University of Virginia Library

The Relation Between the Second Quarto and the Folio Text of Hamlet
by
Harold Jenkins

I

Although the older editors of Hamlet preferred to base their texts on the Folio rather than the Second Quarto, it is now twenty years since Professor Dover Wilson demonstrated the superiority of Q2 His view, elaborated in The Manuscript of Shakespeare's 'Hamlet' (1934), was that we have in this, the first good quarto, a text of the play deriving from Shakespeare's foul papers. It is a view that now needs modifying to allow for some use of the bad Q1 by the printers of Q2, but otherwise it has generally been accepted.[1] The present-day editor of Hamlet, therefore, is in no doubt about what his chief authority is to be.[2] But he has the major problems of what use he is to make of Q1 and F. Q1 in this article I propose to leave aside. F, though farther from Shakespeare's autograph than Q2, has somewhere behind it a good manuscript and it can serve to correct some of Q2's many errors. If it could be taken to rest entirely on a manuscript which was independent of Q2: F would also serve to corroborate Q2 where they agree and the editor would know that readings which occur in both could not be rejected unless they could be explained convincingly as common errors. The question of the relation between Q2 and F is therefore a matter of first editorial importance. Professor Dover Wilson, of course, maintained that F was independent of Q2: "It is clear that the 1605 Hamlet or its reprints, the Smethwick quartos, do not at any point come into the pedigree of the F1 text."[3] The opposite

theory, namely that F derived, with modifications, from Q2? had occasionally been put forward, notably by Hendrik de Groot, who held that a copy of Q2 became the promptbook at the Globe and that this promptbook, altered, corrected, and augmented, served as copy for F.[4] But Professor Dover Wilson's skillful analysis convinced all the leading authorities, from whom the following are some representative opinions:

This is the view that has now been challenged by Dr. Alice Walker,[5] who maintains that Hamlet, like five other plays in which the Folio text diverges very considerably from the quartos that preceded, was nevertheless set up in the Folio from a quarto that had been corrected on collation with a manuscript. Her argument as to Hamlet rests on erroneous readings found in both F and Q2; anomalous spellings common to both; and mistakes of punctuation in which F is thought to have been misled by Q2.

If Dr. Walker is right, then the whole editorial position is changed. Agreement between Q2 and F in any particular reading need not constitute a dual authority but may arise from an error in Q2 which the collator failed to notice or at any rate to correct.

The object of the present article is, then, to review the evidence for the theory that F was printed from a corrected copy of Q2.

Two explanations are in order at the beginning. In theory of course F could derive from Q2 indirectly by way of one of the reprints Q3 and Q4; but in fact some of the crucial spellings disappear from Q3. My own inspection confirms Dr. Walker's judgment[6] that any significant resemblances F has with the quartos are greatest with Q2 itself. It is therefore to a comparison between F and Q2 that I here confine myself. For F I have used the Devonshire copy in the Oxford facsimile: the possibility of variant readings in other copies has of course to be allowed for, but this margin of error, when the evidence is considered as a whole, is certainly a small one.

II

Common errors. Dr. Walker reminds us that Professor Dover Wilson lists eleven identical errors in Q2 and F.[7] She rejects one of these, but adds four others. Her suggestion of course is that F took these errors from Q2. The alternative is to suppose that the two texts derived them independently (Q2 directly and F via one or more transcripts) from Shakespeare's own manuscript: either they stood as errors in the autograph or the autograph was twice independently misread. I give the list of fourteen with the Globe line-numbering:

                           

I.	i.94	Q2 desseigne, F designe (for designd)
I.	iii.74	of a most select
I.	v.43	wits
II.	ii.612	Q2 of a deere, F of the Deere (omitting father)
III.	ii.295	Q2 paiock, F Paiocke
III.	iii.18	somnet
III.	iv.121	haire
IV.	v.119	Q2 browe, F brow
V.	i.71	Q2 ô the time, F O the time
V.	ii.29	Q2 villaines, F Villaines (for villanies)
I.	ii.209	Whereas (= Where as)
II.	ii.510	A rowsed (= Aroused)
III.	iii.66	cannot (= can not)
IV.	vii.126	Q2 indeede, F indeed (= in deed)

