Picture of Gabriel Egan G a b r i e l   E g a n  .  com

Year's Work in English Studies 2002: Shakespeare: Editions and Textual Studies

Firstly, a correction. In last year's review, the introduction to John D. Cox and Eric Rasmussen's Arden Shakespeare 3 Henry 6 was criticized for offering the example of Q1 King Lear as a 'bad' quarto that had, since the heyday of New Bibliography, been rehabilitated (p. 164); this reviewer complained that no-one had ever made the claim that it was a 'bad' quarto (p. 265). In fact, Leo Kirschbaum claimed it as a 'bad' quarto in 1938 (1938) and W. W. Greg concurred in 1942 (Greg 1942, 49-101). The claim has resurfaced occasionally since then, as I should have known.

    Books outweighed journal articles this year: five major scholarly editions appeared and Brian Vickers published two door-stopper monographs on the topic, 2002b and 2002a. For The Oxford Shakespeare, Simon Burrow edited 2002e, Charles Whitworth edited 2002d, and Roger Warren edited 2002b. For The Arden Shakespeare, Charles Forker edited 2002c and David Scott Kastan edited 2002a. Burrow's is a huge book and the hardbound review copy was not well printed, using cheaper paper than is normal for this series and showing the kind of horizontal discontinuities produced by a worn-out office inkjet printer. Happily, a copy consulted in the Shakespeare Institute library had neither of these flaws, although it did have a number of pages of either poor registration or cavalier cropping that reduced either the top or bottom margin to almost nothing. At 158 pages Burrow's introduction is longer that usual for the series, but then it has much more to cover than a single play edition would. The edition includes poems that Burrow thinks are probably not by Shakespeare but which his early readers thought were his, and Burrow describes the "chief aim" of the edition as being to ask "What sort of poet was Shakespeare?" (p. 1). This strikes an odd note as a raison d'être, since editors usually think of that question as secondary to a chief aim of presenting the works to modern readers.

    By putting the narrative poems in the same covers as the Sonnets, Burrow hopes to encourage readers to think of them together as an oeuvre and he suspects that Heminges and Condell left them out of the 1623 Folio because they were men of the theatre and because the poems were selling well in their own right. Perpetuating this split in the canon, eighteenth-century editions of the complete works tended to make the poems an add-on in the last volume of a series. In the poems taken as an oeuvre Burrow, sees a thematic connection, for they "repeatedly meditate on the perverse effects and consequences of sexual desire, on sacrifice and self-sacrifice, on the ways in which a relationship of sexual passion might objectify or enslave both the desirer and the desired, and they repeatedly complicate simple binary distinctions between male and female" (p. 5). Moreover, they not only show another side to the theatrical Shakespeare with which we are familiar, but in fact show "the foundational thought which underpins his dramatic works"

    Burrow's introduction is divided into a section for each major poetic work, starting with Venus and Adonis and proceeding chronologically. Richard Field was an obvious choice as printer of Venus and Adonis for he specialized in high-quality reproduction of poetical works and what we would call literary theory. Q1 of 1593 is the only substantive printing--all others are derivative--and paradoxically, the success of the poem made it peripheral to the canon: as it was too popular in its own right to be included in the Folio (p. 7). Burrow interprets the flood of Stationers' Register entries of plays in 1594 as the players "attempting to use print to realize their assets" during plague closure (p. 9) without responding to Peter W. M. Blayney's now often-quoted objection that the flood comes a little too late for that and must in fact be part of an advertising campaign to announce the reopening of the theatres after the plague subsided (1997, 385-86). Burrow surveys the dedication and concludes that Shakespeare ceased to have anything to do with Southampton in 1594, who was not the W.H. of the Sonnets dedication (pp. 10-15); one minor slip here is that Alan Bray's name is misspelled "Allan Bray" (p. 14n2). Burrow sees Venus and Adonis as part of a new tradition of standalone erotic poems made from tales in Ovid's Metamorphoses and gives a highly intelligent literary critical reading of it (pp. 20-40). Lucrece was also printed by Richard Field, but not for himself but for John Harrison, and like Venus and Adonis it was dedicated to Southampton. Burrow reports of the 1594 quarto that "There are two different states of several gatherings of Q1, one unrevised (Qa) and the other which contains a number of 'press variants' (Qb) . . ." (p. 42), but presumably he means different states of several formes (individual sides of a sheet, not the whole sheet) since, as R. B. McKerrow long ago pointed out (Barnes 1904, xiii-xviii), the forme was the unit of press-correction. If it so happens that both sides of a sheet show the same state of press correction (either unrevised or revised), as Burrow seems to imply by writing about "different states of several gatherings" and that the Bodleian and Yale copies "retain earlier readings in sigs. B, I, and K; other copies contain early readings in sigs. I and/or K" (p. 42n3), Burrow ought to explain how he thinks such an unlikelihood occurred. It would seem to require that after the first forme was wrought-off, the sheets containing the revised state of that forme were kept separate from the rest in order that they might be perfected using only the revised state of the forme printed on the other side. Burrow lists the lines affected by press correction (24, 31, 50, 125, 126, 1182, 1335, 1350) and importantly concludes that some might have been made with reference to authorial copy.

    Burrow notes that the Argument differs from the poem itself in details of the story, and focusses on the Republican outcome while the poem ends with the personal: "Tarquin's everlasting banishment". Of course, that distinction is partly created by the editing, for no possessive apostrophe appears in the quarto's last line ("Tarqvins"), and one could argue as Katherine Duncan-Jones did in an article reviewed here last year (2001) that editors should end the poem with "Tarquins' everlasting banishment" on the evidence of the Argument itself. For Burrow, however, the differences between the Argument and the poem derive from Shakespeare's interest in the multiple perspectives one can take on a single story, encompassing the personal and the social (pp. 48-50). Burrow's literary criticism of Lucrece is excellent throughout and draws on the important scholarship of Heather Dubrow and Nancy Vickers, and he documents well the Elizabethan laws and prejudices about rape--a personal violation, and separately an attack on another man's property--and the extraordinary fact that conception 'proved' consent; thus a victim's suicide could be a way to forestall others concluding from conception that she had consented (pp. 66-68).

    Burrow sees The Passionate Pilgrim (1599) as William Jaggard's attempt to persuade the buying public that he had "hitherto hidden works of Shakespeare the poet" (p. 5) and outlines four possible ways that a couple of Shakespeare's sonnets (numbers 138 and 144) got into this collection of 20 poems that Burrow reproduces in its entirety (pp. 76-77). We should not, Burrow thinks, simply take Thomas Heywood's word for it (in Apology for Actors, 1612) that Shakespeare was angry at William Jaggard over The Passionate Pilgrim, for this could be just exaggeration motivated by Heywood's ongoing battle with Jaggard over Troia Britannica (pp. 78-79). Burrow surveys the other poems in The Passionate Pilgrim and concludes on purely artistic and stylistic grounds that none of them is Shakespeare. Maintaining the characteristic Oxford tradition, Burrow rejects the familiar name of the poem The Phoenix and the Turtle as something not used until 1807 and instead adopts its first line ('Let the bird of loudest lay') as its title (pp. 82-90). The poem appears as fifth in a group of 14 poems by Marston, Chapman, Jonson, and an unknown man, all appended to "a long, digressive narrative poem called Love's Martyr by an obscure poet called Robert Chester" printed in 1601 by Richard Field; the group of fourteen appeared with its own titlepage within the book. Burrow's literary criticism of 'Let the bird' finds that it is dark and elusive and that no-one can fathom it. In an uncharacteristic slip, Burrow here refers to the "prologue to Act 5 of Henry V" (p. 88) when of course he means the chorus.

    The largest section of Burrow's introduction is, naturally, given to the Sonnets (pp. 91-132). To contextualize the publishing history, Burrow claims that for early modern readers "authorial ownership of intellectual property was an emergent concept (if that)" and that "Copy in this period belonged to the printer who entered it as his or hers in the Stationers' Register" (p. 96). As we shall see, Brian Vickers has challenged this claim about the emergent concept of intellectual property (forever emergent, yet never said to have actually emerged), and certainly the claim about 'copy' is contestable. If one means by 'copy' the physical document embodying a work to be printed, that could of course belong to anyone. If one means, however, 'these words in this order' then, as we shall see, Vickers has strong arguments to show that this always attached to authors. However, if one means 'the right to print these words in this order' (or indeed, any other words that come close to the same meaning or story) then, yes, Stationers' Register entry was a way to prove priority in connection with the Stationers' Company's 'first-come-first-serve' regulations. Burrow thinks that we should read the cryptic dedication of the Sonnets in the light of contemporary poetic play with anonymity and address: "Poems do not target their addressees with exact precision in this period. They were addressed to the person who found themselves addressed by them . . ." (pp. 101-2).

    Burrow gives a useful survey of stylometric analysis of Sonnets, especially MacDonald P. Jackson's work on early/late rare word occurrence (pp. 103-6). From the differences between numbers 138 and 144 as they appear in The Passionate Pilgrim and as they appear in Sonnets, Burrow concluded that, like other sonneteers, Shakespeare revised his poems, and Burrow chooses to print a manuscript version of sonnet number 2 as well as the quarto version, just in case the former represents an early authorial version, although he does not really think it does (p. 106-7). Burrow is particularly animated by names, observing that the so-called "young man" of the Sonnets is never called that in them (p. 123) and that nor is the Dark Lady. The latter is only called that "by the prurient in order to make her sound both sexy and upper-crust, and (within careful racially determined limits) exotic" (p. 131). This seems a bit harsh: until not long ago polite children were told to address any woman they did not know as 'lady'. Regarding sexual orientation, Burrow offers the standard post-Foucauldian line (". . .  no one in the period would have sought to define their identity by their sexual activity", p. 125) supported by citations of Alan Bray and Bruce Smith, but since one of his themes is his introduction is the multiplicity of readings they stimulate Burrow might have mentioned the increasingly popular counter-view that we have all be misled by Foucault on this point. (Another small slip in connection with this: Jonathan Goldberg's book is called Queering the Renaissance not Queering the English Renaissance, p. 126n2.)

    Burrow thinks that A Lover's Complaint rightly occupies its place after the Sonnets in the 1609 quarto, although a recent article by Brian Vickers in the Times Literary Supplement to be reviewed next year claims that it is not by Shakespeare at all but by John Davies of Hereford. An entire section of Burrow's introduction is given over to "Poems Attributed to Shakespeare in the Seventeenth Century" (pp. 146-58) in which he explains why some non-Shakespearian works are included in this book, albeit relegated to an appendix. The point is "to show the kinds of poems which Shakespeare was thought to have written by his contemporaries and near-contemporaries" (p. 147). Weighing the evidence, Burrow observes that while printers obviously stood to gain from putting a false name on a book, "scribes and copyists of manuscript miscellanies did not have a financial stake in misascription". Surely they did if there was a market in circulating manuscripts, as Vickers points out in connection with Gary Taylor's claim for 'Shall I die?'. Burrow comes clean and admits that he thinks A Funerall Elegye and 'Shall I die' are not by Shakespeare, but only in the case of the former is he convinced strongly enough to exclude it from the book. Of course, A Funerall Elegye was not actually attributed to Shakespeare--Burrow's criterion for admittance--but rather to "W.S.", and with commendable frankness Burrow admits that ". . . the criteria for including poems in this edition have been determined in this way partly because I find the attribution of A Funerall Elegye to Shakespeare improbable" (p. 154). Burrow surveys reasons to think A Funerall Elegye is by John Ford and observes that the tests (specifically, Ward E. Y. Elliott and Robert J. Valenza's) that reject it also reject A Lover's Complaint, and comments that this "may be an indication of the difficulty of adjusting such tests to take account of inflexions to a dominant style which can result from genre and conscious pastiche" (p. 155). This rather misses the fundamental point of such tests, which is to exclude that which varies with genre and imitation and detect those things that do not change. Burrows also thinks that stylometry is a limited tool when "the theoretical pool of authors is not limited, and where the full extent of that pool is not knowable", to which one might well respond that unless the author of A Funerall Elegye published nothing else, his work will be somewhere in Chadwyck-Healey's Literature Online (LION) database. It would be worth someone's time to check how may poets in LION only ever published one work, although presumably one would be hampered by the difficulty of determining whether many things currently attributed to 'Anon' are in fact by the same writer.

    In particular editorial choices Burrow is generally unsurprising. Lines 1013-14 of Venus and Adonis he renders as "Tells him of trophies, statues, tombs, and stories / His victories, his triumphs, and his glories", whereas Q has a comma after "stories". Taking it out, Burrow following Malone and agreeing with Charles Jaspar Sisson (1956, 207) and the Oxford Complete Works, makes "stories" a verb. The press corrections in Lucrece Burrow treats individually and on their particular merits. Thus his line 31 is "What needeth then apology be made", which is the Qa reading, while Qb has "apologies". Since the 'correction' to "apologies" could be attributed to the press corrector, Sisson thought there "no need for it". Likewise at line 50 Burrow gives "When at Collatium this false lord arrived", the Qa reading, instead of the press-corrected Qb reading of "Collatia", the classically more-correct form; Burrow clearly agrees with Sisson that it was an educated press-corrector's alteration. Sisson thought that accepting this point entailed rejecting other Qb corrections, including those in lines 125-6 that Burrow renders as "And every one to rest themselves betake, / Save thieves, and cares, and troubled minds that wake", in which "themselves betake" and "minds that wake" are the Qb readings and Qa has "himself betake" and "minds that wakes". Sisson thought Qa's "minds that wakes" typically Shakespearian and nothing for editors to get upset about, while finding Qb's "every one . . . themselves" awkward, but Burrow (here departing from the Oxford Complete Works) pointed out that "Shakespeare often takes indefinite pronouns (here every one) to imply more than one person" (p. 250) and hence the press-corrected text might also reflect what Shakespeare wrote. Press corrections are a tricky business, and arguably the criterion should be 'does this correction look clearly more likely to be authorial than compositorial?' Since press correction was usually done without reference to copy, the uncorrected state is, on average, more likely to be right and one should, Fredson Bowers argued (1952), only use the corrected state when one is certain that it resulted from consultation of copy.

    In sonnet 39 Burrow follows Q in printing "Which time and thoughts so sweetly dost deceive", as Sisson argued and against Malone and the Oxford Complete Works in which singular present indicative "dost" was changed to the plural "doth" to suit the plural subject of the verb, "time and thoughts".  Sisson argued that "O absence" three lines earlier is the antecedent of an understood (not expressed) 'thou' that is the singular subject of the verb "deceive" (hence singular "dost deceive") of which "time and thoughts" are the object. Elsewhere a couple of recent suggestions are not actioned, so that in sonnet 69 Burrow prints "Then, churls, their thoughts" and does not mention N. F. Blake's suggestion (1998) that this is a genitive form equivalent to "churls' thoughts", and in sonnet 106 he prints "They had not skill enough your worth to sing" rather than Q's "still enough" that Kenji Go has defended (2001). Burrow's sonnet 146 begins "Poor soul, the centre of my sinful earth, / Spoiled by these rebel powers" where Q notoriously begins the second line by repeating the last three words of the first. Burrow credits "This edition" as the first to use his emendation, and attributes the idea to "(conj. Spence)", but there is no Spence in the list of works cited and of abbreviations, yet elsewhere--such as Lucrece line 1662 "wreathèd ] DYCE 1866 (conj. Walker)"--one finds the person, here W. S. Walker, in the list of abbreviations.

    The introduction to Charles Whitworth's edition of The Comedy of Errors for The Oxford Shakespeare is short and mostly concerned with the genre and the sources of the play (pp. 1-79). Whitworth finds the play "too well plotted" to be very early Shakespeare (such as The Two Gentlemen of Verona and The Taming of the Shrew and the Henry 6 plays) and observes that its preservation of the unities would have appealed to the classically-educated audience at the presumed Inns of Courts performance and to the queen, who also would have liked his recently-published poetry (pp. 6-7).  Whitworth summarizes internal evidence for the date of composition using the stylometric tables in the Textual Companion to the Oxford Complete Works and, like Burrow (see above), takes the standard line that plays were sold to publishers when a company was short of cash (pp. 8-11). Whitworth suspects that it was the plays' availability in clean Ralph Crane manuscript copy that probably made Heminges and Condell put The Tempest, The Two Gentlemen of Verona, The Merry Wives of Windsor, and Measure for Measure first in the 1623 Folio, since these would thereby form a good opening impression upon the reader. Then came The Comedy of Errors, copy for which is "generally accepted" to be "authorial manuscript, or holograph (foul papers)" (p. 12). The general reader probably should be told that Whitworth's word "or" here means "also known as". Whitworth acknowledges Paul Werstine's demurral from the view that 'foul papers' were something too untidy to perform from, but he falls back on the "orthodox opinion of Chambers, McKerrow and Greg" that there is no evidence that authorial papers were played from. (In truth, finding quotations that show those three men agreeing on this matter would not be straightforward.) However, since The Comedy of Errors was performed "only on a few special occasions", the authorial papers could have served "as promptbook on those rare occasions", and Whitworth gives the standard New Bibliographical reasons for seeing the Folio copy as 'foul papers' (p. 13).

