Vocabulary, Chronology, and the First Quarto (1603) of Hamlet.
"All four editors agree that Q1 offers a bad text, concocted by actors who relied on their memories of the play as they had performed it." (1) I wrote those words--in "Editing Hamlet in the 1980s: Textual Theories and Textual Practices"--nearly thirty years ago, and the four editors were Harold Jenkins (Arden), Philip Edwards (Cambridge), G. R. Hibbard (Oxford), and Gary Taylor (Oxford Complete Works). (2) I included the consensus about the origin and nature of Q1 (the Quarto of 1603) among "well-established beliefs" that were not "likely to be overturned." (3) I was wrong. Theories that invoke "memorial reconstruction" to account for the features of Hamlet Q1 and other dramatic texts once labeled "bad" have fallen into disfavor. (4) Once, nearly all Shakespeare scholars believed in them. Now hardly any do. But although the beliefs have been overturned, it is less than perfectly clear that the evidence formerly sustaining them has been successfully countered. At least, the possibility that the printer's copy for Q1 bore a text that had been partly or largely transmitted by memory cannot yet safely be dismissed.
Theories about the manuscript source of Q1 have been of two basic kinds: (a) that it is ultimately derivative from the Quarto of 1604/5, widely believed to have been set from Shakespeare's autograph, with the lightly cut and altered First Folio version of Hamlet--thought to represent a later stage in the theatrical evolution of the play--as probable intermediary; and (b) that it predated the longer play preserved in Q2 and F. Adherents to theory "a" invoke various combinations of (i) theatrical abridgement of the canonical Hamlet, with or without Shakespeare's approval or acquiescence, and (ii) reporting, either by one or more actors who reconstructed a text from their imperfect memories of the play in performance, or by one or more shorthand or longhand note takers who attended performances. Subdivisions of theory "b" are that Q1 preserves (i) Shakespeare's first draft, perhaps partially revised by him or (ii) an earlier staged version of Hamlet, wholly or partly by Shakespeare. Variations on these conflicting hypotheses have combined features from two or more of them. (5) But the crucial issue is whether the manuscript behind Q1 antedated or post-dated the manuscript behind Q2.
The view that the Ur-Hamlet first alluded to by Thomas Nashe in 1589 was by Shakespeare and that it is preserved in Q1 has recently been argued with flair and conviction by Terri Bourus in Young Shakespeare's Young Hamlet. (6) Hers is a far more considered and thorough presentation of the case than that by Eric Sams two decades before, and several other scholars--Steven Urkowitz prominent among them--have published articles in support of the belief that Q1 affords a Shakespearean draft or version of Hamlet earlier than Q2's. (7) Bourus and Urkowitz both draw on extensive theatrical experience. They are alert to "the systematic codes of theatrical script-writing as they appear in Shakespeare's plays" and the variant early editions of them. (8)
Bourus brilliantly contests 1980s orthodoxy. She mounts a spirited argument against the relevance to Q1 of piratical publishers, actors, or theatergoing reporters. She offers a fascinating discussion of the theatrical and political consequences of an inferred difference in age between the Q1 and Q2 Hamlets, and of the implications for the protagonist's relationships with the other characters. And she provides the fullest consideration yet of the various pre-1600 allusions to a Hamlet play and the catch-cry "Hamlet revenge!" while also showing how an early Shakespearean Hamlet and the canonical Q2/F version would fit into Shakespeare's life and play writing career. She sees F as presenting Shakespeare's first seventeenth-century revision of his late-1580s play, preserved in Q1, and Q2 as incorporating further revision. Gary Taylor and Rory Loughnane, in the chapter on "Canon and Chronology" in the New Oxford Shakespeare Authorship Companion declare themselves convinced by her argument. (9) But there are features of Q1 that, as this article aims to show, are insufficiently accounted for by her theory as it stands.
Scholars who classified Q1 as a "bad quarto," derivative from the fuller Q2/F play tended to concentrate on the bad passages and advance reasons for their poor quality. (10) But here my primary focus is on the best bits, because what any theory of the relationship between Q1 and Q2/F also has to explain is why, when the texts run closely parallel, Q1 is stylistically so good and so exemplary of Shakespeare's mature style. In making this claim, I am disagreeing with Bourus, who expresses more than once the view that passages in which Q1 and Q2/F are most nearly identical "were there from the [pre-1589] beginning and did not change." (11) She believes that "the kind of line that remains invariant is... something functional and ordinary, rather than something extraordinary or memorable." (12) She cites some of the least colorful "invariant" lines. But substantiation of my own claim will, I think, place obstacles to accepting that Q1 preserves, in any straightforward way, a Shakespearean Ur-Hamlet of the late 1580s. Some of the evidence I shall adduce suggests that Q1 is unlikely to have been based on an early draft or "first sketch" of any kind.
Take scene 1, lines 42-131, in the modernized text of Q1 of Arden-3, beginning "How now, Horatio, you tremble and look pale." This all seems to me, as it has seemed to numerous scholars since a copy of Q1 was first discovered in 1823, to be the verse of the mature Shakespeare at his best--vigorous, vivid, and exact, with (in the longer speeches) sentences of some complexity deployed across lines of which a high proportion are run-on. Even though the few, mainly isolated, differences from Q2, do make Q1, 1.42-131, inferior to the Q2 counterpart, it contains three doublets of the kind that George T. Wright, in a much admired article on Shakespeare's use of the rhetorical figure known as hendiadys, showed were especially characteristic of Shakespeare during the Hamlet period: "the sensible and true avouch / Of my own eyes" (1.46-47); "the thought [grosse Q2] and scope of my opinion" (1.57); "ratified by law and heraldry" (1.75); Wright added a supplementary list of "phrases that, if not hendiadys, are close, or odd," among them "this same strict and most observant watch" (1.60), "Of inapproved [unimproved Q2] mettle hot and full" (1.80), "For food and diet to some enterprise / That hath a stomach in't" (1.83-84). (13) Shakespeare's first ten plays afforded Wright only fourteen examples of hendiadys between them, with none having more than three, but the incidence rising to a peak of sixty-six in Hamlet. (14) Allied to this stylistic mannerism is Shakespeare's fondness during "the Hamlet period and after" for joining Latin and English synonyms, as in "It is an eager and a nipping air" in Q1, 4.2 (Q2, 1.4.2). (15)
Typical of Shakespeare in his prime are, in fact, the whole of Marcellus's query and Horatio's answer about preparations for war (1.59-85); Horatio's speech on the "cock, that is the trumpet to the morning [morn Q2]" (1.104-12); Marcellus's speech on "that season... / Wherein our Saviour's birth is celebrated" (1.113-120); Horatio's speech noting that "the sun [morn Q2] in russet mantle clad / Walks o'er the dew of yon high mountain top" [eastward hill Q2] (1.121-29)--this all reads like the work of Shakespeare writing around 1600 at his full poetic strength. (16) This is, of course, a subjective critical judgment. But it can be corroborated by objective evidence.
Throughout Q1 are, as has long been recognized, scattered passages that are textually almost identical with their counterparts in Q2 and read like mature Shakespearean verse. Voltemand's (Voltemar's in Q1) speech at Q1 7.31-51 is, as commentators have noted, almost word-for-word the same as in Q2 2.2.60-80. Here too the sentences are quite complex and half of the lines are run-on (if we ignore the last as necessarily being end-stopped); and the pauses within lines are variably placed. Q1 also closely matches Q2 at: 2.110-25; 2.155-64 + 166-69; 4.15-32; 5.8b-18; 5.41-53; 7.340-48. These cover all other verse passages from scene 2 onwards in which Q1 and Q2 are in fairly close agreement over at least nine consecutive lines. I have included 2.166-69, which complete Hamlet's speech and end the scene. I have ignored the extrametrical exclamation "O, I have it!" at the end of 7.341.
Adding these passages to 7.31-51, I calculate that 47 of the 97 lines that do not end speeches, or 48.5 per cent, are run-on in the Arden-3 edition. (17) This is high even for Shakespeare in the early seventeenth century, but the Arden-3 editors may be more sparing of punctuation than their predecessors on whose editions existing metrical tables are based. A count of the 97 Q1 lines based on Irace's edition of Q1 yields a percentage of 39. The standard tables remain those collected by E. K. Chambers in William Shakespeare, where the percentage of blank verse line with "overflows" (or that are "run-on") in Hamlet as whole is given as 23. (18) But this figure includes lines that end speeches and so must be end-stopped. Frederick J. Pohl, more reasonably, calculated percentages that excluded such lines and obtained 27.2 for Hamlet. (19) Only in plays from Antony and Cleopatra (47.7 per cent) onwards did Pohl find Shakespeare reaching percentages comparable to those for the chosen 97 Q1 lines, as derived from Arden-3 (48.5), and Macbeth (38.3) was the first play to approach the 39 percent derived from Irace's edition of Q1. The New Oxford Shakespeare dates Antony and Cleopatra 1607, Macbeth 1606. Pohl gave no counts for Arden of Faversham or Edward III, but for the other six plays that begin the New Oxford chronology his percentages are: The Two Gentlemen of Verona 18.4, Titus Andronicus 16.4, 2 Henry VI 15.4, 3 Henry VI 13.3, The Taming of the Shrew 16.2, Richard III 17.4. The relative infrequency of run-on lines characterizes all these plays, regardless of whether or not they are judged by the New Oxford editors to be collaborative. Pohl based his counts on punctuation in the nineteenth-century Cambridge edition (1863-66) and acknowledged that later editions yielded somewhat higher results. (20) But it is clear that this aspect of the verse of substantial Q1 passages closely matched in Q2 associates it with Shakespeare's maturity, not with his beginnings as a dramatist.