Little need be said of Dr. Walker's four additions. Even if 'cannot' is an error, which is not certain, she herself admits it likely enough that in these wrong word-divisions two compositors could have made the same mistake. Equally of course such readings could have been in Shakespeare's autograph. Of Professor Dover Wilson's original list, the most striking is the reading of I.iii.74, which is certainly obscure, though the trouble may lie, as Professor Dover Wilson admits,[8] not in 'of a most select' but in 'chiefe' (F cheff) later in the line. Q1 agrees in both readings. This may therefore be a case where a good text has been contaminated by a bad; but it is also possible that the three texts agree because their readings are the right ones, though we do not understand what Shakespeare meant. Malone and the N.E.D., by variously glossing 'chiefe', have tried to explain the line as it is. In V.i.71, 'ô', to represent the grave-digger's breath-pause,[9] is anomalous only in comparison with 'a' three times in this line and the next, and could easily go back to Shakespeare. So of course could 'wits', 'haire' and 'browe'. The plural form of the first, echoed by the singular 'wit' in the next line, seems to have been influenced by the 'gifts' with which it is associated, but is as likely to have been due to the author as to a compositor. 'Haire', subject of two plural verbs, and 'browe', governed by 'betweene', are both unsatisfactory in the singular; but the objection is in each case to grammar only, not to sense, and I see nothing beyond what Shakespeare could in haste have written.[10] One must not give too much weight to such anomalies. "The Q2 passage may well be an unresolved tangle", explains Dr. Walker herself on one occasion,[11] and we have to allow for the presence in Shakespeare's foul papers of a number of these, which may have been transmitted independently to F. F's 'designe', 'Paiocke', 'somnet' and 'Villaines', all indubitably wrong, are more difficult to explain. But e for d, i for c, mn for mm, in for ni are very easy misreadings, and if Shakespeare's handwriting led the Q2 compositors to make them, as it apparently did, it may easily have given rise to the same errors in the scribe who copied Shakespeare's foul papers for the playhouse-as Professor Dover Wilson indeed suggests.[12] One might, it is true, have expected 'designe' and 'somnet' to be corrected before they reached F, but it should be noted that 'Villaines' is satisfactory to sense and condemned by metre only. As the most significant identical error we are left with the omission of the word 'father', which Q1 supplies, at II.ii.612. The capital for 'Deere' in F shows that this word was mistaken for a noun and that 'father' was therefore absent from the F copy; but the erroneous substitution of the definite for the indefinite article in F might be against its dependence on Q2 even here.

Altogether the identical errors on which one might argue F's dependence on Q2 make a very small bag for a text of the length and difficulty of Hamlet. Nor can it be much enlarged from the possibles that Dr. Walker from time to time suggests: I.ii.198, wast, where Q1 reads 'vast'; I.iii.130, bonds ( ? bawds); II.ii.397 hand saw (F Handsaw); III.ii.262, mistake, where Q1 has 'must take'; III.ii.269, ban, where Q1 has 'bane'. Of these 'mistake' has been well defended[13] and all are certainly defensible. F's running 'Handsaw' into one word does not suggest copying from Q2

Anomalous spellings. Among the spellings in which F agrees with Q2, Dr. Walker finds a score or so which are anomalous either in relation to contemporary custom or to the Folio practice elsewhere. Some of these do not seem to me to be anomalous at all, viz: I.i.55, Q2 ont, F on't; I.ii.204, Q2 distil'd, F bestil'd; II.i.99, Q2 adoores, F adores;

III.ii.117, Q2 O ho, F Oh ho; III.ii.144, Q2 for ô, for ô, F For o, For o. F's 'adores' is due to a misreading, but a compositor could have committed this quite as easily in working from a manuscript as from Q2.

