University of Virginia Library

The Compositors of Hamlet Q2 and The Merchant of Venice
by
John Russell Brown [*]

In The Manuscript of Shakespeare's Hamlet (1934), Professor J. Dover Wilson made one of the first attempts to analyse the habits of a compositor in an Elizabethan printing house, and his conclusion was that James Roberts employed a bungling and inexperienced workman for the 1604/5 Hamlet. The picture Professor Wilson drew of this compositor was detailed and paradoxical. He was a novice and, left to himself, would have worked letter by letter, reproducing the spellings of his copy, but he was rushed in his work, and sometimes omitted lines, half-lines, phrases, words, or letters. In moments of confidence, he abandoned his plodding method and let his eye 'race ahead of his fingers,' so that he might transpose a word in error from the end to the beginning of a line. This picture of Roberts' compositor has been generally accepted and has greatly influenced the repute of the second quarto of Hamlet; Sir Walter Greg thought it "was printed from Shakespeare's autograph copy, though there was little that was foul about it and the chief trouble is the incompetence of the printing."[1]

Since Professor Wilson's early study, other techniques have been developed in order to trace the work of compositors. The acquisition of these techniques has been recorded in the pages of The Library and Studies in Bibliography, and the first large scale application of them has appeared in Dr. Alice Walker's Textual Problems of the First Folio (1953). The trail has been clearly blazed: editors of Elizabethan texts should no longer speak vaguely of compositorial error or sophistication. Many old textual

problems now need to be re-examined, and among them is the nature of the copy for Q2 Hamlet (1604/5). This paper is a beginning of this reexamination; it presents evidence for believing that two compositors were responsible for Hamlet, the same two who shared the work for The Merchant of Venice which was printed by Roberts for Thomas Hayes in 1600.

I

Professor Fredson Bowers has already questioned whether Q2 Hamlet was set by one compositor who was rushed in his work. From a preliminary analysis of skeleton-formes used for this quarto, he has shown that

Later, he was more definite: "This general alternation involving the use of four skeleton-formes is inexplicable for printing with one press; yet if we hypothecate two presses it follows that there must have been more than one compositor."[3] Fortunately, the evidence for two compositors suggested by the presswork can be made specific from the results of a spelling test. The work of these two men is distinguished by the following spellings:                              

Compositor X.	Compositor Y.
their	theyr/their
deare, dearely, etc.	deere, deerely, etc.
sweete	sweet
farwell	farewell
sayd(e), etc.	said(e), etc.
houre(s)	howre(s)
madam(e)	maddam(e)
being	beeing
receaue, receaued, perceaue, etc.	receiue, receiued, perceiue, etc.
honor(s), honor'd, dishonor, etc.	honour(s), honour'd, dishonour, etc.
mooue, mooued, prooue, etc.	moue, moued, proued, etc.
sodaine	suddaine
choise	choice
reuendge, reuendgeful, etc.(There are 3 Y forms in Act I)	reuenge, reuengeful, etc.

In addition, words which in modern English terminate in - consonant+ow and -consonant+ew and are pronounced similarly to know and knew, are spelt by Y, almost exclusively, -ow and -ew, whereas X has a large number of -owe and -ewe forms. The evidence may be tabulated; and from the evidence of these tables, it appears that the two compositors divided the work as follows:
   

Compositor X	B1-D4v,F1-4v,I1-4v, L1,L4v,N1-O2
Compositor Y	E1-4v,G1-H4v,K1-4v,L1v-4,M1-4v

It is noteworthy that Y started work when the new skeletons were constructed. The evidence is reasonably clear except for H3v, L1, and N1-1v. Some further spellings support these allocations. It is probably safe to assume

