"Leaden Contemplation": Ambiguous Evidence of Revision in Q1 Love's Labour's Lost Kenneth B. Steele University of Toronto Shakespeare Association of America 1991 Annual Meeting (Vancouver) Seminar 1: "Shakespeare's Quartos: Text, Performance, Memory." [This text file has been prepared from a Microsoft Word for Windows file, and has regrettably been stripped of all formatting codes, including italic, superscript, and boldface. The footnotes and bibliography have been appended, but the Appendices have not. These footnotes, indicated with square brackets, often form an important part of the text, and should be read with it. This paper is exploratory rather than conclusive, offering suggestions, pointing out theoretical difficulties, and raising questions for which I genuinely hope others can supply answers. Because the paper arises from my thesis research, comments and suggestions are particularly welcome. ] [1.] Intratextual Revision Sparked by the compelling example of Q1/F1 King Lear, scholarly attention has recently been focused on the evidence for authorial revision in Shakespeare's multi-text plays.[2] The current ascendency of performance criticism - the now widespread recognition that Shakespeare's own priority was most certainly the stage and not the study - has facilitated the acceptance of these revision studies, many of which examine the theatrical implications of alternate texts in terms of modified staging, pacing, or dramatic effect. The anti-theatrical bias of much traditional scholarship, and the editorial fallacy of a definitive text,[3] have both been struck a mortal blow by this partnership, and the change is as healthy as the results have been enlightening. The class of "bad" quartos is being re-examined, and Shakespeareans effectively have twice as much territory to explore as ever before. The study of Shakespeare as an actor and playwright should not, however, completely displace the study of Shakespeare as a poet, particularly in his early work: the multi- text plays are not the only texts which Shakespeare may have revised. Quite literally less dramatic than the intertextual evidence of the multi-text plays, is the subtler intratextual evidence of authorial process found primarily in the "foul paper" texts. Here, "fossilized" evidence of revision abounds,[4] including repetitions, redundancies, ambiguities, contradictions, inconsistent speech prefixes, and even what appear to be alternate drafts of entire speeches. These clues suggest currente calamo corrections, marginal and interlinear additions, and second thoughts during or shortly after composition - with authority considerably less controversial than the intertextual variants, which cannot be so readily placed chronologically, and which cannot be attributed to Shakespeare with much certainty at all. The implications of intratextual revisions are generally more poetic than dramatic: often the variants have few ramifications for performance or the overall interpretation of the play, but they offer tantalizing glimpses of the very process of composition, demonstrating Shakespeare's self-critical faculties in action and indeed bringing us "as close as we can ever come to Shakespeare at work."[5] Many of these intratextual fossils are exceedingly brief, and the alternative readings they offer sometimes seem indifferent - but E.A.J. Honigmann's warning is worth noting: calling variants "indifferent" is using "a word which can as aptly describe the beholder as the thing observed."[6] Authorial second thoughts often highlight first attempts which were somehow unsatisfactory, and through a process of critical triangulation, we may be able to determine the direction of change, and extrapolate Shakespeare's ultimate poetic aims. It must be recognized, however, that both intertextual and intratextual revision are purely hypothetical constructs based upon observable phenomena and patterns; no completely irrefutable proof of Shakespearean revision can ever be established without authorial manuscripts or testimony, neither of which is likely to be forthcoming. The surviving textual evidence provides an incomplete set of clues which are often tenuous, usually ambiguous, and occasionally self- contradictory. Even when intertextual variants appear deliberate, rather than accidental, the author is only one of many possible intervening agents, including scribes, compositors, censors, book-keepers, players, and unidentified playwrights hired to "mend" plays for revival.[7] Intratextual evidence is considerably less controversial, because currente calamo corrections, false starts, and second shots are presumably authorial, and presumably products of the initial act of literary creation. The conventional explanation for the reduplications in Love's Labour's Lost, and elsewhere, is that Shakespeare marked his first drafts for deletion with some form of theatrical bracket and proceeded immediately to his second draft, the foul papers were used as copy in the printing house, and the deletion was somehow overlooked or misunderstood, resulting in the printing of both drafts. This theory is one possible explanation for these reduplications, but the evidence is ambiguous: repetition with variation can be a deliberate authorial strategy (as it most certainly is throughout Love's Labour's Lost); alternate versions of speeches or scenes may be rough papers towards an effect which would be finalized only in performance (as perhaps, for example, the duplicate dawn speeches spoken by Romeo at the end of 2.2, and Friar Lawrence at the opening of 2.3 in Romeo & Juliet Q2 - D4v); and compositors or proofreaders could conceivably have made such a mess of a passage that two "drafts" of essentially the same material could stand in type where only one appeared in the manuscript copy. There can also be no certainty as to which of two consecutive "drafts" actually came first, nor as to which Shakespeare ultimately may have preferred. All of these alternative theories and uncertainties must be kept in balance, through a sort of Keatsian "negative capability," in any exploration of either variety of Shakespearean revision. Lyrical Foul Papers and Shakespeare "the Poet" It is hardly surprising that fossil revisions are found most prominently in those Shakespearean texts deemed to be printed from "foul paper" copy, because such false starts, currente calamo corrections, and consecutive drafts have always been taken as strong evidence of an underlying authorial manuscript, usually in some form preceding fair copy. This logic is dangerously circular, however, because these textual clues are ambiguous until the nature of the printer's copy has been determined: in "prompt copy" texts, or the much-maligned "bad" quartos, similar textual phenomena are often dismissed as playhouse interpolation, textual corruption, memorial error, or crude approximation.[8] This paper sets about to explore the ambiguity of the evidence in the least-contested Shakespearean "foul paper" text of all: Q1 Love's Labour's Lost (1598). The objective is not to challenge the "foul paper" designation, but to demonstrate the ambiguities in even the best-established evidence for revision, and perhaps to raise some questions worthy of consideration. Although the presence of fossil revisions in "foul paper" texts is unsurprising, it is striking that the majority of the Shakespearean examples are concentrated in four texts with very similar printing histories: the "good" quartos of Titus Andronicus (1594), Love's Labour's Lost (1598), Romeo & Juliet (1599), and A Midsummer Night's Dream (1600).[9] Although printed over six years, these two comedies and two tragedies were all composed within three years of each other - between 1592 and 1595, during and after London's worst plague, and probably at much the same time that Shakespeare wrote most of his non-dramatic poetry: Venus & Adonis (1592-3), The Rape of Lucrece (1593-4), and the earliest Sonnets (1593-1603).[10] Intricate thematic and poetic links connect Titus Andronicus with the Rape of Lucrece, the Sonnets with Love's Labour's Lost, and Romeo & Juliet with A Midsummer Night's Dream. The four plays are all inherently lyrical, distinctly rhetorical, and self-consciously "poetic," making greater use of verse and rhyme than most of the canon, and even presenting that most undramatic of forms, the sonnet, onstage.[11] Apparently, during this formative period Shakespeare's artistic focus tended toward dramatic poetry, rather than poetic drama: the surviving textual evidence seems to suggest that in revision the early Shakespeare tinkered with the prosody and rhetoric of his plays, whereas he later engaged in wholesale "theatrical" revision of the major tragedies, such as Hamlet, Othello, King Lear, and Troilus & Cressida. "Polynomials" in Love's Labour's Lost [12] The title page of Q1 Love's Labour's Lost seems to announce its value as a record of Shakespearean revision when it declares itself "Newly corrected and augmented | By W. Shakespere."