    In the section "Farce, City Comedy and Romance" (pp. 42-59) Whitworth makes much of the frequency with which certain words occur, and here some problems emerge. For example, ". . . the words conjure and conjurer together occur six times in Errors, more than in any other plays in the canon (including all cognate forms of the words)" (pp. 41-42). I count five occurrences in the dialogue of The Contention of York and Lancaster (1.2.76, 1.2.99, 2.1.173, 4.2.92, 5.1.197) if we include conjuration as cognate, and two more in stage directions (1.4.0, 2.3.0). Perhaps Whitworth should have made clear if he meant just in dialogue. Still, if we include conjuration then Henry 5 has six dialogue occurrences too (1.2.29, 2.1.52, 3.1.7, 5.2.286, 5.2.290, 5.2.291) and so does Romeo and Juliet (2.1.6, 2.1.16, 2.1.17, 2.1.26, 2.1.29, 5.3.68) and clearly such assertions depend in part upon the choice of edition used for the counting; mine are all from the electronic edition of the Oxford Complete Works. Whitworth observes that "witch and witches together occur six times in the play" (p. 42), with which I agree but notice that there are the same number in 1 Henry 6 if we include bewitched (1.7.6, 1.7.21, 2.1.18, 3.4.3, 3.7.58, 5.4.5), and seven in The Contention of York and Lancaster (1.1.155, 1.2.75, 1.2.91, 2.1.173, 2.3.7, 2.3.8, 3.2.116) even if we exclude stage directions and speech prefixes and exclude an occurrence of witched (3.2.119), and fully nine just in the dialogue in The Merry Wives of Windsor (4.2.78, 4.2.89, 4.2.158, 4.2.164, 4.2.171, 4.2.178, 4.2.179, 4.5.109, 4.5.113). On the same page, Whitworth claims that "sorcerer(s)/sorceress [occurs] three times (more than in any other play except The Tempest)" but in fact 1 Henry 6 has three occurrences of sorcerers/sorceress (1.1.26, 3.4.3, 5.6.1) plus one of sorcery (2.1.15), and that "Devil/devil's occurs seven times". This last claim is true, but what is its relevance when there are 24 occurrences in the dialogue of 1 Henry 4 (1.2.6, 1.2.113, 1.2.116, 1.2.118, 1.2.120, 1.2.122, 1.3.114, 1.3.123, 1.3.251, 2.5.225, 2.5.340, 2.5.371, 2.5.452, 2.5.492, 3.1.54, 3.1.55, 3.1.56, 3.1.59, 3.1.66, 3.1.153, 3.1.226, 4.1.58, 4.2.19, 4.2.50) and 27 in the dialogue of Othello (1.1.91, 1.1.111, 2.1.114, 2.1.227, 2.1.245, 2.3.277, 2.3.289, 2.3.290, 2.3.301, 2.3.342, 3.3.481, 3.4.42, 3.4.134, 3.4.181, 4.1.6, 4.1.8, 4.1.42, 4.1.145, 4.1.240, 4.1.244, 4.2.38, 5.2.140, 5.2.142, 5.2.227, 5.2.284, 5.2.293, 5.2.307)? On page 43 Whitworth claims that "Beat, beaten and beating, always in the primary sense of physical blows (as opposed to the beating of the heart, for example, or of the sea upon the rocks), occur a total of fourteen times in Errors, more than in any other play in the canon". Whitworth seems to be including stage directions here ("There is further vigorous actions in . . ."), in which case Coriolanus has the highest count, 18 occurrences in all (1.3. 48, 1.4.29, 1.5.7, 1.5.12, 1.5.13, 1.7.40, 1.11.8, 2.3.216, 3.1.228, 3.1.242, 3.3.83, 4.5.50, 4.5.52, 4.5.122, 4.6.56, 5.4.49, 5.6.110, 5.6.150), and even if he means only in dialogue, 13 of these examples stand. I suspect that Whitworth took his numbers from Marvin Spevack's 'Harvard' Concordance since it agrees with all his figures quoted while Bartlett's Concordance, for example, gives 11 for beat/beaten/beating rather than Whitworth's claimed 14. Of course, Spevack's concordance is based on the Riverside Shakespeare edited by Gwynne Blakemore Evans, which is considerably different in hundreds of substantive readings from the Oxford Complete Works which one would expect a book in this series to be using.

    The foregoing quibbles do not constitute an attack on the substance of Whitworth's introduction, which ranges across an admirably broad collection of materials germane to the play. In particular editorial choices too Whitworth is incisive. At 1.1.17 he prints "Syracusan" instead of F's "Syracusian", explaining this as being "in the interest of consistency with the noun form Syracusa" and points out that an -ian ending can be one or two syllables. At 1.1.41 Whitworth restores the correct name to the city of Epidamnus (the Latin nominative form), rejecting F's "Epidamium" and Pope's--and indeed Sisson's (1:88) and the Oxford Complete Works'--declined form "Epidamnum". In an explanatory note Whitworth remarks that Shakespeare probably wrote "Epidamium" or "Epidamnium", so clearly his aim is not simply to try to recover the name of the town that Shakespeare thought he had set his play in but rather its true name. At 1.1.54-5 Whitworth prints "A mean young woman was deliverèd / Of such a burden, male twins, both alike" where F has "A meane woman was deliuered / Of such a burthen Male, twins both alike". Whitworth's "mean young" is own solution to the line being metrically short and other editors have proposed their own, none of which has much to recommend it over the others. For the punctuation of "burden, male twins, both" Whitworth points out that the phrase "burden male" has no examples in OED, and indeed I have checked LION too and can confirm that it seems unique.

    At 2.1.110-4 Whitworth prints "I see the jewel best enamellèd / Will lose his beauty; and though gold bides still / That others touch, yet often touching will / Wear gold, and so no man that hath a name / By falsehood and corruption doth it shame" where F has "I see the Iewell best enameled / Will loose his beautie: yet the gold bides still / That others touch, and often touching will, / Where gold and no man that hath a name, / By falshood and corruption doth it shame". This mostly relies on Theobald, but Whitworth makes the pragmatic observation that it does not much matter what an actress says here "so long as she conveys Adriana's anguish at her husband's imagined infidelity and his apparent unconcern about his sullied reputation". Indeed, Whitworth suggests that one might simply cut these lines in performance, as directors have on occasion done. For the speeches at 5.1.347-52 Whitworth gives a long note defending the ordering in F against claims that the Duke should not interpose himself between the Abbess's entreaty ("if thou be'st the same Egeon, speak") and Egeon's response ("If I dream not, thou art Emilia").

    Finally for The Oxford Shakespeare, Roger Warren edits 2 Henry VI and in "A Note on Titles" (p. x) clarifies that he reserves the titles of 1, 2, and 3 Henry 6 for the Folio texts and reserves The Contention of York and Lancaster and Richard Duke of York/True Tragedy for the 1594 quarto and 1595 octavo respectively. This in essence is a gentle remonstrance for the Oxford Complete Work's muddying of the waters by printing essentially the Folio texts under their original stage titles. In his introduction Warren explores the phenomenon that each of the three Henry 6 plays stands on its own yet is enriched by being read or seen in relation to the others (pp. 1-6) and his expected tour of the stage history (pp. 6-26) is enlivened by personal experiences of working in the theatre. A section on "Origins" (pp. 27-32) reports how the raw materials appeared in Hall and Holinshed, but the section on "History and Pseudo-history" (pp. 32-36) strikes an odd note with the claim about Shakespeare that ". . . in the early 1590s, as he was starting out his writing career, there was no such genre" as the history play (p. 35). It is hard to think what else but 'history plays' would be an adequate description of The Famous Victories of Henry V (first performed 1583-8), the anonymous Richard 3 (first performed 1588-94), Marlowe's Edward 2 (first performed 1591-3), the anonymous Jack Straw (first performed 1590-3), the anonymous Richard 2 Part 1 (first performed 1591-5), or Heywood's Edward 4 Parts 1 and 2 (first performed 1592-9). If Warren rejects the generic and/or dating claims that I have here drawn from Alfred Harbage's Annals of English Drama, it would be interesting to hear why. Indeed, in a ground-breaking study of one company Scott McMillin and Sally-Beth MacLean wrote of "the innovation of carrying English history into the popular theatre" that "the Queen's men appear to have been first" (1998, 167), and listed amongst their repertory The Famous Victories of Henry V (Stationers' Register entry 1594, printed 1598 with the Queen's men's name on the titlepage), The Troublesome Reign of King John (printed 1591 with the Queen's men's name on the titlepage), and The True Tragedy of Richard III (printed 1594 with the Queen's men name on titlepage), all of which contradicts Warren's claim the history play genre did not exist when Shakespeare started his career in the early 1590s.

    An interesting intersection of criticism and editing occurs in relation to the Jack Cade rebellion. Because the Bakhtinian carnivalesque (as made relevant to the rebellion by the scholarship of Stephen Longstaffe) is universally mocking, even of its own participants and the leader of the rebellion, Warren reports (p. 52) that he decided not to put the rebels' mocking comments on their leader at 4.2.31-58 into asides, as editors usually do. Warren's section on "Date and Chronology" (pp. 60-74) is typically thorough, although unusually he seems to treat Greene's Groatsworth of Wit as though it were written by Greene himself. Warren notes that the play that Henslowe recorded as "ne" on 3 March 1592 was "harey the vj" and "since one motive for publication [of Richard Duke of York] must have been to cash in on the play's success, it could only have done that if it had used the title under which it was performed--which cannot therefore have been Harry VI" (p. 61). The logic is faulty here, for Henslowe was under no obligation to record the proper stage title, and indeed we know of several plays where his form of the title does not match the one used in publication, as Charles Edelman shows in relation to The Battle of Alcazar/Muly Molocco (see below). Since Henslowe's play cannot, by Warren's logic, be Richard Duke of York it cannot by the same token be The Contention of York and Lancaster, so it must be 1 Henry 6. But "ne" does not necessarily mean 'new', Warren points out, and "perhaps sometimes [indicated] a newly revised or even newly licensed" play. Hence, he argues (pp. 65-71), the order of composition might have been 1, 2, 3 Henry 6, whereas most scholars think that 1 Henry 6 was "ne[w]" in March 1592 and hence there was not enough time to compose 2 and 3 Henry 6 after the first one and in time for the entry of Greene's Groatsworth (with its allusion to 3 Henry 6) in the Stationers' Register in September 1592. He is, of course, quite right to be suspicious of the claim that Henslowe's "ne" meant 'new', for it might just as well have indicated that a particular play was performed at Henslowe's Newington Butts theatre, as Winifred Frazer observed (1991). Warren skilfully deals with the remaining objections against 1, 2, and 3 Henry 6 being the order of composition, including the one that trilogies just were not done, two-parters being the rule (pp. 71-74). But he leaves out the fact that the preparations for the printing of the 1623 Folio included the first-time entry in the Stationers' Register on 8 November 1622 of "The thirde parte of Henry the sixte". This cannot be what we now call 3 Henry 6 because this had already been printed in 1595 as Richard Duke of York, so it is most likely 1 Henry 6 considered the third part of the series in order of composition, not in order of historical events. Likewise on 19 April 1602 Millington transferred his rights in "The first and Second parte of HENRY the VJt[h]" to Pavier who later printed The Contention of York and Lancaster and Richard Duke of York, which suggests that what we now call 2 Henry 6 was originally the first part of a two-parter and 3 Henry 6 was its completion.

    Warren's "Textual Introduction" (pp. 75-100) is of greatest interest to this survey, and in it he points out that Steven Urkowitz (like others who attack the theory of memorial reconstruction) "creates an entirely false antithesis" in suggesting that memorial reconstruction implies a perfect original desecrated by the reconstructors (p. 78). The strongest reason for believing that the 1594 quarto of this play was a memorial reconstruction is York's garbling of his own genealogy in a way that simply cannot be attributed to an author, and Warren observes also that Q's version of the queen's speech to Suffolk (1.3.43-65) contains several bits that in F are in different places. It is inherently unlikely that Shakespeare subsequently took this speech apart and highly likely that we see here an actor undertaking "a piece of cobbling, attempting to reconstruct a dimly remembered speech with material from disparate places" (p. 81). The illogical action of Q's version of the rest of that scene (including an unmotivated exit and reentrance for duke Humphrey of Gloucester) when read against the perfectly logical F version makes it clear that it is a garbled recollection, and likewise Q's version of the witchcraft scene has "borrowings of phrases from other plays known to the actors", namely Marlowe's Doctor Faustus and Tamburlaine (pp. 81-83). In Q's version of 3.2 Margaret says that wherever in the world Suffolk is, she will "haue an Irish that shall find thee out", whereas F has the obviously correct "an Iris". Warren reckons the reporter is recalling a messenger's news about Irish rebellion in the previous scene: "It is inconceivable that an author could have written this as an alternative version". But Q does not simply report the version that was eventually printed in F, for there must also have been revision between composition (reflected in F) and performance (reflected in Q). For example, F gives little space for Somerset and Buckingham's part in the plot against duke Humphrey and in planning the response to the Irish rebellion: they are present in the latter half of 3.1 but say almost nothing. Q seems to fix this, to integrate them into the scene by having the queen address them by name and by giving the otherwise silent Buckingham something to say. 

    More complicatingly still, F also contains signs of subsequent authorial revision. Warren offers the example of Q's "Elnor What hast thou conferd with Margery Iordane, the / cunning witch of Ely" against F's "Elia. What saist thou man? Hast thou as yet confer'd / With Margerie Iordane the cunning Witch". If F simply represents authorial papers and Q performance, who would have put in the historically correct detail of where Jordan was from? As McKerrow argued, it is hard to imagine a playwright sprinkling that sort of detail into a play already written, since it would take extensive digging in his sources. Warren thinks it more likely that "of Ely" was cut to make F's "absolutely regular iambic pentameter" and "more incisive" line; so here F is the revised text (p. 88). Likewise, when Cade orders Lord Saye's execution, Q contains historical details from Hall (about "the Standard in Cheapside" and "Mile End Green") that F omits. Unless we think (as indeed William Montgomery did for the Oxford Complete Works) that someone went to Hall and put such details into the performances that underlie Q, we must suppose instead that someone (and Warren thinks it Shakespeare) omitted them from the text underlying F. Another example is a reference to the historical fact of Cardinal Beaufort's bastardy, present in Q and absent in F. Q shows that it got performed, so either it was added into the play during rehearsal (most unlikely) or it was removed from the manuscript underlying F. This might have happened through censorship but Warren thinks it happened when "Shakespeare, revising and/or copying his manuscript, cut the passage to shorten a scene that contains quite enough bickering between Humphrey and the Cardinal" (p. 90). Once we accept that F contains post-performance authorial revisions, all the Q/F differences come up for question: should we follow Q on the grounds that it is what got first performed or F because that is what Shakespeare later decided he wanted? Warren surveys the Q/F differences (pp. 90-97), trying to determine which might be things that got added-to/cut-from the play during rehearsal (and hence into Q) and things that Shakespeare changed when copying up his authorial papers to make the manuscript that eventually produced F. Here Warren reminds the reader of one possible occasion for the authorial revisions (mentioned earlier, pp. 73-4, but only here explicitly now tied to the argument about revision in F): a revival of the Henry 6 plays around the end of the 1590s to run with the Henry 4 and 5 plays.

    Thankfully, Warren offers a summary of his views on the complex situation (pp. 98-100): Q and F show "signs of two different kinds of revision of the play as originally written. Q seems to report changes to it made in rehearsal, F to reflect changes made in a later revision, probably for a revival". But the authorial revisions reflected in F do not incorporate those made during rehearsal (indeed, that is how come we can detect them at all in Q) so Shakespeare seems not to "have thought sufficiently well of them to include them in the revision detectable in F". Warren's New Bibliographical thinking is apparent: "The Folio text was printed from an authorial manuscript, as its vague and 'permissive' stage directions make clear: for example, Jack Cade's first entry is accompanied with the 'infinite numbers' (4.2.30.2 / TLN 2351) characteristic of a dramatist writing impressionistically rather than with a strict regard for theatrical practicalities". Honigmann showed that in making his own fair copy a dramatist might introduce all sorts of minor changes, and indeed the deletion of chronicle details that are in Q and not in F could easily be explained this way. But against the idea that F was printed from an authorial fair copy is the fact that Henry calls his queen Margaret "Nell", which is in fact Humphrey's pet name for his wife Eleanor; surely Shakespeare would have spotted his slip and fixed it? Unless, Warren argued in an article reviewed here two years ago (2000, 206), Shakespeare did his revision on his original manuscript (rather than copying it out fair) and did not even look at the bits he knew were alright such as Act 3. Warren reports that "The text in this edition is based firmly on that in the Folio, as representing Shakespeare's latest thoughts on the play, and unlike the Oxford Complete Works only incorporates material from the Quarto to correct obvious errors (for example, to supply the missing line at 4.1.49), with one important exception: since the Quarto reports an Elizabethan performance, its interesting and often extensive stage directions are incorporated into this edition where they supplement the rather meagre information given in the Folio ones (for example, at the opening of the play)".