that H3v was set by Y because the form 'musique', occurring twice, is only found elsewhere on E2 and H4 (twice), both Y pages; 'musickt' is also found on a Y page (G3), but 'musicke' is restricted to I4 and O2 pages set by X. The allocation of N1 and 1v is uncertain because 'their' was set by both X and Y, and the houre/howre distinction is not very clear. Here the spelling leasure helps; it is found on D1 (set by X), N1v (twice) and N2, and in Roberts' 1600 quarto of The Merchant of Venice which was probably set by the same two compositors as his Hamlet, this is characteristic of X, Y using the form leysure.[4] The allocation of L1 is the most difficult of all; with L4v, it is an exception to the division of the work by sheets,[5] and deare is the only word on it which is usually significant.[6] But there is further support for the X allocation in 'noise' which occurs twice on this page, whereas on the one immediately following (L1v) 'noyse' is found twice. Elsewhere in the play, 'noise' is only found on pages set by X (viz. I2v, and O1v), whereas 'noyse' is found only on pages set by Y (viz. G4, K1v and M1). Noise/noyse strengthens the case sufficiently to justify assigning L1 to X on the basis of the available evidence.

The evidence of the spellings for Hamlet becomes more impressive when similar tests are applied to The Merchant of Venice quarto of 1600 Here again, there is evidence of more than one set of skeletons. The verso running-title 'The comicall Historie of' is sometimes approximately 4.4 cms. in length and sometimes approximately 4.6 or 4.7. The two lengths alternate regularly:

   

short running-title	A2v:-B4v,D1v-4v,F1v-4v, H1v-4v,and K1v.
long running-tide	C1v-4v,E1v-4 v,G1v-4v,and I1v-4v.

This alternation is reflected in the spelling tests, which show that the following forms are again significant:                  

Compositor X.	Compositor Y.
their	theyr/their
deare	deere
farwell	farewell
sayd(e), etc.	said(e), etc.
houre(s)	howre(s)
madam(e)	maddam(e)
being	beeing
sodainly	suddainely

This is a shorter list than for Hamlet: the mooue/moue distinction, by no means constant in Hamlet, does not hold for The Merchant; the Y form of receaue/receiue etc. is not found; the X form of honor/honour etc. occurs only twice, once on a page set by Y (F2v); and of reuendge/reuenge etc., only the Y form is found-on two pages set by X (E2v [twice] and E3). While X again used choice , Y used the form choyse. The -owe,-ewe / -ow,- ew distinction is irregular: it works almost perfectly for the often recurring word Jew but for other words, the Y forms predominate, -owe or -ewe being found only 7 times on pages set by X, and 5 times on pages set by Y. Similarly, the Y form sweet is often found but X's sweete four times only, thrice on pages set by X and once on a page set by Y.

In addition to the spellings significant in Hamlet, X set leasure(s) where Y set leysure(s). The form Ile is found throughout, but on pages set by X ile is often found. This last peculiarity is probably due to a shortage of capitals which is a noticeable feature of the quarto, verse lines frequently beginning with lower case letters.[7] The evidence is as follows:

For The Merchant of Venice there is further evidence: (1) Compositor X normally set the whole of the entry directions in italic type, while Y used italic only for proper names, setting the rest in Roman, and (2) Compositor

Y occasionally indented single lines, or parts of lines. These peculiarities reinforce the spelling tests:
   

X-italic type only	C1,1v,3v, 4,4v,E1& v,2,3v,G1v, 2,3,4v, and I1,1v,2,3,3v.
Y-Roman type except for proper names	A2,2v,4v,B2,2v,4& v,D1, 1v2,2v,3v,4v ,F2v,4,4v,and H1.

On C2v, C4, and E2v there are entry directions with proper names in Roman and the rest in italic type.
 

Y-indented lines

A2,4v,B2,3v,4,D1,1v ,4,E2v, and F4,4v.

The indentation of the last line on E2v is the only detail of this bibliographical evidence which breaks a regular alternation. It seems safe to assume, on the accumulated evidence of running-titles, spellings, entry directions, and indentations, that the work was divided between the two compositors as follows:    

Compositor X	A1(title-page),C1-4v,E1-4v,G1-4v, and I1-K2.
Compositor Y	A2-B4v,D1-4v,F1-4v, and H1-4v.

A1 seems to be the only page about which there can be any doubt.