I see nothing odd that cannot have descended from Shakespeare's original manuscript in F's 'Illo, ho, ho . . . Hillo, ho, ho' (I.v.115-6) or 'ennactors' (Q2 ennactures, III.ii.207). Slightly, but not very much, more significant, may be I.i.40, of (for 'off', which some copies of F have); I.i.73, brazon; IV.v.206, Q2 colaturall, F Colaterall; V.i.310, Q2 cuplets, F Cuplet. Whatever significance attaches to 'smot' (I.i.63) and 'sent' (= scent, I.v.58) comes from their occurring also in Q1, from which it is suggested that Q2 may have taken them. But both are common spellings in the period, and 'sent', which is the F spelling of the corresponding noun in The Shrew and Twelfth Night, may be regarded as normal. For 'smot', however, it is to be observed that it occurs in the same line as 'pollax' (see below).

Interesting certainly are II.i.3, Q2 meruiles, F maruels (for 'marvellous'); and IV.v.100, Q2 impitious, F impittious. But if the odd Q2 spelling is Shakespeare's own, this could equally well have been retained by a playhouse transcriber of his autograph and so transmitted to give the F readings without any contamination from Q2. (Cf. 'adores' above.) Dr. Walker notes, however, that among 20 other instances of marvellous in A, 'maruel's' occurs but once and that when the F spelling derives from Q. The F spelling 'impetuositie' in Twelfth Night is less pertinent.

The striking common spellings seem to me to be the following:

• I.iii.73 ranck, with a c unique in F in the spelling of the noun and rare even in adjectival use, though its occurrence is to be noted in The Merchant of Venice and Henry VIII;
• II.ii.531 ore-teamed, which contrasts with 10 instances of 'teem(ing)' in F;
• II.ii.566 dosen, instead of 'dozen', which is found in all 33 other instances of the word in F;
• II.ii.578 fixion, contrasting with 'fiction' in Twelfth Night and Timon;
• V.ii.322 how, as a cry, which, though normal enough in Elizabethan spelling, tends in F to be replaced by 'hoa' (see, e.g., Hamlet, III.ii.57; III.iv.22,23; IV.iii.16);

and most remarkable of all,

• I.i.63 pollax (F Pollax), for 'Polacks'.

This last seems inexplicable as a reading of any manuscript and has to be attributed to a phonetic rendering of something said and heard but not understood. It occasions no surprise in the reported text Q1, from which it is reasonably held to derive in Q2 and to have been transmitted through to F.

To what is suggestive about these spellings Dr. Walker adds the not negligible weight of two anomalous apostrophes common to Q2 and F: I.v.162, can'st; I.v.167, dream't. These in themselves are not remarkable, but since they occur also in Q1, that may have been their source, in which case they must have been transmitted to F through Q2. There seems, then, enough evidence from spelling to suggest that some, though not necessarily extensive, use was made of Q2 in preparing or in printing the text of F.

Punctuation. The errors in punctuation common to Q2 and F are, again, not numerous, and the three examples to which Dr. Walker gives prominence could find other explanation. They could derive independently in each text from Shakespeare's practice of underpunctuating, but it would be foolish to deny any significance to them and especially to the second. It is Dr. Walker's view that in all three cases "the F1 compositor either followed or was led astray by, the pointing in Q2."[14]

• I.ii.17 Q2 'Now followes that you knowe young Fortinbrasse. . .' F's introduction of a comma after 'followes' and no point where one is required,after 'knowe', shows an obvious misunderstanding.
• II.ii.420 Q2 'nor Plautus too light for the lawe of writ, and the liberty: these are the only men.' It must have been some such unintelligible punctuation as this that led the F compositor to print 'nor Plautus too light, for the law of Writ, and the Liberty. These are the onely men.'
• III.i.6off. Q2's omission of pointing after 'sleepe' in line 60 and 'die' in line 64 is repeated in F.

