"A response to the Arden3 Hamlet" by Gabriel Egan
The current Arden edition of Hamlet offers three fully-edited and modernized texts, based on the 'bad' quarto of 1603, the 'good' quarto of 1605, and the Folio of 1623. Editors used routinely to conflate the early printings to produce a single modernized Hamlet, but since the 1980s the arguments against such a policy have achieved dominance. If we were sure that all three early editions derived from a single version of the play and were different only because of errors made in copying the script, then it would be acceptable for modern editors to produce one modernized edition from these three sources, choosing the best words and lines from each as they saw fit to produce a text as close as possible to that lost original. But because they are not sure that all three early editions derive from a single version of the play--because they suspect that there might have been different versions of the play existing in Shakespeare's life, perhaps because he revised the play--editors respect the independence of each early printing and make a modernized edition of it. Hence the new three-text Hamlet.
But what is the evidence that the three early printings are distinct versions? We can leave aside the bad quarto of 1603 for a moment, since it is much unlike the other two, the Q2 of 1604-5 and Folio of 1623. These two early editions differ in that F lacks about 230 of Q2's lines and Q2 lacks about 70 of F's lines. In the lines they have in common--which is the vast majority of them--there are hundreds of small differences of wording. As far as I can tell, the first person into print with an argument that these differences were not the random variations cause by errors in transmission but where systematic, even artful, alterations was Paul Werstine, who published his findings in an article 21 years ago (Werstine 1988). Werstine asks 'what is it that makes Hamlet apologize to Laertes for his graveside behaviour?'. The apology appears in Q2 and F, and here it is from the Q2 version [SLIDE]. In Q2 this happens because a lord tells Hamlet that Gertrude wishes it, [SLIDE] but this lord is absent in the Folio version and instead there's an Folio-only moment [SLIDE] where Hamlet tells Horatio that he regrets insulting Laertes because he feels sympathy for Laertes because they have both lost a father. Thus the Q2 and F texts give distinct motivations for Hamlet's apology: in the earlier text Hamlet apologizes because his mother tells him to, and in the later text he apologizes, without promptiong from his mother, because he feels sorry for Laertes.
Such a difference might just come about by the random distortions of textual transmission, but it smacks of conscious revision of the script. If we find a whole series of such differences that seem in each case to be best explained as revision, then there is no way to conflate the two versions. If Shakespeare changed the original version in which Hamlet is told by his mother to make an apology into a version in which he decides for himself to make an apology, then a conflation of the two versions makes for nonsense edition that Shakespeare would not have recognized in which Hamlet is told by his mother and he decides for himself to do it. Werstine found a stack of such differences between Q2 and F that seem like conscious revision. Only in Q2 does Hamlet mock Osric's mode of speaking about Laertes (which mockery also necessarily mocks Laertes), so if you combine F's version of 'why does Hamlet apologize' (answer: because he sympathizes with his loss) with Q2's mockery of Osric, you get a contradiction: Hamlet feels sorry for Laertes and then mocks him. Taken separately, each of Q2 and F makes sense of Hamlet's attitude towards Laertes, but conflated they don't. Moreover, looking back over the whole play, the Folio presents a Laertes more deserving of Hamlet's regard than Q2's Laertes. For example, in pleading to be allowed return to France, Folio Laertes is allowed to speak more in his own right while Q2 Laertes is described as rather begging his father to be let go. Upon seeing the mad Ophelia, Folio Laertes generalizes a principle about human nature from the specific instance before him, rather as Hamlet does elsewhere in the play, so Folio Laertes more Hamlet-like than Q2 Laertes. In seducing Laertes to murderous revenge, Claudius in Q2 is much more flattering and also taunting (and more manipulative), and thus Q2-Laertes more the dupe for buying it, than is Folio Laertes, who seems rather to have his own motivations.