Another feature of Shakespeare's verse that lends itself to computation and has proved efficacious in helping determine the order in which his plays were written is the position of pauses within lines, as distinct from at their ends. Working with the early printed quarto and folio texts of a wide range of early modern plays, Ants Oras furnished comprehensive data about pauses marked by (a) any form of punctuation, (b) punctuation heavier than a comma, and (c) a change of speaker within a single pentameter line. He counted the total numbers of pauses after each syllable, from the first to the ninth, within all Shakespeare's blank verse. (21) Patterns in the placement of all three categories of pauses were remarkably similar. Shakespeare's plays evinced a progressive tendency for pausation to shift to the second half of the line (positions six to nine).
I therefore checked the pause patterns of the passages cited above as analyzed for run-on lines, consulting Q1 for its punctuation and imitating Oras's methods, but accepting the few Arden-3 relineations. If I had kept with the Q1 lineation, the results would not have been significantly different. Nor would they if I had worked from the Arden-3 punctuation, since its placement of stops within lines is similar to Q1's. It was necessary to include all pauses (Oras's type-A), in order to obtain enough data for a clear pattern to emerge. Pauses were distributed as in Table 1. Since this pattern is based on only 53 pauses, there is scope for the operations of pure chance, but it is most improbable that any blocks of text of comparable length from an early Shakespeare play would yield anything remotely similar. In all Shakespeare's early plays Oras's figures show far more A-type pauses after syllable 4 than after syllable 6. The first plays in which there are marginally more pauses after syllable 6 than after syllable 4 are Twelfth Night (dated 1601 by the New Oxford Shakespeare) and the Q2/F Hamlet (1602-3). The profile for the Q2-matching Q1 passages more closely resembles that of slightly later plays. Oras showed that the percentage of pauses in the first half of Shakespeare's pentameters decreases over time. For the Q1 passages the percentage (calculated, as by Oras, by excluding pauses after syllable 5) is 31.8, which is lower than for Hamlet as a whole (47.8) and closer to the percentages for plays usually dated several years later, such as Macbeth (35.3), Shakespeare's share of Pericles (31.4), Antony and Cleopatra (29.6), and Coriolanus (29.8). The indications of a date this late are obviously due to the smallness of the sample, since Q1 was printed in 1603, but these Q1 passages, closely agreeing with their counterparts in Q2, are metrically unlike early Shakespearean plays, which yield the following percentages: The Two Gentlemen of Verona 64.1, Titus Andronicus 69.2, 2 Henry VI 63.5, 3 Henry VI 67.3, The Taming of the Shrew 68.2, Richard III 62.5.
The metrical evidence is confirmed by that of rare vocabulary. Eliot Slater compiled information about all word types occurring in at least two Shakespeare plays but not more than ten times in the whole dramatic canon. (22) His concern was with "dictionary words," in which inflexions, declensions, conjugations, and so on are grouped under one head, but parts of speech and graphic units with entirely separate meanings are differentiated. So the noun "arrest" was treated as different from the verb "arrest," and "bat" as a mammal as distinct from "bat" as a piece of wood. Slater demonstrated that the proportions of these lexical items shared between any two plays correlated with the degree of proximity of the dates at which they were thought to have been written: the closer they were in chronological order of composition, the more rare words they were apt to share, relative to the numbers of link words in each. Sameness of genre tended to enhance the effect, and difference of genre to reduce it, but the chronological factor was paramount. It was strongest for words occurring in the dramatic canon not more than six times. Slater's results were for plays in the First Folio (1623) plus Pericles, which he divided into the non-Shakespearean acts 1-2 and the Shakespearean acts 3-5: he ignored co-authorship in the Folio plays and excluded The Two Noble Kinsmen. For his tests, Slater began with an order of composition derived from Chambers, but with some minor adjustments. Refinements of his methodology, restricted to his data for "two-six-fold link words" (as Slater called those found in two plays but not more than six times in the canon), have since resulted in tables that have contributed to the formation of the New Oxford Shakespeare chronology, with its modifications, mainly minor, of that of Chambers. (23)
In an appendix to his book, Slater listed the words used for his analyses, grouping them under "two-words," "three-words," "four-words," and so on. I entered all these words into a computer file and amalgamated them into a single alphabetical list created by the concordance program Micro-OCP, while attaching to each word coded information about its part of speech as given by Slater and the category of rarity to which Slater had assigned it. With the aid of this mini-concordance, it has not been too difficult to check Hamlet Q1, 1.42-131; 2.110-25; 2.155-64 + 166-69; 4.15-32; 5.8b-18; 5.41-53; 7.31-51, for "two-six-fold words" in Slater's appendix that appear in not more than five other plays. Through reference to Spevack's Concordance and Schmidt's Lexicon (where homophones and parts of speech are distinguished) it was possible to identify in which canonical Shakespeare plays they occurred. (24) There were 107 items in all--too few to allow comparison between the actual numbers for individual plays and the numbers to be expected on a random distribution, but enough to form a telling pattern when plays are grouped according to the order of composition determined in the New Oxford Shakespeare.
Six such groups of six plays each may be formed so as to divide the dramatic canon into a conveniently workable number of roughly equal chronological sections: from The Two Gentlemen of Verona to Richard Third, from The Comedy of Errors to A Midsummer Night's Dream, from King John to Henry V, from Julius Caesar to Othello, from Measure for Measure to Antony and Cleopatra, and from Pericles, III-V, to All Is True (Henry VIII). Table 2 gives the number of links in "two-six-fold words" between the Q1 passages and each of these groups. Hamlet itself was not included in any group, though naturally there were several links to other parts of the play. Nor were Arden of Faversham and Edward III, which The New Oxford Shakespeare accepts as partly by Shakespeare. For each chronological group is also shown the New Oxford Shakespeare span of dates for the group and its number of "two-six-fold words." The New Oxford editors argue that the Henry VI plays, Titus Andronicus, and possibly The Taming of the Shrew, all falling into Group 1, contain the writing of other playwrights; that the Henry VI plays were "adapted" in 1595, and the Group 5 plays Macbeth and All's Well That Ends Well in 1616 and 1622 respectively; and that Timon of Athens (Group 5) and All Is True (Henry VIII) (Group 6) were collaborations with Thomas Middleton and John Fletcher respectively. Some of these conclusions have for some time been orthodox, while others are based on new evidence. But none drastically affects the obvious interpretation of the above figures, which strongly link the selected Q1 passages to the Group 4 period, 1599-1604, and definitely not to Shakespeare's pre-1596 dramatic output.
It might be suspected that collaboration could have reduced the number of Q1's lexical links with the first chronological group of plays by reducing the volume of Shakespearean text. But the findings cannot be explained in this way. Each play in the first group, regardless of whether or not it has been deemed collaborative, is most closely linked, in terms of "two-six-fold words," to early Shakespeare plays. The Two Gentlemen of Verona, not known to be collaborative, is most significantly linked to Richard III, while Richard III, of unchallenged single authorship, is most significantly linked to 3 Henry VI. Moreover, if we were to add to the six Group 1 plays the next three whole single-authored plays in the Mew Oxford chronology--namely The Comedy of Errors, Love's Labour's Lost, and Richard II--the number of Q1 links to this expanded first group would rise only to 15, still far short of the 28 with Group 4 plays.
There are 45 links with the first three groups, 67 with the last three, despite the former's having 774 more "two-six-fold words." Moreover, the links pinpoint with some precision a likely position in the sequence of composition. When all thirty-six plays--Hamlet still being excluded--are ordered as in the New Oxford Shakespeare chronology, the three adjacent plays that share easily the largest number of "two-six-fold words" with Hamlet Q1's Q2-matching passages are Troilus and Cressida (1602), Othello (1604), and Measure for Measure (1604) with 19, and this trio continues to stand well clear of any other when figures for all combinations of three consecutive plays are expressed as rates per 3,500 "two-six-fold words." (25) It is between Troilus and Cressida and Othello that the New Oxford Shakespeare places the creation of Hamlet in its familiar Q2/F form, dating it 1602-3.
This exact dating of the Q1 passages cannot be taken too seriously. The Textual Companion (1987) to the Oxford Complete Works of 1986 dated Hamlet 1600-01, followed by Twelfth Night (1601), Troilus and Cressida (1602), Measure for Measure (1603), and Othello (1603-4). (26) Arden-3 editor E. A. J. Honigmann argued strongly for 1602 as the most probable year of the first performance of Othello. (27) But what we can say with confidence is that the Q1 passages tested not only closely match Q2/F textually but were evidently composed at much the same time as the full canonical play.