Stage-directions. Dr. Walker also notes one or two correspondences between Q2 and F in stage-directions. There is nothing much perhaps in F's retention, at the opening of I.ii, of Enter Claudius, [F Claudius] King of Denmarke, Gertrad [F Gertrude] the Queene. A little stronger-cumulatively at least-are: I.i.18, Exit Fran.; III.iii.26, Exeunt Gent.; and IV.i.31, Enter Ros. & Guild. The striking direction at I.v.149, Ghost cries under the Stage, probably came from Q1, but the irregular entry for Osric at V.ii.360 may have originated in the autograph. The F speech-heading Qu. at III.ii.238 when the Player Queen's other speeches are in F headed Bap(t). is clearly due to an oversight in the copy; but it permits no deduction about what the copy was.

Even after Dr. Walker's very thorough scrutiny of the texts, it is possible that a few further significant resemblances may be detected. It is very unlikely that she has overlooked either the orthographic interest of 'soop- stake' (F Soop-stake) at IV.v.142 and 'vnsinnow'd' (F vnsinnowed) at IV.vii.10 or the probability that in both cases the Shakespearian spelling

has simply been retained.[15] She hints herself at further instances in F of punctuation derived from Q2, and one may note such minutiae as the comma after 'vs' at I.i.25, the absence of a comma after 'them' at I.ii.258, and the period after 'loue' at I.v.23. None of these is surprising, but none is quite satisfactory, and they may all have been taken by Q2 from Q1. In addition to the stage-directions noticed above there might be significance in some of the following: I.i.1, Enter Barnardo, and Francisco, two Centinels (F omits the commas); II.ii.169, Exit King and Queene (F & Queen); IV.v.96, A noise [F Noise] within (set up in each case to the right of the text); V.i.1, Enter two Clownes (plus the regular use of the speech-heading Other for the second).

Altogether it does not seem possible to deny that F may depend in some measure upon Q2 and that an occasional reading may therefore occur in F simply because it was already present in Q2. What a survey of the resemblances between Q2 and F does not justify is a whole-hogging theory that Q2, however much corrected, rather than a manuscript served as the principal copy for F.

III

Against such a theory stand at once two highly significant facts, one acknowledged by Dr. Walker and one ignored. No clear typographical links between Q2 and F have been discovered and considerable passages in Q2 are omitted from F. These omissions, mostly explicable as theatrical cuts and natural enough in a playhouse manuscript prepared for performance,[16] are difficult to account for if the printers of F, which aimed at giving Shakespeare's plays "perfect of their limbes", were working from a copy of Q2 in which these passages were included. Moreover, while Dr. Walker has painstakingly collected evidence which would support her case, she has neglected items, even in her own categories of spelling and punctuation, which go against it.

Divergences between Q2 and F need perhaps even more careful handling than similarities. Anomalies common to two texts can sometimes establish that one was printed from the other; but in the nature of things it is difficult to prove the negative. This is especially so when divergences can be explained as (1) emendations made by a corrector of Q2 based upon the manuscript with which he was collating, as well as (2) the inevitable errors and alterations of the F compositors. Even so the divergences

between the two are very great. It may be thought that Dr. Walker relies too heavily on compositors' errors as explanations of corruptions; and there are, I think, many features of the F text which the two causes of divergences are together insufficient to explain.