[13] This quarto is perhaps the least-disputed example of a Shakespearean text printed from authorial holograph, although it remains manifestly uncertain whether the text preserves second thoughts from the original process of composition (as E.K. Chambers and W.W. Greg argue), or revision for later revival or occasional performance (as John Dover Wilson and Richard David assert).[14] The text preserves a veritable smorgasbord of textual treasures, from false starts to "ghost" characters to factual confusions and contradictions.[15] Speech prefixes are almost always inconsistent, often ambiguous, and occasionally missing altogether.[16] Furthermore, the play is saturated with what Randall McLeod has somewhat whimsically labelled "polynomials": characters with more than one designation in speech prefixes and stage directions.[17] In fact, only a handful of characters are consistently identified in Q1: the three lords of Navarre, Boyet, and three minor characters (Marcade of 5.2, the "Lord" of 2.1, and the Forester of 4.1). The speech prefixes for all other characters vacillate between a number of alternatives, and while this is not necessarily incompatible with prompt copy,[18] it does suggest underlying holograph copy, and seems to reveal a great deal about Shakespeare's composition of the play - although often the evidence is decidedly ambiguous.[19] For example, Dull, Costard, Armado, Nathaniel, and Holofernes are more often than not identified by their functions in Q1 speech prefixes: Constable, Clown, Braggart, Curate, and Pedant.[20] In the first scene, Dull and Costard are initially identified by function, but when they introduce themselves in dialogue, their speech prefixes briefly change to "Antho." and "Cost." With only four exceptions (3.1.143, 3.1.145, 4.2.142, and 4.3.196), Costard is thereafter always "Clowne," a total of 78 times. (Dull tends to be either "Const." or "Dull." - which may be in itself a functional designation.) In their first scene (4.2), Nathaniel and Holofernes are identified almost entirely by name, although with some tendency to forgo "Holo." for "Ped."; in contrast, both 5.1 and 5.3 label the Curate, Pedant, Braggart, Page, and Clown exclusively by function. Perhaps the speech prefixes in 4.2 tend toward proper names because the characters are being introduced for the first time, but clearly also because of the dialogue itself - Holofernes and Nathaniel volley addresses back and forth like a tennis match: Curat.Nath. Truely M. Holofernes, the epythithes are sweetly varried like a scholler at the least... (4.2.8-9; Q1 D4r) Holo. Sir Nathaniel, haud credo. (4.2.11; Q1 D4v) Holo. Sir Nathaniel, will you heare an extemporall Epytaph on the death of the Deare... (4.2.49-50; Q1 E1r) Nath. Perge, good M. Holofernes perge, so it shall please you to abrogate squirilitie. (4.2.52-3; Q1 E1r) Ped. Sir Holofernes, this Berowne is one of the Votaries with the King ... (4.2.134-5; Q1 E2r) Ped. ... But to return to the Verses, Did they please you sir Nathaniel? (4.2.147-8; Q1 E2r) This intensity of name-dropping is not repeated; Nathaniel is never named in dialogue again, and Holofernes is named only once more, at 5.1.107.[21] The evidence suggests (always assuming that the Q1 speech prefixes reflect Shakespeare's manuscript) that he vastly preferred function names for a surprising number of characters, including Holofernes, Nathaniel, Armado, Moth, Dull, and even Costard - with the exception of those lines in which they identify themselves or each other by name.[22] Elsewhere, however, variations in Armado's speech prefixes seem to indicate an underlying irregularity in the copy. In 3.1, the "Braggart" and his "Boy" are so identified, consistently, until 3.1.66, at which they promptly become "Armado" and his "Page" for the remaining 112 lines for which they are on stage (with the solitary exception of "Boy." at 3.1.102). Significantly, this discontinuity in speech prefixes is matched by a discontinuity in content: in the first section of the scene, Armado and Moth are engaging in a lengthy duologue on the subject of love; their prefixes change the moment Moth returns with Costard - in what is essentially a new and independent episode. The simultaneity of the shift in content and in labels significantly increases the likelihood that the origin of the variation was Shakespeare himself, rather than a scribe or compositor. It seems at least possible that the separate and essentially detachable episode with Costard was written separately, either as an afterthought or an earlier thought for another (or undecided) context. The decisive change in speech prefixes suggests that, at the very least, some time elapsed between the composition of the first and second sections of the scene: perhaps only a short break in sequential writing, but perhaps something more significant. Although "Ferdinand K. of Nauar" enters in the play's initial direction (1.1.0sd; Q1 A2r), he is identified as simply "Ferd." or "Fer." throughout the scene. In 2.1, Boyet and the Princess have already referred to him as "Nauar" three times before he enters as "Nauar" (at 2.1.7, 22, and 81), and Boyet's final line once the Duke is onstage is "Heere comes Nauar" (2.1.89; Q1 C1r). It would seem that the foreigners' natural and repeated use of his geographical title influences his speech prefixes throughout the scene: the Duke is consistently "Nau." in speech prefixes - except for lines 2.1.127-175, where he is "Fer." once again. These 48 lines interrupt the flirting stichomythia of Berowne and Katherine[23] to discuss politics: the matter of Aquitaine and a hundred thousand crowns. Nothing about the Duke's formal and confrontational manner could possibly suggest that the more intimate prefix, "Fer.," reflects his function here; more likely, it may indicate that these lines were written separately (either earlier or later - perhaps chronologically closer to the composition of 1.1), and were then inserted into 2.1. Notably, the Duke is suddenly "Nau." again in his exit line, which peaceably ignores the disagreement of the preceding lines: "Thy owne wish wish I thee in euery place" (2.1.177; Q1 C2v). And then, just as suddenly, Berowne is engaging Rosaline in a battle of wit. Did Shakespeare originally write a scene in which Berowne went directly from Katherine to Rosaline, and then insert the material on Aquitaine? Or was one bout of stichomythia part of the insertion, perhaps intended to replace the other? Berowne's dialogue with Katherine ends with "Nay then will I be gon" (2.1.126; Q1 C1v), just as he leaves Rosaline with "I cannot stay thankes-giuing" (2.1.191; Q1 C2v): both seem equally good exit lines. In view of the other revisions to Berowne's part, discussed in detail below, this possibility merits some consideration. Just as the Duke's first entrance direction is distinctly different from his speech prefixes throughout 1.1 (perhaps because of some later annotation, authorial or otherwise), so the Princess of France is proleptically identified as "Queene" when she first speaks to Boyet in 2.1 (2.1.13), but is consistently "Princess" throughout the rest of the scene, and indeed for the first half of the play.[24] At 4.1.1, however, the "Princess" suddenly becomes "Queen," and she retains this title, without exception, for the remainder of the play. Simultaneously, the newly-minted Queen transforms the Duke into a King, coyly enough by questioning his identity: she asks, "Was that the king that spurd his horse so hard[?]" (4.1.1; Q1 D2r; emphasis mine). Prior to this line, the Duke has been many things, but he has never been "King." From this point on, for the last two acts of the play, both the Duke and the Princess are identified as King and Queen in all of their speech prefixes, and many times in the dialogue (in fact, the word "Duke" does not occur at all after 2.1, although the Princess is still occasionally called "Princess" as late as 5.2). The consistent change in the speech prefixes of the Duke and the Princess, accompanied as it is by somewhat less consistent changes in the dialogue, argues against mere compositorial or scribal error or meddling, and suggests an authorial cause. If the death of the King of France, announced by Marcade in 5.2, was not firm in Shakespeare's mind until the composition of the last two acts, the inevitability of the Princess' inheritance may then have been reflected in her speech prefixes, and by association the Duke may have become her match, in both romance and rank. The fact, however, that the Princess is named ambivalently in later dialogue, while the "King" is not, would suggest the reverse: that the Duke's transformation was not secondary, but primary, and that the Princess/Queen was renamed in his wake. Perhaps the "King/Queen" material more directly reflects some lost source, or an earlier draft which Shakespeare "Newly corrected and augmented." Certainly, by Act 4 the political details of 2.1 have been long forgotten; perhaps Shakespeare was no longer paying close attention to the titles of his primary characters. The discontinuity in these character designations might also indicate a discontinuity in Shakespeare's manuscript, and possibly in the process of composition. Perhaps in combination with additional research and a wider variety of evidence, it will be possible to draw firmer conclusions. The ladies of France are variously named, numbered, titled, and unidentified throughout the text, in ways which may reflect the inner workings of Shakespeare's mind in composition, and shed light on the process of composition and revision in the play. When they first appear, in 2.1, the ladies of France are identified simply as "1. Lady," "2. Lad.," "3. Lad.," or even simply "Lad." (when Lady 1 is speaking in continuing dialogue). Shakespeare has neither named nor distinguished their characters, as they respond to the Princess' query about the lords of Navarre. Suddenly, however, Berowne enters into private discussion with one of the ladies, whose speech prefixes become "Kather.," although neither he nor the audience obtains her name until 2.1.208.