University of Virginia Library

Compositor B's Role in The Merchant of Venice Q2 (1619)
by
D. F. McKenzie

We can be reasonably certain that Q2 Merchant of Venice (1619) was set from Q1 (1600) by Jaggard's compositor B,[1] so that in this instance, with certain reservations, we know to a large extent what copy B was given to set and what he did with it. We must of course know a lot more of B than his handling of printed copy in 1619, but the following analysis, limited as it is, may be thought to offer some useful intermediate information about his working practices.

The total number of variants introduced into Q2 is something like 3,200. Whether all of these are in fact compositorial is not easy to determine. Some apparently new readings may derive from variant forms in the copy of Q1 used by B but no longer extant (IV.i.73 and V.i.300 are sufficient warning of this danger). Similarly, it may even be that a Q2 proof-reader introduced some variants and removed others, although there is no evidence of this. But we meet by far the greatest difficulty in deciding whether any of the new readings are the work of an independent reviser. Clearly no reviser is going to worry much about the accidentals of the copy. Most features of presentation certainly, spellings almost certainly, and punctuation very probably, were the compositor's concern. From an examination of Q2's expansions

and contractions of Q1 forms, of elisions and of changes in line-division, I am sure that these too must be attributed to the compositor. A few of the truly substantive alterations may be revisional, but again, apart from changes in stage directions and speech prefix forms, very few of these changes are beyond the compositor; many of them are inexplicable as the deliberate emendations of a reviser. In fact, as we shall see, it is neither difficult nor unreasonable to attribute all the so-called revisions in the text itself to compositor B.

I consider first the changes in substantive readings and then the changes in the accidentals of spelling, punctuation, and so on.

Substantive alterations: Since they appear to have all the characteristics of planned rather than impromptu editing, I shall assume that substantive changes in stage directions and speech prefix forms were not made by compositor B. For the rest, the variant readings fall easily into several groups: literal errors made or corrected; the omission, addition or substitution of single letters; and the omission, addition, transposition or substitution of words or phrases. Of the 134 changes made under these headings, 11 represent literal errors (foul case and so on) and 7 represent the correction of literal errors in Q1. We are therefore left with 116 effective changes or about one to every 23 lines — and this in the most straightforward copy imaginable. Chambers, Wilson, Greg and Brown have assumed that most of these had already been made in the copy supplied to B. Admittedly this is a remote possibility but on the evidence I believe it to be unlikely. On the other hand this group of changes does for the most part show the same combination of misdirected ingenuity, deliberate tampering and plain carelessness which Miss Walker has elsewhere found in the work of B. To establish the probability that most of these changes were B's would require examination of every one in context. For reasons of space alone this is obviously impossible and I hope instead by selective quotation to suggest both the kinds of changes made and the likelihood that they were compositorial. Changes considered later as simple contractions or expansions are not duplicated in this section.

It seems fairly certain that B took upon himself the minor job of deleting, adding or changing a single letter, especially final -s, when he thought it might improve the grammar or the sense. To take some examples:

(a) Omissions:

• 1. I.iii.70 Me-thought (s) you said, you neither lend nor borrow
• 2. IV.i.65 I am not bound to please thee with my answere (s)
• 3. V.i.114 We haue bin praying for our husband (s) health

---

77

(b) Additions:

• 4. III.i.6 if my gossip (s) report be an honest woman of her word
• 5. III.iv.50 render this / Into my Cosin (s) hands
• 6. IV.i.349 If it be proued against an (y) alien
• 7. V.i.32 For happy wedlock (es) houres

(c) Substitutions:

• 8. I.i.95 O my Anthonio, I do know of those [Q1 these]
• 9. II.v.30 And the vile squeaking of the wry-neckt Fife [Q1 squealing]
• 10. II.vi.58 [to Gratiano and Salarino] on gentlemen, away [Q1 gentleman]
• 11. II.viii.39 Slubber not businesse for my sake Bassanio [Q1 slumber]

The variants under these headings actually number 27, made up of 11 omissions, 12 additions and 4 substitutions. Only 5 of these changes are accepted by the editor of the New Arden edition, and all 5 could easily have been compositorial. We certainly cannot assume that Slubber, for example, because it is not now in current use, would not present itself to B in 1619 as a most elementary correction. Many of the changes, although they might be considered conscious corrections, may in fact be memorial errors or just slips of the eye (this could be the explanation for I.ii.47 and III.i.40 below) but many others I believe do show an often wrong-headed emendation for grammar and a weak sort of sense.