Erroneous readings. If the F text was in fact set up from a corrected copy of Q2, then the manuscript used by the hypothetical corrector of Q2 was certainly at one and possibly at more removes from Shakespeare's autograph; and it had also undergone alteration by or for the players. It had therefore plenty of opportunity for the corruption and vulgarization which have given in F so many inferior readings. But the question that arises is how far such corruptions would have been transferred by the collator to the copy of Q2 on which he worked. It is a necessary postulate of Dr. Walker's theory that the corrector sought mechanically to bring the Q2 text into line with the manuscript he was using, exercising no editorial discrimination; but granted that the aim was to secure a better text for F, I find it difficult to accept correction of so extremely mechanical a pattern as it would be necessary to infer.[17] A mechanical corrector might conceivably have introduced such nonsense as the notorious 'or Norman' for 'nor man' at III.ii.36 or 'our Nation' for 'the Nation' at IV.vii.95-if indeed these are not attributable to compositors' errors-perhaps even 'their corporall' for 'th'incorporall' at III.iv.118. But was 'inobled' deliberately introduced for the 'mobled' Queen? Without Q2 to guide him either a transcriber or a compositor might easily have so misread a manuscript copy; but, with 'mobled' twice in front of him in print, one would have expected a collator to pause before insisting on this reading and a compositor not to make this error three times over. Other nonsensical readings which could have arisen from the misreading of manuscript copy but which a corrector might have been expected to refrain from introducing into Q2 include: II.ii.580, warm'd (Q2 wand, = wann'd); III.i.48, surge (Q2 sugar); III.i.99, then perfume left (Q2 their perfume lost); IV.iii.7, neerer (Q2 neuer); IV.vii.143, I but dipt (Q2 that but dippe); IV.vii.156, commings (Q2 cunnings); IV.vii.183, buy (Q2 lay). It would have to be a very mechanical correction indeed which substituted readings such as these. And again one cannot well impute them all to simple errors of a compositor working from a printed copy. Some of them might of course have arisen from compositor's errors which, as we know often happened, a proof- corrector emended by guesswork without reference to copy. The complete collation of the Folger Folios which is now being undertaken might throw some light on this. But in the present state of my

knowledge I see no reasonable explanation of F's numerous errors of this order on the assumption that F was set up from Q2.[18]

Stage-directions and speech-headings. In the matter of stage- directions, although a few small similarities have been noticed, the most obvious thing that emerges from a comparison of Q2 and F is the remarkable nonconformity of the two texts. A corrector of Q2 who achieved this result would have had to carry out a very thorough-going revision. This of course cannot be ruled out, for something of the kind was apparently done with Q6 of Richard III, if scholars are right in holding that F printed from that quarto.[19] But it is worth noting that the Hamlet alterations embrace a number of quite pointless variations. Among these I cite: I.ii.159, Q2 Enter Horatio, Marcellus, and Bernardo F Enter Horatio, Barnard, and Marcellus; II.ii.221, Q2 Enter Guyldersterne, and Rosencraus, F Enter Rosincran and Guildensterne; III.ii.52, Q2 Enter Polonius, Guyldensterne, & Rosencraus, F Enter Polonius, Rosincrance, and Guildensterne. In the last two a transcriber might easily have given Rosencrantz the priority normally accorded him elsewhere, but a deliberate transposition of the Q2 order, either by collator or compositor, is unlikely. At II.ii.39, Q2 has Exeunt Ros. and Guyld., and it is difficult to see why a corrector or a compositor with plenty of room should have wished to replace this explicit direction with the single word Exit. I do not think the correction of Q2 from a manuscript is a sufficient explanation of these changes.

A comparison of speech-headings yields similar results. My own fragmentary observations confirm the conclusion of Dr. Philip Williams that in plays set from quartos the Folio compositors tended to be guided by the speech-headings in their copy.[20] But of course the speech-headings of a quarto might reach the F compositors already heavily altered by a collator, as Richard III, again, suggests. So perhaps not too much should be made of the discrepancy in III.iv. and IV.i. of Hamlet, where Q2 heads the Queen's speeches, with one exception, Ger., while F sticks steadily to Qu. The replacement of short by longer forms of speech-heading requires a different explanation. A compositor familiar with the play, if we could assume such, might, I suppose, have substituted such longer forms as

Polon. for the Pol. which Q2, after an initial Polo., invariably uses. The occurrence of Polon. in F in I.iii, II.i and II.ii coincides with the work of compositor A, who always uses this form, while B overwhelmingly prefers Pol. [21] But Polon. turns up in one small block (III.ii.390 ff.), as well as sporadically elsewhere, in the work of B; and it is probable that this form of the speech-heading was the one commonly found in the copy, which A, according to his wont, tended to follow more exactly. Rosin., rather than the Ros. of Q2, would seem to have been the normal speech-heading for Rosencrantz in F's copy:[22] its use in F is quite consistent except for IV.ii.5, where Ro. occurs for reasons of space, and it runs through the work of both compositors. F's single instance of Marcell. at I.v.148, in a scene where both texts are otherwise content with Mar., is, on the theory that F was set up from Q2, more than a little odd.