    Warren accepts that F was set directly from a quarto (probably Q3) at certain points, and thinks these three to be particularly clear examples: "2.1.112-149.2 / TLN 858-904; 2.3.58.1-2.4.0.2 / TLN 1115-70; and 4.5.0.1-4.6.0.2 / TLN 2598-2614" (p. 99). This is a shorter list than that produced by William Montgomery for the Oxford Complete Works (TLN 63-79, 795-6, 858-904, 1115-70, 2598-2639) but is wholly a subset of it. The important point is that where F was printed from Q, the latter is the only substantive text and should be any editor's control text, but Warren unaccountably departs from this principle. To be fair, Montgomery did too, claiming that he had decided to take Q as the "only substantive text" for these passages and "treat any departures from it--including those we adopt from F--as emendations" (Textual Companion p. 176). But in the detailed textual notes Montgomery suddenly introduced the idea that F's copy was not simply an example of Q3 but "an annotated copy", which would explain why F during these passages does not always follow Q1 or Q3. Montgomery did not tell the reader where he thought these annotations came from, although in his notes to 2.3.76-83 and 2.3.91-103 he wondered aloud whether it was "perhaps corrected by recourse to foul papers". To give such credence to F's substantive differences from Q3 is in fact not to treat it has his "only substantive text" at all and casts doubt on Montgomery's procedure of joining those individual occasions where Q3 seems to influence F into whole stretches where Q3 is supposed to be F's copy. Warren certainly prefers F over Q even in those sections that he declares are "particularly clear" cases of F being printed from Q, as when he prints "But cloaks and gowns before this day, a many" (2.1.113) where Q1-3 have "ere" instead of F's "before". Again at 2.1.132 Warren gives "O master, that you could!" following F instead of Q's "I would you could". In the second of Warren's three stretches of clear use of Q copy for F, he nonetheless has Peter leave his hammer to Will and his apron to Robin (as F does) instead of the other way around (as Q does).

    The third of Warren's three stretches of Q copy for F begins with the opening stage direction of 4.5 that he renders as "Enter Lord Scales [and Matthew Gough] upon the Tower, walking. Then enters two or three Citizens below". In a note to this Warren writes "Since the texts (and even the layout) of F and Q3 are almost identical at this point, F was probably set from Q3 here (see Textual Introduction, pp. 99-100); neither mentions Scales by name. The reporter of Q may have omitted a line or phrase identifying Scales, or he may simply have known who Scales was through rehearsal discussion, while the audience may not even have needed to know his name: his essential dramatic function is to be an authority figure defending the Tower". (Warren actually means that no-one mentions Scales's name in dialogue.) The evidence that F was set from a quarto (not specifically Q3, as Warren implies) is indeed the closeness in the phrasing ("upon . . .  walking") of a stage direction, which a memorially reconstructed text and an authorial copy are unlikely to agree upon by chance. But the evidence that the quarto used for F was specifically Q3 comes from something else not mentioned by Warren: had it been Q1 or Q2 it would be bizarre to change 3-4 citizens (the Q1/2 reading) to 2-3 citizens (the F reading). Q3 omits to say how many citizens, and since this is obviously an omission (a citizen almost immediately speaks) the Folio compositor probably rectified this himself, guessing at the number of citizens and guessing (from the dramatic context) that they enter below. This was the argument in Montgomery's PhD thesis and repeated in the Textual Companion (p. 177). Warren keeps the F reading, which by this argument is of course a non-Shakespearian interpolation and even keeps the un-Shakespearian phrasing 'enters' that Montgomery pointed out only occurs elsewhere in the parts of the Shakespeare canon thought to have been printed from scribal copy or not written by Shakespeare. (Actually, Montgomery garbled this point, writing that 'enters' occurs in "quarto and Folio Contention 2.4.83.3 / 1019.3" but he must have meant 2.3.58.3 and in any case the important point is the ungrammatical use of this word, here at the beginning of 4.5 and previously in "Enters a Messenger" near the end of 1.1 in Q1, corrected in Q3.) For the rest of this scene, 4.5, that Warren thinks was printed in F from Q, he nonetheless follows F for substantive variants, giving "The rebels have essayed to win the Tower" (4.5.8) where Q1/3 have "attempted to win".

    Generally Warren's editorial choices are properly explained, but not always. At 1.3.92 he prints "That she will light to listen to their lays," where F has "listen to the lays". Warren, following Rowe and the Oxford Complete Works, here departs from his copy (F) in a matter of substance yet there is no mention of it in the collation, nor an explanatory note. At 3.1.260 he prints "As Humphrey, proved by reasons, to my liege", adding commas around "proved by reasons" that are not in F. Warren says that he punctuates lightly (p. 102) but these commas shut down a possible different interpretation. In Warren's reading Humphrey has shown himself ('proved') to be an enemy to the king just as the fox proves himself an enemy to the flock (two lines earlier), whereas Sisson argued (2:77) that Humphrey had earlier proved something to the king (3.1.168-71), namely that reasons will be found to get rid of him and that "'A staff is quickly found to beat a dog'", and that Suffolk is here "sardonically" recalling that moment. Sisson's reading might be thought somewhat strained, but by punctuating even more lightly that he has Warren might have left the matter open.

    For The Arden Shakespeare, Charles Forker gives King Richard II a huge introduction covering "Politics" (pp. 5-55), "Language" (pp. 55-90), "Afterlife" (pp.90-111), "The Date" (pp. 111-120), "Probable Venues of Performance" (pp. 120-123), "Sources" (pp. 123-164), and "Text" (pp. 165-169). This last is just a summary, however, and there is a full appendix (pp. 506-541) for the detailed "Textual Analysis". Under "Politics" Forker considers Ernst Kantorowicz's reading of the play as "a tragedy of royal christology" about the dual nature of king and man, but fails to make the connection necessary for those unfamiliar with these concepts: Christ's dual nature as deity and man. Perhaps it seems too obvious to Forker, and indeed sometimes his choices of what needs explaining are curious. For example, under "Afterlife" he lists places that Richard 2 has been performed: "Athens, Avignon, Berlin, Bonn, Bratislava, Braunschweig, Bucharest, Dublin, Hamburg, Kampala (Uganda), Marseilles, Milan, Munich, Verona and Zurich" (p. 100). What kind of reader does Forker expect will recognize the German spelling of the small town of Brunswick (Braunschweig) but needs to be told which country contains the much larger capital city of Kampala? Under "Probable Venues" Forker makes a slight slip in writing that the Swan drawing shows no central opening between the left and right doors "such as the Globe possessed" (p. 122) when in fact we do not know whether the Globe possessed one, and the Swan drawing is one of the major pieces of evidence in that particular puzzle. Also unreliable is his comment that "According to the computer-based findings of Donald Foster" Shakespeare played Gaunt and the head Gardener and an anonymous lord in 4.1 and maybe also the Groom, for which Forker cites Foster's article in Shakespeare Newsletter in 1995 that, in the light of Foster's subsequent exposure as a charlatan, nobody should trust. A final small slip is Forker's repetition of the familiar error of claiming that the opening moments of Hamlet dramatize a violation of military protocol because the relieving sentinel challenges the man on duty (p. 126n1). As Charles Edelman has privately pointed out, even if the protocol were strict on this matter (and it is not), no-one watching the opening moments of the play could spot which man is relieving which.

    In his summary of the textual situation, Forker reports that Q1 (printed 1597) is the "basic text" except for the deposition scene (4.1.155-318) that first appeared in a reliable form in F, Q4-1608's version of it being unreliable. Thus there is no early text of roughly equal authority to the one chosen as Forker's control text, and hence this edition does not include (as some Ardens do) a facsimile reprint of the one not chosen. Forker writes that "Q is apparently the text closest to Shakespeare's holograph" (p. 165) and perhaps was printed directly from it. F seems to be printed from a copy of Q3 (printed 1598) that had been extensively annotated by reference "to the theatre promptbook and augmented from the same source by a manuscript insertion of the missing abdication episode". Forker thinks that the deposition scene was always performed, but "was considered too dangerous to print in 1597". For bits where F's wordings not in Q have been included in this edition, Forker uses the convention of wrapping superscripted letter 'F' around them (Fxxx xxxxF) just as R. A. Foakes did for his Arden 3 King Lear (p. 167). Forker gives an intelligent explanation of his decision to retain Richard's speech prefix as "King Richard" right to the end of the play, even though Bolingbroke changes to "King Henry" in 5.3: one of the play's themes is precisely the question of whether both can be king at once, or is the office indivisible? (p. 168).

    Forker's Appendix 1 goes into the detail underlying this summary of the textual situation, such as the nature of the copy underlying F. Was it a whole copy of Q3 or a one whose missing final leaves were supplied from a copy of Q5? Everyone agrees that some promptbook influence on the Q3 used to print F occurred, since F's version of the deposition is not simply a reprint of the one in Q4-5, and F restores a number of good Q1 readings that got corrupted in Q2 and Q3, and that it has cuts that seem theatrically authoritative. Regarding terminology, Forker acknowledges William Long's objection to use of the word 'promptbook' and although he agrees that 'playbook' is better, ". . . the older term has been so commonly used by scholars that it is difficult to avoid it" (p. 508n1). Of course, if anyone has the power to change misleading terminology it is the editors of major critical editions, whom one might hope would take a more adventurous position than Forker's. Additional reasons for thinking that a promptbook influenced F is that it improves on its quarto copy's speech prefixes and stage directions in ways that we can imagine occurring during rehearsal, as when it is realized that Gaunt is needed at the beginning of 1.2 so he had better exit before the end of 1.1. Corroborating the theatrical influence is F's substitution of safe words (such as "heaven") for its quarto copy's oaths (such as "God") in line with the 1606 act that applied to performance but not to printing. In all, F has what Jowett and Gary Taylor picturesquely termed "sprinklings of authority" (p. 509-15).

    Forker believes that Q1 simply omitted to print the deposition scene that was in its manuscript copy, although he footnotes David Bergeron's argument that it might in fact have been written later than the rest of the play. Q1 seems to be authorial in its imprecise and theatrically pointless stage directions (of the kind "enter xxxx with news"), and in its variant speech prefixes; was the copy authorial papers or scribal copy of them? We cannot tell: certain features (such as the use of abbreviations) seem authorial, while others (such as the preference for Oh instead of Shakespeare's habitual O) seem scribal (p. 518-20). What of the nature of the text used to annotate Q3? Forker reports Jowett and Taylor's hypothesis, based on the clustering of Q5 readings at the end of the Folio text, that the last leaf of the promptbook went missing and was replaced with a transcript from Q5, and that this promptbook was used to annotate an example of Q3 to make copy for F. A problem, Forker observes, is the unlikelihood that "the annotator of Q3 would trouble to introduce a cluster of such unimportant readings from Q5, even if these had somehow got into the promptbook" (p. 523). Forker praises Jowett and Taylor's work on just how heavily, and where, the annotator of Q3 (using the promptbook as his authority) did his annotating: evenly in stage directions and speech prefixes but in dialogue he was more interventionist in the first third or so of the play. For a modern editor, then, the problem when using Q1 as the basic text is which of the F variants to prefer on the grounds that they have promptbook authority (p. 524). Indeed, how authoritative is a promptbook? Obviously there is bound to be an element of subjectivity, and Forker lists the major F variants from Q3 that he adopts--that is, where he lets F, differing from its copy Q3, overrule his basic text Q1--chosen on the criterion of having "garnered the approval of a significant number of modern editors" (p. 525). (Where F differs from Forker basic's text Q1 but follows Q3, that is probably because F is simply reprinting Q3.) Forker's appeal to a consensus of editors is most odd: a kind of editing-by-committee. Lest that seems harsh, I quote Forker being quite explicit: "The 'corrections' of Q3 discussed in the preceding paragraphs (including modifications of the SDs and the introduction of dubiously authorial act and scene divisions) represent something like a consensus of twentieth-century editorial judgement on which F readings should be admitted into a modern critical editions of Richard II" (p. 528). Although Forker mentions a couple of moments where he breaks with tradition, especially when he is unconvinced that just because something got in the promptbook and thence into F, it had authorial sanction. He is not one of the "cultural contextualists" for whom a play is inherently an overdetermined collaboration (p. 529).

    Alan Craven's PhD on the printing of Q1 revised Charlton Hinman's compositor attribution, giving more of the play to compositor A (57½ pages) and less to compositor S (15 pages). Compositor A did all of the reprint Q2, and from his departures from his copy (Q1) in that task we can determine his habits, which add up to a lot of unauthoritative departures from copy (pp. 537-8). For this reason, Craven advocated greater editorial intervention in Q1 (when used as the basis of a modern edition) than one would otherwise want to perform, on the grounds that Q1 probably has lots of compositor A's errors in it.  But Peter Davison pointed out the slenderness of the evidence on which we make the compositorial identifications, and did not think the errors in Q2 were as many as Craven saw: some are defensible alternatives, others may have come from the Q1 being used as copy having uncorrected sheets in it. Still, observes Forker, whether compositor A or S is responsible for Q1's errors, there are enough errors in Q2 (printed from known Q1 copy) to suggest that the printing of Q1 brought in considerable compositorial error, and for this reason he incorporates six of Craven's 13 suggested departures from Q1, moments when the compositor probably erred from his copy (pp. 539-40). 

    At 1.3.26 Forker prints "Marshal, ask of yonder knight in arms" where Q and F have (modernized) "Marshal, ask yonder knight in arms". Forker's note surveys some possibilities, but having accepted that ". . . Shakespeare occasionally introduces short lines for variety" Forker gives no reason for adding the word of (which is his own) rather than following what Q and F agree upon; and this is not one of his six moments of claimed compositorial slip. At 1.3.239-42 Forker has Gaunt say "O, had it been a stranger, not my child, / To smooth his fault I should have been more mild. / A partial slander sought I to avoid, / And in the sentence my own life destroyed". These four lines are in Q1-3 but omitted in F, and the Oxford Complete Works omitted them too on the grounds that the comparison of Q3 with the promptbook (to prepare Q3 to be copy for F) showed that they had been cut in performance. Forker's approach conflates Q3 and F, accepting F's reading wherever it adds something to the play (such as fuller stage directions) but ignoring its evidence where it takes material away. This is not really reasonable use of F, for it treats the text as a pot to be filled as much as possible rather than a dramatic product where sometimes 'less' (textually) is 'more' (dramatically). The same point is true of the other passages that the Oxford Complete Works prints as "Additional" and Forker incorporates in his main text. At 2.1.114 Richard cuts off the dying Gaunt's words in a celebrated dramatic moment: "[GAUNT] And thou -- / KING RICHARD A lunatic lean-witted fool". A modernized form of Q1 would be "And thou / KING RICHARD A lunatic . . ." while F has "And -- / KING RICHARD And thou . . .". Forker is following Capell's emendation (that is, using Q's words but adding a dash), but for no other reason than that it "clarifies the dramatic interruption". Yet F makes sense and since F here seems to emend Q3--presumably by reference to the promptbook--the logical thing to do would be follow F.

    Forker's interventions are occasionally justified by judgments about what Shakespeare ought to have written, so that at 3.2.35 he has "[AUMERLE] Grows strong and great in substance and in power" which is the Q1 reading, even though F here departs from its copy to have "in friends"; this Forker rejects because it seems "distinctly weaker". Likewise, at 3.2.37-8 Forker has Richard say "That when the searching eye of heaven is hid / Behind the globe and lights the lower world", where Q and F agree on "eye of heaven is hid / Behind the globe, that lights the lower world", meaning 'the searching eye of heaven . . . that lights the lower world'. Forker uses Hanmer's emendation of "that" to "and" because it "clarifies the syntax and makes the sentence more readily intelligible in the theatre". This seems to go beyond the editor's job, for both textual authorities agree on a syntax that is awkward to us but entirely acceptable by Elizabethan standards, so it is hard to see the need to emend. A further example is at 3.2.84 when Forker has Richard "Awake, thou coward Majesty, thou sleepest!", which is essentially Q's reading, even though F departs from its copy Q3--presumably on promptbook authority--to give "sluggard Majesty". Forker's reasons for not following F here is that it "weakens the speech". It is hard to find an authority strong enough to shake Forker's faith in his own judgment, since at 3.2.202-3 he prints "And all your southern gentlemen in arms / Upon his party", the Q1 reading, rather than F's departure from Q3 of "Upon his faction". Forker admits that F's reading "presumably comes from the promptbook" but he still will not adopt it because it "seems more specific and perhaps more limiting". In general the convention of wrapping superscripted 'F' around words taken from the Folio works well, although for consistency Forker puts the entire deposition scene (4.1.153-318) inside such a pairing; across 21 pages the convention fails to signal its meaning to the reader. The only literal fault I could find was at 5.6.25n where Q3's "reverend" is given as "rverend".