II

Once it is accepted that the same two compositors set Q2 Hamlet (1604/5) and The Merchant of Venice (1600), certain theories about the text of the former must be examined afresh: can it still be said that the 'chief trouble' is incompetent printing? Certainly it is no longer possible to blame an 'untutored dolt' for any of its short-comings, for The Merchant is a remarkably clean text throughout. It would seem that either these workmen were abnormally rushed while setting Hamlet-for which no certain evidence is yet known-or else their copy was abnormally illegible.

Before accepting either of these explanations, much more must be known about the printing of the quarto: its imperfections must be reexamined and their distribution studied, the press-work must be considered, and a more detailed knowledge about Roberts' printing house must be acquired. In the meantime, the text of Hamlet must be treated with added caution. For example, Professor Wilson thought that a considerable number of half- lines, phrases, and words found in the Folio text (generally thought to derive from a prompt-book),[8] and not found in Q2, had been present in the copy for Q2 but had been omitted through the incompetence of the compositor.[9] Professor Wilson's 'omissions' are from pages set by both compositors, neither of whom is known to have skipped numerous lines and words in other work. Until it is proved that they were both guilty of such incompetence, or that they were rushed in setting Q2 Hamlet, we must presume that they were both defeated by a peculiar illegibility of their copy, or else that the Folio Hamlet contains additions to Shakespeare's original foul papers. A glance at Professor Wilson's lists of 'omissions' will show how likely the latter explanation is; few of them contain any hard words which do not occur in neighbouring lines, and many of them involve mere repetition or the addition of particles. In the present state of knowledge, it seems possible that the 'omissions,' accepted as authoritative by Professor Wilson, have no stronger authority than that of a scribe or the players' prompt-book.

New knowledge about the compositors responsible for these two quartos will also illuminate the problem of the copy used for The Merchant of Venice. It is usually argued that this copy was closely related to Shakespeare's foul papers or autograph manuscript.[10] Obviously it cannot have been 'foul' to the same degree as the copy used for Hamlet and some other plays, and it is therefore risky to claim that it was in fact Shakespeare's autograph copy. But if the same compositors set both Hamlet and The Merchant and if we are agreed that the copy for Hamlet was autograph, it should be possible to compare the accidentals of the two texts and come to a more definite opinion about the copy for The Merchant.

First of all, it will be necessary to know how much the compositors habitually altered the spelling and punctuation of their copies, and what spellings are normal in texts of their composition. To get this information we must examine other books printed by Roberts about the same time as Hamlet and The Merchant.

III

The choice of books for comparison is not easy, for the practice of compositors certainly altered with varying circumstances. We have seen how a shortage of capitals probably led to the recurrence of a form in The Merchant which is not found in Hamlet, and we must presume that the fount of type with which they were working might have predisposed compositors in favour of certain forms. For our purposes, therefore, books printed in the same type as the two quartos (Roman '82') will be especially valuable. The spelling of their copy was perhaps the largest single factor which affected the work of compositors. We can check this, to some extent, by examining books reprinted from earlier editions which came from other printing houses, but the check is not fully adequate, for compositors could work far more rapidly from a printed text than from a manuscript and this inevitably affected their faithfulness to copy. Likewise, the legibility of manuscript copy would influence the speed and workmanship of compositors, and we have little hope of making any satisfactory allowance for this. Another factor is the amounts of prose and verse in a book; compositors sometimes departed from their normal spellings and punctuation in order to justify a line, and there would be a greater tendency to do this in a prose work than in a verse one. The format of a book might similarly influence their work; a duodecimo book in prose might present justifying difficulties every half dozen words or so. Over and above these factors, there must be the human factor; a compositor

might favour one spelling for a year, or perhaps for only a few days, or he might have to work for a short time in a bad light, or be rushed in his work. The study of compositorial habits is obviously full of difficulties, but nevertheless it is possible, at this early stage, to present some kind of answer to our enquiry and thereby add to our knowledge of the two Shakespeare texts. I have chosen to examine books which represent most of the different circumstances which I have just described in the hope that something may emerge about the wider problems involved in the study of compositorial habits.