Punctuation. A consideration of the punctuation provides more clues. It is true that there are passages showing a high degree of correspondence between the two texts. Dr. Walker has cited the 'To be or not to be' soliloquy, and another example might be the speech of Claudius which ends IV.iii. But the opposite situation is notorious in the 'What a piece of work is a man' speech, and whatever one's view of the alternative methods of punctuating this, the one thing that is clear is that the F punctuation did not derive from Q2.

It is Dr. Walker's observation of the Folio compositors that they "normally reproduced the majority of the parentheses in their quarto copy".[23] This does not suggest that their copy for Hamlet was Q2, where, discounting passages of text which are not present in F at all, I count eleven instances of parenthesis and find as many as nine of these not reproduced in F. Yet the F Hamlet gives plenty of evidence of compositor B's recognized penchant for parenthesis, which occasionally leads him into error. A good example occurs at II.ii.140 with '(my yong Mistris)', which is absurdly taken for a vocative. Yet in the same speech this compositor missed an obvious opportunity in the line which Q2 points

(II.ii.133) His omission of any point after 'that' suggests that he never even understood that the second half of the line was parenthetical. He can hardly have had Q2 before him either then or when he failed to recognize the parenthetical phrase at III.i.192: Q2 F These examples go against those three which Dr. Walker cites where F's erroneous punctuation seems to derive from Q2. And there are many similar ones. In the following cases F has been led into wrong, or at least inferior, punctuation precisely through not following Q2, and the natural inference is that Q2 did not serve as copy:                

I.ii.202	Q2	Goes slowe and stately by them; thrice he walkt		F	Goes slow and stately: By them thrice he walkt,
II.iv.56	Q2	With thoughts beyond the reaches of our soules		F	With thoughts beyond thee; reaches of our Soules
II.i.41	Q2	Marke you, your partie in conuerse,		F	Marke you your party in conuerse;
II.ii.145-6	Q2	she tooke the fruites of my aduise: And he repell'd, a short tale to make,		F	she tooke the Fruites of my Aduice, And he repulsed A short Tale to make,
IV.iii.24-6	Q2	your fat King and your leane begger is but variable seruice, two dishes but to one table, that's the end.		F	Your fat King, and your leane Begger is but variable seruice to dishes, but to one Table that's the end.
IV.v.112	Q2	Where is this King? sirs stand you all without.		F	Where is the King, sirs? Stand you all without.
IV.vii.45-6	Q2	to see your kingly eyes, when I shal first asking you pardon, there-vnto recount		F	to see your Kingly Eyes. When I shall (first asking your Pardon thereunto) recount
IV.vii.58-9	Q2	And how should it be so, how otherwise, Will you be rul'd by me?		F	as how should it be so: How otherwise will you be rul'd by me?

The third and fourth examples in this list are spoken by Polonius, whose involved syntax frequently caused the F compositors trouble, in which the lightly punctuated Q2 gave little help. But I judge it to have offered more guidance in these two instances than F availed itself of. Again, of course, one must not neglect the possibility that an erroneous F reading might have arisen in proof-correction when a corrector tried to tidy up some compositorial muddle without reference to copy; and this should particularly be allowed for when, as in the second and fifth examples, an eccentric spelling ('thee' for 'the', 'to' for 'two') points to a likely source of confusion. But it would be very surprising if the proof-correction could be shown to account for all the misunderstandings I have cited.

Two further passages deserve more detailed consideration. The first is I.iii.8-10:

   

Q2	lasting, The perfume and suppliance of a minute No more.
F	lasting The suppliance of a minute? No more.