Hamlet himself is considerably different in Q2 and F. Q2 has Hamlet aware that Rosencrantz and Guildenstern are to accompany him to England and has Hamlet vow to turn their knavery back on them ("hoist with his own petard"), but F omits this knowledge and vow. Q2 Hamlet has a knack for using his suspicions to smell out what people are up to without being told, as when he first meets Rosencrantz and Guildenstern and correctly figures out that they were sent for. In the Folio, Hamlet doesn't smell them out but rather Rosencrantz and Guildenstern give themselves away by denying that Denmark is a prison and accusing him of ambition, which only a supporter of Claudius would say. Thus Folio Hamlet is less smart at figuring out his enemies than Q2 Hamlet, and hence F lacks the "hoist on his own petard" speech (which is his smelling out of Rosencrantz and Guildenstern's treachery again) and lacks the "How all occasions do inform against me" speech that shows he knows he's being sent to England to die and vows to react against it. In general F show Claudius in doing better in his manoeuvres against Hamlet than Q2 does. Q2's "How all occasions" speech shows Hamlet using the notion of honour to chivvy himself up for revenge, whereas F cuts this but enlarges the later conversation with Horatio in which Hamlet uses religious justification for his revenge ("is't not to be damn'd | To let this canker of our nature [ie Claudius] come | In further evil?"). Regarding the whole trip towards England, the Folio has Hamlet basically ignorant of what's going on, and finding the hand of providence in his being delivered back to Denmark, where Q2 has this be more Hamlet's own doing. In short, F is providential, Q2 secular. As foils, Fortinbras is compared to Hamlet in Q2 and the point is essential secular: it's about how one conceives honour. In the Folio Hamlet is rather more insistently compared to Laertes and the point is essentially religious: it's about being willing to depart from Christian precepts, even to damn oneself eternally, to revenge a wrong.
Oddly enough, having just produced copious evidence that the differences between Q2 and F are meaningful artistic reshapings of detail, Werstine refused to consent to privileging their separateness over "the aesthetic forms that critics have produced from their reading of the combined Q2/F text" (Werstine 1988, 23). He said this because each of Q2 and F on its own has material that threatens the same harm that conflation entails; that is, each has its own internal contradictions too. There is plenty of the religious matter in Q2 as well as F, and in both (not just Q2) Hamlet is canny enough to suspect Claudius and Polonius of moving against him ("where's your father?"). The problem is that we can tie these separate Q2 and F integrities to Q2 and F as historical documents, but we can't tie them to the author Shakespeare because Q2 and F have lots of trivial variants too and have lots of non-trivial variants that don't fit the analysis offered by Werstine. E. A. J. Honigmann showed (and John Kerrigan confirmed) that authorial revision produces minor variants as well as major, and we know that transmission produces minor variants only, and Kerrigan showed that non-authorial revision produces major variants only. So when you have a case of many minor and many major variants (as with Q2 and F Hamlet) you've got too many potential agents to choose from. This is almost a textbook illustration of the problem that Marxists call over-determination: there are too many causes for the number of effects. Scholars would like to tie all the agency to Shakespeare, but as Werstine pointed out a look at extant dramatic manuscripts shows many hands in lots of them and you just can tell who's revising and who's just copy. Where do you turn in such a case? Well, Werstine was writing in 1988 and back then the fashionable person to turn to was Michel Foucault and his unjustly celebrated work on the impossibility of finding authorial origins.