The most interesting evidence comes from James Roberts' reprint of Titus Andronicus for Edward White in 1600 The previous edition was by John Danter in 1594 and in this first quarto the following spellings predominate: -ow, -ew, their, deare, farewell, said, houre, being, honour, moue, reuenge, Ile. In the Roberts reprint, farewell, said, and reuenge were not altered, but sometimes the others were, as follows:

             

-ow, -ew became -owe, -ewe	B2,D3v,I2v.
their became theyr	A1,2,4(3),4v,D1(2),E1(2),1v,F1v(2), 2,4v,G1v(2),I1,1v(3), 4v(3),K1,1v(3),2v.
deare, etc. became deere, etc.	A4(2),C1,D2,E2v,3v(2),4,F1,1v, 2,I1v,K1,2v.
being became beeing	D1v.
honour, etc. became honor, etc.	A2,B1,1v,2.
moue, etc. became mooue, etc.	B4v,D2v.
Ile became ile	D3,E3,4v,I3(2),3v.

A few irregular forms were changed:          

-owe, -ewe became -ow, -ew	D4,F3,G4v,I1v,2v.
howres became houres	C4v.
honorable became honourable	K2.
mooued became moued	F1v.
ile became I'le	I1.

Danter's first quarto varied the following forms: sweete/sweet, madam/ maddam, receaue/receiue, sodaine/suddaine, and choise/choice/choyse. The Roberts reprint sometimes altered these, as follows:        

sweete became sweet	A3v,4,B2v,4,C1v(2), D1(2),E2v, 3v(2),4,F1,G2,H2,3,I3v,K1.
maddam(e) became madam(e)	D1(2),2v,F3,I4v.
receiue, etc. became receaue, etc.	E3,I2v.
choise became choyse	F3v.

In addition to these spelling changes (which Hamlet and The Merchant of Venice suggest may be significant), this was changed to thys on C1v(2),2(2),3v,F4(2), and I1v.

From this evidence, it would seem that the two compositors who set Hamlet and The Merchant of Venice also set Titus Andronicus, and certain pages can be allocated to each:

   

Compositor X	B1,1v,2,4v,C4v,D2& v,3,3v,E3,4v,F3, I2v,3,3v.
Compositor Y	A1,2,3v,4,4v,B2v,4, C1,1v,2,3v,D1, 1v,2,4,E1,1v,2v,3& v,4,F1,1v,2,3v,4,4 v, G1v,2,4v,H2,3,I1,1v, 4v,K1,1v,2,2v.

Very few details conflict with this allocation: X's honorable replaced Y's form on A2, but the honor/honour distinction does not hold for The Merchant, and in Hamlet, X's -or form is sometimes found on pages set by Y; X's madam replaced maddam on D1(2) and I4v which were both set by Y, but there are a few irregularities in the madam/maddam distinction in both Hamlet and The Merchant; Y's -ow, -ew forms replaced X's - owe, -ewe on F3 and I2v and sweet replaced sweete on I3v, but the shorter forms are often found on X's pages, especially in The Merchant. The allocation of pages A3v B2v, 4, D4, F3, G2,4v, and H2,3 must be considered doubtful, for they rely on -ow, -ew, sweet, or madam spellings which are not entirely trustworthy.

Judging from Titus Andronicus it would seem that if Compositor X found -ow and -ew in his copy he did not often bother to change them to the -owe and -ewe forms. Nor did he always change honour to honor, and moue to mooue. He did not once change farewell to farwell, reuenge to reuendge, or sweet to sweete. Compositor Y did not change houre to howre and was not consistent in changing being to beeing, their to theyr, and madam to maddam. Both compositors sometimes changed sweete to sweet and -owe, -ewe to - ow, -ew, and both sometimes retained uncharacteristic forms of receaue/receiue, etc. These tendencies help to divide other books between the two compositors.