    The final edition reviewed this year follows historically and dramatically from Forker's: David Scott Kastan's King Henry IV Part 1. By comparison with Part Two that was not reprinted between its first printing in 1600 and the 1623 Folio, Part One went through nine editions between 1598 and 1630; it was a bestseller (pp. 1-2). Kastan notes that when it was published, what we call the first part of Henry 4 was simply called Henry 4 and only the second part got explicitly called the second (p. 18). That is true, but it is not quite as surprising as Kastan seems to think: Henslowe followed this practice and Hollywood does today (Rocky, Rocky II). Kastan's account of Falstaff/Oldcastle is much as in one of his recent books (1999, 93-108) and addresses the tricky question of why Shakespeare might mock a protestant martyr. The answer is that by the 1590s the Lollards were seen as extremists of the Puritan kind and hence mocking them could be a moderate protestant thing to do (pp. 51-62). In an uncharacteristic slip, Kastan reproduces the wrong picture for his caption "Frontispiece of Kirkman, The Wits, 1662" (p. 82) using the coarse copy of 1672 rather than the original, as can be seen by comparison with John Astington's reproductions in an article that got to the bottom of this matter (1993).

    With admirable succinctness, Kastan keeps his discussion of "The Play on the Page" to 13 pages (pp. 106-18), noting that Q2-8 and F "are all derived from the 1598 quartos" (that is Q0, existing only in a fragment, and Q1) and that F was printed from Q5 of 1613 (p. 111). There must have been some outside influence on F, however, for it removes Q5's oaths in line with the 1606 act. Jowett and Taylor saw in this the unmistakable influence of a playhouse manuscript, for only there would the oaths have been altered to meet the act's requirements, but Kastan wonders if perhaps Heminges and Condell might have editorially removed the oaths when preparing Q5 as the F copy, perhaps "to bring the printed text in line with the play as it had been performed" (p. 113). Indeed, had a promptbook been used to annotate an example of Q5, it is a wonder that the stage directions were not tidied up more than they are in F. Crucially, to Kastan's mind, F gives the "But this our purpose is a twelue month old" (1.1.28), which regularizes Q5's faulty meter of "But this our purpose is twelue month old". Were a promptbook or other authoritative manuscript used to do the annotating of Q5 to make it ready to be copy for F, we would expect F to have the correct reading that we find in Q1: "But this our purpose now is twelue month old" (pp. 114-15). Thus F is merely a derivative text with no independent authority, and hence Q0 is the substantive text for what little of the play it has, and Q1 is substantive for the rest. However, F's departures from Q5 "do demand careful consideration" because Heminges and Condell knew the play in performance (p. 116). Comparing Q1 to its copy Q0, we can see that Q1 was carefully printed, but the quartos' stage directions are "insufficiently specific to cue performance" so the quartos do not derive from "the theatre company's prompt copy" (p. 117). Like almost everyone these days, Kastan notes Long's objections to the term 'promptbook' and his article showing that stage directions were not necessarily tidied in theatrical manuscripts (1985), but then continues as though these things do not matter. Firmly in New Bibliographical territory, Kastan decides that the manuscript used to print Q might have been authorial papers but finds that the lack of Shakespearian preferences--betwixt where he preferred between, prithee where he preferred pray thee--suggests rather that it was a scribal copy of authorial papers (p. 118).

    Kastan's section on "Editorial Procedures" (pp. 119-31) is somewhat longer than one would expect, mostly because it is taken up with a detailed study of the effects of modernization (pp. 119-26). Commendably, Kastan writes "I have . . . departed from Q0 and Q1 only when they are evidently in error", which he thinks makes him "more conservative than many editors" (p. 119). Much concerned with the forms of names, Kastan retains the anglicization of the name Glendwr as Glendower because "for Shakespeare 'Glendower' exists as he is written", already inserted in this form into English history. Kastan uses the name Bardoll because he is never Bardolph (or Bardolfe) in a sixteenth- or seventeenth-century quarto of this play. Only in F does he get the 'f' ending and apparently that was a matter of "bringing it into line with the spelling in 2 Henry IV, Henry V and Merry Wives". In Shakespeare After Theory Kastan made an innovative argument against the Oxford Complete Work's restoration of Oldcastle's name in 1 Henry 4: that might well have been Shakespeare's first intention for this character, but it was also his intention to have the same name in both parts of the play (1999, 95). I suppose one might argue that Bardolph deserves the same uniform treatment rather than being Bardoll in Part One and Bardolph in Part Two. There is little to be said about Kastan's particular editorial choices, based as they are on a sound conservative principle of following Q. In a note to line 3.1.226 about the music that Glendower conjures up, Kastan writes that "Hotspur . . . is presumably aware, as is the audience of the play, that the musicians in the theatre company are most likely sitting behind a curtain in the music gallery and so might indeed be said to 'Hang in the air', 221". Actually, Richard Hosley long ago showed that in stage directions for music the location changed from 'within' (that is, inside the tiring house) to 'above' around 1609, presumably when the King's men altered the practices at their open-air Globe playhouse to match the practices at the indoor Blackfriars (1960). Unless we think they altered their practices twice, the strong likelihood is that at the Theatre and the Curtain (where 1 Henry 4 was first performed) the music room was also inside the tiring house, not in "the music gallery" as Kastan supposes. On a couple of occasions Kastan might have stuck with a quarto reading but does not. At 4.1.20 he has a messenger say "His letters bear his mind, not I, my lord" where Q1 has "His letters beares his mind, not I my mind". Sisson (2:36) pointed out that the quarto reading make perfect sense: it is not my job to carry my own opinions to you, just letters. Likewise, at 5.3.22 Kastan has Douglas say to slain Blount "A fool go with thy soul, whither it goes!" where Q1 has "Ah foole, goe with thy soule whither it goes"; Sisson pointed out (2:38) that Q1 makes perfect sense: you and your soul, dead Blount, go together to wherever you are going.

    Brian Vickers's Shakespeare, Co-Author: A Historical Study of Five Collaborative Plays is divided in two parts, moving from the general ("Elizabethan Drama and the Methodology of Authorship Studies", pp. 1-134) to the specific ("Shakespeare as Co-Author", pp. 135-500). In a brief survey of early printings of Shakespeare, Vickers claims that "A 'booke' was the term used to describe a playhouse manuscript or prompt-book, prepared by or for the 'book holder', who acted as stage manager and prompter" (p. 6). Actually, Henslowe frequently referred to what he bought from a dramatist as "the booke", and such a thing certainly had yet to be theatricalized; there are 9 examples within two pages of Foakes's edition (1961, 96-97). Vickers calls Henslowe's Diary "a business record" (p. 20), which it is, but it also has spells, medical treatments, and card tricks in it. Occasionally an error might be due to over-zealous copy-editing, such as the failure to capitalize the name of The Theatre in Shoreditch (p. 7). Vickers shows that naming a sole dramatist was, in the early part of the period, considered unimportant, and naming joint dramatists even less so, but over time both kinds of naming became more usual (pp. 10-17). Collaborations were less likely to reach print, but by the time of the 1623 Folio it was not unusual to acknowledge co-authorship. Perhaps in constructing 'Shakespeare' in F, Heminges and Condell decided not to acknowledge his collaborations, but we do not have to follow them Vickers points out. Vickers thinks that Shakespeare demonstrably collaborated "in at least seven surviving plays" (p. 19) and it would have been a courtesy to here name them for the reader. By G. E. Bentley's calculation about 15-20% of all plays were collaborative, and of the professional plays as many as 50% "incorporated the writing at some date of more than one man" (that is, including those that got subsequently reworked by someone other than the original writer) and in Henslowe's Diary the proportion reaches 66%.

    Vickers uncritically repeats W. W. Greg's view of what a playhouse 'plot' was for (p. 21) and surveys the scant evidence for the unit of division between collaborators, usually the 'act' as frequently mentioned in Henslowe's Diary (pp. 27-29). The discussion here really ought to have addressed the problem that before 1609 the open-air amphitheatres, for which many of the dramatists were writing, did not use act intervals. Did they think in acts anyway? In a chapter on "Identifying Co-Authors" (pp. 44-134) Vickers surveys methods of making attributions from internal evidence, including pause pattern tests and proclitic and enclitic micro-phrases. Vickers gives a scathing summary of the faults in Gary Taylor's Textual Companion to the Oxford Complete Works concerning function word frequencies as a means of determining authorship, and does the same to the faults in the stylometric methods of A. Q. Morton and his followers G. Harold Metz and Thomas Merriam, and ending with the off-the-spectrum craziness of Barron Brainerd. Against all this Vickers contrasts the entirely reasonable work of Ward Elliott and Robert Valenza and the Claremont McKenna 'Shakespeare Authorship Clinic' and the sociolinguistics of Jonathan Hope that tracks changes in usage over time and between different social groups, as in 'you' instead of 'thou' being educated, urban youth-speak more likely to come from a Fletcher than a Shakespeare (pp. 116-19). A particularly good marker is 'auxiliary do', the absence of which (the 'unregulated' form, as in 'went you home?' instead of our 'did you go home?') was old-fashioned. Shakespeare's rate of regulation never exceeded 84%, while six other dramatists Hope compared him to never fell below 85%. (Unaccountably in the middle of discussing Hope's work on the choice of thou/you, Vickers starts referring to "T/V choice" (p. 124); I wonder what "V" stands for.) Vickers has a high opinion of the work of Gregor Sarrazin on the words that Shakespeare dropped from and added to his vocabulary over his career, which showed that each play really does have most 'rare word' links with other plays composed at the same time, and MacDonald Jackson's sophistications of this (pp. 129-32), as reviewed here last year.

    Turning to Shakespeare, Vickers revises his count of co-authored plays to eight by adding Edward 3. Of these, the collaboration with Middleton on Timon of Athens (pp. 244-90) is not of itself going to surprise anyone, nor is the collaboration with Wilkins on Pericles (pp. 291-332), nor with Fletcher on Henry 8 and The Two Noble Kinsmen (pp. 333-432). Vickers begins with his most surprising claim, that Peele wrote part of Titus Andronicus (pp. 148-243). The date of Titus is no longer set by the 'ne' record in Henslowe for 23 [read 24] January 1593, since Frazer (discussed above in connection with Roger Warren's edition of 2 Henry 6) showed that it could mean 'performed at Newington' (p. 149). Vickers accepts June Schlueter's argument taking the Peacham picture away from Shakespeare's play, but neglects to mention Stanley Wells's immediate objection (1999) nor Richard Levin's subsequent demolition of Schlueter's position (2002); the latter probably came too late for the book. Vickers gives a brief history of more than 300 years of critical dislike of the play (pp. 150-55) and moves on to the tangible evidence. T. M. Parrott compared the rate of feminine endings in Shakespeare (8-12% on average) with other dramatists who used it much less, and noticed that Act 1 of Titus has only about 3.6%; Parrott also noticed verbal parallels between Act 1 and the works of Peele, and the rest with Shakespeare (pp. 156-60). Peele coined the word 'palliament', meaning robe, in his poem The Honour of the Garter (1593) and it appears also in Act 1 of Titus but nowhere else in Shakespeare or indeed anyone else. Philip Timberlake's 1931 book on feminine endings showed that differing definitions of this phenomenon give different counts, but that Titus 1.1, 2.1, and 4.1 have far too few, howsoever counted, to be Shakespeare and the rest of the plays has far too many to be by anyone else (p. 163). Because Hereward T. Price misrepresented Timberlake's findings, no-one paid serious attention to his work until MacDonald Jackson confirmed it in 1979. Vickers is a master of rhetoric and sometimes needs to be watched closely, as when he admits that a dramatist might alter a co-author's work, but ". . . all the historical evidence reviewed in Chapter 1 indicates that co-authors normally contributed whole acts, or at least whole scenes, and the piecemeal over-writing [Shakespeare redoing Peele] that Dover Wilson claimed to discover seems improbable" (p. 181). Actually, there was little historical evidence about the unit of collaboration in chapter 1, and it was all confined to pages 27-34. 

    Vickers is particularly scathing about scholarly avoidance of the fairly obvious (and forever evidentially growing) fact that Titus was co-authored with Peele. Eugene Waith for The Oxford Shakespeare pretended that it was not and the Oxford Complete Works did not mention it, although Taylor had become convinced and said so in the Textual Companion. (In fact, the electronic edition of the Oxford Complete Works, published in 1989, does mark the editors' doubts about Shakespeare's authorship of act 1 and scenes 2.1 and 4.1, but the reader has to know that that is what is meant by the arcane electronic tag "<?A Shakespeare>".) Vickers roundly castigates Alan Hughes's New Cambridge Shakespeare and Jonathan Bate's Arden 3 Shakespeare editions of Titus, the latter especially for relying on A. Q. Morton's discredited work, and Sonia Massai's New Penguin edition for simply reporting that "Bate distances himself" from the view that Peele wrote Act 1 (pp. 192-210). Vickers is occasionally so rude about other scholars' ignorance that it is tempting to respond in kind, and opportunities certainly present themselves. To claim that permissive phrasing "has long been understood as typifying an author's stage directions in its vagueness (since those involved in theatrical productions need to specify just which actors/characters are involved)" (p. 121) is to show ignorance of an entire field of textual-theatrical scholarship led by Paul Werstine and William Long that is firmly ranged against what Vickers thinks "long . . . understood". As an example, Vickers offers the permissive direction for the entrance of "as many as can be" in the first act of Titus and in Peele's Edward 1 that Jackson thought unique in the canon of English Renaissance drama from 1576 to 1642. When Vickers repeated this claim on Hardy Cook's SHAKSPER email discussion list (archived at www.shaksper.net) I supplied him with references to five further examples, two anonymous, and one each by Armin, Chettle, and the Fletcher/Massinger collaboration (posting SHK 14.0994 on 20 May 2003).

    Vickers's chapter "Timon of Athens with Thomas Middleton" offers nothing new but a summary of the arguments that he thinks have been underexamined by the scholarly community. An odd comment occurs in a footnote that claims he has  "twice attempted to obtain Dr [R. V.] Holdsworth's permission to read his thesis, but without success" (p. 280n30). I should have thought anyone could just go to the University of Manchester library where it resides and read it there if the authorities will not let it be posted elsewhere. The chapter on Pericles follows the same trajectory, but here Vickers lays into the New Cambridge Shakespeare edition by two professors of drama, Doreen Del Vecchio and Anthony Hammond, for its ignorance of much of the authorship scholarship and for their misrepresentation of the little they knew. The implication is that the shift towards stage-centered thinking in Shakespeare studies has been at the cost of some English Literature specialist skills (". . . the stage is not in any case a court of appeal in authorship studies, for quite different concerns are involved" p. 329), but one could argue with equal force about Vickers's ignorance of developments in thinking about matters theatrical. At the start of the sixth chapter, on collaboration with Fletcher, Vickers tacitly admits that the book has become formulaic ("a by now familiar pattern" p. 333) in its movement from nineteenth-century pioneering work on authorship, later consolidated, denied or overlooked by the "Shakespeare 'conservators'", and now reconfirmed by fresh scholarship. Here, however, Vickers is hamstrung by the lack of opposing voices--even E. K. Chambers could see that Henry 8 was collaborative--so he starts to invent them. A sustained attack on Gordon McMullan's Arden 3 edition of Henry 8 is pure 'straw man', as is the complaint that Literary Theory (Vickers's capitals) have banished the concept of the author. The final chapter, "Plot and character in co-authored plays: Problems of co-ordination" (pp. 433-500) is a survey of traditional scholarship on organic unity and shows scant familiarity with recent critical and historical work, which failing in others is one of Vickers's chief bugbears. For example, Vickers quotes without demur the view of J. Q. Adams that "Henslowe continually oppressed by hard dealings" the Lady Elizabeth's men and had a policy of keeping them in debt (p. 433), which no-one familiar with recent work on the subject--or even passingly familiar with Foakes's scholarship--would accept.

    The second appendix of Vickers's book is called "Abolishing the author? Theory versus history" (pp. 506-41) and in it he exercizes his familiar antipathy to theory will making a number of substantial points that could in fact be usefully corrective of it. Michel Foucault's claim (in "What is an Author?") that from the seventeenth century scientific texts were essentially anonymous is not true: Kepler, Galileo, Descartes, and Newton were scientific authors, and likewise his claim that before then poetic texts were effectively anonymous is in complete ignorance of the classical world's concept of the author (pp. 509-17). 90% of all books published in England 1475-1641 (30,000+ items) have authors specifically thanking their patrons, which is not the behaviour of 'anon' and publishers often boasted that the author's permission had not been received, hence an unauthorized publication depending on the notion of  an author who will be displeased. Vickers offers a most useful survey of the evidence that plagiarism was recognized and detested (p. 522-27), and then gets personal. Jeffrey Masten is excoriated for swallowing Foucault's nonsense about the invention of the author and moving it back to Shakespeare's time and for not engaging with the practicalities of collaboration at all; rather (and this complaint has some merit) Masten simply repeats the mantra that we cannot separate out who wrote what, and that collaboration preceded single authorship. It is fair of Vickers to complain that Masten is  evasive about just when the 'author-function' emerged: in Masten, it is forever "emergent" but never comes out (p. 531). Somewhat unfair, however, is the claim that Masten entirely imposes homoerotic interest on the texts that he reads and links them to collaboration without justification. Appropriately, Vickers ends with a strike that rebounds on him, quoting Masten repeating the word "homosocial" and commenting "Michel Foucault, thou art mighty yet" (p. 535). Vickers seems not to know that the term is not Foucault's but Eve Kosofsky Sedgwick's. Physician, heal thyself.