Despite Werstine's refusal to go the last mile and insist that Q2 and F Hamlet are distinct plays, the idea that they are distinct has caught on and Neil Taylor and Ann Thompson's admirable Arden3 Hamlet--that splits the formerly unified play into three versions--is the logical outcome of this insight. In the second half of this talk I want to see what happens if we go further and look at differences within each of these now distinct versions. How could there be differences within each version? Because the way early-modern books were made. I have been speaking of Q1 (the bad quarto), Q2, and F as though each were one thing, but of course they are merely books surviving from the early-seventeenth century. Of Q1 there are in fact just two surviving exemplars--two actual copies of the book--one in California (missing its last leaf) and on in London (missing its title-page). Thus there is no complete book of Q1 and it exists merely as mental construct, a conflation of these two physical objects. Of Q2 there are 7 surviving exemlars, 3 in England, 3 in America, and 1 in Poland. Of the Folio, the most recent census indicates that there are 228 surviving exemplars across the world (West 2002), with the single largest collection being the 79 complete exemplars at the Folger Shakespeare Library in Washington DC. I said that each of these editions is internally varied, and while that is obviously true of the two mutilated exemplars of Q1, one missing its last leaf and one its title-page, this is only an accidental variation: the publisher intended each to be whole. But in the case of Q2 there is another kind of difference between the 7 exemplars, and it happens because early-modern printers would stop the printing press during a run, make corrections to the type and then restart the press. Sheets printed before the correction were not discarded but rather included with the corrected sheets, and hence any early-modern book is likely to be a random mix of uncorrected and corrected sheets. The only way to discover that correction has taken place is to compare each exemplar with every other one and look for differences between them. This process is known as collating for press variants.
Collating the 7 exemplars of Q2 Hamlet, Ann Thompson and Neil Taylor identified 26 press variants (Shakespeare 2006, 479n1, 524-55), 10 of them clustered on one side of one sheet, known as N(outer) for reasons that we need not go into. Listed here according to which of the 7 exemplars contains each combination of variants:
| F, HN, Y2, Wro | thirtie | pall | sellingly | dosie | yaw | neither in | too't | reponsiue | be hangers | sir |
| L | thereby | fall | sellingly | dazzie | raw | neither, in | doo't | reponsiue | be might hangers | so sir |
| C2, VER | thereby | fall | fellingly | dazzie | raw | neither, in | doo't | responsiue | be might hangers | so sir |
Since the 7 exemplars of Q2 show this this part of the book, N(outer), in 3 states there must have been two rounds of correction, and the above table shows Wilson's conjectured order of correction, accepted by Thompson and Taylor, with improvement running down the page Thus the forme was first typeset as shown in the Folger, Huntington, Beinecke, and Wroclaw exemplars and the sheets that ended up in those copies (plus more, presumably) were wrought off. Then the press was stopped and 8 corrections were made: thirtie > thereby, pall > fall, dosie > dazzie, yaw > raw, neither in > neither, in, too't > doo't, be hangers > be might hangers, and sir > so sir. The press was restarted and the sheet that ended up in the British Library copy (plus more, presumably) was wrought off. Then the press was stopped again and two more corrections were made (sellingly > fellingly and reponsiue > responsiue) and the the press restarted to produce, amongst others, the sheets that ended up in the Cambridge and Oxford university exemplars.
Let us look at each of these press corrections in detail and see if you agree that it what Wilson and Thompson and Taylor call the corrected version is indeed the corrected one. According to Thompson and Taylor (Shakespeare 2006, 479), there is clear of improvement in turning thirtie into thereby here:
[CLAUDIUS]
This graue shall haue a liuing monument,
An houre of quiet thirtie shall we see
Tell then in patience our proceeding be. Exeunt
(F, HN, Y2, Wro, sig. N1)[CLAUDIUS]
This graue shall haue a liuing monument,
An houre of quiet thereby shall we see
Tell then in patience our proceeding be. Exeunt
(L, C2, VER, sig. N1)
In the top reading, Claudius might be specifying one hour and then instantly revising it to 30 hours (in modernized form, "an hour of quiet--thirty--shall we see"), which would be unusual but acceptable. If the choice is limited to thereby versus thirtie (that is, so long as the Folio is ignored) a case could be made either way.
