Mistaken Methodology

Letter to the Editor

London Review of Books, 14 June 1990

May I renew my unanswered protests (LRB, 6 March 1986 and 19 May 1988) about modern Shakespeare editing? The latest Oxford volume, The Merry Wives of Windsor, exemplifies the mistaken methodology I indicted. Its editor, T. W. Craik, who teaches English literature, is far from impressed by the credentials of the critic John Dennis or the poet laureate Nicholas Rowe. Indeed, he calmly calls them both liars, in effect, just for claiming in 1702 and 1709 respectively that the play was commissioned by Queen Elizabeth. John Dryden and William Davenant fare no better: if they were the sources of the story, 'either of them may have invented it'.

   Professor Craik, having thus spurred early evidence as invention, proceeds to embrace late invention as evidence. 'It is now accepted that [the 1602 Quarto] is a corrupt text reconstructed from memory [of the 1623 Folio version] and, where memory failed, from invention.' In fact, every operative word of that sentence is false. 'Corrupt', 'reconstructed', 'memory', and the pre-existence of the 1623 text, are themselves inventions. Even the 'invention' is an invention.

   So is 'accepted'. In fact, this theory has been rejected ever since Walter Greg announced it in 1910. Even Greg himself rejected it, in 1928. By 1942 he had withdrawn his notion that the 'reporter' was the actor who played the Host. T. W. Craik is silent about this damaging decapitation of 'memorial' reconstruction, which is still running about in Oxford circles. He inadvertently defines its fatal fallacy, as follows: "The relatively accurate reporting of the Host's speeches in particular gives good reason for believing that the actor who played this role was either the sole reporter or the principal one.' Even without my italics, everyone can see that the assumed reporter is assumed from the assumed reporting, and conversely.

   We are then told that 'there is no need to set down all the evidence' for this theory: fortunately for it, since none exists. It also defies the historical facts, the rules of reasoning, and three hundred years of consensus to the contrary. Prima facie the 1602 version ofMerry Wives represents a comedy by William Shakespeare, played by his company before his sovereign, as its title page tells us. There are no factual grounds whatever for asserting that it was a corrupt travesty botched up by piratical actor-reporters for sale to dishonest publishers, and thence to a gullible public, as the quite different play that had been presented on the London stage. This is not just invention but fairy-tale fantasy.

   It is also refuted by the textual facts. The 1623 edition has over 2700 lines, as against only 1600 in the earlier version, which contains some five hundred otherwise unknown. Only 120 are the same in both sources. So the hypothetical Elizabethan memoriser of a Jacobean text completely forgot most of it, misremembered almost all the rest, and added five lines of his own - which is absurd. Shakespeare is, however, well-known to have revised his own work. It is time to get his dates right and give him his plays back, including such typical descriptions of Falstaff as 'an unreasonable woolsack' or 'a bladder of iniquity' which modern editors attribute to an imaginary reporter.


Eric Sams

Sanderstead, Surrey