[25] Likewise, a second lady is identified as "Ros." the moment Berowne begins sparring with her at 2.1.179, although Dumaine and the audience are told her name only at 2.1.193. It seems reasonable to theorize that the ladies of France, particularly Rosaline and Katherine, became individuated for Shakespeare in this scene (although modern editors cannot distinguish them here), and that the speech prefixes reflect this changed attitude.[26] Strangely, however, once the Duke and lords leave the stage and the ladies begin to engage Boyet in witty discourse, their designations shift back to their earlier ambiguity. At first, the shift is subtle (or perhaps it has been corrected by partial annotation): "Lady Maria" and "Lady Ka." speak. When Katherine continues her dialogue with Boyet, however, her prefixes revert to the barely sufficient "Lad.," and suddenly "Lad. 2" and "Lad. 3" return. Clearly, Shakespeare used the ambiguous "Lad." when he felt that an ongoing dialogue made further distinctions unnecessary, but the return to numbered ladies seems quite extraordinary. Coming as it does at the moment that Berowne leaves the stage, it suggests that Shakespeare composed lines 2.1.213-56 separately from the central passage, perhaps simultaneously with the earlier lines in which the ladies are also numbered rather than named. Throughout the remainder of the play, the ladies of France are always identified by name in their speech prefixes.[27] Although Jaquenetta, as Costard observes in 1.1, can be labelled in a great many ways, Q1 uses only two. Costard speaks of her by name three times in the very first scene (at 1.1.199 and again, twice, at 1.1.299); the only other occurrence of her name in the first scene is in Armado's letter, read by the Duke, and here the very act of naming her is a self-conscious one: "For Iaquenetta (so is the weaker vessel called) ..." (1.1.261; Q1 B1r). When Jaquenetta actually appears onstage, however, in 1.2, she is consistently designated "Maid," as if she had not been introduced in the first scene (or as if the first scene had not yet been written). Furthermore, the only use of her name in the scene is in the problematic line, "Clo. Come, Iaquenetta, away" (1.2.138; Q1 B3v) - again, spoken by Costard.[28] With the single exception of 4.2.140, where her speech prefix lapses to "Mayd," Jaquenetta is identified by name everywhere except 1.2. This would seem to suggest that Shakespeare first created her character in 1.2, uncertain of her name until Costard's line, and then either wrote 1.1 later, or later inserted the dialogue involving her name, and perhaps all the dialogue involving Costard (as mentioned above, the dialogue involving Costard in 3.1 may well have been written separately).[29] If Jaquenetta's on-stage lines in 1.2 were indeed written earlier than 1.1, it might also provide a tempting argument that Shakespeare intended Armado to meet Jaquenetta before his lengthy lovesickness in the earlier half of 1.2. Unfortunately, although these explanations may have some validity, there is another which rather thoroughly undermines the likelihood that Jaquenetta's speech prefixes can disclose much about Shakespeare's sequence of composition: just as the Duke enters as "Nauar" in 2.1, apparently in response to the preceding dialogue, the lines in 1.2 which identify Jaquenetta as "Maid" seem to follow as a consequence of Armado's address to her: "I do betray my selfe with blushing: Maide," to which the "Maide" naturally enough replies, "Man" (1.2.126-7; Q1 B3r), and remains "Maide" for the rest of the scene. If the variant speech prefixes in Q1 Love's Labour's Lost do indeed reflect underlying holograph copy, they occasionally reveal (through a glass darkly) hints of the authorial process underlying the text, suggestions of the chronology of composition, and clues to otherwise unsuspected duplications (like the two wit matches in 2.1). More commonly, however, it would seem that the catalyst for Shakespeare's "polynomials" is his own dialogue - an observation which should prove considerably unsettling to those who wish to interpret the vague prefixes of the "bad" quartos as evidence of a reporter's ignorance. It would seem that the author himself, just as a "reporter," can appear unaware of a character's name until it arises in dialogue. The Fossil Revisions in Love's Labour's Lost As controversial as evidence of intertextual revision is, and as tenuous as the implications of the "polynomials" discussed above may be, modern editors and textual critics almost unanimously agree that two passages in Q1/F1 Love's Labour's Lost represent indisputable intratextual evidence of authorial revision. It would appear that only one "draft" of each of Berowne's two duplicated speeches - "O we haue made a Vow to studie, Lordes" (4.3.293-340; Q1 F2r-F3v; see Appendix A), and his penance, imposed by Rosaline, to "iest a tweluemonth in an Hospitall" (5.2.804-53; Q1 K1r-v; see Appendix B) - was intended to stand in the play, and therefore most editions either relegate the "first" drafts to an appendix, the textual apparatus, or editorial brackets. Although the redundancy of these passages, and the repetition of words, phrases, and lines between "drafts," seems strong evidence, for the moment it is wise to maintain a healthy skepticism of the "repetition bracket" theory. I believe that the following examination of the apparent alterations between "drafts" does indeed support the theory of authorial revision in these passages, but I will close this paper with an exploration of the ambiguity of such evidence in Love's Labour's Lost. Passage A: "O we haue made a Vow to studie, Lordes" "Passage A" is perhaps the most solid example of twin drafts in the quarto. Lines 8-30 are apparently the first attempt at Berowne's speech, which Shakespeare seems to have immediately rewritten in lines 31-78.[30] As the parallel text in Appendix A visually illustrates, eight lines are repeated or substantially echoed between the two drafts, which also have a great many individual words and ideas in common. Molly Mahood suggests that, by the end of the "first" draft, "the rhythm flags, the diction begins to creep, the logical distinctions of the speech grow blurred," and so Shakespeare slept on it (71). Clearly the final lines of the first draft were not composed without effort; J.C. Maxwell points out that line 29, "With ourselves," is probably the compositor's attempt to make sense of an authorial interlineation, in combination with two unnoticed deletions in line 30.[31] Maxwell convincingly suggests reading: with our selues Do we not ^ (likewise) see (our) learning there? These final lines of the first draft are not merely flagging or blurred; they also begin toying with the very features which are expanded in the second draft, and epitomize several of the elements in the first draft which Shakespeare seems to have abandoned in the rewritten version. This intersection of the two drafts in lines 26-30, more than any other evidence, asserts the sequence of the drafts as printed, and corroborates the argument that we are, in fact, dealing with a revised passage. Lines 8-16, the opening nine lines of Shakespeare's "discarded" draft, evidently remained largely satisfactory to him, although in the "second" draft they are divided (and in the case of lines 12-13, they are scattered among other lines). Minor repetitions and echoes of the original draft also appear throughout the second draft: the "fierie Numbers" of line 35 may have been suggested by the "true Promethean fire" of line 16, which evidently persisted in Shakespeare's mind until he restored it as line 64. "Teaches such beautie" (line 25) may have suggested "beautis tutors" (36); the "motion and long during action" (19) may have suggested the "motion of all elamentes" (42); and the "Authour in the worlde" (24) may be echoed in "all the worlde" (66), and "the authour of these Women" (72). Both drafts begin with delight in the paradox that the vow to study is a vow to avoid women's eyes, the true "Bookes." Significantly, the