The presence in those Q1 passages of instances of hendiadys, beyond those already mentioned in 1.42-131, reinforces this conclusion: "the dead vast ['waste' Q2, 'wast' F] and middle of the night" (2.112), "Angels and ministers of grace" (4.15), "ponderous and marble jaws" (4.26), "knotted and combined locks" (5.13). At 7.50, Q1's plural "allowances" in "On such regards of safety and allowances" is a simple error for Q2's "allowance" (2.2.79). Wright's supplementary list of doublets that do not quite qualify as instances of hendiadys includes "my weakness and my melancholy" (7.431).
Might Q1 nevertheless represent a Shakespearean Ur-Hamlet that, in the passages so far discussed, includes revisions that Shakespeare made ten or more years later? Without attempting the formidable task of tracking down all Slater's "two-to-six-fold" words in the whole of Q1, we can employ another set of lexical data to help us answer the question. Slater's research was to some extent anticipated in the nineteenth century by Gregor Sarrazin, who, working from Schmidt's Lexicon, compiled lists for each Shakespeare play of the words occurring only once or twice elsewhere in Shakespeare's work. (28) He demonstrated that the shared "dislegomena" and "trislegomena," as he called them, were most frequent between plays of more or less the same date, and that the statistics placed each individual work in the right period. In Studies in Attribution (1979) I showed that a vocabulary index, constructed from the number of each play's Sarrazin links with the seventeen plays from As You Like It onwards, as a percentage of its links to all thirty-seven plays (Sarrazin gave no data for The Two Noble Kinsmen) placed plays in an order correlated to standard chronologies. (29) 2 Henry VI came first with a percentage of 29.8, Antony and Cleopatra last with a percentage of 63.7. For All Is True (Henry VIII), the last play to be written, the percentage was 63.2.
At that time I used Karl Wentersdorf's chronology, but the same plays fell within the first twenty and last seventeen as in the Oxford Textual Companion's chronology of 1987. (30) I have sifted through Sarrazin's dislegomena and trislegomena for Hamlet, finding those that are in Q1. Q1 contains 80 instances of these rare-word links to other Shakespeare plays, 33 to the twenty plays up to Julius Caesar, 47 to the plays from As You Like It to All Is True (Henry VII). (31) This (47 as a percentage of 80) gives an index of 58.8. In my Studies in Attribution table, the first Shakespeare plays in the New Oxford Shakespeare chronology with indexes over 55.0 are the canonical Hamlet (dated 1602-3) with 60.8 and Othello (dated 1604) with 57.2. The New Oxford Shakespeare shifts The Merry Wives of Windsor from the first twenty plays into the last seventeen, so that As You Like It moves from the last seventeen to the first twenty. This, however, does not change the figures for Q1, since both re-categorized plays share with it only a single Sarrazin word. Nor would it change my old figures for the Q2/F Hamlet, since each of the re-categorized plays shared seven Sarrazin words with the canonical play.
Sarrazin's tables included dislegomena and trislegomena that a play shared with itself. So "unction," which appeared twice in the canon, each time in a different scene of Hamlet, was included in his chart for Hamlet as one link to Hamlet. The verb "comply," which occurred twice in Hamlet and once in Othello was recorded in the chart for Hamlet as one link to Othello and one to Hamlet. This can become rather confusing, and Sarrazin occasionally confused himself. He made some errors, and although I tried to correct those that I observed, perfect consistency and accuracy is hard to achieve. (32) Sarrazin's Hamlet was, of course, the canonical Q2/F version: my calculations of Hamlet-to-Hamlet links use only those in Sarrazin's table that also occur in Q1.
But despite Sarrazin's occasional slips, and the likelihood that I have made some too, it seems impossible to reconcile the evidence from Sarrazin's data with a theory that Q1 essentially preserves a play written by Shakespeare in the late 1580s. And if we wish to postulate that what Q1 prints is a partial revision of such a play, it appears, from these results, that such a revision would have to have gone well beyond the insertion of the several passages identified above as late-Shakespearean writing that has few variants from the corresponding passages in Q2. Seventeen instances of the Sarrazin words occur within the Q1 passages tested for their prosody and by Slater's "two-six-fold words," but fifty are scattered over the rest of Q1.
Again, the presence in early plays of material by playwrights other than Shakespeare cannot account for these results. The Studies in Attribution indexes for the earliest six Shakespeare plays are: The Two Gentlemen of Verona 37.0, Titus Andronicus 29.8, 2 Henry VI 31.4, 3 Henry VI 32.5, The Taming of the Shrew 34.5, Richard III 35.6. If Q1 represented a Shakespearean play of the late 1580s, its index should fall within or below this range, not be as high as 58.8. The majority of Sarrazin words that Q1 shares with Q2/F are linked to Shakespeare's mature plays. The pattern is incompatible with their having survived into Q2 from a Hamlet written by Shakespeare at the very beginning of his playwriting career. (33)
Wright's instances of hendiadys also extend beyond the passages previously selected as nearly identical in Q1 and Q2: "purged and burnt away" (5.6), which inverts Q2's superior "burnt and purged away" (1.5.13), "knitted and combined locks" (5.13), "thin and wholesome blood" (5.54), "wild and whirling words" (5.104), "the whiff and wind of his fell sword" (7.358). Q1's "chronicles and brief abstracts" (7.381) inverts the F hendiadys, listed by Wright, "abstracts and brief chronicles" (2.2.520; "abstract" Q2 2.2.462).
Q1's Deviations from Q: Evidence of Textual Corruption
The evidence adduced above seems to prove that if Q1 is based on a Hamlet composed by Shakespeare about 1589, it cannot preserve the early play in unmodified form, but must incorporate the extensive beginnings of a process of Shakespearean revision and expansion that was to result, very soon afterwards, in the canonical Q2/F play. Bourus does at one point in her book postulate a transcript behind Q1: "Stylistically, the language and verse style of the first edition makes sense as a play written in the late 1580s, transcribed and printed in the early seventeenth century." (34) But mere transcription would not have created the metrical and lexical patterns to which I have drawn attention, and Bourus's repeatedly implied, and several times stated, position is "that 1589 Hamlet was Shakespeare's, and that the text is preserved in the 1603 first edition." (35) She invites us to read Q1 "as though it were the first play that young Shakespeare wrote without a collaborator." (36) But the invitation, as it stands, must be declined.
What of the alternative theory that dates Shakespeare's first draft, supposed to be preserved in Q1, to immediately before the revision that resulted in the Q2 version? This would require a draft that mixed typical mature Shakespearean writing with rough, poetically lackluster jottings, still to be worked up into the brilliant dialogue of Q2. This is, in fact, what Urkowitz envisages: in variant Shakespeare texts such as Q1, Q2, and F Hamlet "We can learn about dramatic art from closely observing how he worked and re-worked those [variant] lines, speeches, scenes, characterizations." (37)
That a quarto should mingle the products of different phases of a play's evolution is not, of course, impossible. But we must recognize what this would entail for the details of particular passages over which the wording of Hamlet Q1 and Q2 is markedly divergent. In Stolne and Surreptitious Copies (1942), Alfred Hart discussed eleven representative passages in Hamlet Q1 where "the text is defective or without meaning or suggests some omission of matter relevant to the context or even the plot." (38) His second example will serve as illustration. Claudius begins scene 2 as follows:
Lords, we here have writ to Fortenbrasse, Nephew to old Norway who, impudent And bedrid, scarcely hears of this his Nephew's purpose. And we here dispatch You, good Cornelia, and you, Voltemar, For bearers of these greetings to old Norway... (2.1-6) (39)
The Arden-3 editors have emended Q1's "Yong good Cornelia" to "You, good Cornelia" and moved "Norway" up into line 6, which ends in Q1 with "olde." Claudius, King of Denmark, has written to Fortenbras, (40) the nephew of the King of Norway. The words "here have writ" imply that Claudius holds and displays a letter or packet, though Hart does not make this point. The uncle, an invalid, knows little of his nephew's "purpose." This purpose is not specified, though in the opening scene we have learned that the Danes are arming themselves against a perceived military threat and that Fortenbras has gathered a band of "lawless resolutes" for "some enterprise" (1.79-84). Claudius then sends two ambassadors to bear "these greetings" to the King of Norway. A physical object, conveying the message, must be shown and now handed over. Surely it must be the letter that Claudius has "writ"--but that was "to Fortenbrasse," not the King of Norway. Thompson and Taylor conclude lamely that Claudius "seems to have written to both the nephew and the uncle." Yet the deduction that the reference is to a single letter is supported by one's natural inference that "we here have writ" leads to "we here dispatch," and nobody is charged with delivering a letter to Fortenbras. When the ambassadors return, it is clear that their transactions have been with the King of Norway alone, and that he has responded to Claudius's missive by duly disciplining his nephew (Q1, 7.30-51; Q2, 2.2.58-80).
Q1's speech contains further anomalies. Its "impudent" could be a compositor's misreading of Q2's "impotent" (validated by Q1, 7.37; Q2, 2.2.66), though it seems more like a mishearing, and the fact that only the last two of the six quoted lines are, in the Arden-3 edition of Q1, respectable iambic pentameters seems to confirm a diagnosis of more pervasive corruption. The fifth line is regular if "Cornelia" is given a trisyllabic pronunciation. The sixth (which has a feminine ending) lacks "Norway" in Q1's setting out of the verse.