    Vickers other book this year, 'Counterfeiting' Shakespeare: Evidence, Authorship, and John Ford's Funerall Elegye, is longer still and much better. The 'counterfeiting' of his title is not the creation of false documents but the venial sin of "presenting anonymously authored work as Shakespeare's" (p. xii). The subtitle makes clear that Vickers is concerned with Donald Foster's since-retracted claim that A Funerall Elegye is by Shakespeare, but in fact the retraction came while the book was in production. Indeed, in a letter to the Times Literary Supplement on the day that this review was completed (15 January 2004) Vickers claimed that Foster used bullying tactics to hold up Cambridge University Press's publication of this book in order to be able to concede defeat to Giles Monsarrat (whose article is reviewed below) rather than to Vickers. In a stop-press note (p. xxi), Vickers mentions that Foster announced his submission in a posting to the SHAKSPER email discussion but gives the wrong date (it was 12 June 2002 not 13 June) and the wrong SHAKSPER index number (it was 13.1514 not 13.1519). Limbering up for his attack on Foster, Vickers begins with the relatively easy target of Gary Taylor's attribution of 'Shall I die' to Shakespeare made using just printed concordances, whereas these days full-text computer databases would be employed (pp. 1-53). Like Shakespeare, Co-Author, 'Counterfeiting' Shakespeare is divided into two parts: the first (pp. 55-260) on what is wrong with Foster's attribution of Funerall Elegye to Shakespeare, and the second (pp. 261-464) on why we should think it was really by Ford. Key pieces of evidence that Vickers reviews are by now familiar: the poem gets wrong the length of William Peter's marriage at the time of his death (three years, not nine), Gilbert Shakespeare was buried the same week as William Peter--so it is hard to imagine Shakespeare taking time to honour a man he hardly (if at all) knew--and that the initials "W.S." should raise immediate suspicions for they never appeared on a genuine Shakespeare work and were used several times on things we know were not his. The chapter "Parallels? Plagiarism?" (pp. 80-99) shows that none of Foster's claimed links between Funerall Elegye and Shakespeare stand up: they are all commonplaces or simple misreadings of the texts, and the next chapter shows that Foster's grasp of "Vocabulary and Diction" (pp. 100-120), especially Shakespeare's, was woefully inadequate to the task. Foster chose a list of 'function words' (ones that add no poetic content and the relative frequencies of which are stable across the Shakespeare canon), starting with and, but, not, so, that, by, in, to and with, but he made the fatal error (pointed out by MacDonald Jackson in 1989) of paring this list down by eliminating the last four and so entirely distorted the picture (pp. 114-9)

    Foster's database of 'representative' English Memorial Verse 1610-13 was horribly unrepresentative, and hence Foster convinced himself that using 'who' for an inanimate antecedent (as in, "hopes . . . who") when personification is not being used ("stars . . . who") was distinctly Shakespearian; it is not. Foster was also wildly inaccurate in summarizing the findings of A. C. Partridge and Charles Barber (pp. 121-138). Charles Bathurst established that three aspects of Shakespeare's prosody changed over his career: he moved the caesura, he made fewer pauses at the end of a line, and he used more double endings (p. 139), and by contrast Foster's analysis of the poetic structure of Funerall Elegye (and its likeness to Shakespeare's Sonnets) is a catalogue of half-truths (p. 141-4). To make his claims about Funerall Elegye's enjambment, Foster blamed George Eld's compositors for heavily punctuating W.S's lightly pointed text, but Vickers thinks that there is no reason to suppose that the copy was lightly pointed and that most evidence indicates that compositors usually respected copy punctuation (p. 146). This is not quite true, since Joseph Moxon advised that compositors should improve punctuation if necessary, and Vickers appears to concede this point in a footnote (p. 151n16) that cites D. F. McKenzie showing that Pavier's compositor B added heavy punctuation, especially at the ends of lines. Vickers objects to Foster's claim that "'As a measure of style any reasonable definition is adequate so long as the same criterion is used throughout'", thinking rather that one needs sensitivity to historical differences, specifically here between syntactic enjambment and non-punctuated lines (p. 147). This objection Vickers seems to have forgotten 350 pages (!) later when he writes in support of F. E. Peirce's claim that ". . . 'the value of any criterion depends largely on the consistency with which' it is used . . ." (p. 499), which is much the same thing as Foster claimed. One reason why there are so many pages in Vickers's book is that the publisher has permitted slack writing such as "wrote a particular poem in Anno Domini 1612" (p. 147); the words I have emphasized could go without hardship.

    Foster edited Funerall Elegye to remove most of the end-of-line punctuation, and then showed that in not having it (that is, in having enjambed lines) the poem was like late Shakespeare. Put the punctuation back in, as Vickers has done, and its rate of enjambment looks like Shakespeare in 1599, which is of course impossible given the date of Peter's death (p. 152). Piling up the evidence, Vickers shows that the poem's rate of feminine endings, 11.6% of lines, is far below the 30+% of late Shakespeare, that the frequency of hexameter lines, 0.35%, is far below the 2% average of late Shakespeare, and that the pause-patterns too are markedly different from late Shakespeare (pp. 153-9). The pioneer of pause-pattern analysis was Ants Oras who "showed that Shakespeare's early plays favour a caesura after the fourth foot; by 1600 the fourth and sixth positions are equally favoured, while in the later period the sixth position is dominant, with the unstressed seventh position gaining importance" (p. 156). The point is a good one, although Vickers surely means the fourth syllable not foot; pity the proof-reader of a 568-page monograph. Driving the final quantitative nails in, Vickers shows that study of proclitic and enclitic micro-phrases (that is, an unstressed monosyllable leaning forward to the stressed syllable following it, or backward to the stressed syllable preceding it) puts Funerall Elegye impossibly far from Shakespeare (pp. 160-2). The rest of this section gilds the lily with chapters showing that the hendiadys Foster saw in Funerall Elegye just is not there and that he did not really understand the rhetorical term properly (pp. 163-88) and that Foster is terrible at statistics (pp. 189-203). Most of the second part of the book, showing that Ford wrote Funerall Elegye, is not directly relevant to this review. In chapter 10, however, Vickers shows that many habits that Foster claimed were distinctly Shakespearean--such as the coining of un- words, the making of compound words, the use of 'very' as a restrictive adjective ('the very man'), and elliptic use of 'can' (as in 'I can not more')--are found in abundance in Ford, and Vickers adds more things of his own that Ford and Funerall Elegye share (pp. 302-62). Chapter 11 (pp. 363-431) completes the argument with a list of all the Ford parallels that Vickers found in Funerall Elegye.

    In an 'epilogue' with the surprisingly trendy title of "The politics of attribution" (pp. 422-65), Vickers considers the fascinating question of how Foster's nonsense was taken so seriously for so long. Or, what is wrong with the academy? Here Vickers examines the kudos Foster earned from identifying Joe Klein as the author of Primary Colors--there were hardly a lot of candidates in that case--and most shockingly his shameful intrusion in the JonBenét Ramsey murder case, first blaming the victim's half brother, who was never a suspect, then asserting the innocence of the mother, then asserting the guilt of the mother (pp. 458-62). Vickers finds a way to link the themes of this book--the academy's systemic failure in the case of Funerall Elegye--to his particular bugbear of "the 'social agendas' of race, gender, and class" and uses John M. Ellis's false analogy with Darwinian 'fitness': adverse criticism picks off the weakest elements of an intellectual species and "keep[s] it strong", without which "the species degenerates" (p. 463). The analogy is false because species do not degenerate without predators, and this view of 'species health' has disturbing overtones of social Darwinism. Also, the opposition of political criticism with pure disinterested research is false: many who consider themselves politically motivated in their criticism feel that their politics is best served by truthful models of the world, and that those schools of criticism that deny that they are political just want everyone to treat their politics as simple statements about reality and not a form of politics at all. In a final Vickersian paradox, he seems to accept a form of this argument, that what we call truth is often just whatever the powerful have got everyone to accept as truth, quoting C. S. Peirce making precisely this point (p. 465)

    On to the year's articles. The journal Studies in Bibliography seems to arrive at most libraries at least a couple of years after the date on the spine, so it has not been noticed before within this reviewer's stint for YWES that began with work published in 1999. Volume 52 (for 1999) is now available but contains nothing of relevance to this review. Volume 53 (for 2000) begins with, of all things, a new essay by R. B. McKerrow (2000) that is of great historical interest but does not fall within the scope of this review of work newly done. As reported last year, the journal Analytical and Enumerative Bibliography has closed. The 2002 issues of the journal The Library contained nothing of interest to this review. Volumes 14 and 15 of the journal TEXT allegedly contain work published in 2001 and 2002 and hence relevant to this review, but the industry of the librarians of the Shakespeare Institute of the University of Birmingham (United Kingdom) has failed to elicit a copy of either from the publishers (University of Michigan Press) and review of this work will be held over until such time as it may be seen. Likewise, the most recent Shakespearean International Yearbook was dated the end of the last millennium and more recent volumes will be noticed when (if?) they ever appear.

    Two articles in Review of English Studies were relevant this year. The first is Giles Monsarrat's demonstration that A Funerall Elegye is not by Shakespeare but by Ford (2002). This we can tell from certain phrases that are seldom or never used by Shakespeare but are common in Ford; presumably "W.S." was a man who wanted to honour William Peter so he got Ford to do it as a job of work. In the second article, Randall Martin expands upon the arguments in his Oxford Shakespeare edition of 3 Henry 6 (reviewed here last year), and specifically that memorial reconstruction and revision separate the octavo and Folio versions (2002b). Neither phenomenon on its own can account for the O/F relationship and the case for memorial reconstruction is clinched by the misreporting of the marriage regarding Lord Scales. In O the complaint is about Scale's marriage to the daughter of Lord Bonfield and in F it is the marriage of Scales's daughter to the new queen's brother and Bonville's daughter's marriage to the new queen's son. As Peter Alexander observed, the whole point is the king's favouring of the queen's relatives, so O's version makes no sense (pp. 10-11). Steven Urkowitz defended O as still making theatrical sense, but overlooked Alexander's crucial point that Bonfield does not exist in the chronicles, and Martin thinks the name came from another play (accidentally remembered by the actor), George a Greene (p. 12). Likewise Lord Cobham getting the (historically incorrect) personal name Edmund Brooke in O makes no sense other than as an actor's mistake or interpolation, and there are examples in O of characters anticipating what they come to know only later in the play, which is just the sort of trick memory plays (pp. 13-14). Here Martin gives more evidence than he presented in his play edition: O (sig. A7r) has York say he is going to St Albans and staying at Sandall castle, all in one speech, and the Keeper spots deposed Henry 6 in disguise before he has spoken (which is impossible) while F has him rightly overhear the ex-king and learn who he is from his own mouth (pp. 14-16). Some inartistic internal repetitions in O also betray its memorial-reconstruction origins and Martin here repeats his argument--rather more clearly than in the play edition--that O used mostly Hall for its history and F used mostly Holinshed. Even bits that we cannot relate to Hall or Holinshed show a general pattern of artistic reshaping of MSO to make MSF, and O and F are really different plays separated by authorial revision.

    The most important article this year is by Lukas Erne and appeared in Shakespeare Quarterly (2002). The substance of it is that we have no reason to suppose that Shakespeare was indifferent to the publication of his plays: companies appear to have favoured printing about two years after composition, but they were at the mercy of the market, which Blayney has shown was not good for playtexts. Alexander Pope started the myth that Shakespeare did not care about publishing his plays ("And grew Immortal in his own despight") and it is still often repeated, although in 1965 E. A. J. Honigmann suggested we think again. In the light of Blayney's fresh view of the economics of publishing plays, we can revisit E. K. Chambers's assertion that a company would fear that publication would give their plays to rivals and A. W. Pollard's that publication would hurt playhouse attendance. After all, The Spanish Tragedy was much printed and apparently stayed popular in performance. Richard Dutton suggested that it was the "contracted 'ordinary poets'" work that the companies wanted to keep out of print, but if so they failed miserably: only eight non-Shakespearian Chamberlain's men plays from 1594-1603 reached print, while in the same period 12 of Shakespeare's plays for that company were printed. Of course, judging whose plays were printed from the evidence of what has survived is unsafe: Shakespeare quartos may well have been valued more, so had a better chance of surviving. My list of the Shakespeare plays printed between 1594 and 1603 is Titus Andronicus, 2 Henry 6, 3 Henry 6, Richard 2, Richard 3, Romeo and Juliet, 1 Henry 4, Love's Labour's Lost, Henry 5, 2 Henry 4, Much Ado About Nothing, A Midsummer Night's Dream, The Merchant of Venice, The Merry Wives of Windsor, and Hamlet, a total of 15. Erne excludes Edward 3 but includes the lost Love' Labour's Won, so how does he get his figure of 12 plays? Presumably by excluding Titus Andronicus, 2 Henry 6, and 3 Henry 6 as not being written for the Chamberlain's men; life would be easier if the reader did not have to guess such details. Even discounting the so-called 'bad' quartos, Erne points out that there are a lot of Shakespeare printings to account for. Chambers and Andrew Gurr accounted for the publications around the turn of the century with the need for money to build the Globe and Fortune, but Neil Carson's analysis of the Henslowe Diary shows that while £5 or £8 might be paid for a script, £20 or £30 might be paid for costumes and other necessaries, and Blayney reckons that about 30 shillings were paid by a stationer for a play to print. In any case, as Erne argues, Shakespeare and his fellows were not hard up in the late 1590s to judge by Shakespeare's purchases and James Burbage's will.

    Erne sets himself the task of examining the first 12 plays Shakespeare wrote for the Chamberlain's men (p. 6), but frustratingly he does not state what he thinks those 12 plays are. Erne excludes The Comedy of Errors because he thinks it predates Shakespeare's joining the company, but does not explain how come Titus Andronicus and perhaps the early histories are not excluded for the same reason. In case the reader has retained a firm grasp of the lists that Erne is juggling, he switches at this point (p. 7) to the 'bad' quartos of plays written in the 1590s (Romeo and Juliet, The Merry Wives of Windsor, Henry 5, and perhaps Love's Labour's Lost) and argues that surely Shakespeare would have preferred good texts to get printed. Well, answering himself, Erne observes that in the cases of Romeo and Juliet, Hamlet, and Love's Labour's Lost the 'bad' quartos were indeed followed by good. (If you are wondering what Hamlet has to do with this--were we not confining ourselves to the 1590s for a moment?--then that makes two of us.) Erne wisely does not rest too heavily on this post hoc ergo propter hoc argument: in the case of Romeo and Juliet the stationer owning the good manuscript may have got it before the bad quarto appeared even though he published his good manuscript after the bad, although Erne offers no reasons to prefer this 'may have' to its opposite. In the case of Hamlet, the guess that the good manuscript underlying Q2-1604 changed hands before the printing of Q1 is "a strong, indeed the strongest, possibility" because James Roberts (who published Q2) entered Hamlet in the Stationers' Register on 26 July 1602, before Q1--printed by Valentine Simmes, Nicholas Ling, and John Trundle, without Stationers' Register entry--appeared in 1603.

    Actually, Roberts did not publish Q2 Hamlet, for its titlepage says "Printed by I[ames] R[oberts] for N[icholas] L[ing]", which means that Ling published it and Roberts was just the printer; Erne goes on to account for this. Erne finds that one theory fits these facts best:

. . . Ling and Trundell seem to have licensed but not entered their manuscript [that is, MSQ1] and had it printed without anyone realizing that Roberts had once entered a different version. Having found out about Ling and Trundle's unintentional breach, Roberts could have caused them trouble but may have preferred to negotiate an advantageous deal with his neighbors in Fleet Street, selling to Ling and Trundle his longer and better manuscript and having them pay him to print it. (p. 7)

In the case of Love's Labour's Lost, there is no extant bad quarto, but Q1's titlepages implies that one once existed by describing the text as "Newly corrected and augmented". But that claim is also made on the titlepage of Q3 Richard 3 where it is "demonstrably misleading" (Textual Companion p. 270) and Werstine has shown that the Q1 Love's Labour's Lost's copy appears to have been print, not manuscript, so there was a good Q0. So, in each of these three cases (Romeo and Juliet, Hamlet, and Love's Labour's Lost) it seems that the Chamberlain's men sold a good manuscript before the first edition, rather than that they reacted to the bad quarto by selling a good manuscript to supersede it. At least, that is what Erne thinks he has shown, but in fact he offers no evidence for his chronology in the case of Romeo and Juliet, he has a workable (but not the only) hypothesis for Hamlet, and has shown that there was no bad quarto of Love's Labour's Lost in the first place; this does not amount to showing that a good manuscript was sold before a bad quarto came out. What of the 'bad' quartos which were not superseded by good ones, Henry 5 and The Merry Wives of Windsor? Erne knows that it will not do to argue that the non-superseding of these shows the company's indifference to printing, for it may rather have been that they sold poorly and no publisher had reason to invest in a subsequent edition. Having dealt with four of the 12 plays (Romeo and Juliet, The Merry Wives of Windsor, Henry 5, Love's Labour's Lost), Erne comments that "Of the eight other plays Shakespeare is likely to have written for his company from 1594 until close to the turn of the century"--yes, yet another way of defining an unstated list--Love's Labour's Won (if it ever existed) cannot be discussed because lost and King John did not get printed until 1623 (pp. 8-9). Perhaps in the case of King John the company feared infringing the Troublesome Reign quarto of 1591.