first draft uses second-person pronouns consistently in lines 3-13 and 21-25: "Consider what you first did sweare vnto," "Can you fast?," "And where that you haue vowd to studie," "In that each of you haue forsworne his Booke," "Can you still dreame and poare and thereon looke," "when would you my Lord, or you, or you ...." [32]. It cannot be accidental that, at line 26, the first draft seems to realize the mistake and begins the first-person plural in earnest, before the draft itself is abandoned: Learning is but an adiunct to our selfe, And where we are, our Learning likewise is. Then when our selues we see in Ladies eyes, With our selues. Do we not likewise see our learning there? It would seem that, even before Shakespeare began the second draft, he had firmly decided that Berowne should not insulate himself from those he criticizes, but implicate himself as well. Lines 31-2 assert that "we haue made a Vow to studie, Lordes, | And in that Vow we haue forsworne our Bookes." The rhetorical question initiates a series of second-person pronouns in lines 33-6, and intriguingly they resurface in lines 68-9 (perhaps supporting Staring Wells' argument about these lines), but the first person returns in lines 74-5, which sound an unmistakable note of unity: Lets vs once loose our othes to finde our selues, Or els we loose our selues, to keepe our othes. The rhetorical questions themselves are a technique which fills the first draft; Berowne asks a total of five questions in the first version (lines 6, 10, 13, 25, 30), including the troublesome final lines which evidently brought the first draft to an end. In the second version, however, Berowne asks only three questions (at lines 33 or 36, 53, and 73), the first of which is carried over from the first draft, and the second of which hardly seems to qualify as a question ("For valoure, is not Loue a Hercules"). The shift of pronouns which implicates Berowne along with his companions also seems to do away with the first draft's strategy of repeated rhetorical questions, ultimately leaving only one new question in the second draft, in the final lines. This is the question which prompts the Duke's energetic response, no longer a rhetorical question but one which elicits precisely the answer Berowne seeks. The first draft emphasizes the importance of the woman's "face" (13, 21), but the second remains firmly fixed in the "eyes" (35, 40, 63), the windows of the soul and a fertile poetic image, being both organs and objects of sight, both transparent and reflective. The first draft makes use of these qualities in the final lines (25-30), but for the second draft Shakespeare abandons most of the metaphysical image, retaining only a hint in "shew, containe, and nourish all the worlde" (66). It is particularly intriguing that the four references to "eyes" in the first draft (lines 14, 22, 25, and 28) seem to spring from the line, "From womens eyes this doctrine I deriue," an aurally-complex and resonant line which is in fact precisely retained in the second draft. The second draft demonstrates the reverse pattern, in which the five references to "eyes" (lines 35, 40, 46, 47, and 63) gradually build up to the repeated line. If, for a moment, we take the "first" draft as the later draft, and the "second" as the earlier, the pattern would suggest the following: Shakespeare began the lengthier passage, in which the eyes feature as one of the five senses elaborated in lines 45-52, and finally hit upon their use in the line which both drafts share. When he turned to rewrite the passage, he began with the repeated line, and then followed it up with several references before the climactic metaphysical image in the final lines. If, however, the printed sequence is also the chronological one (as other evidence would suggest), Shakespeare clearly decided to save the key line until the end of Berowne's speech, making it a logical development from the sensory imagery that precedes it. The first draft names women "the Ground, the Bookes, the Achadems," but the second modifies this to "the Bookes, the Artes, the Achademes." "Ground" seems connected to line 12 of the first draft, while "Artes" seems to have been influenced by line 37 of the second. The second version of these lines also substitutes the more concrete "They sparcle still the right promethean fier" for the earlier abstract "From whence doth spring the true Promethean fire"; the sparkling eyes are a much more visual image, and make the connection between fire and eyes more coherent. The reversal of lines in the second draft makes possible the sublime conclusion to the thought in line 66, "That shew, contain, and nourish all the worlde." The first draft of this passage seems to depend heavily upon triplings and triple patterns, whether in syntax, alliteration, or prosody. Syntactic triplings include "To fast, to study, and to see no woman," "still dreame and poare and thereon looke," "you my Lord, or you, or you," and "the Ground, the Bookes, the Achadems." In contrast, the second draft depends upon quadruple patterns and pairings, particularly in structural terms. Four of the five senses are described in lines 46-52, the first three getting two lines each, and the last one two balanced half-lines. Several rhetorical schemes in the second draft balance four words or elements. The four lines of rhetorical gradatio (70-73) develop from "Wisedome" to "Loue" to "Men" to "Women," a quadruple pattern which is emphasized by the compositor's use of capitalization. Notably, this four-line scheme is neatly framed by two fourfold repetitions: the chiasmus of lines 68-9 and of lines 74-5. The triple pattern is not wholly excised in the second draft; in fact, it is evident in precisely those lines which have been reused intact from the first draft (the syntactic tripling of lines 33 and 65, and the rhythmic triad of "From womens eyes this doctrine I deriue" at line 63), and in the line which expands upon them, "That shew, containe, and nourish all the worlde." The triple patterns of the first draft are insistent, obvious, and emphatic; in contrast, the second draft is subtler, more fluent, and more convincing, drawing in additional examples and bringing additional rhetorical guns to bear. The first draft proceeds to discuss, in biological terms, the poisonous effects of excessive study. In the second draft, "vniuersall plodding" has become "leaden contemplation," perhaps a deliberate pun on the leaden type used to produce books (as it is in my title). The second draft, however, moves beyond the negative effects of the "Other slow Artes" to discuss the positive consequences of "Loue" for the mind, senses, and virtues of the lover: "Loue ... with the motion of all elamentes, / Courses as swift as thought in euery power." It seems highly significant that the first draft refers to "study" four times (at lines 4, 8, 12, and 23) , but never to "love," whereas the second draft refers only once to "studie," in its first line (31), but eleven times to "love" or "louers" (at lines 40, 47, 48, 50, 52, 53, 57, 60, 70, 71, and 78). In rewriting the passage, Shakespeare's emphasis seems to have shifted from learning itself, through female beauty ("For where is any Authour in the worlde, | Teaches such beautie as a woma[n]s eye"), to the power and divinity of Love itself ("when Loue speakes, the voyce of all the Goddes, | Make heauen drowsie with the harmonie"). Shakespeare seems to have found mythological allusion irresistible in the second draft, adding somewhat gratuitous allusions to Venus, Bacchus, Hercules, the Hesperides, the Sphinx, and Apollo, all within a mere five lines (52-6). Shakespeare's "fatal Cleopatra," the pun, also sneaks into the second draft, with a quibble on "braine" and "barraine" at lines 37-8 (a pun I also believe to be found at Twelfth Night 1.5.85, and Troilus & Cressida 1.3.327). In the second shot, Shakespeare seems to have consciously added a self-reflexiveness to his poetry, describing poetry as the highest form of virtue to be gained from Love (59- 62), before returning again to the doctrine Berowne derives from women's eyes. Such poetic self-reflexiveness also arises in the revised passage at 5.1.1-22 in A Midsummer Night's Dream, suggesting not only that Shakespeare was conscious of his role as a poet in these early plays, but that the very act of revision focused his attention still more on that role. Aural effects, obvious in the first draft, become still more pervasive in the revision. The sole end-rhyme proper in the first draft, "Booke | ... looke" (9-10), disappears entirely in the second (where there are no end-rhymes). Half-rhymes in the first draft, such as "found ... ground" (12), are surpassed in the second by such as "eare ... heare" (48), and the three-way "Hercules | trees ... Hesperides" (53-4). Alliteration appears in the first draft, such as "vniuersall plodding poysons vp" (17), "Learning likewise" (27), and "likewise ... learning" (30), but to nowhere near the extent it can be found in the second: "haruest ... heauie" (39), "Loue ... learned ... Ladies" (40), "suspitious . . . stopt" (49), "soft and sensible" (50), "Subtit ... Sphinx ... sweete" (55), "heauen ... harmonie" (58), "Poet ... pen" (59), "all ... ought" (67), "were ... women ... forsweare" (68), and "Wisedomes ... word" (70). Consonance appears likewise, such as "Appolos Lute" (56), or "rauish sauage" (61), as well as many other aural effects too complex to catalogue here. While the first draft ends in an image of metaphysical complexity (the lovers and their learning held in stasis within the beloveds' eyes), the second ends in a crescendo of rhetorical virtuosity. Chiasmus ("fooles ... forsweare | ... sworne ... fooles" - 68-9) moves through paradox and antonym ("fooles ... Wisedome" - 69-70) into gradatio and anadiplosis which climaxes in repetition ("loue ... | Loues ... men ... | Mens ... Women ... | Womens ... Men ... Men" - 70-3), and ends once more in chiasmus and paradox ("othes ... selues ... | selues ... othes" - 74-5). Finally the revised passage ends with scriptural manipulation (Romans 13:8) and wordplay. Passage B: "And what to me my Loue? and what to me?" "Passage B," the second major passage of fossilized revision in Q1 Love's Labour's Lost, occurs at 5.2.805-53 (Q1 K1r-v). Lines 1-6 are a considerably abbreviated version of lines 22-56, but their redundancy cannot be ignored; it seems quite clear that both drafts cannot have been intended to stand in the final text. James Cunningham argues that Shakespeare began the passage with the Princess' "sentencing" of the Duke out of respect for his social position, and then followed his own interest as it turned to Berowne and Rosaline. He then realized that, "by all the laws of interest these two must come last," and proceeded to the other lovers first (Cunningham 107). It is also possible, in view of the steady increase in Berowne's importance apparent through the other passages of revision in the text, that Shakespeare rewrote this passage at a late stage of composition, heightening Berowne's centrality to the play beyond that of even the Duke. The mere six lines of the first draft of this passage (fewer than spent on the sentencing of Dumaine!) could never be considered sufficient for the character that Berowne has become once these revisions have occurred. Shakespeare seems to have repeated Berowne's first line with minor alteration and assigned it to Dumaine (line 7), but it seems to me that there is considerable confusion in the Katherine/Dumaine exchange as it stands in the text: Duma. But what to me my Loue? but what to me? Kath. A wife? a beard, faire health, and honestie, With three folde loue I wish you all these three. Duma. O shall I say, I thanke you gentle Wife? Kath. Not so my Lord, a tweluemonth and a day, Ile marke no wordes that smothfast wooers say, Come when the King doth to my Lady come: Then if I haue much loue, Ile giue you some. Duma. Ile serue thee true and faythfully till then. Kath. Yet sweare not, least ye be forsworne agen. Editors tend to assign Katherine's first words at line 8 to Dumaine, but they suggest that Dumaine's second line (10) actually belongs before her first (line 8): Duma. O shall I say, I thanke you gentle Wife? Kath. A wife? a beard, faire health, and honestie, With three folde loue I wish you all these three. The variation on Berowne's line may simply result from compositorial eyeskip, although the confusion in lines 11-16 is hard to resolve. The first, brief draft of the Berowne/Rosaline passage is unadorned with her lengthy monologues, and does not make explicit the appropriateness of the punishment to Berowne's crime - tending to the sick seems a perfectly traditional Christian work of mercy. It is intriguing, however, that the original version seems to contain most of the essentials which are expanded in the second draft.[33] Berowne's eager and rather self-centred questions ("And what to me my Loue? and what to me?") are replaced in the revised draft by a lengthier, self-consciously poetic speech, in which Berowne invokes cliches like "the window of my heart" (23), and the Petrarchan idiom ("humble suite," "seruice" - 24-5). Responding to this effusion of hollow poetry, Rosaline replies with even harsher criticism in the second draft. The religious implications of the original version ("purged," "sins" - 2) are postponed until her final word ("reformation" - 54), and "faults" (3) resurfaces only in lines 51 and 53.[34] Two lines of criticism expand into six lines of caustic commentary on Berowne's reputation. Images of "weed[s]" and "Wormewood" (32) are applied to Berowne's humour, and the personification of "the worlds large tongue" is introduced at line 27. The teasing irony of Rosaline's original third line, "Therefore if you my fauor meane to get," persists in the revised line, "And therewithall to win me, if you please ..." (33). Rosaline's final lines in the first version (5-6) are expanded into 35-9, and in some sense permeate lines 43-54 also. An impulse to alliteration seems to structure much of the revision: "A tweluemonth shall you spend, and neuer rest" (5) becomes "You shall this tweluemonth terme from day to day, | Visite the speechlesse sicke" (35-6). The first version is remarkably clear of aural effects (with the exception of the potential pun, "seeke ... sicke," at line 6), but alliteration or consonance abounds in virtually every line of the revised passage: "attends ... answer" (24), "Impose some seruice" (25), "haue ... heard" (26), "man ... mockes" (28), "Full ... floutes" (29), "estetes ... execute" (30), "within ... wi[t]" (31), "weed ... Wormewood" (32), "therewithall ... win" (33), "Without ... which ... won" (34), "tweluemonth ... terme ... day ... day" (35), "speechlesse sicke ... still" (36), "With ... wit" (38), "pained impotent" (39), "Mirth ... moue" (42), "Why ... way" (43), "begot ... grace" (44), "him ... heares" (47), "Deaft ... deare" (49), "will ... withall" (51), "finde ... fault" (53), and "Right ... reformation" (54). Perhaps the most intricate sound effects are created by the line "A tweluemonth? Well: befall what will befall" (55), which also contributes to the sole rhyming couplet (55-6).[35] The preceding analysis of the two major passages of fossil revision in Q1 Love's Labour's Lost, while acknowledging occasional ambiguities in the evidence and attempting to consider a variety of explanations for the evidence, essentially supports the fossil revision theory. The rational, consistent, and deliberate alterations between "drafts" distinctly suggest Shakespearean revision, rather than compositorial error or deliberate repetition. In both cases, if the printed sequence is also the chronological order, Shakespeare expanded his first effort, rather than condensing it. The second drafts indicate a heightening of rhetorical and aural special effects, subtle shifts in content and emphasis, and perhaps most significantly, an expansion of Berowne's significance in the play.[35a] In Passage A, Berowne turns from criticism of his companions to justification of himself, from an emphasis on learning through beauty to an emphasis on the divine supremacy of love. In Passage B, Rosaline's brief description of a standard Christian penance, or act of mercy, is expanded to become distinctly appropriate to Berowne, whose rhetorical flamboyance and poetic artificiality returns in the second draft. In Love's Labour's Lost, Shakespeare's revisions seem geared to increase poetic artifice and self-consciousness; the extravagant poetry of the play, like that of Berowne himself, was quite deliberate and achieved with some effort. Fortunately, the first quarto has preserved invaluable evidence of these poetic labours almost lost. Revision or Repetition? Q1 Love's Labour's Lost, then, contains some of the least controversial evidence for Shakespearean revision. It also, however, presents some of the toughest challenges to that evidence. Most critics and editors agree regarding Passages "A" and "B," but more controversial is, for example, the Princess's "Holde Rosaline, this Favour thou shalt weare" (5.2.130-4; Q1 G3r), the supposed first draft of which the Oxford Complete Works consigns to the appendix, while George Hibbard's Oxford edition and the Riverside Shakespeare retain both as a single draft, on the grounds that the repetition is not redundancy but clarification: Holde Rosaline, this Fauour thou shalt weare, And then the King will court thee for his Deare: Holde take thou this my sweete, and giue mee thine, So shall Berowne take me for Rosaline. And change you Fauours two, so shall your Loues Woo contrarie, deceyued by these remoues. (5.2.130-3; Q1 G3r) Here, it is primarily the repetition of "Hold" which suggests two drafts; the content of the lines is essentially complementary rather than redundant. If lines 3 and 4 of this passage were meant to replace lines 1 and 2, though, it would suggest an intriguing shift in emphasis, from the Princess' delight at the thought of the Duke's confusion, to an eager anticipation of being courted by Berowne. Such