In Q2 Claudius begins his much longer speech (1.2.1-49) by addressing the theme of his marriage to his recently deceased brother's widow, and then devotes ten lines to Fortinbras' belligerent demand that Norwegian lands lost in single fight between his father and the late King Hamlet be returned, before continuing "we here have writ / To Norway, uncle of young Fortinbras," urging the King to restrain his nephew, and finally sending away Cornelius and Voltemand as bearers of the relevant documents. The speech throughout is metrically regular, rhythmically and syntactically energetic and varied, and enlivened by apt imagery--and it makes perfect sense.
An explanation of the difference between Q1 and Q2 that identified Q1 as derivative and, in 2.1-6, corrupt would acknowledge the probability that the opening of the scene has been subject to deliberate abridgement, but argue that the true text's "we have here writ / To Norway, uncle of young Fortinbras" has been mistakenly turned back to front in Q1 as "we here have writ to Fortenbrasse, / Nephew to old Norway," through memorial anticipation of "nephew to old Norway, Fortinbras" in Q2, 4.4.13, rendered as "Fortenbrasse, nephew to old Norway in Q1, 11.2; and that Q1's inversion at 2.1-2, together with the absence from Q1 of Q2 matter, has distorted Shakespeare's meaning and meter, while also creating the clumsy repetitions (absent in Q2) of "old Norway" and "Nephew." The "young" of Q2's "young Fortinbras" is in Q1 applied to "Cornelia" ("Yong Cornelia"), though this may well be a simple misreading for "You," to which the Arden-3 editors emend. If, on the other hand, the manuscript behind Q1 antedated Q2 and F, either Shakespeare had drafted this speech most ineptly and incompletely, leaving it inchoate, or the manuscript was at this point so untidy as to be barely decipherable.
Hart's eleven examples of Q1/Q2 divergence of this kind were offered merely as a sample, which could easily be enlarged. I want to discuss one that Hart ignores. Bourus gives an illuminating account of her own production of Q1 Hamlet, in which she also played Gertrude. (41) She describes the theatrical potential she found in Q1's text of the Queen's speech about Ophelia's death. Q2's version at 4.7.160-82 is choric, lyrical, and elegiac--akin to an operatic aria. The plangent blend of the emblematic and the realistic is perfectly "designed in all its details to provide the Ophelia we have seen with her most appropriate end," as Jenkins's sensitive analysis demonstrates. (42) In Q1's version, but not in Q2's, Gertrude twice calls Ophelia "young," and Bourus, stressing Ophelia's youth, developed an intimacy between her and Gertrude, a surrogate mother, protective towards the girl. Bourus discovered that Q1 here "has much more immediacy, and creates a far more personal sense of grief." The Queen enters "in the first shock of bereavement," her state of mind being captured in utterance that "is rushed, barely grammatical." (43)
The apparent disjointedness of the Queen's narrative in Q1 can thus, by an astute director, be given psychological point in performance and be taken as a clue to characterization. But some features of the Q1 speech resist explanation except as the result of textual corruption:
O, my lord, the young Ofelia, Having made a garland of sundry sorts of flowers, Sitting upon a willow by a brook, The envious sprig broke. Into the brook she fell And for a while her clothes, spread wide abroad, Bore the young lady up and there she sat Smiling even mermaid-like 'twixt heaven and earth, Chanting old sundry tunes uncapable, As it were, of her distress. But long it could not be Till that her clothes, being heavy with their drink, Dragged the sweet wretch to death. (15.40-50)
Laertes completes the last verse line with "So--she is drowned?" This is Q1's only use of "drowned," whereas five instances of the word reverberate in Q2. The scene is not set, as it is in Q2's "There is a willow grows askant a brook / That shows his hoary leaves in the glassy stream." (44) Ophelia has "made a garland of sundry sorts of flowers." In Q2 they are itemized by name and carry symbolic suggestiveness. In Q1 they are not named, merely generalized as "sundry," an adjective clumsily repeated for the "tunes" that she sings after she has fallen into the brook. The accident happened when a "sprig" unaccountably broke while she was "Sitting upon a willow," whereas in Q2 the disaster more plausibly and pointedly occurred while she was "clambering" to hang the garland on the tree, as convention prescribed for the lovelorn. In Q1, as in Q2, Ophelia's clothes spread out and bore her up, but Q1 alone adds that "there she sat." How would John Everett Millais have pictured this? Ophelia has fallen into the water and her clothes have kept her from sinking, but she is floating in a sitting position as she smiles like a mermaid. The simile seems less than apt, since mermaids are not especially noted for their smiles, except when beguiling seafarers to their doom, an unwanted association in this context. In Q2 "mermaid-like" has obvious point: Ophelia's garments buoy her up as though she were the mythical creature, half woman and half fish, and so "native and endued / Unto that element" into which she has fallen (Q2, 4.7.177-78). In Q1 Ophelia is, moreover, "'twixt heaven and earth," as though she were still sitting on the willow or dangling from a branch that does not overhang the stream.
An explanation of Q1's text here as the creation of a struggling memory would suggest that "'twixt heaven and earth" has been recollected from the Nunnery scene, where Hamlet exclaims "What should such fellows as I do crawling between heaven and earth?" (Q2, 3.1.126-27), and the use of "heaven and earth" at 1.1.123, 1.2.142, and 1.5.165 (Q2, the instance at Q2, 1.1.123 being absent from F); and that a reporter has concocted the bizarre "sat / Smiling" from vague recollection of the King's "This gentle and unforced accord of Hamlet / Sits smiling to my heart" (Q2, 1.2.123-24). Other Q1 deviations from the fuller text hint at the same process, though the plethora of commonplace tags and discourse markers accumulated in the Queen's speech but sprinkled over other parts of the Q2/F play need not have specific sources. (45)
The irregularity of the verse--especially lines 40, 41, 44, 47, and 48--seems unlikely to be due to Gertrude's distress. When in Othello, 4.1, the protagonist is reduced to incoherence that culminates in his falling into a trance, the medium is prose. Nor can a beginner's lack of expertise be reasonably postulated. Even in Sonnet 145, probably composed in the early 1580s, Shakespeare proved himself able to write correct iambic verse--in tetrameters, but metrical. (46) The nerveless string of present participles heading the Q1 lines--"Having... Sitting... Smiling... Chanting"--betrays a lack of invention, not normally associated with Shakespeare; and "being" is characteristic of the syntax of the person responsible for Q1: on fifteen occasions it is used in Q1 where its wording diverges markedly from Q2's. (47)
Q1's repetitions include not only "young" and "sundry" but also "brook" and "clothes" and the way "Sitting" is taken up in the incongruous "sat." Q2 does have two instances of "brook," but these are nine lines apart, not obtrusively placed in consecutive lines as in Q1, and when Ophelia tumbles into it, it is in Q2 "the weeping brook," and it is also "the glassy stream." Q2 varies "clothes" with "garments." At no time in Q2's telling of her demise does Ophelia sit. In Q2 the Queen describes a credible sequence of events. In Q1 she does not, and the muddle cannot, in my view, be attributed solely to her distraught state.
If Q1 was based on an early Shakespearean Hamlet or first draft, that precursor of the canonical play must surely have here been represented by Q1 most imperfectly. Q1's "'twixt heaven and earth" might be defended as meaning that Ophelia was suspended in the water between the sky above and the mud of the river bottom, the idea being more neatly encapsulated in Q2's "muddy death" where Q1 has the unmodified "death." But it is hard to believe that Shakespeare ever composed, as a forerunner to the Queen's lyrical masterpiece, an unmetrical speech in which Ophelia was sitting on a willow branch, which broke so that she fell into a brook but remained in a seated posture ("there she sat") while floating and singing with a smile on her face, before sinking and drowning. Moreover, the speech is like an epitome of the whole Q1 text, in that within the blandness, crudity, and confusion are interspersed words and phrases that both seem typically Shakespearean and are also in Q2, as in the adjective personifying "the envious sprig" (Q1's unmetrical variant for Q2's "the envious sliver"), the idea of Ophelia's garments becoming "heavy with their drink," and the compound "mermaid-like," misapplied in Q1. Even "But long it could not be / Till that" sounds like the kind of Shakespearean formulation that might impress itself on a hearer's mind. Are these expressions not more likely to be vestiges of the canonical Q2/F play than fragments accepted into it? (48)
There is doubtless scope for conflicting answers to this rhetorical question. Also open to interpretation is the phenomenon described by G. I. Duthie in his detailed examination of passages in Q1 of tolerably regular blank verse that conveys roughly the same content as Q2/F but in a very different style and wording. He sought to show that these "consist simply of numerous stray fragments of text gathered together from various points scattered throughout the full Shakespearian versions" and "welded into presentably metrical, though generally dull and flat, blank verse," and that those fragments had often been recalled through association with contexts that reminded the actor-versifier (whom he postulated) of the one he was attempting to paraphrase. (49)
Laurie E. Maguire discusses Duthie's methodology under the head "Internal repetition" and within her strictures on critics who "cite repetition of single words as evidence of faulty memory." She objects that words Duthie relies on in his attempt to demonstrate that within Q1's two lines at 14.10-11 "practically everything can be traced to other passages"--such as "looks," "perceive," and "villainy"--have many entries in Shakespeare concordances and that the phrase "sugar o'er" is similarly valueless for Duthie's purposes, since the association between sugar and false friends is also frequent in the Shakespeare canon. (50)
Duthie analyses the whole Q1 scene, which is absent from Q2/F. (51) The two lines singled out by Maguire follow on from Horatio's revelation that Hamlet "escaped the danger / And subtle treason that the King had plotted" and run as follows:
Then I perceive there's treason in his looks That seemed to sugar o'er his villainy. (14.10-11)
The searchable electronic database Literature Online finds only one other use in drama of 1576-1642 of the expression "sugar o'er," and that is by Shakespeare in the canonical Hamlet's scene in which Polonius and Claudius eavesdrop on Ophelia's encounter with Hamlet. (52) Polonius, having equipped his daughter with a book, presumably a prayer book, to read so as to "colour" her "loneliness," reflects "We are oft too blame in this--/ 'Tis too much proved that with devotion's visage / And pious action we do sugar o'er / The devil himself (Q2, 3.1.45-48), prompting Claudius to exclaim in an aside "How smart a lash that speech doth give my conscience!" and to liken his outward show to the "harlot's cheek beautified with plastering art" and to contrast his ugly "deed" to his "most painted word" (3.1.49-52). None of this is in Q1.
