine words dated 1594, we experts can infallibly infer the necessary pre-existence c.1590 of an entire five-act play, equate it with Shakespeare's Henry VI, Part II, as first published in 1623, i.e. antedate it by 30 years, explain its complex interconnection with the 1594 version, and identify the cause as aural transmission by a professional Tudor actor who was also a born fool.

     The same argument also proves, beyond all doubt, that the Oxford editor William Montgomery cannot exist, because no real person could have reasoned thus. In fact, all that hap­pened was that the Quarto writer confused two sons, a slip duly corrected in later editions. He may be forgiven for fail­ing to notice his error at all, let alone foresee the fantastic inferences that would be forced out of it 400 years later by sheer pressure of preconviction. Even if the Quarto character is naively imagined as a real-life historical personage, in­stead of words on a page, he cannot really be claiming descent from the wrong son. That would vitiate the entire play, not just one speech. Nor in fact is any Duke of York necessarily the ancestor of any later Duke of York, so no such descent was necessarily entailed at all. Such notions could only occur to one text-editor echoing another. Yet it is the Tudor playwright who is unkindly called a dim parrot.

     The second linchpin is a little longer, but no whit less wob­bly. The 1595 version of Henry VI, Part III, includes (line 1,898) a complaint that King Edward’s allocation of heiresses favours his wife's family instead of his own brothers: “Lord Scales did . . . have the daughter of the Lord Bonfield”. This, we are instructed, “mangles history and has no point of its own”. Thus spoke Alexander. So, as before, we experts can infallibly infer everything about two plays from one line in one of them. But again this miraculous haul of hypothesis consists solely of red herrings. Of course the phrase has a point, namely exemplification. The alleged mangling of his­tory merely means that the Tudor dramatist has failed his Oxford examination, for writing his own chronicle play instead of the history textbook required by Alexander, at whose jaunty piping generations of editors have disappeared in the wrong direction, far beyond the reach of reason.

     They must now, at whatever loss of face or revenue, retrace their steps and start again, in the opposite sense. As an accredited authority on logic patiently pointed out: "How often have I said to you that when you have eliminated the impossible, whatever remains, however improbable, must be the truth?”12 MRA is impossible. Therefore the only rational alternative must prevail: the plays fashionably rejected as "Bad Quartos", whether we like them or not, represent Shakespeare's own early versions, which he later revised. As I have shown in my earlier Encounter article ("The Oxford or the Stratford Shakespeare?", June 1987), he is still the Stratford Shakespeare identified by his first biographers and editors, not the Oxford Shakespeare invented by their modern successors. He began with popular plays, such as the early Hamlet lampooned by Nashe in 1589, and The True Tragedy of Richard Duke of York parodied by Greene in 1592. Their thousands of unfamiliar lines define his earliest style, which can also be identified in other plays such as The Troublesome Reign of King John and Edward III, both c.1590.13 The course of his development can thus be charted in detail, from historical documents instead of literary opinion. For that journey, the Oxford Works will be abandoned as useless baggage, and the Companion sent packing.

     The new routes and landmarks will call for new books and editions, some of which have already been published, or are in course of preparation, by professionally qualified Shakespeareans.14 Other such studies will, I hope, be stimulated by this essay, on which I shall be grateful to receive critical comment, whether in public or in private, from all interested parties  and especially readers, playgoers, actors and directors. Shakespeare is far too serious a subject to be left to editors.






1William Shakespeare: The Complete Works,general editors Stanley Wells and Gary Taylor (Oxford University Press, 1986). William Shakespeare: A Textual Companion. By Stanley Wells and Gary Taylor, with John Jowett and William Montgomery. Oxford University Press, £60.00.


2Notably The Arden Shakespeare (Methuen), general editor Richard Proudfoot; The New Cambridge Shakespeare (Cambridge University Press), general editor Philip Brockbank; The New Penguin Shakespeare (Penguin Books), general editor Stanley Wells; The Oxford Shakespeare (Oxford University Press), general editor Stanley Wells; and The Signet Classic Shakespeare (New American Library), general editor Sylvan Barnet.


3Edmund Ironside: Shakespeare's Lost Play, ed. Eric Sams (Fourth Estate, 1985; reissued with a new preface, Wildwood House, 1986). E. B. Everitt, The Young Shakespeare: Studies in Documen­tary Evidence, Anglistica II (1954). Charles Hamilton, as quoted by Don Corathers in "Much Ado...", Dramatics no. 58 (1986), pp. 14­19; Charles Hamilton, In Search of Shakespeare: A Reconnaissance into the Poet's Life and Handwriting (1985).