When we ask, as we must, why any reviser should have made most of the changes in the next group — words omitted, added or transposed — we again find ourselves at a loss for any reasonable explanation. These changes number 49, made up of 20 omissions (including one whole line), 16 additions and 13 transpositions. Again, taking omissions and additions first, some examples will best show what B was about:

(a) Omissions:

• Prose
• 1. I.ii.35 J but one who (you) shall rightly loue
• 2. I.ii.47 J that he can shoo (him) himselfe
• 3. III.i.40 I say my daughter is my flesh and (my) blood
• Verse
• 4. III.ii.82 Some (marke) of vertue on his outward parts
• 5. IV.i.400 To bring thee to the gallowes, not (to) the Font
• 6. V.i.305 That I were couching with the (Doctors) Clarke

(b) Additions:

• Prose
• 7. I.ii.140 J with so good (a) heart
• 8. II.v.9 J was wont to tell me, (that) I could
• 9. III.i.92 J (O) / would shee were hearst
• 10. III.i.113 J (in) one / night
• Verse
• 11. II.ii.186 I haue (a) sute to you. / You haue obtain'd it
• 12. IV.i.346 Ile stay no longer (heere in) question. / Tarry Iew
• 13. IV.i.423 Not as (a) fee: grant me two things I pray you

The first two examples betray elementary confusions of which only a very careless reviser would have been guilty, and 3, like 5, 7 and 11, shows the sort of unconscious normalisation which it is reasonable to expect of a compositor. It is very likely that 9 and 10 illustrate the way in which B might alter the copy for his own ends. In each of these lines the word added comes at or near the end and in neither case could the first word of the next line be fitted in or conveniently broken: it would seem that B was here justifying his line by addition. Again then we have the familiar mixture of accidental changes (4 and 6), probably unconscious normalisations and deliberate emendations. The New Arden editor accepts 2 of Q2's omissions and 2 of the additions — 4 changes out of 36 — but none needs a reviser's hand.

Excluding one that occurs in a stage direction, there are 4 instances of transpositions in prose and 9 of transpositions in verse. There is little point in quoting the prose examples. Two of them, one also involving an omission, quite obviously must be attributed to a momentary mental blindness on B's part, and the other two, although they give a slightly more natural reading, are more likely to represent unconscious normalisation than deliberate emendation either by B or a reviser. The 9 verse transpositions also point strongly to corruption, probably unconscious, by B. I quote four to illustrate:

• 1. I.i.24 Q2 What harme a winde too great at sea, might do
• Q1 what harme a winde too great might doe at sea
• 2. IV.i.434 Q2 There's more then this depends vpon the valew
• Q1 There's more depends on this then on the valew
• 3. V.i.65 Q2 Doth grossely close in it, we cannot heare it
• Q1 dooth grosly close it in, we cannot heare it
• 4. V.i.213 Q2 And suffer'd him to go away displeased
• Q1 and sufferd him to goe displeased away

None of Q1's readings is at all awkward either in expression or sense and again it is difficult to imagine a reviser changing them. Two of the transpositions moreover completely destroy the meaning (2 and 3).

Of the 40 remaining substantive changes, 36 represent the substitution of single words and 4 of phrases. Some of them are clearly compositorial: for and if (twice) for and, a for of, of (twice) and on for a, and equall for egall. Others, being obviously wrong, are most unlikely to be those of a reviser regularising the text: Hath for Haue (in "virgins of our clime / Hath. . ." at II.i.11) and That for till (at V.i.305). Others again are less clearly but certainly wrong, and since

they are unnecessary and stylistically bad they are not easily explained as the wilful alterations of a reviser: although for albeit, the for this (twice), and foole for food.

Others are more difficult to attribute. Eight could reasonably be compositorial: com-/mand for commaundement at II.ii.32 is not unusual and the business of justification would have come in the middle of setting the word; range (III.ii.113) may be a variant of reigne or reine and Q1's ambiguous raine; it may be doubted whether roofe for rough (III.ii.206) is merely a spelling variant, but in any case it is an obvious correction; the for her (I.i.33), one for man (I.i.78), his for the (III.iii.29), presently for instantly (IV.i.281) and And for hee (IV.i.339) are hardly beyond the compositor and could even be accounted for as memorial errors; ill for well (II.ii.21 J), conclusions for confusions (II.ii.39) and their for the (II.ix.81) introduce difficulties which even a reasonably intelligent reviser would have wished to avoid. Even out-stare for ore-stare (II.i.27), while it may be revisional, could also be a variety of foul-case error. There are 5 readings which, it could be maintained, may be truly revisional: incarnall for incarnation (II.ii.27 J); fledg'd for flidge (III.i.31 J); To view for I view (III.ii.62); then for it (III.v.82); tis for as (IV.i.100). However, the very smallness of the group seems a strong argument against a reviser, and in any case it is a little uncharitable to give a reviser credit for these few near-misses and blame B for the obvious wides. But these changes at least, even if misguided, may be thought sufficiently important and plausible to justify a reviser's hand.