    The other six Shakespeare plays (Richard 2, A Midsummer Night's Dream, The Merchant of Venice, 1 Henry 4, 2 Henry 4, and Much Ado About Nothing--that is, the remaining six of the first 12 that Shakespeare wrote for the Chamberlain's Men--were in fact printed between 1597 and 1600. For these Erne thinks we should look at the underlying copy of the printing to see if the players sold their manuscripts to the printer. The copy is uncertain for Richard 2 and The Merchant of Venice (authorial manuscript, or faithful transcript of it), for 1 Henry 4 is probably scribal transcript of authorial manuscript, and probably author's manuscript for A Midsummer Night's Dream, 2 Henry 4, and Much Ado About Nothing (pp. 9-10). Erne acknowledges recent objections to the New Bibliography that gives us these conclusions, but points out that there is nothing to "contradict the interpretation that any one of them [that is, the copy manuscripts] may (though not necessarily all of them must) have been in the possession of the Lord Chamberlain's Men and/or their playwright before being sold to a stationer".  Looking at the likely dates of composition and of entrance in the Stationers' Register, Erne finds roughly a two-year wait in each case. Leaving aside the corrupt Pericles, only two more of Shakespeare's plays were printed in his lifetime--Troilus and Cressida and King Lear--and again there is a two-year gap between composition (1601 and 1605, respectively) and Stationers' Register entry (1603 and 1607, respectively). One could quibble with some of the dates here, but mostly not by more than a year or so. Erne sees the danger of circularity in his method: some of the datings are dependant on assumptions about the unlikelihood of the players letting the printers get the plays, but he admits that in the case of A Midsummer Night's Dream "there's nothing beyond style to suggest a particular date" (p. 11). He has to say that, for the usual dating of 1595 is a full five years before the first printing, bucking his trend and prompting him to wonder if its being written for private wedding and not publicly performed until some time later solves the problem.

    Erne sums up crisply: ". .  of Shakespeare's first dozen or so plays written for the Lord Chamberlain's Men, not a single one that could legally have been printed remained unprinted by 1602" (p. 12) and the typical vector was the company selling a manuscript to a printer two years after first performance. Why wait two years? Because that was about the time to publicly promote a revival of a play (pp. 14-15). For some reason the printing of Shakespeare plays fell off after 1600: 13 plays in 24 editions from 1594 to 1600 (more than 3 a year) whereas only five plays in 19 editions from 1601 to 1616 (just over one a year). Blayney suggested that perhaps the market was glutted around 1600 (27 plays entered in the Stationers' Register between May 1600 and October 1601) and publishers were finding that they did not sell as hoped. The remainder of Erne's article deals with small objections that might be made to his main thesis and he concludes by observing, as Blayney does, that we have been looking at play publication from the wrong end (writers and companies) rather than from the end of stationers, publishers, and booksellers. The essential error, revealed by Blayney's scholarship, was that we assumed that demand for printed plays exceeded supply, but in fact it did not. Once we refocus our attention in the light of this, we can address some old problems with fresh insight: why are plays too long to be performed in two hours, what lies behind the 'bad' or short quartos, and how was a play text 'socialized'?

    One of those questions is addressed with fresh evidence by Jesus Tronch-Pérez who shows that a Spanish case of memorial reconstruction of a play gives comfort to the theory that Q1 Hamlet was made this way (2002). Memory man Luís Remírez signed his reconstructed copy of Lope de Vega's play La dama boba and we have Lope's autograph copy and a published version. Folio Hamlet deviates from Q2 Hamlet in much the way the memorial reconstruction of La dama boba deviates from Lope's holograph, and Q1 Hamlet is quite different again, so at first it looks like Folio Hamlet is a memorial reconstruction. But with a study of how the variants do their differing, this changes: whereas 45% of Folio Hamlet differences from Q2 are 'indifferent', 36.5% change the meaning, and 18.5% are uses of synonyms, in the memorial reconstruction of Lope's play nearly half the substitutions are "words of related meaning and paraphrases", 'indifferent' variants count for 36.5%, and 15% actually change the meaning. Compared to Q2, Q1 Hamlet too mostly changes that preserve the meaning and least numerous are those changes that alter the meaning; thus the Lope memorial reconstruction is like Q1 Hamlet, and this is especially true of those parts of act 1 involving Marcellus and, to a slightly lesser extent, his allegedly doubled role of Voltemar. The memory man Remírez sometimes brings forward a word that should occur later (sometimes much later) in the play, and Folio Hamlet does this in respect of Q2 and so does Q1, and Q1 repeatedly uses stock phrases in relation to a certain idea ("the cause and ground" of Hamlet's madness), and so does Remírez (pp. 44-46). Tronch-Pérez concludes: 

Of the textual features I have analysed in all three texts, Remírez's version of La dama boba has many more in common with the First Folio Hamlet than with the First Quarto Hamlet. . . . This shows that Remírez's memorial reconstruction was of such good quality that it resembled more the variant textual versions of Shakespeare's multiple-text plays such as King Lear, Richard III and Hamlet than the 'bad' quarto versions, and that it would not be judged a 'bad' or suspect text unless we had the external documentary evidence that indicates the contrary. (p. 52)

But the Spanish memorial reconstruction and Q1 Hamlet do have things in common: "In three kinds of textual alteration, both DB and Q1 interestingly show higher frequency than F1 does: synonymic and near-synonymic substitutions (from single words to paraphrase of several lines), internal repetitions of single words and phrases (including quasi-formulaic expressions), transposition of single words, phrases and lines" (p. 53). Importantly, Q1 Hamlet resembles the Spanish memorial reconstruction "in those segments in which the alleged reporter Marcellus intervenes". This does not prove that Q1 Hamlet is a memorial reconstruction, nor that the parts of Q1 Hamlet unlike the Spanish memorial reconstruction are not due to memorial reconstruction: the internal evidence is simply inconclusive and in the Spanish case we have reliable external evidence that tells us it is a memorial reconstruction. Thus, "Q1's explanation as a memorial reconstruction, based on internal evidence alone, remains a probable, but unproved, hypothesis" (p. 54), and importantly the recollections of a man consciously attempting to memorize a play would necessarily be different from the memory of a man performing in it and only later attempting to recall the text.

    Three articles from Shakespeare Survey were relevant this year. In the first Richard Knowles argues that Shakespeare might have seen King Leir in the 1590s but it did not much affect his writing until it was published in 1605, whereupon it became a source for Shakespeare's King Lear (2002). Henslowe's Diary records "Kinge Leare" performed 6 and 8 April 1594 and it was entered in the Stationers' Register on 14 May 1594 by Edward White and again on 8 May 1605 by Simon Stafford and then the same day transferred to John Wright who had Stafford print it later in 1605. Many commentators have thought King Lear little indebted to Leir, but Knowles lists the similarities of plot and argues that they show Shakespeare's "recent and detailed" knowledge of the source--most of them are not in the other sources--and there are quite a few verbal parallels too (pp. 14-17). McMillin and MacLean thought that Shakespeare was in the Queen's men (who played Leir), but if so it is hard to understand how come his pre-1594 plays (the murky period) were performed by Strange's men (1 Henry 6, Titus Andronicus), Pembroke's men (2 and 3 Henry 6, The Taming of the Shrew, and Titus Andronicus), and Sussex's men (Titus Andronicus). More likely, thinks Knowles, he moved between these companies and never played in Leir (p. 18). Scholarly claims that Leir influenced lots of other Shakespeare plays have been grossly inflated and Knowles sees no such influence at all, and even if accepted, there is no certainty that Shakespeare was the borrower rather than the lender.

    Knowles takes care to demolish claimed echoes of Leir in Shakespeare's plays other than King Lear, for example by showing that malapropism existed well before he wrote it for the Watch in Much Ado About Nothing (pp. 22-27). In any case there is no obvious means by which Leir might have influenced Shakespeare: no edition was printed from 1594 (recorded performance) to 1605, although if Shakespeare was in Sussex's men in April 1594, when they were sharing the Rose with the Queen's men--as Henslowe's column heading of plays performed by "the Quenes men and my lord of Susexe to gether" is usually taken to mean--then he might have seen it. But this would not give the occasion for repeated and sustained influence for the next 10 years that some scholars have claimed. Nor could he have had access to the manuscript: White would have kept the one he registered in 1594. Henslowe wrote that in May 1594 the Queen's men "brocke & went into the contrey to playe", which is a bit ambiguous ('broke' and carried on playing?), but they certainly did not disappear as a company but rather toured successfully for another decade, so they are unlikely to have given up the playbook of one of their most successful plays since they would need the licensed copy if challenged about their authority to play it on tour. Probably what they sold in 1594 was an authorial or scribal copy of their licensed playbook. Henslowe must have had faith in the Queen's men's provincial future, for he lent his nephew Francis Henslowe £15 to buy a share in them. The Queen's men disbanded with their patron's death in 1603 and that is probably when they sold off the licensed playbook of Leir and hence Stafford got it. This cannot have been the same manuscript that Edward White registered in 1594 because i) White did not transfer it to anyone, and ii) White's heirs continued to claim ownership of it long after his death (p. 30). The fact that the 1605 edition of Leir does not mention the company or the venue is also consistent with the copy coming from the disbanded Queen's men; they had not been in London for over a decade and there would be no advantage in harking back to 1594 when they last were. Hence the titlepage's simple claim that the play was "divers and sundry times lately acted".

    On the other hand, what got printed does not seem to be a battered and presumable written-over licensed playbook (marked for a reduced company on tour) but rather a clean pre-theatrical text for a full-sized company. And what of the fact that Wright was formerly White's apprentice? Wright had only just got his freedom, so he is hardly likely to have risked his whole future on a piracy of his master's possession. Knowles's ingenious hypothesis is that Stafford got hold of a manuscript from the disbanded Queen's men, went to register it and found that White had registered the play 11 years before. Stafford went to White who did "two good deeds at once": he allowed Stafford to print it, if he made Wright (White's newly-freed apprentice) the publisher; then White handed over his own manuscript (the one he registered in 1594) and that is what got printed and hence the 1605 printing is not like what we would expect from a manuscript that had been used on tour for a decade. Or perhaps White just let Stafford's compositor look at his 1594 manuscript to check readings. There is no evidence that any of the newly-formed Chamberlain's men of 1594 came from the Queen's men, so in all of this theatre and textual history there is no obvious means for Shakespeare to get hold of a manuscript of Leir. The obvious impetus for his doing his Lear play was simply the publication of Leir in 1605.

    The second piece from Shakespeare Survey is Sonia Massai's claim that the Third Folio copy for the Fourth Folio was editorially annotated, probably by Nahum Tate and hence Rowe should not be counted the first editor of Shakespeare (2002). Massai lists occasions when F4 speculatively emends its copy, F3, to produce really rather good (even F1) readings in place of bad; but this was not done by reference to F1 since "on other occasions mistakes first introduced in F2 or F3 are not emended according to F1" and Massai points out that the F3/F4 corrections she has noted could not happen during stop-press correction (pp. 260-61). An example: F4 wrongly calls a character in Coriolanus "Titus Lucius" where F1-F3 had the correct "Titus Lartius"; the change probably happened because in F1-3 this man is addressed (wrongly) as "Titus Lucius". It is unlikely that a compositor of F4 adopted "Lucius" (which is what he set) throughout his work, and against his F3 copy, on the basis of this one line of dialogue. More likely his F3 copy was annotated with all the "Lartius" readings changed to "Lucius" readings. Similarly, elsewhere in F4 Coriolanus speech prefixes are altered (that is, the F3 speech prefix overruled) to make them match names as spoken in dialogue.

    Some F4 plays were much more altered (in respect of their F3 copy) than others, which also suggests not correction in the three printing houses (which would tend to be uniform across plays printed in each house) but editorial intervention at the level of F3 copy and varying according to "the editor's familiarity with a specific text, or the intrinsic quality of his copy-text". Confirming this is the "consistency of procedures" across the division of labour in printing F4 between the three printing houses: compared the F3 copy, commas are used to improve syntax, apostrophes are used to mark contractions and possession, and spelling is modernized. To show that this is not simply what anyone printing in the 1680s would do with copy from the 1660s, Massai looked at The Northern Lass (printed 1663 and 1684) and A Jovial Crew (printed 1661 and 1684) and found that the changes between F3 and F4 are much more numerous than the changes in these two Brome plays. It is clearly intentional editorial change, not a shift in the general climate (p. 262). Looking at other printings of belles-lettres there appears to be a nascent 'editor' function in the sixteenth and seventeenth centuries even though it did not become a distinct job title until the early eighteenth century. As the literary-author-as-sole-originator-of-his-meanings came to be constructed in the seventeenth century, there was some reluctance to admit that editing had taken place, and Massai quotes prefaces that refer to the need to make sure that printers do not mangle the author's words (p. 265). Writers tended not to mention editorial annotation of copy prior to printing because it was "ideologically controversial" although necessary, and increasingly printing houses built up communities of scholarly free-lancers who worked on copy prior to printing, as when Nicholas and John Okes hired Thomas Heywood (whose plays they had printed) to work on others' plays, including making choice of copy, making explanatory notes, and indices. Henry Herringman, one of the F4 publishers, probably had the services of John Dryden, and another Herringman reprint, Cutter of Cole-man Street (1663 and 1693), has levels of editorial intervention similar to those that Massai has found between F3 and F4, and in particular such features as making speech prefixes match what someone is called in dialogue (p. 268). The likeliest candidate for the secret editor of F4 is Nahum Tate, for Dryden was at the time working for another publisher, had in past passed work on to Tate, and we know that Tate subsequently did editorial work for Herringman. F4 Coriolanus has "Commons" where F3 has "Commoners", which Massai thinks would be a typical intervention from Tate, for he does not characterize the populace as a rabble in his own adaptations of Shakespeare.

    Finally from Shakespeare Survey comes Michael Cordner's argument that modern editions are not as stage-aware as their creators would like to think and are far too quick to close down meanings rather than explore the multiple performance possibilities latent in the words (2002). Standards for annotation in a Shakespeare edition are, Cordner observes, almost never discussed and being stage-centered should not just mean thinking about action but also should include thinking about words as actions. Cordner catches Rene Weiss and Philip Edwards offering glosses that assert the superiority of one possible interpretation of a particular moment over another without saying why it should be preferred. Shakespeare probably wrote knowing that the full semantic possibilities of the words he was using were not available to his imagining let alone his control: the actors' craft (including intonation, stress, timing, blocking) brings out different ones, often differing between performances. In particular Cordner object to R. A. Foakes's glossing of Antipholus of Syracuse's statement that in looking for his mother and brother he will "lose myself" as perhaps meaning 'lose my wits', on the grounds that at that this stage in the play (1.2) there is no reason to suppose he will lose his wits, and comments:

To a spectator already acquainted with the play, the phrase may acquire proleptic irony, given the strange experiences which await the character, and which will in due course lead him to doubt both his own and the Ephesians' sanity. But that is a layer of dramatic implication quite distinct from anything the actor of the Syracusan Antipholus can here represent his character as consciously intending; and those layers should be systematically distinguished by the annotator. (p. 190)

On close inspection, this comment is awfully carefully phrased (a model for editors?), for it allows that the actor might convey the proleptic irony--why should not the dramatist and players aim to entertain those who see the play more than once?--but that this conveyance is not the same as what the actor "represent[s] his character as consciously intending".

    F. Elizabeth Hart argues that there is no missing music scene in the quarto of Pericles, as has often been claimed, because what the sources have has been replaced with a dancing scene instead (2002). The morning after the revels, Simonides thanks Pericles for making beautiful music but the audience did not see him do this or hear it. The sources have the hero and his future wife, the king's daughter, in a harp-playing contest, and the Oxford Complete Works used Wilkins's 1608 novella to put in a bit of Pericles playing, but the New Cambridge editors think this hubris. Hart agrees because Shakespeare makes the female body itself be a musical instrument (Antiochus's daughter as a "faire Violl" in the first scene), there are father/daughter pairings at the beginning (Antiochus and daughter), middle (Simonides and daughter), and end (Pericles and Marina) of the play. There are other symmetries between the beginning of the play and its middle--rather than incoherences derived from dual authorship--such as enflaming, tournamenting, and the playing of a women like the playing of an instrument, discordant in the case of Antiochus and harmonious in the case of Simonides "arranging" his daughter for Pericles. Where some editors think there is a missing scene there is a dance, and this Hart thinks is deliberate: instead of a divisive competition of producing music there is a unifying consumption of it. This is a "condensation of energy" and a "conservation of energy" by which "Musical mastery is condensed into metaphor, in which form it offers rich meaning but never threatens to monopolize stage dynamics" (p. 9). Thus Hart sees no missing scene, although one cannot help wondering why Simonides makes a reference to playing music, so calculated to make readers, audiences, and editors think that something has fallen out.