The second variant is evenly balanced:
[HAMLET]
Our indiscretion sometime serues vs well
When our deepe plots doe pall
(F, HN, Y2, Wro, sig. N1)[HAMLET]
Our indiscretion sometime serues vs well
When our deepe plots doe fall
(L, C2, VER, sig. N1)
Both readings make perfect sense, so the variant is utterly indifferent. The third variant, involving the praise of Laertes, might at first seem straightforward:
[COURTIER]
to speake sellingly of him, he is the card or kalender
(F, HN, Y2, Wro, L, sig. N2v)[COURTIER]
to speake fellingly of him, he is the card or kalender
(C2, VER, sig. N2v)
The word fellingly has generally been taken as an alternative spelling of feelingly, and Thompson and Taylor's preference for sellingly is one of the stimulating surprises of their edition. The word feelingly was certainly Shakespearian--he used it in As You Like It, King Lear, Measure for Measure, Twelfth Night and Lucrece--but the Literature Online database shows that it was also common in literary writing of the period, appearing in Chaucer's Canterbury Tales and poetry by George Gascoigne and Edmund Spenser, and plays preceding Shakespeare's such as Thomas Kyd's The Spanish Tragedy as well as others of his time and shortly after. In its favour, sellingly occurs nowhere else in Literature Online, so if accepted it seems a genuinely Shakespearian coinage.
The fourth, fifth, and sixth variants are also about Laertes and occur within two lines, so they may be taken together:
[HAMLET]
to deuide him inuentorially, would dosie th'arithmaticke of
memory, and yet but yaw neither in respect of his quick saile
(F, HN, Y2, Wro, sig. N2v)[HAMLET]
to deuide him inuentorially, would dazzie th'arithmaticke of
memory, and yet but raw neither, in respect of his quick saile
(L, C2, VER, sig. N2v)
If we think that neither dosie nor dazzie makes much sense, their equal obscurity renders this variant indifferent. However, for the Arden2 Hamlet Harold Jenkins defended dosie as a rare verb meaning to bewilder, to stupefy (Shakespeare 1982, 5.2.114n). Thompson and Taylor considered the corrected state's dazzie to be an improvement that did not quite go all the way, and they completed the correction by emending to dazzle. To see the variant as it stands being anything but indifferent requires such a subjective preference. The meanings of yaw and raw are also obscure, which of course is the point of the speech: Hamlet is mocking obscure and convoluted courtly affectations. As a nautical term, yaw (meaning to point away from the direction of heading) has at least the merit of agreeing with the metaphorical saile, but again the choice is subjective. The comma between neither and in is entirely indifferent.
The seventh variant is also indifferent:
Hora[tio, to Courtier] Ist not possible to vnderstand in another tongue, you will
too't sir really.
(F, HN, Y2, Wro, sig. N2v)Hora[tio, to Courtier] Ist not possible to vnderstand in another tongue, you will
doo't sir really.
(L, C2, VER, sig. N2v)
The expression to it is implicitly accompanied by the verb to go (so, go to it), which in this context is as acceptable as do it. The eighth variant is straightforward, as one reading seems to be nonsense:
[COURTIER]
the carriages in faith, are very deare to fancy, very reponsiue to
the hilts
(F, HN, Y2, Wro, L, sig. N2v)[COURTIER]
the carriages in faith, are very deare to fancy, very responsiue to
the hilts
(C2, VER, sig. N2v)
As reponsiue is not a known word, the adding of an s seems a clear correction, whereas it is hard to see why the printers would go the other way and change responsiue to reponsiue. (As we shall see, however, a case can be made for it.)
The ninth variant is tricky because we have to weigh what might have been an improvement if it had not been bungled:
[HAMLET]
I would it be hangers till then
(F, HN, Y2, Wro, sig. N3)[HAMLET]
I would it be might hangers till then
(L, C2, VER, sig. N3)
Hamlet is referring to the word carriages that the Courtier (later identified as Osric) had used for the leather and metal straps by which rapiers are suspended from the belt, which Hamlet thinks more appropriate if cannons, not swords, hung at a man's hips. Until they do, he says, let the word be hangers. If the word might were essential to Hamlet's meaning, we could hypothesize a press correction that was intended to put might before be but mistakenly put it after, as indeed Wilson argued (Wilson 1934, 126-27). However, the word might is not essential to the meaning, for the subjunctive mood is already clear from Hamlet's I would. Indeed, one could argue that there is more sense in seeing correction going the other way, from the ungrammatical be might hangers to the perfectly acceptable be hangers. The tenth and last variant on this forme is indifferent:
Hora[tio] This Lapwing runnes away with the shell on his head.