an increase in the stature and centrality of the figure of Berowne would also be consistent with the two major fossils in the text. More unsettling than inevitable editorial disagreement over this ambiguous fossil, however, is the play's heavy reliance on repetition, redundancy, interruption, and "false starts" as part of a deliberate authorial strategy. In Berowne's first speech of the first scene, variations on the line "the which I hope is not enrolled there" occur three times (Q1 A2v, 1.1.38, 41, 46). Duke Ferdinand is interrupted by Costard when he attempts to read Armado's letter, and the result is a clearly deliberate false start, a stuttering of "So it is ... So it is ..." (1.1.220, 227; Q1 A4v-B1r). A self-interruption, which also looks distinctly like a compositorial omission or false start, belongs to Nathaniel at 4.2.100: "Vnder pardon sir, What are the contentes? or rather as Horace sayes in his, What my soule verses" (Q1 E1v). Of course, Holofernes and Armado both depend heavily on the use of rhetorical copia and synonyms. Armado's letter to Ferdinand epitomizes his diction, which is repetitive and redundant by nature, using triplings and quadruplings of epithets and synonyms: the ebon coloured Incke, which here thou viewest, beholdest, suruayest, or seest ... that low spirited Swaine, that base Minow of thy myrth, (Clowne. Mee?) that vnlettered smal knowing soule, (Clow. Mee?) that shallow vassall (Clown. Still mee.) which as I remember, hight Costard, (Clow. O mee). (1.1.237-49; Q1 B1r) [36] In the second act, Katherine and Berowne repeat the same question so precisely that the type seems exactly duplicated:[37] Berowne. Did not I dance with you in Brabant once? Kather. Did not I dance with you in Brabant once? (Q1 C1v; 2.1.113-4) The effect is much like that of compositorial eyeskip, or perhaps an authorial reassignment of a speech, but the surrounding dialogue makes it clear that this duplication is deliberate and necessary to the sense of the passage. In 5.2, Boyet plays middleman in a deliberate strategy of repetition which finally tries even his own patience: Boyet. What would you with the Princes? Berow. Nothing but peace, and gentle visitation. Rosa. What would they, say they? Boy. Nothing but peace, and gentle visitation. Rosa. Why that they haue, and bid them so be gon. Boy. She saies you haue it, and you may be gon. King. Say to her we haue measurd many miles, To treade a Measure with her on this grasse. Boy.They say that they haue measurd many a mile, To tread a Measure with you on this grasse. Rosa. It is not so. Aske them how manie inches Is in one mile? If they haue measured manie, The measure then of one is easlie tolde. Boy. If to come hither, you haue measurde miles, And manie miles: the Princesse bids you tell, How manie inches doth fill vp one mile? Berow. Tell her we measure them by weerie steps. Boy. She heares her selfe. (5.2:179-195; Q1 G4r) It would seem that, throughout Love's Labour's Lost, Shakespeare is challenging us to define unnecessary or unintentional repetition, deliberately revelling in rhetorical tropes of copia and repetition. Faced with the ambiguity of such passages, we can turn to additional evidence to reinforce our conclusions about the underlying copy. Often fossil revisions can be identified with more certainty when variant texts are collated. Although the independent authority of the F1 text of Love's Labour's Lost is questionable at best, it does make a number of cuts in passages which do indeed seem repetitive. Armado's "Fox, Ape, and Humble-Bee" jig, for example, appears with a triply-repeated refrain in Q1: A. No Page, it is an epilogue or discourse to make plaine, Some obscure presedence that hath tofore bin saine. I will example it. The Fox, the Ape, and the Humble-Bee, Were still at oddes being but three. Ther's the morrall : Now the lenuoy. Pag. I will adde the lenuoy, say the morrall againe. Ar. The Foxe, the Ape, and the Humble-Bee, Were still at oddes, being but three. Pag. Vntill the Goose came out of doore, And staied the oddes by adding foure Now will I begin your morrall, and do you follow with my lenuoy. The Foxe, the Ape, and the Humble-Bee, Were still at oddes, being but three. Arm. Vntill the Goose came out of doore, Staying the oddes by adding foure. (Q1 C4v, 3.1.78-94, emphasis mine) As with so many reduplications, there is little in the text to indicate redundancy with any certainty, but F1 seems to have attempted to "correct" this passage by deleting nine lines of Armado's dialogue: (plaine, Ar. No Page, it is an epilogue or discourse to make Some obscure precedence that hath tofore bin saine. Now will I begin your morrall, and do you follow with my lenuoy. The Foxe, the Ape, and the Humble-Bee, Were still at oddes, being but three. Arm. Vntill the Goose came out of doore, Staying the oddes by adding foure. (F1 L4v, TLN 855-68, 3.1.78-94) Although the F1 alternative certainly eliminates the repetition of the Q1 version of this passage, the confusion of speech prefixes in F1 certainly suggests some kind of unusual compositorial eyeskip rather than deliberate editing. Ultimately, then, this is unlikely to represent an unintentional duplication, or fossilized revision. Likewise, F1 omits Holofernes' admiration for Mantuan at 4.2.98, which may be redundant (it does express much the same thought as the Italian tag immediately preceding it in Q1): Nath. Facile precor gellida, quando pecas omnia sub vmbra ruminat, and so foorth. Ah good olde Mantuan, I may speake of thee as the traueiler doth of Venice, vemchie, vencha, que non te vnde, que non te perreche. Olde Mantuan, olde Mantuan, Who vnderstandeth thee not, loues thee not, vt re sol la mi fa: Vnder pardon sir, What are the contentes? or rather as Horrace sayes in his, What my soule verses. (4.2.91-100; Q1 E1v) But in Nathaniel's ruminations, rather as in the mad scenes of King Lear, how can we really distinguish between deliberate and accidental inconsistency, repetition, redundancy, or incoherence? Summary Less controversial than intertextual evidence for Shakespearean revision in the major tragedies is intratextual evidence in his early plays, particularly the quartos apparently based on authorial holograph. In particular, four "foul paper" plays (Titus Andronicus, Love's Labour's Lost, Romeo & Juliet, and A Midsummer Night's Dream) contain the majority of "fossilized" revision: inconsistencies, repetitions, and redundancies. This intratextual revision is more poetic and less theatrical than the later, intertextual variety, although equally hypothetical and ambiguous: consecutive "drafts" may actually represent deliberate repetition, compositorial or editorial error, or alternative options which Shakespeare left to be decided in rehearsal. Furthermore, much of the evidence in the "good" quartos looks dangerously like that in the "bad" ones: an examination of variant speech prefixes, or "polynomials," in Q1 Love's Labour's Lost suggests that Shakespeare, like a supposed "reporter," could often appear unaware of character names unless they were prominent in dialogue. Foul-paper "drafts" and confusions also resemble some confusions in early "bad" quartos like Q1 2 Henry VI. The "polynomials" in Q1 Love's Labour's Lost also seem to suggest some discontinuities in the underlying copy, and may help identify passages which were written out of their published sequence, or at some chronological remove from their contexts - particularly when they occur simultaneously with consistent changes in the content of the dialogue. Close readings of the two best-known textual fossils, "O we haue made a Vow to studie Lordes" (LLL Q1 F2v-3v; 4.3.293-340) and "And what to me my Loue?" (LLL Q1 K1r-v; 5.2.805-53), reveal deliberate and consistent revisions in rhetoric, content, and poetic effects. Shakespeare, it would seem, has deliberately heightened the poetic artifice and self-consciousness of his lines, increasing Berowne's centrality to the play and further developing his character. Yet this solid evidence for authorial revision appears in a text in which inconsistency, repetition and redundancy are part of Shakespeare's comic strategy, and hence in which apparent "revisions" are rendered unusually ambiguous. Notes 1. This paper has evolved from ongoing research for my thesis, "The Second Heat Upon the Muse's Anvil": Poetic Revision in Shakespeare's Early Plays. I am pleased to acknowledge the support of the Social Sciences and Humanities Research Council of Canada for my course of study. 2. The growing critical acceptance of revision theory has been in large measure due to the impact of two books: Steven Urkowitz's Shakespeare's Revision of King Lear (1980), and the collection edited by Gary Taylor and Michael Warren, The Division of the Kingdoms: Shakespeare's Two Versions of King Lear (1983). The arguments posed in these works were firmly entrenched by the Oxford Complete Works (1986), which prints edited texts of both Q1 and F1 King Lear. 3. Random Cloud insists that we must look "away from the editor's ideal single version - the so-called 'definitive text' - to the author's actual multiple versions: an infinitive text" (111). The Oxford Shakespeare's two texts of King Lear, or perhaps better, Michael Warren's parallel text facsimile, are steps in this direction, which may ultimately lead to an electronic hypertext edition, which could offer continual choice between facsimile, edition, and editorial apparatus. 