Samuel Harsnet's A Declaration of Egregious Popish Impostures (1603) is in Roman '82', the type used for the three Shakespeare quartos.[11] It is a quarto of prose with 36 gatherings, not counting signature A which contains preliminary matter in italic type. The copy for this book probably had very few, or no, -owe, -ewe or sweete forms; I have noted only 5 -owe or -ewe forms (B3v, N2v, T2, V1, & Dd3) and two sweetes (E3 & R4v). Similarly I have noted honor only once (R2), deare once (L2v), and sayd eight times (Q1, T4v, V2,4v, Y3v, Dd1v, & Ll3v); in each case the Y forms predominate and probably did so in the copy. I have not found ile, farewell or farwell, madam or maddam, leisure or leasure; houre is infrequent and invariable; thys I found once (M1) and reuenge

once (L1); choise and choyse I found twice each (C3, M2 & G1v, R1), and choice once (Aa1). The work of the two compositors is distinguished by their/theyr, receaue/receiue etc., being/beeing, sodaine/suddaine, and, less reliably, mooue/moue etc. The following pages may be allocated with some confidence:    

Compositor X	C3,3v,D2v,El-3v,F4- G2,H2v- 3v,I3v-K1,L2,2v,M1 v,2,3,Q4,R2,4v,T1v, 2,4,X2,Z1v,Aa3,3v,Bb3v,4,Cc2v- 4,Dd2v,3,Ee2v- 4,Ff4,4v,Hh3,3v,Kk2,2v,3v,Ll3v- 4v,Mm2,Nn3, 3v,
Compositor Y	B1v-C2v,C4v-D1&v,D3- 4v,E4,4v, F1v-3v,G3v-H1v,H4v- I3,K2v-L1,L3-M1,M4-N1v, N2v,N3v-O1v,O4,4v,P2,3v,Q1 v-3v,R2v-4, S2v,3v,4,T2v,3,4&v,V1v,V3- X1,X4,4v,Y2-4v, Aa1,1v,Aa4v-Bb2v,Cc1-2,Dd1v,Dd4-Ee2,Ff1v,Ff2v-3v,Gg2v,3,Gg4-Hh2v,Ii4-Kk1v,Kk4-Ll2v,Mm1,3,Mm4v- Nn2v,Nn4,4v.

Again only a few of the distinctions noticed in Hamlet and The Merchant are significant in A. Warren's The Poor Man's Passions (1605). This is a quarto of verse, so the use of the longer forms would not be greatly influenced by the need to justify. Once more, the copy probably had few -owe or -ewe forms, for these occur in the printed book only on B1,4v, D3, E3 and F1. It might, however, have read sweete a number of times; I have found it 5 times (C1v, E2, F2v, & G1,1v) but sweet only 3 times (C1v, 3, & H4). Honour, houre, choise, and farewell are invariable, while reuenge, said, suddaine and leasure occur only once each (G3v, I1v G4v, & C2). I found being on I2 and beeing on G4; deere is on G4v and deare on D3v, E1, H3v(2), and K1; both the mooue and moue forms are found throughout. It is chiefly through the their/theyr distinction that the book can be divided between the two compositors. Their is found throughout, but theyr only seven times on G4(2),4v(3)and I4v(2). These spellings, together with the deere, beeing, reuenge, said and suddaine already noticed, and the absence of - owe or -ewe forms from sheets G and I, suggest that the whole book may have been set by Compositor X, with the exception of parts of sheets G and I which were probably set by Y.

The copy for G. Estie's A Most Sweet and Comfortable Exposition upon the Ten Commandments and upon the 51 Psalm (1602) almost certainly had a preponderance of -ewe and -owe spellings. It is a small octavo of prose with a comparatively large type, so that there are only a few words to each line, and justifying must have been a frequently recurring difficulty. Again Y's theyr spellings are especially significant: their is found throughout the book, but theyr occurs some 41 times up to K2v and thereafter is not found at all. There is a sprinkling of X's -owe, -ewe spellings from the first (I have noted them on B3, 4v, C3v,6,6v, D3v,7, E7, F1(2),