From the extreme rarity of "sugar o'er" we may confidently conclude that either the Q2/F instance was transferred to a different scene of Q1 or vice versa. The two contexts are linked by concern with the King's hypocrisy. It seems not unreasonable of Duthie to suppose that a reporter borrowed from a dim memory of Q2/F as he concocted Q1-only scene 14, which is unsatisfactory in other respects, as we shall see. Since "perceive," "his looks," and "villainy" are in Q2 all similarly associated with the King's crime and the play-scene's scheme to expose his guilt, Duthie plausibly regards these words in the Queen's Q1 lines as further recollections of the relevant Q2 passages.
The Queen's two lines are very loosely phrased, with redundancy in "looks" and "seemed" or in "seemed" and "sugar o'er," while "treason" (an echo of Horatio's "subtle treason") is not only inexact, because Claudius is himself the monarch, but feebly varied in "villainy." The wording and imagery of Polonius's speech in Q2's 3.3.45-48, in contrast, exhibit the vividness and concentration typical of mature Shakespearean verse. That "sugar o'er" was not transferred from the Queen's speech in Q1 to Polonius's in Q2 and that even a youthful Shakespeare would not have written 14.10-11 are subjective literary judgments, but it is certainly true that if we maintain that Q1's scene 14 as a whole rests on a pre-Q2 version by him, of whatever date, we are forced to believe either that he wrote, in a drab and primitive style, an account of Hamlet's voyage so elliptical as to make little sense, or that Q1's text of the postulated pre-Q2 scene suffers from serious textual corruption, of a kind not associated with regular textual transmission.
The charge that Q1's brief scene 14 is "unsatisfactory" and that Horatio's version of Hamlet's adventures is confusedly "elliptical" requires substantiation. Horatio reports to the Queen on Hamlet's escape from the death in England that Claudius had plotted for him. There is no such scene between Horatio and Gertrude in Q2/F. In Q1 Horatio tells the Queen that Hamlet "is safe arrived in Denmark." Horatio has just received a letter from him,
Wherein he writes how he escaped the danger And subtle treason that the King had plotted. Being crossed by the contention of the winds, He found the packet sent to the King of England, Wherein he saw himself betrayed to death... (14.3-7)
Hamlet will give his mother details when he next sees her. Q1's punctuation, which the Arden-3 editors retain, implies a connection, left unexplained, between Hamlet's being "crossed by the contention of the winds" and his finding the packet, but if the full-stop at the end of line 4 and the comma at the end of line 5 were transposed, line 5 could be understood as meaning that the King's plot had been thwarted by unfavorable winds. Hamlet has asked Horatio to meet him "on the east side of the city / Tomorrow morning" (14.16-17), though no reason is given for the delay and it is unclear why Hamlet did not immediately upon landing seek out Horatio at court. Asked by the Queen what became of Rosencrantz and Guildenstern, Horatio replies:
He being set ashore, they went for England. And in the packet there writ down that doom To be performed on them 'pointed for him. And, by great chance, he had his father's seal-So all was done without discovery. (14.26-30)
Horatio is as bad at telling a story as Gertrude in Q1's scene 15. Lines 27-28 stand free of any syntactical connection with line 26, though transposition of the first and second halves of line 26 and deletion of the full stop would provide the semblance of one. Hamlet can only, in this account, have been "set ashore" in Denmark, which might mean that the "contentious winds" had driven the vessel back to the shores from which it set forth, so that the King's so-called treasonous intent was "crossed." (53) But it is highly improbable that Shakespeare composed a draft in which Rosencrantz and Guildenstern, who had been charged with escorting Hamlet to England, allowed him to disembark in his home country and themselves made a second attempt to sail to England without him.
Because Horatio's account in Q1 of Hamlet's adventures lacks any mention of the skirmish that left the prince on a pirate vessel, separated from his escorts, as detailed in Q2, 4.6, it sketches such a disordered sequence of events as to be barely intelligible. Duthie and Jenkins have offered alternative explanations for Q1's narrative incoherence, the latter arguing strongly that everything that Horatio says in Q1's scene 14 derives solely from Q2/F, garbled in the telling. (54) Another oddity in the scene is that in her concluding speech the Queen says "Horatio, once again I take my leave," despite not having taken her leave before. The Arden-3 editors suggest that she was about to exit after Horatio's couplet at lines 23-24, but turned back with the question "But what became of Guilderstone and Rossencraft?"(25). From "once again" we would thus be invited to infer an implicit stage direction at line 25, like "Nurse offers to go in and turns again" (modernized) in Q1 (1597) Romeo and Juliet (G[2.sup.r]). (55) Whatever the reasons for Q1's deficiencies in scene 14, it seems to me unlikely that the text here faithfully reflects Shakespeare's writing at any stage of his career.
Many Q1 speeches discussed by Duthie appear to gather phrases and collocations from other parts of the play than the corresponding, patently Shakespearean speeches in Q2. Rosencrantz's four-line response to the King's request that he and Guildenstern seek "the cause and ground" of Hamlet's "distemperance" runs as follows in Q1, 7.9-12:
My lord, whatsoever lies within our power Your majesty may more command in words Than use persuasions to your liegemen, bound By love, by duty and obedience.
This mixes words also in Q2's version of the same speech (Q2, 2.2.26-29) with "lies within our" from Q2, 2.2.18, and "duty and obedience" from Q2, 2.2.106. (56) Neither phrase is used anywhere else by Shakespeare. The word "liegemen" is absent from the Q2 version of Rosencrantz's lines, but occurs prominently in Q2 and F at 1.1.13. Here in Q1 the writing is again slack, the antithesis between commanding and persuading being blurred by "in words," since the "persuasions" to which Rosencrantz alludes have also been verbal. Guildenstern's confirmation in Q1 that "what we may do for both your majesties" the pair will endeavor "all the best we may" (7.13, 15) gauchely repeats "we may." More significantly, its putative appropriation of "both your majesties" from Rosencrantz's speech in Q2 (2.1.25) would appear to confirm that here Q2 is the source and Q1 the debtor, because in Q2, the Queen has added her voice to the King's (Q2, 2.2.18-26), but she has remained silent in Q1.
The point is that although Q1's numerous repetitions and apparent "anticipations" or "recollections" of scraps of dialogue in other portions of the canonical play need not in themselves constitute evidence of some form of reporting, they are almost always associated with other signs of textual disruption. Scholars who believe that Q1 preserves intact a Shakespearean version antedating the Q2/F play, are forced to suppose that words and phrases that Shakespeare originally included in pedestrian, clumsily repetitive, and sometimes barely meaningful verse were, as he revised, dispersed to be incorporated into eloquent speeches in other parts of the play. An alternative theory that Q1 often misrepresented a Shakespearean pre-Q2 draft would seem to have little advantage over Duthie's more economical hypothesis--accepted by editors Jenkins, Edwards, Hibbard, and Taylor in the 1980s--that the text misrepresented in Q1 was essentially the canonical Q2/F Hamlet itself.
Of course, revising authors are apt enough to shift material around and their revisions are usually improvements. Inevitably, subjective notions about what Shakespeare could and could not have written affect assessment of Duthie's arguments. Urkowitz is scornful of scholars who believe "that passages that appear to be incoherent in the Q1 text... could not have been part of an authorial draft" or "prompted by confusion in an authorial manuscript." (57) But a Shakespearean draft of Hamlet that was little more than half the length of the Q2/F play shortly to be completed (as I have shown in the opening section of this article) and that was so confusing as to provoke the defects from which Q1 suffers would have been Shakespeare's personal property, and it is hard to see why he would release it for publication in print.