4. The term "fossil revision" seems to have been coined by Fredson Bowers (cited by Honigmann, 22). 5. Mahood, 84. Molly Mahood explores process in Shakespeare's art, including a variety of alternate versions, such as ghost characters, conflicting plot details, and other evidence of Shakespeare's "inspired carelessness" (70). 6. Honigmann, 167. 7. The term is that used repeatedly by Philip Henslowe in his business records. (See Foakes & Rickert, eds., Henslowe's Diary.) 8. I am thinking in particular of Q1 The First Part of the Contention betwixt the two famous houses of York and Lancaster (the "bad" quarto of 2 Henry VI), which contains a number of redundant passages which look more than a little like those in Q1 Love's Labour's Lost; and of Q1 Hamlet, in which speech prefixes show some of the same variations I observe below in a "foul paper" text. 9. Although it is not possible to explore the specific passages in all of these texts, a good indication of their locations is provided in the Oxford Shakespeare Textual Companion. In A Midsummer Night's Dream, the "fossils" themselves are hypothetical, based on mislineation which suggests marginal revisions. Of course intratextual revision also occurs in conjunction with intertextual revision in the later tragedies, but my focus here remains on the early plays. 10. The chronology used here is that presented in Gary Taylor's "The Canon and Chronology of Shakespeare's Plays," in the Oxford Shakespeare Textual Companion. 11. George Hibbard observes that "no fewer than 43.1 per cent of the total lines in Love's Labour's Lost and 45.5 per cent of those in A Midsummer Night's Dream are rhymed. No other of Shakespeare's comedies comes anywhere near to approaching them in this respect. Their closest rival is The Comedy of Errors, with 21.5 per cent... . Romeo and Juliet, with 16.6 per cent, [has] more than any other of the tragedies" (Introduction, 43-4). 12. Quotations from Love's Labour's Lost will be taken from the Q1 facsimile in the collection edited by Michael J.B. Allen and Kenneth Muir. Obsolete typographical features must regrettably be normalized here. Lineation is keyed to George Hibbard's new Oxford edition, and is supplemented by references to Q1 signatures, where appropriate. Quotations from other works of Shakespeare are referenced to the lineation of the Riverside Shakespeare. 13. Q1 A1r. This introduces a new level of ambiguity to the evidence, of course; if the revised passages are literally "new" in 1598, they would have to represent revision after performance, perhaps for revival - what would be essentially intertextual revision rather than currente calamo corrections in the original foul papers. George Price argues that "the omission of a period after 'augmented' should be attributed to a mere oversight by the publisher or compositor rather than to an implication that Shakespeare was only the reviser, not the author" (406), but the presence of such a period would support Guy Lambrecht's argument that a non-Shakespearean reviser was involved. Price's investigation of casting-off errors in Q1 contradicts E.K. Chamber's suggestion that Q1 was set from an annotated earlier (and non-extant) Q0. 14. Price, 409. 15. Consider Jaquenetta's blatant error (and self-contradiction) at 4.2.128: "I sir from one mounsieur Berowne, one of the strange Queenes Lordes" (Q1 E2r). 16. For example, no speech prefix is supplied for "The vvordes or Mercurie, are harsh after the | songes of Apollo" (5.2.911; Q1 K2v), and "The partie is gone" reads more like a line of dialogue than a stage direction (5.2.661 ;Q1 I3r). The prefix "B." is ambiguous before "Ver begin" (5.2.874 ;Q1 K2r) - Armado's speech is immediately before, identified as "Brag.," so perhaps Berowne is intended. Likewise, "Lady" (in its various states of abbreviation) variously refers to Rosaline, Katherine, Maria, and perhaps even the Princess of France. And although care is usually taken to identify Moth as "Page." rather than "Boy." whenever Boyet shares the stage ("Boy." refers 43 times to Armado's page and 28 times to Boyet), at 5.2.701 (Q1 I3v) it is significantly unclear who speaks "True, and it was inioyned him in Rome for want of Linnen: since when, Ile be sworne he wore none, but a dish-cloute of Jaquenettaes, and that a weares next his hart for a Fauour": either Boyet continues to mock Armado, or Moth's speech prefix has suddenly changed and he is reporting actual past events. (I am inclined to the former, although most editions choose the latter, interpretation.) 17. McLeod, 49. 18. In "The Psychopathology of Everyday Art," Random Cloud has put forth a convincing argument that prompt copy is often inconsistent (using the examples of George Bernard Shaw, Shakespeare, and John Barton), and that editors "wrectify" texts in obliterating these variations: "however we try to explain it, we must not explain it away" (142). In "Stage Directions: A Misinterpreted Factor in Determining Textual Provenance," William Long observes that "regularity and consistency of theatrical marking is most emphatically what these manuscripts [Elizabethan promptbooks] do not demonstrate" (134). 19. Another level of ambiguity, which I have not been able to incorporate in this paper, stems from Price's argument that the abbreviation of speech prefixes vary with compositors in this text (Price, 418). I have not yet been able to consider the evidence in relation to the compositor stints he suggests. 20. It is conceivable that Shakespeare might have thought of his characters in generic terms before assigning them names, or that the names were not so significant as the functions to him. It is also possible that a playhouse scribe, annotating or transcribing the play, might smooth out inconsistencies by assigning functional labels to these characters. It is even possible that a reporter might describe characters by function out of ignorance. What is not likely, however, is that a compositor would impose generic speech prefixes if proper names stood in the copy text; if the compositor was indeed setting from "foul papers," these speech prefixes are very likely authorial. For the purposes of this survey, variant abbreviated forms (e.g. "Ferd.," "Fer.," and "Ferdinand") will not be distinguished. 21. Although strangely, Q1 here has "Ped." speaking to "Sir Holofernes," as at 4.2.134 above. Hibbard's edition emends the dialogue in each case to "Sir Nathaniel," but this solution is not completely satisfying. In any event, this is the only occurrence in dialogue of one of their names after scene 4.2. 22. It is particularly surprising, however, that Costard is consistently "Clo." throughout the lines in 3.1 at which Moth puns on "a Costard broken in a shin" (3.1.67; Q1 C4v). It is only 76 lines later, once Armado and Moth have left the stage and Costard is speaking to Berowne, that his speech prefix suddenly becomes "Cost." once more (3.1.143; Q1 D1v). 23. Although editors unanimously agree that Berowne must be flirting with Rosaline here, and therefore alter Katherine's speech prefixes, I see no reason why the readings of Q1 cannot be allowed to stand. Indeed, it seems to me that the humour is heightened if Berowne unsuccessfully attempts to court both ladies in this scene - and this would make thematic sense, in connection with the masquing confusion of 5.2. Q1 and F1 present a similar situation in the masquing scene, in which Maria seems to attract the attention of both Dumaine and Longaville (5.2.242-55; Q1 H1r-v). Grace Ioppolo suggests too that these passages could stand in the text: ...it does not seem unusual that Berowne, chafing at his vow to avoid women, would speak to Katherine, particularly since he apparently met her in the company of Rosaline at the home of the Duke of Alencon. (191) 24. The editors of F1 may have attempted to correct the Princess' speech prefix at 2.1.13 with a marginal note, which the compositor clearly misunderstood: F1 retains the "Queen" designation, but inserts a "Prin." speech prefix at 2.1.20, in the middle of the speech. 25. In both Q1 and F1, Berowne asks "Whats her name in the capp?," to which Boyet replies "Katherin by good happ" (2.1.208, Q1 C2v). Modern editions emend this line to "Rosaline by good hap" - erroneously, I would argue. 26. Of course, if one accepts the argument that promptbook annotators would seek to correct ambiguous speech prefixes, it is also possible that such a non-authorial annotation is responsible for the naming (or mis-naming) of the ladies in these lines. (But William Long refutes that argument.) 