H1v,2,3v,7, & K4v), but since Y's distinctive theyr is also found on many of these pages, the longer form has probably been retained to help justification. Again, a change comes at sheet L in which I have noted only 2 -ows or -ews but a cluster of longer X spellings (on L3,3v,4(3), & 6). On sheet M there are a further 7 X spellings (on M1,1v,2,4,5v(2), & 8v). There are only 7 of these spellings in the remaining 6 sheets, but since there is no recurrence of Y's theyr, X is probably still responsible but has used the shorter form occasionally to accommodate the text to the short lines. It is noteworthy that the pages are signed 1 to 3 up to and including sheet K, L is signed 1 to 5, and the following ones, 1 to 4. Moreover other spellings suggest that X took over the composition from Y at L1; suddaine etc. is found on D3 and sodaine etc. on M2v,8, and O7; deere is found on A5 and deare on M2; choyse is found on D2 only; thys is found on A2v(2),4v, E3, G1, and K1,7 only; sweete and sweet are both found up to F7, thereafter the word occurs only once, on S4 in the X form. There are some anomalies: reuenge occurs infrequently but is invariable; Y's honour is the normal form occurring on L1v and following pages, and X's honor is found on N7v and O8 only; said occurs throughout and X's sayd on A5v and N8v, the former having a theyr characteristic of Y; the receaue/receiue distinction does not hold (the X form occurs up to Is only) nor does the less frequent mooue/moue. These irregularities need not greatly influence the allocation of pages between X and Y; neither receaue/receiue nor mooue/moue held for The Merchant, which is fairly close to this book in date, and the conservatism of both compositors as demonstrated in Titus Andronicus sufficiently explains the rest. Y probably set up to the end of sheet K of Estie's Exposition, and X the remainder.

H. Smith's Four Sermons (1602) is a quarto in eights written in prose and therefore with a recurrent need to justify. The allocation of work between the two compositors seems to be complicated, but Compositor Y probably set most of the pages of signature A, and Compositor X most of those of signature B. The significant spellings are:

After signatures A and B, for C, D, E, and F, the workmen seem to have changed more frequently, the two, on occasion, setting alternate pages. It is possible that a third compositor was employed. Deere is invariable and said is so until C8v, after which a few X forms occur. Being/beeing, receaue/receiue etc., mooue/moue etc., sodaine/suddaine, and sweete/ sweet are seldom helpful.

England's Helicon (1600) is probably another book for which the same two compositors were at least in part responsible, but for many of its pages there is no certain evidence. This might well be due to its copy being written in many different hands. All that may be claimed here is that signatures C, F, and O seem especially to belong to Compositor X, and S, Aa, and Bb to Compositor Y.

IV

It is time to return to the question which led to the examination of these books printed by Roberts: what evidence is there for saying that the copies for Q2 Hamlet (1604/5) and The Merchant of Venice were in the same handwriting? One answer is fairly straightforward. It has been shown from Roberts' reprint of Titus Andronicus and from other books first printed by him, that the two compositors whose work has been identified sometimes retained uncharacteristic forms from their copy. Since Hamlet and The Merchant share a larger number of significant spellings than any other two books which have been examined-and this in spite of a gap of some four years-it follows that the spellings of their copies were much alike. In particular, it appears from Titus Andronicus that if the copy contained only -ow and -ew forms, very few -owe or -ewe forms would be found in the printed book. This view is supported by Harsnet's Declaration and The Poore Man's Passions. Now in both Hamlet and The Merchant the longer spellings occur not only in pages set by X but also in those set by Y. It therefore seems highly probable that the copy for both plays had a considerable number of -owe and -ewe forms. It is worth

noting that Addition 'D' of the manuscript of Sir Thomas More (believed by many scholars to be in Shakespeare's handwriting) contains 4 -owe or -ewe spellings,[12] and not one -ow or -ew.

Similarly, it appears from Titus Andronicus that Compositor X did not always change farewell to farwell, and this is borne out by A Poor Man's Passions. Smith's Four Sermons suggests that he did not often change deere to deare, and also, with Harsnet's Declaration and Estie's Exposition, that he did not often change said to sayd. Titus Andronicus shows that Compositor Y did not often change houre to howre, and this is borne out by A Declaration. These spellings were significantly varied in Hamlet and The Merchant, and it therefore seems likely that the copy for both plays contained a preponderance of farwell, deare, sayd, and howre forms.