Hart's meticulously compiled lexical data showed that all six texts that he stigmatized as bad quartos, (58) including Q1 Hamlet, share from 86.5 to 93.3 per cent of their word-types with their canonical counterparts, the conflated Q2 and F play of editorial tradition in the case of Hamlet, and that this degree of overlap is far greater than between King John or King Lear and their acknowledged source plays The Troublesome Reign of King John and King Leir. (59) Although this vocabulary evidence alone suffices to rule out the theory, sometimes advanced in the past, that Q1 Hamlet represented a non-Shakespearean Ur-Hamlet, it leaves unscathed the theory that Q1 was based on an early draft by Shakespeare himself. But, as Hart insisted, if Q1 preserves the playwright's own draft, "Shakespeare, after omitting 400 lines, corrected syntax, diction, sense, verse and order of what was left of Q1 with minute exactitude and fidelity; he then inserted nearly two thousand lines of poetry, unnecessary to the story or action and of little dramatic value." Hart asserted that "This method of mending old plays was not used by any known dramatist" and declared that "the poet who, his fellows said, 'never blotted out line' does not fit the picture of dull and patient drudgery which such line-by-line revision suggests." (60) Hart's formulation "of little dramatic value" is ill-considered, begging the question of how this should be defined, but it is true that the postulated revision had minimal effect on the development of the plot.
Any convincing theory of Q1's origins must account for its extreme stylistic variability. Much of its material is undoubtedly contemporaneous with that in Q2, which it often matches verbatim, or nearly so. But it also contains dialogue that is linguistically insipid, rhythmically limp, and syntactically rudimentary. Further, there are many passages that make no pertinent sense, and several of these also suffer from metrical breakdown, usually associated with clumsy repetition. Ultimately, attempts to account for the origins of Q1 are inextricable from critical judgments on it, as literature and as drama. Experienced directors and actors have stressed the potential of the Q1 version of Hamlet for performance as a fast-paced action play. With a little tidying up, the script has formed the basis for successful theatre productions. Critics who consider Shakespeare's artistry as a dramatist to be intimately related to his artistry as a poet may nevertheless continue to doubt that he composed, at any stage of his career, those Q1 passages in which the wording diverges markedly from Q2's. This is especially so now that attribution studies have sharpened our sense of Shakespeare's earliest style in dramatic verse by establishing that the monotonous first act of Titus Andronicus and the jerky first act of 1 Henry VI are not by him but by Peele and Nashe respectively. (61) The theory held by editors of the 1980s that Q1 presents an abridged and corrupted text ultimately derived from Q2 did allow for Q1's theatrical merits, its often humdrum language, its definite errors, and the Shakespearean quality of its verse when it closely agrees with Q2's.
Bourus argues strongly against both the hypothesis that an actor who played Marcellus (and doubled Lucianus) constructed the Q1 text from his memory of performances and the alternative hypothesis that note-taking members of an audience recorded the Q1 version. (62) But Tiffany Sterne's revival of the theory that printer's copy for Q1 had been compiled by theatergoing note takers has received a boost from John Jowett's examination of the career of James Roberts, who not only entered Hamlet in the Stationers' Register on 26 July 1601 and printed Q2 for Q1's publisher Nicholas Ling but also held a monopoly, which he was soon to lose, on the manufacture of "writing tablets." (63) Of course, memory would have been involved in the kind of textual transmission envisaged by Sterne: the performing actors had memorized their lines and the auditors would have placed some reliance on their memories as they tried to piece together their notes. In another article Jowett provides a new argument, quite independent of mine, for believing that a substantial block of Hamlet text unique to Q1 belongs to 1602, rather than the late 1580s. (64)
The aim of the present article is not, however, to advance a fully fledged theory about the genesis of Q1. (65) It does present reasons for doubting that the manuscript behind Q1 existed before the manuscript behind Q2. But its main claim is to have demonstrated, beyond reasonable doubt, that where Q1's wording is closest to Q2's, that wording was not present in an Ur-Hamlet of the late 1580s but was first set down around the same time as the rest of Q2. (66)
(1.) MacD. P. Jackson, "Editing Hamlet in the 1980s: Textual Theories and Textual Practices," Hamlet Studies 11.1-2 (Summer and Winter 1989): 60-72, at 60.
(2.) Harold Jenkins, ed., Hamlet (London: Methuen, 1982), Arden Shakespeare; Philip Edwards, ed., Hamlet, Prince of Denmark (Cambridge: Cambridge University Press, 1985); G. R. Hibbard, ed., Hamlet (Oxford: Clarendon Press, 1987); Gary Taylor, ed., Hamlet, in Stanley Wells and Gary Taylor, gen. eds., William Shakespeare: The Complete Works (Oxford: Clarendon Press, 1986), 735-77; Taylor's responsibility for Hamlet is clear from his initials being appended to the textual introduction to the play in Stanley Wells and Gary Taylor, with John Jowett and William Montgomery, William Shakespeare: A Textual Companion (Oxford: Clarendon Press, 1987), 396-402.
(3.) Jackson, "Editing Hamlet," 60.
(4.) Laurie E. Maguire's Shakespearean Suspect Texts: The "bad" quartos and their contexts (Cambridge: Cambridge University Press, 1996) consolidated a growing skepticism. See Gabriel Egan's chapter, "Intermezzo: the rise and fall of the theory of memorial reconstruction," in his The Struggle for Shakespeare's Text: Twentieth-Century Editorial Theory and Practice (Cambridge: Cambridge University Press, 2010), 100-128. Influential arguments against theories of memorial reconstruction have been put forward by Paul Werstine, notably in "A Century of 'Bad' Shakespeare Quartos," Shakespeare Quarterly 50 (1999); 310-33.
(5.) G. I. Duthie outlined all theories advanced before his own book was written, in The "Bad" Quarto of "Hamlet": A Critical Study (Cambridge: Cambridge University Press, 1941); more recent surveys are by Kathleen O. Irace, ed., The First Quarto of Hamlet (Cambridge: Cambridge University Press, 1998); Ann Thompson and Neil Taylor, eds. Hamlet and Hamlet: The Texts of 1603 and 1623, 2 vols. (London: Arden Shakespeare, 2006), especially the first of these two volumes, 74-94; Terri Bourus, Young Shakespeare's Young Hamlet: Print, Piracy, and Performance (New York: Palgrave Macmillan, 2014), hereafter Bourus; and Zachary Lesser, Hamlet After Q1: An Uncanny History of the Shakespearean Text (Philadelphia: University of Pennsylvania Press, 2015). I quote the Hamlet texts from, and give line-references to, Thompson and Taylor's Arden-3 editions; these are in modern spelling, which I have preferred to use, so as to prevent Q1 (1603) from appearing unnecessarily alien in comparison with familiar modernized texts of Shakespeare's plays, but I have, of course, checked the original quartos and First Folio and note their spellings or variants when relevant. I use the term "report" or "reporting" for the creation of a text without reference to an authentic manuscript.
(6.) Bourus, as in note 5. Her fullest discussion of Nashe's allusion in Menaphon (1589) is at 159-66.
(7.) Eric Sams, "Taboo or not Taboo? The Text, Dating and Authorship of Hamlet, 1589-1623," Hamlet Studies 10 (1986): 12-46; The Real Shakespeare: Retrieving the Early Years, 1564-1594 (New Haven, CT: Yale University Press, 1995), 125-35. Steven Urkowitz, " 'Well-sayd olde Mole': Burying Three Hamlets in Modern Editions," in Georgianna Ziegler, ed., Shakespeare Study Today: The Horace Howard Furness Memorial Lectures (New York: AMS Press, 1986), 37-70; "Good News about "Bad" Quartos," in Maurice Charney, ed., Shakespeare: Revaluations of the Shakespeare Canon (Madison, NJ: Fairleigh Dickinson University Press, 1988), 189-206; "Back to Basics: Thinking about the Hamlet First Quarto," in Thomas Clayton, ed., The "Hamlet" First Published (Q1 1603): Origins, Form, Intertextualities (Newark: University of Delaware Press 1992), 257-91; William Davis, "Now, Gods, Stand Up for Bastards: The 1603 'Good Quarto' Hamlet," Textual Cultures: Text, Contexts, Interpretation 1.2 (2006): 60-89; Charles Adams Kelly, The Evidence Matrix for the 1st Quarto of Shakespeare's "Hamlet" and Echoes & Shadows in the Texts of Shakespeare's Hamlet (Ann Arbor, MI: Triple Anvil Press at Mystery Ridge. 2008); Margrethe Jolly, "Hamlet and the French Connection: The Relationship of Q1 and Q2 Hamlet and the Evidence of Belleforest's Histoires Tragiques" Parergon 29.1 (2012), 83-105.
(8.) Urkowitz, "Burying Three Hamlets," 45.
(9.) Gary Taylor and Rory Loughnane, "The Canon and Chronology of Shakespeare's Works," in Gary Taylor and Gabriel Egan, eds., The New Oxford Shakespeare Authorship Companion (Oxford: Oxford University Press, 2017), 416-602, at 547-48.
(10.) By "Q2/F" I mean the traditional conflated Hamlet. Usually, I quote this "canonical" version (as I sometimes follow others in calling it) from the Thompson and Taylor Arden-3 text based on Q2, but include references to the Arden-3 Folio-based text, where it offers significant variation.
(11.) Bourus, 62.
(12.) Ibid., 61.
(13.) George T. Wright, "Hendiadys and Hamlet," Publications of the Modern Language Association of America 96 (1981): 168-93. Thomson and Taylor call Q1's "the thought and scope of my opinion" "a different example of hendiadys" from Q2's "the gross and scope of my opinion," though Wright might not have allowed it. Q2 reads "In what particular thought to work, I know not, / But in the gross and scope of mine opinion / This bodes some strange eruption in our state" (1.1.66-68). Q1 omits "thought" from the first of these lines, transferring it to the second, where it replaces Q2's "gross." This looks like the result of corruption in Q1, rather than of revision in Q2.