Ham[let] A did sir with his dugge before a suckt it
(F, HN, Y2, Wro, sig. N3)Hora[tio] This Lapwing runnes away with the shell on his head.
Ham[let] A did so sir with his dugge before a suckt it
(L, C2, VER, sig. N3)
Of these 10 variants, then, about half (depending on your point of view) are utterly indifferent and all but one (reponsiue > responsiue) could be argued either way with more or less conviction.
Rather than adopt or reject all 10 readings as a set, Thompson and Taylor accepted Wilson's deduction of the order of correction shown above (Shakespeare 2006, 524-25) but treated 4 of the corrections as mistaken in one way or another. They preferred uncorrected sellingly (5.2.95) over corrected fellingly, they emended corrected dazzie to dazzle (5.2.99), they preferred uncorrected yaw (5.2.100) over corrected raw, and they emended corrected be might hangers to might be 'hangers' (5.2.142). The danger that comes with asserting that apparent corrections in both rounds of alteration were really miscorrections is that it threatens to undermine the basis upon which was made the original determination of the order of correction. If miscorrections are nearly as common as corrections, it becomes hard to understand what the printers thought they were doing and hence there is little certainty about which state preceded and which followed the stop-press alterations.
The usual procedure is to seek at least one incontrovertible case of good correction to establish the direction of change between variant states of a forme. This determining case need not be good in the sense of authorially correct but must at least make sense of the trouble taken to alter the reading. In the present case it is no use appealing to readings of the Folio to help settle the matter, because the Folio may have its readings from Q3, which was a page-for-page reprint of Q2, or from Q4, a reprint of Q3. The evidence for sporadic consultation of Q3 or Q4 by the Folio compositors is summarized by the editors of the Oxford Complete Works (Wells et al. 1987, 396-402). For the variants we are concerned with here on N(outer), Q3's differences from the Q2 exemplars C2 and VER (supposed by Wilson to be the most corrected exemplars of Q2 for this forme) are changes easily made by a compositor modernizing and correcting as he worked--fellingly > feelingly and be might hangers > might be hangers--so it offers no more independent a perspective on the goodness of those readings than F does.
Before accepting Wilson's conjectured direction of correction--that is, his proposed ordering of the exemplars from least to most corrected state of N(outer)--we should ask if any other order makes as much sense. An unlikelihood we should rule out is that one or more of the variants was created by the first stage of correction only to be undone by restoration of the original reading during the second stage. If such restoration of first-setting readings occurred, it must have reinstated exactly the original spacing too, for superimposition of agreed readings from different exemplars shows perfect alignment of the type. This can easily be seen by making clear-plastic photocopies from the published fascimiles of these books and placing one upon another, as described by Randall McLeod (McLeod 1979), or else by superimposing excerpts from the available digital facsimiles within a 'graphics' software package on a computer. Applying this rule against the perfect undoing by the second stage of correction of the changes made in the first excludes all but one of the permutations. This remaining possibility is simply the reverse of Wilson's order and can be seen by reading the 3-row table on page 00 above from bottom to top: C2, VER > L > F, HN, Y2, Wro. This conjectured arrangement has the demerit of requiring that press correction turned the apparently good reading responsiue into the seemingly bad reading reponsiue but the compensatory merit of supposing that it turned the bad English of be might hangers into the good English of be hangers. Also, if Thompson and Taylor's preferred readings are accepted, it has the merit of changing the faulty reading fellingly into the correct sellingly and the faulty reading raw into the correct reading yaw.