4"Shall I die?" was discovered in the Bodleian Library, Oxford, by Gary Taylor, "after looking in an index under Shakespeare, William" (Sunday Times, 15 December 1985); see also M. C. Crum, First-Line Index of English Poetry in Manuscripts of the Bodleian Library (Oxford University Press, 1969), under Shakespeare, William. John, Lord Campbell, Shakespeare's Legal Acquirements Considered (1859), p. 103; see also J. C. Jeaffreson, "A New View of Shakespeare's Will", The Athenaeum, no. 2,844 (29 April 1882), pp. 539­40, and J. P. Yeatman,Is Shakespeare's Will Holographic? (1901).


5E.g. by P. Alexander, "The Taming of a Shrew", Times Literary Supplement, 16 September 1926; see also his Shakespeare's Life and Art (1939), pp. 35, 43, 246; Complete Works of William Shakespeare (1951, 2nd ed., 1981), p. 154; A Shakespeare Primer (1951), p. 149; Shakespeare (1964), p. 121.


6The Taming of the Shrew, ed. H. J. Oliver, Oxford Shakespeare (1982, 2/ 1984), p. 22: "The analysis of the preceding pages of parts of the text of The Taming of a Shrew has demonstrated [sic] that in some way the play must be a reconstruction, from memory, of another that we know in the form in which it appears in the First Folio as The Taming of the Shrew..."


The Oxford Shakespeare general editor Stanley Wells has repeatedly said that A Shrewis variously, a "derivative" (Times Liter­ary Supplement, 29 October 1982, p. 1,193), an "adaptation" (back cover of the 1984 reissue) or an "imitation" (Complete Works, 1986, p. 29) of the Folio play; i.e. it is not a memorial reconstruction at all.


7P. Alexander, Shakespeare's Henry VI and Richard III (1929).


8George Duthie, Shakespeare's King Lear: A Critical Edition (1949): King Lear, ed. George Duthie and John Dover Wilson ( 1960). pp. 131-39; George Duthie, The "Bad" Quarto of Hamlet I 1941).


The 1597 Richard III is described (TC, p. 229) as a "memorial text" which is "unusually accurate…hence the hypothesis of communal reconstruction".


9The controversy is contained in successive issues of the Oxford Magazine, Nos. 24-28 (1987), under the heading "Revisionist Shakespeare"; the correspondence was then closed.


10Thus the MRA theorist Philip Edwards is reproached by the MRA theorist Gary Taylor for "an intrinsically implausible" hypothesis aboutHamlet (TC, p. 401) and a "self-contradictory, unsubstantiated, unparalleled and unlikely" hypothesis about Pericles (TC, p. 557).


11E.g. C. Greer, "The York and Lancaster Quarto-Folio Sequence". PMLA 48 (1933), pp. 655-705; A. Richardson, The First Part of the Contention (Diss., Yale, 1953); C. Prouty, The Contention and Shakespeare's 2 Henry VI(1954); H. Craig, A New Look at Shakespeare's Quartos (1964); Eric Sams, "Shakespeare's Text and Common Sense”, The Times Literary Supplement, 2 September 1983. pp. 933-4.


12A. Conan Doyle, The Sign of Four (1890).


13Eric Sams, "The Troublesome Wrangle over King John", Notes and Queries, vol. 35, no. 1 (March 1988), pp. 41-44. See The Problem of the Reign of King Edward III, by Eliot Slater,Cambridge University Press, £35.00; yet another contribution to "the consensus of investigators that Edward III deserves a place in the Shakespeare canon", as the Oxford editors now acknowledge (TC, p. 136). They add (TC, p. 137): "If we had attempted a thorough reinvestiga­tion of candidates for inclusion in the early dramatic canon, it would have begun with Edward III". My own edition of that play, now in preparation, stresses its manifold and striking affinities with Edmund Ironside, a resemblance first pointed out in general terms by E. B. Everitt in The Young Shapespeare.


14I. Robinson, Richard II and Woodstock(1988) has an appendix on the case for the latter play's Shakespearean authorship. Steven Urkowitz — in "Well sayd olde Mole", inShakespeare Study Today (ed. G. Ziegler, 1988), pp. 37-70; "I am not made of stone", inRenaissance and Reformation, no. 10 (1986), pp. 79-93; and "Reconsidering the Relationship of Quarto and Folio Texts of Richard III" inEnglish Literary Renaissance, no. 16 (1986), pp. 442-466  advocates the early version, not the MRA explanation, of so-called Bad Quartos.