(14.) The plays with the next highest total instances of hendiadys are, according to Wright, Othello (28), Troilus and Cressida (19), Macbeth (18), Measure for Measure (16), and King Lear (15). The same five plays have the most instances when these are expressed proportionally to the numbers of lines, though the rank order changes. In the New Oxford Shakespeare chronology (summarized in Authorship Companion, 384-85), they fall within the period 1602-6.
(15.) This feature of Shakespeare's style around 1600 is discussed by Frank O'Connor, Shakspeare's Progress (New York: Collier Books, 1961), 114.
(16.) As Jenkins comments on the variants between Q1 and Q2 in the last of these passages, in Q1 "The dawn light appears more vaguely on a mountain top than on an eastward hill, and the metaphor is confused when the russet mantle clothes the sun instead of the morn" (23). Personally, I can (just) believe that Shakespeare might have first written "mountain top" and then revised to the greatly superior "eastward hill," but I find it harder to imagine that a Shakespeare capable of creating the rest of the imagery ("in russet mantle clad / Walks o'er the dew")--with its fusion of an early-rising shepherd and Aurora, goddess of the dawn--would have damaged it by making the walker the sun rather than the morn.
(17.) The appearance of the Ghost within 1.42-131 means that there are many short lines, but the percentage of run-on lines among full pentameters that do not end speeches is 40.5 in the Arden-3 edition of Q1.
(18.) E. K. Chambers, William Shakespeare: A Study of Facts and Problems, 2 vols. (Oxford: Clarendon Press, 1930), II, 401, 405.
(19.) Frederick J. Pohl, Like to the Lark: The Early Years of Shakespeare (New York: Clarkson N. Potter, 1972), 143-54.
(20.) W. G. Clark and W. A. Wright, eds., The Works of William Shakespeare, 9 vols. (Cambridge: Macmillan, 1863-66).
(21.) Ants Oras, Pause Patterns in Elizabethan and Jacobean Drama (Gainesville: University of Florida Press); MacD. P. Jackson, "Pause Patterns in Shakespeare's Verse: Canon and Chronology," Literary and Linguistic Computing 17 (2002), 37-46. Oras's prosodic analysis and the lexical and stylistic indicators of chronology employed in the present article are described in the "Canon and Chronology" section of Authorship Companion, 417-602.
(22.) Eliot Slater, The Problem of "The Reign of King Edward III": A Statistical Approach (Cambridge: Cambridge University Press, 1988).
(23.) MacDonald P. Jackson, "Vocabulary Links between Shakespeare's Plays as a Guide to Chronology: A Reworking of Eliot Slater's Tables," Shakespeare 11 (2015): 446-58.
(24.) Marvin Spevack, The Harvard Concordance to Shakespeare (Cambridge, MA: Harvard University Press, 1973). Alexander Schmidt, Shakespeare Lexicon, 2 vols., rev. Gregor Sarrazin (Berlin: Georg Reimer, 1902). Slater's "two-six-fold words" falling within the specified portions of Q1 are: alley (n.), annual (adj.), apparition (n.), arrest (n.), blazon (n.), canonize (v), cap-a-pe (adj.), competent (adj.), conveniently (adv.), dawning (n.), eruption (n.), exactly (adv.), foreknowing (v.), fretful (adj.), gage (v), garbage (n), glimpse (n.), harrow (v.), hearsed (v.), heraldry/heraldy (n.), illusion (n.), implement (n.), impress (n.), invulnerable (adj.), jelly (n.), juice (n.), jump (adv.), observant (adj.), platform (n.), ponderous (adj.), porch (n.), quicksilver (n.), quill (n.), rugged (adj), rasset (adj.), sable (adj.), skirts (n.), smear (v.), stalk (v.), weary (adj.), truncheon (n.). I have included Slater's identifications of parts of speech. He seems, however, to have been muddled over stalk, registering it as a noun that occurs in three plays, but Schmidt rightly gives under separate headings stalk as the stem of a plant (in Richard III and Pericles) and as a stately walk (in Hamlet); however, stalk as a verb links Hamlet with three other plays, and so I have included these in calculations. Slater wrongly includes an adjectival use of dawning "dawning day" in Titus Andronicus among instances of dawning as a noun: I have discounted this instance.
(25.) Thus Troilus and Cressida, Othello, and Measure for Measure together share 19 of their total of 3,461 "two-six-fold words" with Q1: 19 divided by 3,461 and multiplied by a standard 3,500 equals 19.2. Even after this kind of standardization, no consecutive group of play that excludes these three produces a result as high: Macbeth, Antony and Cleopatra, and Pericles, 3-5 come easily the closest with 10 of their 2,347 "two-six-fold words" shared with Q1, a rate of 14.9 per 3,500.
(26.) Textual Companion, 122-23. Relying in part on a persuasive demonstration that John Marston's Antonio's Revenge was indebted to Hamlet, Jenkins concluded that Shakespeare's play, "as it has come down to us belongs to 1601," but that "the essential Hamlet, minus the passage on the troubles of the actors," was certainly being acted on the stage "in the course of 1600" (Arden Hamlet, 13).
(27.) E. A. J. Honigmann, ed., Othello (Walton-on-Thames, Surrey: Thomas Nelson, 1997), Arden Shakespeare, 344-50
(28.) Gregor Sarrazin, "Wortechos bei Shakespeare," Shakespeare Jahrbuch 33 (1897): 121-65, and 34 (1898): 119-69).
(29.) MacD. P. Jackson, Studies in Attribution: Middleton and Shakespeare (Salzburg: University of Salzburg, 1979), 148-58, 211-12.
(30.) Karl Wentersdorf, "Shakespearian Chronology and the Metrical Tests," in Walther Fischer and Karl Wentersdorf, eds., Shakespeare-Studien: Festschrift fur Heinrich Mutschmann (Marburg: N. G. Elwert, 1951), 161-93.
(31.) Sarrazin's table for Hamlet is in Jahrbuch 34: 128-29. In the following list of dislegomena in Hamlet Q1, with line references to Arden-3,1 give the part of speech only where he did so or where disambiguation seems needed: absent (v.) 7.85, attractive 9.78, baker 13.85, cap-a-pe 2.114, down (as in a-down) 13.90, garbage 5.43, gran? 7.129, horridly 4.31 and 7.347, indict 16.47, implements 1.63, journeyman 9.19, knotted 5.13, lewdness 5.40, mason 6.19 and 21, nickname 7.189, nose (v.) 11.143, observant 161, pastoral 7.296, perturbed 5.150, positively 7.97, potently 7.220, repel 6.47, robustious 9.7, russet 1.122, savoury, 7.335, shipwright 1.64 and 16.19, shreds 11.45, sweaty 1.66, tanner 16.80, tennis 6.20, trick (v.) 7.347, trippingly 9.1, ungalled 9.176, wharf 5.28, wince (v.) 9.141. The following are the trislegomena in Q1: altiude 7.321, bark (v.) 5.56, bodkin (7.131), bray (v.) 4.9, calumnious 3.9, cockle 13.17, dirge 16.133, flush 2.64, foreknowing 1.29, german 17.24, glimpse 4.29, hearsed 4.23, insert 7.396, inward (adv.) 9.205, jelly 2.118, joyfully 7.19, jump (meaning "exactly") 1.54, labourer 1.69, redeliver 7.139, springe 3.59, tragedian 7. 264, unbated 17.93, undiscovered 7.121, university 9.68, voucher 16.49 and 50, whip ("move suddenly") 11.112, wrist 6.47.
(32.) Among trislegomena linking Hamlet to Hamlet, Sarrazin lists calumnious, but this occurs only once in Hamlet (including in Q1), once in All's Well That Ends Well, and once in All Is True (Henry VIII); nevertheless it belongs among trislegomena. In his chart for Hamlet he includes nunnery among trislegemona, but, although Shakespeare does not use the word outside of Hamlet, it occurs five times in Q2/F, 3.1, as many as seven times in Q1, scene 7. Sarrazin wrongly listed unction among dislegomena linking Hamlet to Hamlet: it occurs once in Hamlet and once in Romeo and Juliet. He lists redeliver among dislegomena linking Hamlet to Hamlet, but in the Hamlet tables also lists an instance in Measure for Measure under trislegomena.
(33.) One other item of lexical evidence is worth mentioning. Donald W. Foster, Elegy by W.S.: A Study in Attribution (Newark: University of Delaware Press, 1989), 109-10, noted that Shakespeare used "most" with increasing frequency and that the rate per 1,000 words is roughly correlated with chronological order. The lowest play is 3 Henry VI with 0.2 and the highest happens to be Hamlet, with 2.7, just ahead of The Tempest, with 2.6. Hamlet Q1 has 41 examples in what Bourus (39) gives as 15,983 words. This is a rate of 2.6. For the first six plays in the New Oxford Shakespeare chronology, the rates are TGV 0.5, Tit 0.5, 2H6 0.6, 3H6 0.2, Shr 0.3, R3 1.1. The one play before Hamlet in the Oxford chronology that gets up to or above 2.0 is Love's Labour's Lost (1594), which alone is seriously anomalous, with a rate of 2.5. Foster suggested that "In the case of Love's Labor's Lost, the anomalous frequency of most may be partly explained as Shakespeare's lampoon on the mannered language of academe, which was heavily peppered with superlatives" (109). As Foster remarks, Merry Wives, "written almost entirely in the 'low' style, has a smaller frequency of most than should otherwise be expected" (109). Again, the rate of usage of "most" would place Q1 Hamlet next to Q2/F Hamlet and with plays of 1600 onwards, not with Shakespeare's earliest.