Here the F compositor was baffled by his copy as he should not have been if working from Q2. His running of 'No more' into the same line looks like an attempt to eke out the metre consequent on the dropping of 'perfume and', which was presumably absent from F's copy. The omission of the comma, which suggests that 'the suppliance of a minute' was taken to be the object of 'lasting', also supports this. The other passage, IV.vii.101-3, is also one in which F has an omission:  

Q2

If one could match you; the Scrimures of their nation He swore had neither motion, guard, nor eye, If you opposd them; sir this report of his

F

If one could match you Sir. This report of his

The dropping of two lines is not in itself remarkable. But the usual explanations-a deletion in the copy or a compositor's oversight-are not here available. The attachment of 'Sir' to the sentence which precedes instead of that which follows the omission suggests that the printer's copy itself ran straight on. I think, then, that a study of the punctuation does not assist the view that F was in general set up from Q2.

Spellings. Divergences of spelling are difficult to argue from. Abnormal spellings common to two texts-and Dr. Walker, as we have seen, has revealed a few pertinent ones in Hamlet-may well suggest dependence. But since the compositor was under no obligation to follow the spelling of his copy, the converse is not true. Yet there is one kind of word in which the compositor would normally be guided by his copy. I refer to proper names. Not all of these of course are significant: a compositor will have his own way of spelling Gloucester[24] and the names of other English earls or counties, as well as those of the classical figures who throng the Roman plays. But there are other names which must have been strange to the compositor, who would approach them without predilection. Hamlet has a number of these. The most striking are, to adopt

the modern spelling, Rosencrantz and Guildenstern. These were well-known Danish names in the sixteenth century, and their possible occurrence in the ur-Hamlet and their pre-Shakespearian history in England do not concern us here. What does is the consistent orthographic divergence of Q2 and F. Q2 invariably prints 'Rosencraus', F normally (II times) 'Rosincrance'. The F exceptions-'Rosincrane' three times and 'Rosincran' once-may safely be regarded as erroneous, and in any case they come no nearer than the usual F form to the practice of Q2. Even if one could presume a corrector who changed u to n throughout Q2, I should find it impossible to believe in one who regularly replaced the medial e by i or the final s by ce. The medial i in F also runs consistently through the speech-headings (42 of them). The name of Rosencrantz's companion shows a like discrepancy: it is regularly spelt 'Guyldensterne' in Q2, 'Guildensterne' in F. The variants 'Guyldersterne' (twice) in Q2 and 'Guildenstern' (twice) in F are insignificant. What is significant is that, apart from two instances of i in Q2 (III.ii.376; IV.i.31), the variation between y in Q2 and i in F is consistently maintained throughout text, stage-directions and speech-headings. There are over forty instances.

The F spelling of these two names is, then, consistently different from that of Q2. And it cannot be attributed to a compositor, for both the compositors who worked on the F Hamlet followed exactly the same practice. The inference is inevitable that when they set up 'Rosincrance' and 'Guildensterne', together with speech-headings Rosin. and Guil. or Guild., they were following their copy. And therefore their copy could not have been Q2. This single piece of evidence seems to me conclusive.

Other rare names in Hamlet give confirmation. 'Fortinbrasse' or 'Fortenbrasse' in Q2 becomes 'Fortinbras' invariably in F, both compositors again being involved. F's substitution of 'Gertrude' for 'Gertrard' may be accounted mere normalization, but since it too is quite consistent with both compositors, it cannot be ignored. Q2 always spells 'Elsonoure', but in all three instances in II.ii, shared between the two compositors, F prints the name 'Elsonower'. It is true that at I.ii.174 compositor A spells it 'Elsenour', but even here the medial vowel and the absence of the final e do not suggest that the spelling derives from Q2. Osric, either in full or in abbreviated speech-headings, is in Q2 eight times spelt with a t- 'Ostr(ick(e)'-before appearing twice without it; but 'Osr(icke)' without the t is always used by both compositors in F.[25]