    In an article from Literary and Linguistic Computing, MacDonald P. Jackson uses new statistical analyses of old data to show that the Oxford Complete Works's chronology of Shakespeare's plays is essentially right (2002b). Ants Oras measured where the pause tended to fall in Shakespeare's line (after 1st, 2nd, 3rd, 4th, 5th, 6th, 7th, 8th, or 9th syllable) across his career and found that like his fellow dramatists Shakespeare moved away from the standard pause after the 4th syllable (the dominant pattern until near the end of the sixteenth century) and started to put the pause in the second half of the line. Oras published the findings as a book with charts in 1960, and Jackson has done statistical analysis on it to make a matrix of "Pearson product moment correlation coefficients" that show how similar each play's pause pattern is to each other play's pause pattern. Jackson offers a table that shows for each play its Oxford-assigned year and then the five other plays with the closest pause patterns to it; as we would expect the plays composed around the same time tend to be alike. This strongly confirms the Oxford chronology, although as Jackson notes that chronology was in part determined using Oras's work. Where there is mixed authorship (as in 1 Henry 6, Titus Andronicus, and Timon of Athens) clouding the issue--since Oras took the whole play in all cases except Pericles, Henry 8, and Two Noble Kinsmen--the pattern is not much disturbed. This is because Shakespeare followed the general climate of change (that is, all the dramatists changed together) in respect of pause patterns. King John has strong links with A Midsummer Night's Dream (1595), Romeo and Juliet (1595), Love's Labour's Lost (1594-5), The Comedy of Erros (1594), and Richard 2 (1595), so it can be dated about 1595 if the Oxford dating of these other plays is right; hence Troublesome Reign (printed 1591) was a source not a copy of King John. The Merchant of Venice has links with 1 Henry 4 (1596-7), Much Ado About Nothing (1598), Henry 5 (1598-9), Julius Caesar (1599), and As You Like It (1599-00, not "1599-60" as given here), and hence perhaps it should be dated after the three Falstaff plays not before them.

    The Merry Wives of Windsor oddly has links with a lot of plays widely separated in time, perhaps because it has so little verse (240 lines) and this is a test that relies on verse. Likewise 2 Henry 4 has widely-spread links and is more than half prose, although the averages of the widely-separated dates for these two plays come out fairly near the Oxford chronology figures. But for Troilus and Cressida the average of the link-plays' dates is 1597, which is rather earlier than previously thought and the play might have been misdated. Also oddly, Othello (1603-4) and All's Well that Ends Well (1604-5) are close in the chronology but do not appear in one another's list of the five other plays most like it in its pause pattern. Othello is most like Hamlet (1600-1) so perhaps holding on until Richard Knolles's History of the Turks (1603) came out--because it is a presumed source--is an error: details of the Turkish fleet's movements could have reached England before this book. Similarly, perhaps All's Well should be moved a little later to get within the era of the plays it is most like. All this bears also on authorship: the first two acts of Pericles (Wilkins's alleged work) link to middle-period Shakespeare (1597-1604) whereas the other three acts link to Shakespeare's late plays, just as we would expect if in 1607 he wrote acts 3-5 but not 1-2. Confirming this is the likeness of acts 1-2 with other Wilkins's work of the same period. Likewise, the alleged Fletcher scenes in Henry 8 link with the alleged Fletcher scenes in Two Noble Kinsmen more highly than with any wholly Shakespearian play, although the next closest thing they are like is the Shakespeare scenes in Henry 8, and the same is true of alleged Fletcher scenes in The Two Noble Kinsmen. In short Fletcher and Shakespeare are "clearly distinguishable in their pause patterns" and Oras was probably right in thinking that pause patterns were an unconscious phenomenon. King John still remains a problem: if Troublesome Reign is a borrower from it, then the whole first half dozen of Shakespeare's plays need to be shunted about three years earlier than they currently are reckoned to be.

    Two articles of interest appeared in Papers of the Bibliographical Society of America. In the first James Hirsh argues that the Jaggards, not the King's men, were responsible for the division of the 1623 Folio texts into acts, and that most of the work was carried out by Ralph Crane (2002). The 'act' does seem to be the sub-unit of a play, but not forming one-fifth as we might think, to judge from the evidence of Jacques's "His acts being seven ages" and the division of Pericles by eight choruses. Certainly act intervals spread from the Blackfriars to the open-air theatres after about 1607, but only slowly and moreover "The main venue, and the one from which the shareholders derived most of their profits, remained the Globe . . . "  (p. 224). It would indeed be interesting to see evidence for this statement, as it runs counter to the overwhelming evidence that the actors always wanted to be indoors in the affluent districts; Hirsh offers nothing to support his surprising claim. Hirsh finds no evidence that Shakespeare changed his style to suit division into five units, and thinks that The Winter's Tale is in essentially two parts. So, if Shakespeare did not start to write in acts even after 1607, where did the act divisions in 28 of the Folio's 36 plays come from? It cannot be a theatre person, because some of the divisions are rather inept, and Hirsh conjectures two patterns of division based on very simple principles. Pattern A plays are The Tempest, The Two Gentlemen of Verona, The Merry Wives of Windsor, Measure for Measure, The Comedy of Errors, Much Ado About Nothing, As You Like It, Twelfth Night, The Winter's Tale, 1 Henry 4, 2 Henry 4, Henry 5, Henry 8, Titus Andronicus, Julius Caesar, Macbeth, Othello, and probably The Merchant of Venice (18 plays), while Pattern B plays are Love's Labour's Lost, A Midsummer Night's Dream, All's Well that Ends Well, King John, Richard 2, 1 Henry 6, Richard 3, Coriolanus, and The Taming of the Shrew (nine plays), leaving Cymbeline and King Lear to their own anomalous patterns and seven plays (2 Henry 6, 3 Henry 6, Troilus and Cressida, Romeo and Juliet, Timon of Athens, Antony and Cleopatra, and Hamlet) with no divisions. The rules are mechanical ones of the kind 'look for scene breaks', 'even out the chunks', and 'do not put an interval before an alarum or excursion direction'. Although he probably did not count lines, Divider A's results add up to Hinman Through Line Numbering counts being almost perfectly divided by five, while Divider B was less finicky about numerical proportionality in his divisions and more interested in starting acts with large entrances.

    As one might imagine, Hirsh has to give Divider B responsibility for A Midsummer Night Dream's direction "They sleepe all the Act" since he has already ruled out Shakespeare as its author. Hirsh calculates the chances that the patterns were arrived at by artistic means and that the rules he has constructed just happen to fit as well, and they are small for any one play and virtually nil for so many together. After some categorical statistical statements, Hirsh perhaps anxiously writes that his argument is "at least as reasonable and as firmly grounded in hard evidence as the arguments that scholars have been making for generations about the methods of compositors who worked on the Folio" (p. 244), which is true but not much of a recommendation. Five of the Pattern A plays (The Tempest, The Two Gentlemen of Verona, The Merry Wives of Windsor, Measure for Measure, and The Winter's Tale) are known to have been printed from Crane transcript, and he is the obvious candidate for being Divider A since the chances of all five of Crane's transcripts randomly being assigned to Divider A are small and he certainly did not scruple about "literary embellishments", as Jowett called them. Crane's habits with his literary transcripts of non-Shakespearian drama seem to fit the Divider A pattern too. Hirsh ends by speculating (rather wildly) about the availability of Dividers A and B during the print-run of the Folio in order to explain why some plays were divided and others not.

    In the second article from Papers of the Bibliographical Society of America, Marvin Spevack surveys nineteenth-century attempts to produce cheap single-volume editions of the complete works (2002c). Spevack focusses particularly on James Orchard Halliwell-Phillipps's plans hatched in 1863 for a shillling edition that never materialized, in part it seems because the editor himself had conflictual feelings about the kind of people who would buy it. Others' plans for a shilling-Shakespeare succeeded and Spevack usefully contextualizes them within patrician ideals about bringing all English-speaking men together and within laissez-faire economics that were almost bound to succeed; someone would inevitably work out how to get the thing made for almost no cost.

    Spevack is a veteran of computer applications to Shakespeare and elsewhere argues that the machines have not delivered what we wanted and that literary scholars need to shape what their software providers create (2002b). Although Spevack's concordance was completed nearly 40 years ago, no-one has yet done anything more interesting with this kind of approach. The only intelligent contribution that Spevack thinks he made to the concordance was the algorithm for working out how much context to give for each hit, and he was disappointed that no reviewer commented on this. Modern technology has given us much more data and made it available more quickly, but where we are woefully lacking (because the computer people do not really care about it) is in the organization of data in meaningful ways. (I would take slight exception to that comment, since web-crawler indexing software underlying search engines is remarkably efficient and, incredibly, one generally does find what one wants.) In the early 1970s we thought that we would eventually have a systematized Shakespeare data centre holding all the data organized coherently, but the new technology has not brought this. Rather, there are scraps of data indifferently organized and in lots of different places, and much of the software upon which we rely imposes structures on the data that the users remain unaware of. Spevack discusses the problems of hand-tagging for content and imagines the process being computerized, in the course of which he rather confusingly for most people mentions TEI without glossing this as the Text Encoding Initiative. More importantly, he does not address the argument that some such as Peter Robinson are making the case that perhaps we should not worry about tagging at all but rather make our search engines better at understanding texts that are not tagged. Elsewhere Spevack offers a detailed history of James Orchard Halliwell-Phillipps work getting his sixteen-volume 'Folio' edition of Shakespeare (1853-65) completed (2002a).

    John V. Robinson thinks that Hamlet's "dram of eale" should be emended to "dram of ale" (2002). In Jonson's Bartholomew Fair, Northern says "the eale's too meeghty" ('the ale's too mighty') and although Robinson notices that it is a drunken scene he wonders why no other dialect-speaking character in the play says "eale" to mean 'ale'. (I would hazard the answer that the spelling was meant to indicate a northern English pronunciation, and that only Northern speaks this way.) Robinson thinks "eale" in Q2 is a "compositor's error" for 'ale, and he takes trouble to explain what I would have thought well known, that 'dram' means a small serving of spirits. Thus "the dram of eale / Doth all the noble substance of a doubt / To his own scandle" (Q2, sig. D1v) should be emended to "The dram of ale doth all the noble substance often doubt to his own scandal", meaning that "people doubt [are skeptical of] the power of liquor and it often leads to their ruin". Robinson explains his hypothesized 'ale' to 'eale' compositorial error by pointing out that Shakespeare's handwriting had an 'f' with a final flourish that looked like an 'e', so 'of ale' could easily be read as 'of eale'. He is right that Hand D of Sir Thomas More has this flourish, but although it looks like an 'e' to us, it looks nothing like the 'e' that Hand D writes, whether initially, medially, or terminally. Robinson imagines that the compositor was overworked and inexperienced, but the point of John R. Brown's compositorial study he cites was that the same two men who set Q2 Hamlet had previously printed The Merchant of Venice, so they were not inexperienced.

    David Haley thinks that we should emend Q2 Hamlet to read "the dram of esill / Doth all the noble substance often sour" (2002). Haley uses the example of Cassio (whose honour was lost by drinking) to illuminate Hamlet's "dram of eale" speech, the link between the Ghost (which is what the characters are waiting for) and drinking being that both involve spirits. Wine is the "noble substance" that gets tainted and Haley uses a bit of Nashe's Pierce Pennilesse that he thinks inspired this speech to help make sense of what Shakespeare wrote. The suggestion that "of a doubt" is a compositor's mistake for "often dout" (that is, the bad thing extinguishes the good) is not much help, Haley decides, because the sense is clearly of transforming good to bad, not extinguishing it. Haley considers the palaeographical possibilities for misreading, and then gives them over to suggest that "of a doubt" was in Shakespeare's manuscript "often sour", even though he admits in a footnote that the long 's' that would begin "sowre" (the likely spelling) really could not be misread as a 'd'. It is vinegar that sours wine, Haley asserts, and later in Q2 Hamlet  (the 'eat a crocodile' speech, on sig. M4v) Shakespeare spells it "Esill", and Haley makes a not-unreasonable palaeographic case for "Esill" being set as "eale". In the same book Janus Lull offers what looks like an essay about editing but turns out to be about the resonances of Buckingham's dying words on All Soul's Day and how this relates to Richard 3's parade of dead souls before the battle of Bosworth field (2002). More relevantly, the book also carries Linda Anderson's argument about the cast size for the first quarto of The Merry Wives of Windsor (2002). This is normally assumed to be a touring text and Anderson notices that the stage directions for servants doing things outside their normal duties are accurate and explicit, but those for them doing their usual duties are haphazard and in several cases manifestly wrong. Perhaps "The adapter and the audience may simply have assumed that servants would accompany their masters onstage whether or not they had any part to play in the action" (p. 70). This sounds profligate because we tend to assume that on tour the companies used the minimum possible number of actors for each play, but Anderson makes the incisive point that the minimum needed for the largest-cast play would determine the number of actors actually present on the tour, and other plays in the tour's repertory could use more than the minimum number of actors needed for each play instead of having men idle.

    William B. Hunter wrote the only two articles of relevance from American Notes and Queries. In the first he argues that perhaps Heminges and Condell themselves emended the quartos that became Folio copy, using their knowledge of performance of these plays (2002a). Folio Titus Andronicus differs from the quarto only in scenes in which Marcus is present, perhaps because the actor playing him annotated the Q copy that made F. Hunter tries to apply this principle to Love's Labour's Lost and Much Ado About Nothing, and as one might expect it gives him explanatorial carte blanche: when the 'correction' is right, the actor is remembering what he performed, when 'wrong' he is misremembering. Strangely, Hunter seems to think that the error in the Much Ado About Nothing quarto's direction "Enter Leonato, his brother, his wife, Hero his daughter, and Beatrice his neece, and a kinsman" (sig. B3r) is that of Hero being called Leonato's wife. The problem, of course, is that of Leonato having a wife at all. In his second article Hunter argues that Shakespeare revised A Midsummer Night's Dream for different weddings, creating some of the textual muddles we have, and that Heminges played Egeus and annotated Q2 to make F's copy (2002b). Hunter recaps his own work elsewhere on the play being an occasional piece for a wedding in 1594 and revived for another wedding in 1596, and how its content fits the known time schemes of the performances. Like David Wiles, Hunter thinks that Romeo and Juliet must come earlier because Pyramus and Thisbe is clearly a parody of it, and a parody cannot come before what it parodies. (This is surely a weak argument: both can be different workings of essentially the same source material.) The play's multiple endings--the mechanicals' dance, then Theseus epilogue sending everyone to bed, then the fairies' masque, and finally Puck's epilogue--come from the different performances at court, where Puck's "gentles" would be inappropriate, and on the public stage where it would not. Hunter's best point is that the play as we have it has an oddity in Hippolyta's part. She is in the first scene but says almost nothing (surprisingly for an Amazon in a scene where a woman is being compelled to obey a man) and then she reappears in Act 4 where she speaks a considerable amount. Hunter thinks there was more for Hippolyta and that it has dropped out in revision. The beginning of the play also has Philostrate who in F never reappears (Egeus takes over as Master of the Revels) and this too Hunter reckons to result from authorial revision. Hunter takes a guess--that is all it is--that Heminges did the textual work on the Folio's copy by annotating an example of Q2 using his memories of being in it. Thus it was he who put in the actor's name "Tawyer" and changed all the Philostrate speech prefixes to Egeus (as it was when he played it, the Philostrate name coming in later revision and then getting into Q2) but he missed one.

    Jeremy Ehrlich notes that if, as many believe, Q1 Hamlet is based on a recollection of a markedly different version of the play from that we know from Q2/F, we could for sport put F's words into Q1's structure, character names, and pattern of cuts (2002). To do so would be to eliminate some of the linguistic flaws of recollection and thereby inch a little closer to the lost, different version. After all, the remembering actor is not likely to have invented the order of scenes and the plot of the play he was recalling, and the majority of his errors would have hurt only the language. Noticeably, F's verse lines make neater patterns when Q1's cuts are applied to them than they do in F, particularly by the elimination of unmetrical short lines. Could the adapter who made the good text that is buried under Q1's misrememberings have been trying to tidy the verse? Ehrlich ends with the quite reasonable complaint that "identifying what we do not know about early printed texts" has too often put a brake on "speculative projects" that would be revealing.