Finally among the single words altered there is a small group of substitutions which are difficult to account for as deliberate alterations but which could be unconscious changes: misery for cruelty may have been suggested by the abundance of words at that point containing an s (III.iv.21); cutst for tak'st (IV.i.326) is probably a recollection since cut occurs in the preceding line; health for welfare (in "We haue bin praying for our husband health" — V.i.114) makes the line run a little more easily and could represent revision; apparreld for accoutered (III.iv.63) and far'st for cherst (III.v.75) remain inexplicable unless we treat them simply as unconscious changes.

The four instances of phrase substitution are: are you resolu'd for is hee yet possest (I.iii.65) and he would haue for ye would (I.iii.66) which, since they occur together and are out of character with most other compositorial changes, may suggest an attempt, altogether unnecessary, to clear up the sense; at the length for in the ende (II.ii.83 J — although justification in this case could not have been a factor); and did vphold for had held vp (V.i.214).

This tinker-tailor fashion of allocating changes to B and the inevitable uncertainty about the presence of a reviser make it difficult to be authoritative about B's accuracy. Still it is even harder to be authoritative about his work in a text where the underlying copy remains largely subject for speculation, and it may be taken as an indication of the value of Miss Walker's work that her general conclusions about B are substantially confirmed by this examination of Q2 Merchant of Venice. If most of the changes discussed are B's then it is indeed a matter for grave concern that a compositor who was responsible for setting so much of the First Folio should have some significant error every 23 lines. Over all his work, B's average may not be so high as this, though with heavier copy the number of outright mistakes and of easy solutions to difficult readings might well be expected to rise. Some of B's readings may of course be accepted as desirable and so correct, but the main point is not whether one reading in 11 or 12 happened to be acceptable, but whether it was B's intention faithfully to follow his copy or deliberately to depart from it in order to give what he considered to be a better reading. The evidence from Q2 Merchant suggests more strongly than ever that he was not only subject to numerous lapses even when trying to follow copy but that he might also correct it on no one's authority but his own.

Alterations in accidentals: In what follows I shall, with one or two minor exceptions, consider only the changes made by B against his copy. There is of course an obvious objection to this procedure, namely that important preferences happening to coincide with the copy itself might be overlooked, but it is the only safe way of isolating copy influence and so of arriving at what are therefore perhaps B's more positive preferences. As will be seen, the changes made cover a wide range and while some are naturally more significant than others, all are to a greater or less degree relevant to discussion of the nature of the copy used for F1 plays and to the question of B's consistency and accuracy. Here then I want to consider B's handling of spelling, punctuation, capitalization, stage directions, speech prefixes, typographical abbreviations and proper nouns, his use of italics, and his attitudes to justification, line division, syncope and elision, expansions and contractions and to versification generally.

The spelling changes introduced into Q2 by compositor B offer the most important evidence for his identification in other plays. Some changes are regularly made, many betray great inconsistency, but in either case of course the changes are significant. Besides knowing those spellings habitually used by B against his copy we must also know those

to which he was apparently indifferent. One of the most important features of Q2 is the mixture of A and B preference spellings. It is clear that compositor A could not have been working on this text and yet there are in fact 64 A spellings (17.3 per cent of the total) of words normally used to test B's presence in the Folio. To take some of the more significant spellings: do 103 to doe 11; go 41 to goe 16; heere 62 to here 15; yong 12 to young 13. These figures give us some idea of the measure of inconsistency we might expect in the earlier work of B. Even more significant as an indication of the instability of B's preferences are the A spellings which he gave against the B forms of Q1. There are 13 A forms which could not have derived from the copy and which are most unlikely to have been affected by justification. These are goe 1, chuse 2, deare 3, and here 7. Compositor B then might once in 10 or 11 times put doe under copy influence or to justify a line, but he never once set it against his copy; his preference for go must be considered much less stable than that for do, and he might even occasionally set goe against his copy; chuse and deare are perhaps not so significant as the numbers involved are not large; here however B gave 7 times in unjustified lines and against his copy, so that this word might not be an altogether reliable test for B's presence. Young is also worth mention, for although B never goes against his copy to give an A spelling (unless justification intervenes) he actually set the A form on 13 of the 25 occasions on which the word occurs.

Some other comments on B's spelling are worth making. He seemed to prefer, for example, to hyphenate words like bed-fellow, me-thinks (and me-thought), selfe-same and under-valewed rather than not, and in fact hyphenated such words 25 times whereas he deleted the hyphen in only 9 words. On 15 occasions words divided in Q1 were given by B as one word, and on 10 occasions he made the reverse change. It is quite clear from the changes made that many a Prethee of Shakespeare's is in fact compositor B's, for he 6 times substituted this word for Q1's Pray thee. The evidence of these changes taken together with that of hyphenation suggests a tendency on B's part to contract word forms. Although sometimes the alterations are unimportant the preference for contracted or compound forms can upset verse lines and may possibly and even more seriously provide us with a non-Shakespearian element in low comedy scenes where the temptation to use them would be greater.