When a name is confined to that part of the play set up by one of the

F compositors, the significance of divergent spellings lessens but does not vanish. The 'Reynaldo' of Q2, for example, is 'Reynoldo' in F in the one scene in which he appears; and since F's speech-heading is always Reynol., the crucial middle vowel is vouched for 17 times. 'Pirrhus' in Q2, though it has the variants 'Pirhus' and 'Phirrhus', retains the i in all 10 instances; in F it is invariably 'Pyrrhus'.[26] Q2's 'Ieptha' appears in F in all three instances as 'Iephta'. With Voltemand there is inconsistency in both texts: but where F spells 'Voltemand' in I.ii, Q2 has 'Valtemand', and when Q2 does spell 'Voltemand' in II.ii, F shifts to 'Voltumand'. The coincidence of 'Pollax' (Q2 pollax) at I.i.63 is offset by divergences elsewhere: for the Q2 'Pollacke' or 'Pollack' (II.ii.63, 75; IV.iv.23; V.ii.387), F has 'Poleak' and 'Polake'. Proper nouns or adjectives which have but a single occurrence include Q2 Nemeon, F. Nemian (I.iv.83) and Q2 Lamord, F Lamound (IV.vii.93); and, since it was apparently mistaken for a proper name, one might add Q2 Mallico, F Malicho (III.ii.149). Against all this there is not a single coincidence in the spelling of proper names throughout the play which can be held significant.

IV

Dr. Walker's expert scrutiny of the two texts has revealed, as we have seen, a number of resemblances between them. And when due allowance has been made for those which need not have the significance which she attaches to them, enough remain to make it probable that, in the preparation of F, some use was made of Q2. Such a conclusion is not in itself a new one; scholars have sometimes fallen back on the explanation that the printers of F occasionally "consulted" Q2. Dr. Walker ridicules such an explanation as lacking in logic and realism. It is also very inconvenient, since, if neither the nature nor the extent of consultation can be defined, it leaves the position of the modern editor hazy and insecure. But this he may have to put up with. What is illogical, of course, is the easy supposition that a handful of obvious common errors may be explained as due to consultation at the same time as F and Q2 are held to be otherwise independent. As Dr. Walker has so well insisted, in an important clarification of textual theory, to admit consultation in a few instances is to admit its possibility at any point in the text. For it is not to be supposed that every time Q2 was consulted what was taken from it would be a questionable reading; many readings thought to be above suspicion may have come

into F from Q2. In fact the door is open for a hypothesis of conflation on a considerable scale; and an editor will do well to heed Dr. Walker's warning that "the Folio will not serve as . . . an independent witness to the correctness of readings where the two texts agree".[27]

But Dr. Walker's theory that the actual copy for the Folio Hamlet was a corrected Second Quarto must clearly be rejected.[28] And although this may seem, so long as one accepts that Q2 was used at all, to leave the editorial position materially unchanged nevertheless where so much is uncertain, even the disproving of a hypothesis is something gained.

What remains uncertain is exactly how Q2 was used for F. Professor Duthie has brilliantly shown how Q2 of Romeo and Juliet was set up from a copy containing manuscript and printed leaves,[29] and Dr. Philip Williams has recently argued that a similar composite copy lay behind the F King Lear. [30] In his view the printers set up Lear from a transcript of a promptbook which consisted of a copy of a quarto with some of the leaves replaced by leaves of manuscript; and he suggests that the copy for the F Hamlet may have been of the same kind. I do not know what further investigation may reveal, but I have found no evidence of this. Although significant resemblances between F and Q2 are greater in some passages than in others, notably in I.i.1-94 and II.ii.420-612, these passages also contain significant divergences and it does not seem to me that limits of text can be defined within which F does or does not depend on Q2.

The probability, as I see it, is that when the printer's copy for the Folio was being got together, Heminge and Condell were not satisfied with the Hamlet quarto and, notwithstanding Jaggard's supposed preference for printed copy, supplied a manuscript version. In that case one may tentatively suggest that either the scribe who made a transcript for the printer (Dover Wilson's scribe C) or someone in the printing-house itself made reference to the quarto. The first seems to me the more likely. For everything goes to show that the printers were normally content to work from the copy that was supplied them and make the best they could of it. But I see no difficulty in supposing that a scribe who was charged with preparing a transcript of a manuscript might have a copy of the quarto at hand, or even open, in case of need. How much he may have used it will not be easy to determine.

Notes