    Finally, the round-up of material in Notes and Queries. Andrew Breeze thinks that Ophelia's "the owl was a baker's daughter" (4.5.41-2) comes from a medieval Welsh tale circulating in various forms about a girl punished for sexual betrayal (2002). He has not got any substantial evidence for this other than the fact that there was a Welsh story about a girl turned into a owl; unfortunately she was not a baker's daughter. Charles Edelman shows that Peele's play The Battle of Alcazar is the same as the Muly Molocco that Henslowe's records shows was played 14 times between February 1592 and January 1593 (2002). The quarto of The Battle of Alcazar has two stage directions that call for "chambers" to be "discharged" and the only other plays of the period that use this phrasing are the quartos of Shakespeare's 2 Henry 6 and 3 Henry 6, both Strange's men's plays from around 1591. This link of discharged chambers suggests that The Battle of Alcazar was also a Strange's men's play and that the 1594 quarto of it with these stage directions was based on copy written by the same person who wrote the 2 Henry 6 and 3 Henry 6 directions. Thus The Battle of Alcazar is the same as Muly Molocco that Henslowe records, and the reason why the main character is much more often called by his alternative name of 'Abdelmelec' than 'Muly Molocco' is that it sounds better and scans much more easily.

    Thomas Merriam thinks that Marlowe wrote the Joan-as-witch scene in 1 Henry 6 (2002). The penultimate appearance of Joan la Pucell's (or "de Pucell", as Merriam unaccountably calls her) has her addressing her diabolical helpers and offering her soul, and it shares quite a few words and phrases with Marlowe's works, especially Doctor Faustus. Merriam claims that the appearance of devils onstage in 1 Henry 6 "is unique in Shakespeare", but that is true only if we forget about the conjuring of Asnath in The Contention of York and Lancaster, its sequel. Merriam uses "1st Principal Component" and "2nd Principal Component" without explaining what these terms mean, nor the difference between them, and there is not even a citation telling the reader where to find an explanation. Worse, there is something strangely wrong with the graph that Merriam thinks explains it all: the horizontal access (labelled "1st PRINCIPAL COMPONENT (ALL)")--the one that Merriam says is all that matters for distinguishing authorship--goes in six uniform steps of 0.2 left from 0.8 to -0.4, which is fair enough, but then takes another six uniform steps to get to -1.5 (should be -1.6). Likewise, it goes in six uniform steps (of 0.2?) right from 0.8 to 1.9 (should be 2.0) and then a further six uniform steps to get to 3.0, which should be 3.1, or the previous mark should have been 1.8. In short, this is not a linear scale but has been made to look like one by small tweaks; this does not encourage confidence. David Farley-Hills thinks that the placing of entrance and exit directions shows that The Comedy of Errors was written for indoor hall performance (2002). Mariko Ichikawa showed that on the open-air stages Shakespeare's major characters were usually allowed four lines to enter or exit to or from their place on the stage (front centre), and that minor characters--who tend to stay in the background--usually are allowed two lines. We have always suspected that The Comedy of Errors was written for the smaller stage of an indoor hall performance, where entrances and exits could have been made much more quickly, and indeed the play is anomalous in giving characters less than the usually four lines. Farley-Hills refers to Ichikawa's "unpublished doctoral thesis" but it has since appeared as a splendid book (2002).

    J. J. M. Tobin thinks that Titus Andronicus borrows from the epistle to the reader in Nashe's Christ's Tears Over Jerusalem (1594), which itself responds to criticism of Nashe's Unfortunate Traveller (also 1594), so Shakespeare cannot have written the play earlier than that year (2002). The epistle refers to criticism that Nashe makes up -ize verbs such as 'tyrannize', which appears twice in Titus Andronicus, and that Nashe makes compound nouns, as does Shakespeare. Tobin cites a few, not terribly close, verbal parallels between Christ's Tears and Shakespeare's play, and makes an argument for the direction--Shakespeare borrowing from Nashe--that I cannot understand and that he does not seem to find convincing either: "it is certainly possible" that Nashe borrowed from Shakespeare". Adrian Streete notes that Berowne's "charity itself fulfils the law" (Love's Labour's Lost 4.3.337) is usually taken to be a biblical allusion (Romans 13:8 or 13:10), but in fact it is not directly from scripture but from Calvin's Sermons (1549), which was itself alluding to the bible (2002). As Streete observes, early modern writers could use Calvin without getting partisan about his doctrines. Randall Martin glosses the page's report that Falstaff sups in Eastcheap with "Ephesians . . . of the old church" (2 Henry 4 2.2.142) as meaning 'with Catholics' (2002a). The Geneva bible likened Catholicism to the old cult of Eastern Diana/Artemis worship--connoting fertility not chastity as in the West--at Ephesus, whose adherents resented Paul's first-century mission to replace it with Christianity. Naseeb Shaheen objects to the Arden 3 editor David Daniell finding in Luke 1:8 the source for Cassius's "Will you go see the order of the course" (Julius Caesar 1.2.25), since it comes from North's Plutarch describing the same moment (2002). Shaheen suggests that a faint memory of another line in Luke ("as his course came in order") in a totally unrelated context might also have shaped Shakespeare's phrasing.

    William Poole supports others' recent arguments that the anonymous Tragedy of Caesar and Pompey, performed at Trinity College Oxford, is a source for Julius Caesar (2002). Poole spots a new link: "To grace in captive bonds his chariot wheels" (Julius Caesar 1.1.34), said of Pompey, has "grace[d]", "captive", and "chariot wheels" in common with one of Pompey's speeches from the other play. Tiffany Stern has a new source for the mock trial in Shakespeare's King Lear (2002). Greene's mad-for-jealousy character Orlando Furioso--much more known than Ariosto's Italian or Harington's English versions--was frequently alluded to, and is the only role for which we have a surviving actor's 'part'. This contains a mock trial by mad Orlando that Stern thinks similar to Lear's; could Child Rowland be Orlando? Stern assumes that because the 'part' was "found among Alleyn's effects" it must be his, but in fact David Kathman will shortly publish an article showing that at least one other document in the Dulwich cache, the plot of 2 Seven Deadly Sins, got there decades after Alleyn's death. The editors of Notes and Queries might have tidied some of Stern's phrasing, such as the comment that we have the 'part' and a play quarto "both of which differ significantly from each other"; as opposed to only one of them differing? Stern rightly condemns W. W. Greg's explanation that the 'part' is 'good' and the quarto is 'bad' and wonders if the differences come from revisions associated with Greene's double-selling of the play, first to the Queen's and then to the Admiral's men.

    MacDonald P. Jackson reports that fresh processing of rare-word analysis undertaken by J. A. Fort in the 1930s produces the same links between certain runs of Shakespeare's sonnets and certain groups of Shakespeare's plays that Jackson found and reported in an article reviewed here last year (2002a). Interestingly, Fort himself did not read his data this way and remained convinced that all the sonnets were written 1593-6. Kenji Go notes that "I am that I am" (sonnet 121 line 9) is generally taken to be from Exodus 3:14, but since these are God's words that would seem to make the poet sound "smug, presumptuous, and stupid", as Stephen Booth put it (2002). In fact the phrase also occurs in 1 Corinthians 15:10, which would have been familiar from prescribed pulpit reading, and in which context it shows St Paul's humility, not megalomania; Iago's "He's that he is" (Othello 4.1.270) is a witty parody of Paul's phrase. Bryan Crockett argues not very convincingly that some lines for which we already have Shakespeare's sources actually come from hearing and/or reading the charismatic preacher Thomas Playfere (2002).

Works Cited

Anderson, Linda. 2002. "'Who's In, Who's Out': Stage Directions and Stage Presences in The Merry Wives of Windsor, Q1." 'A Certain Text': Close Readings and Textual Studies in Shakespeare and Others. Edited by Linda Anderson and Janis Lull. Newark NJ. University of Delaware Press. 65-74.

Astington, John H. 1993. "The Wits Illustration, 1662." Theatre Notebook. 47. 122-40.

Barnes, Barnabe. 1904. The Devil's Charter. Ed. R. B. McKerrow. Materialien zur kunde des alteren englischen Drama. 6. Louvain. A. Uystpruyst.

Blake, N. F. 1998. "Shakespeare's Sonnet 69." Notes and Queries. 243. 355-57.

Blayney, Peter W. M. 1997. "The Publication of Playbooks." A New History of Early English Drama. Edited by John D. Cox and David Scott Kastan. New York. Columbia University Press. 383-422.

Bowers, Fredson. 1952. "The Problem of the Variant Forme in a Facsimile Edition." The Library (=Transactions of the Bibliographical Society). 7. 262-72.

Breeze, Andrew. 2002. "Welsh Tradition and the Baker's Daughter in Hamlet." Notes and Queries. 247. 199-200.

Cordner, Michael. 2002. "Actors, Editors, and the Annotation of Shakespearean Playscripts." Shakespeare Survey. 55. 181-98.

Crockett, Bryan. 2002. "From Pulpit to Stage: Thomas Playfere's Influence on Shakespeare." Notes and Queries. 247. 243-45.

Duncan-Jones, Katherine. 2001. "Ravished and Revised: The 1616 Lucrece." Review of English Studies. 52. 516-23.

Edelman, Charles. 2002. "The Battle of Alcazar, Muly Molocco, and Shakespeare's 2 and 3 Henry VI." Notes and Queries. 247. 215-18.

Ehrlich, Jeremy. 2002. "The Search for the Hamlet 'Director's Cut'." English Studies. 83. 399-406.

Erne, Lukas. 2002. "Shakespeare and the Publication of His Plays." Shakespeare Quarterly. 53. 1-20.

Farley-Hills, David. 2002. "The Theatrical Provenance of The Comedy of Errors." Notes and Queries. 247. 220-22.

Foakes, R. A. and R. T. Rickert, eds. 1961. Henslowe's Diary, Edited with Supplementary Material, Introduction and Notes. Cambridge. Cambridge University Press.

Frazer, Winifred. 1991. "Henslowe's 'Ne'." Notes and Queries. 236. 34-35.

Go, Kenji. 2001. "Unemending the Emendation of 'Still' in Shakespeare's Sonnet 106." Studies in Philology. 98. 114-42.

Go, Kenji. 2002. "'I Am That I Am' in Shakespeare's Sonnet 121 and 1 Corinthians 1510:." Notes and Queries. 247. 241-42.

Greg, W. W. 1942. The Editorial Problem in Shakespeare: A Survey of the Foundations of the Text. Oxford. Clarendon.

Haley, David. 2002. "'The Cause of This Defect': The Dram of Eale." 'A Certain Text': Close Readings and Textual Studies in Shakespeare and Others. Edited by Linda Anderson and Janis Lull. Newark NJ. University of Delaware Press. 29-49.

Hart, F. Elizabeth. 2002. "The 'Missing' Scene in Act 2 of Pericles." English Language Notes. 40.2. 4-12.

Hirsh, James. 2002. "Act Divisions in the Shakespeare First Folio." Papers of the Bibliographical Society of America. 96. 219-56.

Hosley, Richard. 1960. "Was There a Music-room in Shakespeare's Globe?" Shakespeare Survey. 13. 113-23.

Hunter, William B. 2002a. "Heminge and Condell as Editors of the Shakespeare First Folio." American Notes and Queries. 15.4. 11-19.

Hunter, William B. 2002b. "New Readings of A Midsummer Night's Dream." American Notes and Queries. 15.4. 3-10.

Ichikawa, Mariko. 2002. Shakespearean Entrances. Basingstoke. Palgrave.

Jackson, MacD P. 2002a. "Dating Shakespeare's Sonnets: Some Old Evidence Revisited." Notes and Queries. 247. 237-41.

Jackson, MacDonald P. 2002b. "Pause Patterns in Shakespeare's Verse: Canon and Chronology." Literary and Linguistic Computing. 17. 36-47.

Kastan, David Scott. 1999. Shakespeare After Theory. New York. Routledge.

Kirschbaum, Leo. 1938. "A Census of Bad Quartos." Review of English Studies. 14. 20-43.

Knowles, Richard. 2002. "How Shakespeare Knew King Leir." Shakespeare Survey. 55. 12-35.

Levin, Richard. 2002. "The Longleat Manuscript and Titus Andronicus." Shakespeare Quarterly. 53. 323-40.

Long, William B. 1985. "Stage-directions: A Misinterpreted Factor in Determining Textual Provenance." Text: Transactions of the Society for Textual Scholarship. 2. 121-37.

Lull, Janis. 2002. "Thirteen Ways of Looking at a Blooper: Some Notes on the Endless Editing of Richard III." 'A Certain Text': Close Readings and Textual Studies in Shakespeare and Others. Edited by Linda Anderson and Janis Lull. Newark NJ. University of Delaware Press. 50-64.

Martin, Randall. 2002a. "Catholic Ephesians in Henry IV, Part Two." Notes and Queries. 247. 225-26.

Martin, Randall. 2002b. "The True Tragedy of Richard Duke of York and 3 Henry VI: Report and Revision." Review of English Studies. 53. 8-30.

Massai, Sonia. 2002. "'Taking Just Care of the Impression': Editorial Intervention in Shakespeare's Fourth Folio, 1685." Shakespeare Survey. 55. 257-70.

McKerrow, R. B. 2000. "The Relationship of English Printed Books to Authors' Manuscripts During the Sixteenth and Seventeenth Centuries (The 1928 Sandars Lectures)." Ed. Carlo M. Bajetta. Studies in Bibliography. 53. 1-65.

McMillin, Scott and Sally-Beth MacLean. 1998. The Queen's Men and Their Plays. Cambridge. Cambridge University Press.

Merriam, Thomas. 2002. "Faustian Joan." Notes and Queries. 247. 218-20.

Monsarrat, G. D. 2002. "A Funeral Elegy: Ford, W. S., and Shakespeare." Review of English Studies. 53. 186-203.

Poole, William. 2002. "Julius Caesar and Caesar's Revenge again." Notes and Queries. 247. 227-28.

Robinson, John V. 2002. "Hamlet's Evil Ale: Hamlet 1.4.36-38." Hamlet Studies. 24. 10-25.

Shaheen, Naseeb. 2002. "Biblical References in Julius Caesar." Notes and Queries. 247. 226-27.

Shakespeare, William. 2002a. King Henry IV Part 1. Ed. David Scott Kastan. The Arden Shakespeare. London. Thomson Learning.

Shakespeare, William. 2002b. King Henry VI Part 2. Ed. Roger Warren. The Oxford Shakespeare. Oxford. Oxford University Press.

Shakespeare, William. 2002c. King Richard II. Ed. Charles R. Forker. The Arden Shakespeare. London. Thomson Learning.

Shakespeare, William. 2002d. The Comedy of Errors. Ed. Charles Whitworth. The Oxford Shakespeare. Oxford. Oxford University Press.

Shakespeare, William. 2002e. The Complete Sonnets and Poems. Ed. Colin Burrow. The Oxford Shakespeare. Oxford. Oxford University Press.

Sisson, C. J. 1956. New Readings in Shakespeare. Vol. 1: Introduction, The Comedies, The Poems. Cambridge. Cambridge University Press.

Spevack, Marvin. 2002a. "A Victorian at Work: Halliwell's Folio Edition of Shakespeare." In the Company of Shakespeare: Essays on English Renaissance Literature in Honor of G. Blakemore Evans. Edited by Thomas Moisan and Douglas Bruster. Madison NJ. Fairleigh Dickinson University Press. 292-313.

Spevack, Marvin. 2002b. "Shakespearecomputer.horizons@." Shakespeare Newsletter. 52. 61, 82-84,86.

Spevack, Marvin. 2002c. "What Price Shakespeare?: James Orchard Halliwell-Phillipps and the Shilling Shakespeares of the 1860s." Papers of the Bibliographical Society of America. 96. 23-47.

Stern, Tiffany. 2002. "The 'Part' for Greene's Orlando Furioso: A Source for the 'Mock Trial' in Shakespeare's Lear?" Notes and Queries. 247. 229-31.

Streete, Adrian. 2002. "Charity and Law in Love's Labour's Lost: A Calvinist Analogue." Notes and Queries. 247. 224-25.

Tobin, J. J. M. 2002. "More Evidence for a 1594 Titus." Notes and Queries. 247. 222-24.

Tronch-Pérez, Jesus. 2002. "A Comparison of the Suspect Texts of Lope de Vega's La Dama Boba and Shakespeare's Hamlet." Shakespeare Yearbook. 13: Shakespeare and Spain. 30-57.

Vickers, Brian. 2002a. 'Counterfeiting' Shakespeare: Evidence, Authorship, and John Ford's Funerall Elegye. Cambridge. Cambridge University Press.

Vickers, Brian. 2002b. Shakespeare, Co-author: A Historical Study of Five Collaborative Plays. Oxford. Oxford University Press.

Warren, Roger. 2000. "The Quarto and Folio Texts of 2 Henry VI: A Reconsideration." Review of English Studies. 51. 193-207.

Wells, Stanley. 1999. "Letter to the Editor: 'The Longleat Drawing'." 23 April. Times Literary Supplement. 5012. 17.