27. Only two abnormalities occur: at 5.2.53 and 5.2.57, Maria is identified as "Marg." (probably a compositorial misreading of "Mary" or "Maria," or less likely an authorial slip for the name of the least prominent lady), and at 5.2.552 an unidentified "Lady." cries out "Great thankes great Pompey." It seems highly unlikely that Shakespeare would revert to this designation so late in the play, after demonstrating almost perfect consistency in the use of proper names for Rosaline, Katherine, and Maria, but it seems still more unlikely that he would use this term for the Princess of France, as modern editors generally claim - nowhere else in the entire text does "Lady" ever refer to the Princess (or Queen, as she is identified in this scene). 28. Q1 and F1 use the speech prefix "Clo.," which refers only to Costard throughout the play. The modern editorial consensus, which reassigns this speech to Dull, is based on the misguided assumption that Dull should speak this line immediately before his exit with Jaquenetta. This overlooks considerable humour in Shakespeare's original staging: it is highly appropriate that Costard should try to exit at this point, with Jaquenetta on his arm, because only such an action would provoke Armado's next line (to stop him short): "Villaine, thou shalt fast for thy offenses ere thou be pardoned" (1.2.140; Q1 B3v). 29. It would be intriguing to attempt to demonstrate that Costard was a late authorial afterthought, added to the Armado/Jaquenetta subplot as a rustic foil, perhaps in the tradition of the Petrarchan pastourelle, in which a courtly man competes with a peasant boy in an attempt to seduce a peasant girl (detailed by Leonard Forster in The Icy Fire, 87). 30. To facilitate concentrated references to these revised passages, which do not appear complete or sequentially in most editions, I will use the simplified through-numbers in Appendices A and B. Staring Wells suggests that lines 68-78 are the completion of the first draft of this passage, and that lines 31-67 represent a marginally-inserted second draft. This explanation may help justify the awkwardness of line 67, but rejects the currente calamo nature of the revision. Guy Lambrechts argues that a mediocre non-Shakespearean revisor attempted to replace the longer passage with the shorter one, which was printed earlier; I dispute that argument. 31. Price's bibliographical analysis suggests that, in casting-off, allowance was made for two and a half lines to follow the "mysterious half line" (line 29), but I find his speculation less convincing than the simplicity of Maxwell's argument (Price 414). 32. Although the second-person pronouns of line 11 are repeated verbatim in the second draft, at line 33, they take on a very different quality there, in the context of first-person plural pronouns, and seem no longer so much accusations as rhetorical questions. (In quotations from the Appendices, as here, emphasis will be added without specific notation.) Grace Ioppolo observes the first alteration of pronouns in this passage (184), but I arrived at this argument independently, in a 1988 graduate paper on which this paper is in part based. 33. It may be that the first version represents a marginal memorandum at the point of breaking off composition. Honigmann detects such memoranda in a number of plays, particularly Timon of Athens, which was set from particularly "foul" papers (148). He theorizes that they might serve as reminders for the next few speeches, jotted down before the author breaks off for the evening or any break in composition. 34. Rosaline's use of the word "periurie" may carry overtones of Shakespeare's use in Romeo and Juliet at 2.2.92-3: "at Louers periuries / They say Ioue laught." 35. It may well be significant that both drafts of the interchange between Rosaline and Berowne are in unrhymed verse, but are found in a fully rhymed context. The Katherine / Dumaine and Maria / Longaville exchanges, found between the first and second drafts of this passage, are fully rhymed, as are the Princess' final lines and the Duke's speech immediately before the passages, and the dialogue immediately following (with the exception of line 5.2.854, also spoken by the Princess). If the move to blank verse was not a deliberate strategy to heighten the gravity of Berowne's offense, it may reflect a discontinuity in composition. 35a. Ioppolo is apparently the first scholar to note the confluence of revision on the character of Berowne -- the implications of which have yet to be fully investigated. (See Ioppolo, 184, 187). 36. All the modern editions consulted (Hibbard's Oxford, the Oxford Complete Works, David's Arden, and Evan's Riverside) obscure the nature of these interjections by removing them from parentheses and treating them as distinct speeches. Costard's interruptions in Q1 and F1 are visually more intrusive, and preserve the textual suggestion that they may have been afterthoughts, like Hamlet's marginal interjections "That's wormwood" and "If she should breake it now" in 3.2 (Hamlet Q2 H2r-v). 37. Again, I argue that the Q1/F1 speech prefixes should not be emended. Bibliography Allen, Michael J.B., and Kenneth Muir. Shakespeare's Plays in Quarto: A Facsimile Edition of Copies Primarily from the Henry E. Huntington Library. Los Angeles: University of California Press, 1981. Barton, Anne. "Love's Labour's Lost." Essays in Shakespearean Criticism. Ed. James L. Calderwood, and Harold E. Toliver. Englewood Cliffs, NJ: Prentice-Hall, 1970. 204-221. Calderwood, James L. "Love's Labor's Lost: A Wantoning with Words." Studies in English Literature 5 (1965): 317-32. Cloud, Random [Randall McLeod]. "The Psychopathology of Everyday Art." The Elizabethan Theatre IX: Papers Given at the Ninth International Conference on Elizabethan Theatre held at the University of Waterloo, Ontario, in July 1981. Ed. G.R. Hibbard. Port Credit, Ont.: P.D. Meany, 1986. Cunningham, J[ames] V[incent]. "'With that Facility': False Starts and Revision in Love's Labour's Lost." Essays on Shakespeare. Ed. Gerald Chapman. Princeton: Princeton University Press, 1965. 91- 115. David, Richard W., ed. Love's Labour's Lost. Arden Shakespeare. London: Methuen, 1951. Draudt, Manfred. "Printer's Copy for the Quarto of Love's Labour's Lost (1598)." The Library 6.3 (1981): 119-31. --------. "The 'Rosaline-Katherine Tangle' in Love's Labour's Lost." The Library 6.4 (1982): 381-96. Evans, G. Blakemore. The Riverside Shakespeare. 6th ed. Boston: Houghton Mifflin, 1974. Foakes, R.A., and R.T. Rickert, eds. Henslowe's Diary. Cambridge: Cambridge University Press, 1961. Forster, Leonard. The Icy Fire: Five Studies in European Petrarchanism. Cambridge: Cambridge University Press, 1969. Hibbard, George R[ichard], ed. Love's Labour's Lost. The Oxford Shakespeare. Ed. Stanley Wells. Oxford: Oxford University Press, 1990. Honigmann, E[rnst] A[nselm] J[oachim]. The Stability of Shakespeare's Text. London: Edward Arnold, 1965. Ioppolo, Grace Janette. Textual Revision in the Plays of William Shakespeare. Ph.D. Diss., UCLA, 1989. Ann Arbor: UMI, 1990. 9008101. Jackson, MacD. P. "Compositors B, C, and D, and the First Folio Text of Love's Labour's Lost." Papers of the Bibliographical Society of America 72.1 (1978): 61-5. Kerrigan, John. "Love's Labour's Lost and Shakespearean Revision." Shakespeare Quarterly 33 (1982): 337-339. --------. "Shakespeare at Work: The Katherine-Rosaline Tangle in Love's Labour's Lost." Review of English Studies 33 (1982): 129-136. Lambrechts, Guy. "'The Brief and the Tedious of it': Note sur le Texte de Love's Labour's Lost." Etudes Anglaises 17 (1964): 269-283. Long, William B. "Stage-Directions: A Misinterpreted Factor in Determining Textual Provenance." TEXT: Transactions of the Society for Textual Scholarship 2 (1985): 121-37. Mahood, Molly M. "Unblotted Lines: Shakespeare at Work." Interpretations of Shakespeare. Ed. Kenneth Muir. Oxford: Clarendon Press, 1985. 69- 84. Maxwell, J.C. "'Love's Labour's Lost': IV.iii.313-4." Notes & Queries 13 (1966): 128. McLeod, Randall. "UN-Editing Shak-speare." Sub-Stance 33/34 (1982): 26-55. Price, George R. "The Printing of Love's Labour's Lost (1598)." Papers of the Bibliographical Society of America 72.4 (1978): 405-34. Shakespeare, William. Love's Labour's Lost. A New Variorum Edition of Shakespeare. Ed. Henry Howard Furness. Philadelphia: J.B. Lippincott, 1904. Urkowitz, Steven. Shakespeare's Revision of King Lear. Princeton, NJ: Princeton University Press, 1980. Taylor, Gary, & Michael Warren, eds. The Division of the Kingdoms: Shakespeare's Two Versions of King Lear. Oxford: Clarendon Press, 1986. Wells, Stanley. "The Copy for the Folio Text of Love's Labour's Lost." Review of English Studies, New Series 33 (1982): 137-47. Wells, Stanley, Gary Taylor, John Jowett, and William Montgomery. William Shakespeare: A Textual Companion (The Oxford Shakespeare). Oxford: Clarendon Press, 1987. Wells, Staring B. "Love's Labour's Lost IV, iii, 285-362." Modern Language Notes 55 (1940): 102-3. Werstine, Paul. "Variants in the First Quarto of Love's Labour's Lost." Shakespeare Studies 12 (1979): 35-47. [This is a test of the new file put procedure.]