(34.) Bourus, 132.
(35.) Ibid., 207.
(36.) Ibid., 212.
(37.) Steven Urkowitz, "Texts of King Lear." Online Posting. 24 May 2017. SHAKSPER: The Global Electronic Shakespeare Conference. Accessed 30 July 2017. https://shaksper.net/archive/2017/790-may.
(38.) Alfred Hart, Stolne and Surreptitious Copies: A Comparative Study of Shakspeare's Bad Quartos (Melbourne: Melbourne University Press, 1942), 170.
(39.) Hart, Stolne and Surreptitious, 214-15.
(40.) Except when quoting Q1 directly, I use the standard spellings of characters' names: Fortinbras rather than Q1's Fortenbrasse, Voltemand rather than Voltemar, Ophelia rather than Ofelia, Rosencrantz and Guildenstern rather than Rossencraft and Gilderstone. I also refer to Q1's Corambis as Polonius.
(41.) Bourus, 116-29.
(42.) Jenkins, Arden Hamlet, 546. His long note on the speech covers 544-47.
(43.) Bourus, 120.
(44.) F's "hore" (hoar) improves the metre (F, 4.3.139). The superfluous syllable in Q2's "horry" (as it is spelt) may have been a compositor's mistake, influenced by "glassy" later in the line.
(45.) Gertrude's opening "O, my Lord," where in Q2/F she begins " There is a willow" could have been recalled from Ophelia's use of the phrase at 2.1.72 or Guildenstern's at 3.2.240 (Q2 line references), while Horatio's "O my dear lord" (Q2, 3.2.52) becomes "O my lord" in Q1 (9.45); Q1's "As it were" (not in the Q2/F version of the speech) may have come from "as 'twere" at 1.2.10, 2.1.13, and 2.1.40); "for a while," where Q2/F has "awhile," from 1.2.191 (Q2) or 2.2.352 (F, not in Q2); Q1's "sweet wretch," instead of Q2/F "poor wretch," from a conflation of addresses or references to the mad Ophelia by the Queen ("sweet lady" Q2, 4.5.27) and King, whose "pretty lady" and "Pretty Ophelia" (4.5.41, 56) become "sweet Ofelia" and "A pretty wretch" in Q1, 13.27, 43).
(46.) Bourus, 102-3, quotes Sonnet 145 as an example of Shakespeare's style before he began "sounding like Shakespeare."
(47.) Instance of "being" in Q1 but not in Q2 are at 2.30, 6.7, 6.23, 7.126, 8.3, 9.236 (in a rhymed couplet), 11.166 (rhyme), 13.61, 14.5, 14.24 (rhyme), 14.26, 15.19, 15.25, 15.49, and 16.123.
(48.) Thomas Clayton has a good brief comparison of the Q1 and Q2 versions of the Queen's account of Ophelia's end in his introduction to The "Hamlet" First Published, 30-32.
(49.) Duthie, "Bad" Quarto, 91.
(50.) Maguire, Shakespearean Suspect Texts, 169. I reviewed Maguire's important book in Modern Language Review 93 (1998): 184-85, suggesting that she may not have taken sufficient account of the ways that various defects that had led the New Biographers to diagnose memorial error in suspect texts combined and interacted.
(51.) Duthie, "Bad" Quarto, 150-68.
(52.) Literature Online (LION), http://literature.proquest.com (accessed 30 July 2017). F, 3.1.48, has "surge ore," which the Arden-3 editors rightly emend (and modernize) to "sugar o'er," with the suggestion that F's reading "could be an error provoked by the possible misspelling 'sugre'." The spelling "suger" was common in the period. Q1's spelling of modern "o'er" is "o're."
(53.) In Q2 the King, telling Hamlet to prepare to travel to England for his "especial safety," adds that "The bark is ready and the wind at help" (4.3.39, 43).
(54.) Duthie, "Bad" Quarto, 186-93; Harold Jenkins, "Hamlet's Voyage," Notes and Queries 226 (1981): 135-36.
(55.) Hamlet Q1, scene 14, has several obtrusive repetitions: "is... arrived" (1), "is arrived" (22); "treason" (4, 10); "fail" (17, 20); "mother's" (18, 33); "being" (24, 26); "blessing" (31), "blessings" (33). While repetition need not in itself indicate textual corruption, here it increases the suspicion, aroused by all the other anomalies, that the whole scene is un-Shakespearean.
(56.) The collocation "love and duty" is used by Laertes in Q1 only, at 2.24.
(57.) Urkowitz, "Back to Basics," 259. Among the varied perspectives on Q1 in The "Hamlet" First Published, Urkowitz's was sharply countered by Sidney Thomas's essay "Hamlet Q1: First Version or Bad Quarto?" 249-56.
(58.) The others were The First Part of the Contention (1594) and The True Tragedy of Richard Duke of York (1595)--versions of 2 and 3 Henry VI, Romeo and Juliet (1597), Henry V (1600), and The Merry Wives of Windsor (1602).
(59.) Hart presents statistics on vocabulary in chapters III-VI of Stolne and Surreptitious, 21-66.
(60.) Both quotations in this paragraph from Hart, Stolne and Surreptitious, 27.
(61.) The evidence for these conclusions is summarized in Authorship Companion, 490-93 and 513-17.
(62.) Bourus, 35-100, where she refers to works cited in notes 4 and 7 above; see also the surveys cited in note 5.
(63.) Tiffany Stern, "Sermons, Plays and Note-Takers: Hamlet Q1 as a 'Noted' Text," Shakespeare Survey 66 (2013): 1-23; John Jowett, "The Writing Tablets of James Roberts" (forthcoming in The Library). I am grateful to Professor Jowett for showing me an early version of this article and also of the article listed in the next note.
(64.) John Jowett, "Whose Hamlet Mocks the Warm Clown?" (forthcoming).
(65.) Some scholars have thought that texts of Early Modern English plays tended to deteriorate through normal theatrical use, without any contribution from failures of memory. Albert B. Weiner, ed., Hamlet: The First Quarto 1693 (Great Neck, NY: Barron's Educational Series, 1962) proposed that "the printer's copy for Q1 Hamlet was prepared from a legitimate abridgement of an early version of the play, and that the abridgement-adaptation was made on Shakespeare's foul papers," any textual corruption having arisen from an illegible manuscript (57). Hardin Craig, A New Look at Shakespeare's Quartos (Stanford, CA: Stanford University Press, 1961) concluded that "Q1 is a somewhat degenerate version of Shakespeare's earliest known play on the theme of Hamlet" (82), corruption resulting from "alterations to the text suffered while it was a repertory play of a traveling company" (81). More recently Paul Metzer, The "Hamlets": Cues, Qs, and Remembered Texts (Newark: University of Delaware Press, 2008), has analyzed in the three Hamlet texts the cues that would have accompanied each of the actors' "parts," from which they learned their lines, and claims that Q1's unknown creators were not aiming to reconstruct Shakespeare's Hamlet, but were engaged on "an independent act of creation," which involved "their memory of earlier versions of the Hamlet tale as well as... Shakespeare's Hamlet"; and that "the manuscript behind Q1 was not a 'performance text' but was solely intended for publication" (21).
(66.) I am grateful to John Jowett for his helpful comments on my first draft of this article.
TABLE 1 Distribution of pauses within substantial passages in which Q1's verse is most nearly identical to Q2's Position: 1 2 3 4 5 6 7 8 9 Number of pauses: 1 4 1 8 9 18 9 2 1 Percentages: 1.9 7.5 1.9 15.1 17.0 34.0 17.0 3.8 1.9 Note: Figures are for pauses after each of the nine syllabic positions within pentameter lines, as marked by punctuation in Q1. TABLE 2 Numbers of Eliot Slater's "two-six-fold words" that verse passages in which Q1 and Q2 are most similar share with chronological groups of Shakespeare plays Group 1 Group 2 Group 3 Group 4 Group 5 Group 6 TGV--R3 Err--MND Jn--H5 JC--Oth MM--Ant Per--H8 1588-1592 1594-1596 1596-1599 1599-1604 1604-1607 1608-1613 9 16 20 28 21 18 6002 5725 6164 5769 5894 5454 Note: Chronological groups and dates are those as in The New Oxford Shakespeare Authorship Companion (2017). The last line shows the number of "two-six-fold words" in each chronological group.
|Printer friendly Cite/link Email Feedback|
|Author:||Jackson, MacDonald P.|
|Publication:||Medieval and Renaissance Drama in England|
|Article Type:||Critical essay|
|Date:||Jan 1, 2018|
|Next Article:||The Boundaries of John Marston's Dramatic Canon.|