A complete check of the spellings in the two plays and other books by Roberts would be a very long task, and beyond the scope of this paper. It would be especially complicated by the possible illegibility of the copy for Hamlet. Nevertheless, it is possible to check a few unusual forms, as, for instance, 'how so mere' in The Merchant (G2v) and 'howsomeuer' in Hamlet (D3v). Both were set by Compositor X and we might expect that he was responsible for the 'm' spellings; but there is more evidence in signature Q of Estie's Exposition where the same man, in all probability, set the usual form 'howsoeuer'. Until a complete check is made, such details are of little value,[13] and the recurrence of -owe, -ewe, farwell, deare, sayd, and howre forms is, at this stage, the strongest evidence that the two copies were in the same handwriting.

Most of the spelling checks I have made have proved of little use in showing that the two copies were in a single hand, but they have shown that many unusual spellings in Hamlet which were listed by Professor Wilson as possible evidence of Shakespearian orthography,[14] are in fact often found in other work by the two compositors responsible for Hamlet. Hether is a clear example: it was set several times in Q2 Titus Andronicus where Q1 reads hither; X set it in The Poor Man's Passions (F4), and Y in Harsnet's Declaration (B2v); it occurs also in England's Helicon and Smith's Four Sermons. Similarly, doost was set several times in Titus where Q1 reads dost (e.g., D3v,4, & E1v). The evidence for the compositors'

predilections is not quite so strong with seauen, but it should make us hesitate before calling the spelling Shakespearian: X set it in Estie's Exposition (M5v,& N1) and both compositors set it many times in A Declaration (e.g., H3v, K3v, N1v, O3v, T2, & Ee3). Tearmes was likewise set by X in The Poor Man's Passions (B1) and by Y in Estie's Exposition (B1v). I hope these few examples will show that no spelling in Q2 Hamlet can be accepted as Shakespeare's own until a complete check of the spellings used by Roberts' compositors has been undertaken. The changes made in reprints will be especially valuable evidence for this.

V

The punctuation of both Hamlet and The Merchant of Venice has been praised as particularly sensitive.[15] It is certainly very light, and the possibility that it represents Shakespeare's own punctuation must be considered. The first question to ask is whether it was typical of Roberts' printing house.

My general impression of the punctuation in the Roberts books I have examined is that the two compositors in whom we are interested, varied their punctuation considerably; this is especially clear in England's Helicon where their copy was possibly in different hands, with as many different kinds of punctuation. However, the lyrics of England's Helicon are perhaps special cases. I will quote some passages from the prose works already examined. First, two passages from Estie's Exposition, the first set by Y, and the second by X:

These two passages are, I believe, representative of the kind of punctuation throughout this book. Occasionally there is a proliferation of commas: Both compositors in Smith's Four Sermons reproduce a peculiar use of the colon to mark direct speech: The examples I have given could be greatly multiplied, but I hope they are sufficient to show that the light punctuation of Hamlet and The Merchant of Venice is by no means the invariable practice of the two compositors who set the plays.

This is borne out by Roberts' Q2 Titus Andronicus; in this book, the compositors made the punctuation heavier than their copy. Not counting question marks, there were some 240 changes which made the punctuation heavier, and some 120 which made it lighter. A good proportion of these must be considered as corrections to the copy; there was a measure of editing in the reprint and the 'editor' may have considered punctuation to be within his scope. Of all the changes, about 80 or so involved semicolons, colons, or full-stops. Titus Andronicus therefore suggests that the compositors of Hamlet and The Merchant tended to add to the punctuation, not reduce it. And despite their changes and some obvious errors, they seem to have been careful and conservative over it; for example, on G2 a comma is kept at the end of a line, where it may imply that the following line breaks out abruptly, stopping the first speaker in the middle of a sentence (IV.ii.80). I have been unable to find any way of treating the copy's punctuation which could be ascribed to one compositor only.

I think this is as far as my investigations will allow me to go in considering the printing of Q2 Hamlet and The Merchant of Venice. Obviously there is a lot left to do. We should like to know more about the spelling of each compositor, and, perhaps more important, we should like to know to what kind of errors each compositor was prone. This paper can only be a beginning, and will have achieved its purpose if it has shown the evidence for two compositors in each play, and suggested ways of tackling the many problems connected with editing their work.

Notes