With so many variants that could be argued either way, acceptance of Wilson's order of correction must depend heavily upon the supposed impossibility of responsiue getting corrected, or miscorrected, to reponsiue. To challenge the apparently unassailable directionality of reponsiue > responsiue one would instinctively seek support from the Oxford English Dictionary, which surprisingly does contain two occurrences of reponsive. However, these appear as the OED's own misprints for responsive (OED need n. 10c [Additions of 1993] quotation from 1960; sense n. 30 *sense-cells quotation from 1908). Accidents happen, but can we be sure that in Q2 Hamlet the word reponsiue is itself an impossible reading? In context, the sense required is 'answering to', or 'matching': "Three of the carriages in faith, are very deare to fancy, very reponsiue to the hilts" (N2v). The cultural context is French courtly excess and flamboyance. Laertes has been in France earning a reputation for his rapier that Claudius says has been bruited in the Danish court by Lamord of Normandy (L4-L4v), and Laertes's side of the wager is "six French Rapiers and Poynards" (N2v), the carriages of which are in question. Danish national honour is at stake, and that Laertes is now on the French side is restated: "six French swords their assignes, and three liberall conceited carriages, that's the French bet against the Danish" (N3).
Miscalling the hangers carriages is but one in a series of the Courtier's linguistic excesses that cause Horatio to invite him to use "another tongue", meaning plain English. If not, he can expect that Hamlet will have to, as it were, check the footnotes of his speeches to make sense of them: "I knew you must be edified by the margent ere you had done" (N2v-N3). The verbs 'to gloss' and 'to gloze' are cognate (from the French gloser), and the mockery of incomprehensible diction here is a rerun of the earlier mockery of Polonius's euphuism and contorted syntax. In this linguistic context, the carriages 'answering to' the hilts might be reponsiue (not responsiue) if this were a coinage derived from the French noun réponse meaning answer. Lexicons of Early Modern English gives no examples of reponsive and Literature Online's earliest example of the French word reponse appearing in English literature is Cornelius Arnold's poem "The mirror for the year 1755". Arnold uses the word as a shard of French embedded in English to connote elaborate flamboyance, and curiously it too concerns a "prating" and "fribbling" dandy:
A Kerchief white then from his Neck [Death] did lease [steal],
Which gave the Beau a Cold, when, sans reponse,
He shrug'd, his Throat grew sore, could hardly wheeze
(Arnold 1757, 159)
It might be argued that reponsive is unlikely to be a Shakespearian coinage because it is found nowhere else but Q2 Hamlet. But that is also true of the variant sellingly that Thompson and Taylor adopted in place of fellingly, for it too is absent from the rest of English literature. In addition to his many coinages that caught on, Shakespeare minted some that did not, as John Jowett argued is the case with the word inductious in Richard 3's opening soliloquy ("Plots have I laid inductious, dangerous"), a Q1 reading that Jowett preferred over the familar Folio reading of inductions (Shakespeare 2000, 1.1.31 and pp. 378-379).
Works Cited
Arnold, Cornelius. 1757. Poems on Several Occasions. London. Privately printed.
McLeod, Randall. 1979. "The Technique of Headline Analysis, with Application to Shakespeares Sonnets, 1609." Studies in Bibliography 32. 197-210.
Shakespeare, William. 1982. Hamlet. Ed. Harold Jenkins. The Arden Shakespeare. London. Methuen.
Shakespeare, William. 2000. Richard III. Ed. John Jowett. The Oxford Shakespeare. Oxford. Oxford University Press.
Shakespeare, William. 2006. Hamlet. Ed. Ann Thompson and Neil Taylor. The Arden Shakespeare. London. Thomson Learning.
Wells, Stanley, Gary Taylor, John Jowett and William Montgomery. 1987. William Shakespeare: A Textual Companion. Oxford. Oxford University Press.
Werstine, Paul. 1988. "The Textual Mystery of Hamlet." Shakespeare Quarterly 39. 1-26.
West, Anthony James. 2002. The Shakespeare First Folio: The History of the Book. Vol. 2: A New World Census of First Folios. Oxford. Oxford University Press.
Wilson, John Dover. 1934. The Manuscript of Shakespeare's Hamlet and the Problems of its Transmission: An Essay in Critical Bibliography. Vol. 1: The Texts of 1605 and 1623. 2 vols. Shakespeare's Problems. 4. Cambridge. Cambridge University Press.