Another fairly pronounced characteristic of B's spelling is his substitution of medial -i- for medial -y-, although this change is not consistently made. We find forfeitures and forfeyture, said and sayes on the same pages together with similar variants retained from Q1. However

B had a three to one preference for the -i- form; he substituted it for -y- 94 times and made the reverse change only 30 times. Moreover many of the changes to -y- concern the same words: lye 5, dye 2, gyrle 3, fayle 2, sayle 2.

A far more consistent alteration was that of au or ou plus nasal to a or o, although almost half the changes relate to the one word Lancelet and although he sometimes followed Q1's au or ou (cf. young). However this change was made 96 times in all. Justification may have been a factor on 32 occasions, but the complete absence of any change to au or ou suggests that B's preference was fairly well established.

The use of final -y for -ie is almost though not quite absolute. B gave -ie for -y 12 times (4 times in verie and 4 times in the not quite comparable monies) but he gave -y for -ie 110 times. Among the words most frequently changed are deny, Lady, twenty, mercy. Again, however, B often kept Q1's -ie forms, giving a mixture of forms not only on the same pages but within the same lines. Only 2 of the -ie substitutions occur in justified lines.

Far less significant is B's use of single for double consonants or vice versa. He chose to set double consonants 80 times and single 77, the number of possibly justified spellings in each group being 21. Most of the variants are insignificant but B did prefer the -esse ending to Q1's frequent -es in nouns (happinesse, heauinesse, businesse, etc) and he also seemed to prefer a single consonant in the third person singular present indicative of verbs ending in -l (tels, cals, fals and, perhaps by analogy, wals).

Compositor B contracted Q1's double vowels to single vowels 104 times and extended them 56 times, but except in a few words (too, given 10 times against copy, lose 7, doing 3, bene 9), the changes were usually haphazard. Again justification probably dictated many (23 words given double vowels and 40 given single ones occur in full prose lines), but the proportion of words affected is about the same in each group. B's treatment of the pronouns he, she, we, me was variable but he preferred the shorter forms (47 as against 19).

Although generally speaking B also preferred to drop the final -e from many words (he did this 368 times) he also added it to a large number (215), and again we frequently find the same word treated in different ways. Some words however are far more often given without the -e than with it and vice versa. There are of course do and go, but also anon, Iew, indeed, forfet, need, wee'l; with the single exception of goe a final -e is not once added to these words against the copy. Words to which it was fairly frequently added are backe, finde, olde, quicke, thanke, thicke, thinke, trickes, waste, wilde and winde; however

beside the 7 occurrences of thinke we have to place one think, given against the copy but in a justified line, and beside the 6 occurrences of thanke we must place one thank, given against the copy in an unjustified line. Apart from these few words there seems to be little governing principle behind B's choices. Although he made 583 changes of this kind, his frequent inconsistency shows that he was quite indifferent to the spelling of many of the words he altered.

The last major group of spelling variants is that composed of words which do not fall within any of the preceding categories. While some of these words may suggest broader preferences (for example, the substitution of -cke for -que), most of them recur too infrequently to justify further classification. There are 301 variants in this group, the words which appear most frequently being diuell 11, heart 22, master 30, mistris 5, musicke 15, shew-shewes-shewing 10, then (for Q1's than) 8. Many of these preferences were no doubt shared by compositors A and E so that their significance as test words may be slight. Nevertheless we still need to know with some certainty the sort of things B was doing to his copy even when we have determined his presence. In the long run of course it is the only reason for determining his presence at all.

A short note is all that is possible about B's changes to the punctuation of Q1. Many of them are haphazard but a large number show deliberate normalisation. There are some 715 changes altogether. The addition of 347 commas makes up by far the largest single group. By contrast, only 23 commas were deleted. Some of the more important classes of additions are those which occur with the vocative and exclamations, with relative clauses, participial clauses, various kinds of adverbial clauses, prepositional phrases, parenthetical statements and phrases, and between the protasis and apodosis of conditional sentences. These are places in which we might expect to find commas, but there are other classes important enough to justify comment. For example, B frequently marked off the subject, either a noun or a noun equivalent, with a comma ("If to do, were as easie as to know what were good to do. . ." at I.ii.13). He would also often add a comma before an infinitive although many of these occur at the end of verse lines and may simply mark the end of line pause ("And these assume but valours excrement,/ To render them redoubted. . ." at III.ii.87-8). And a comma is frequently inserted before a noun clause or to mark the ellipsis of that after verbs of saying or wishing.

Of the commas added to verse lines, 113 come within the line and 186 at the end.

There are of course many other changes in punctuation. As we should expect, most of them give heavier pointing but there are numerous exceptions. The figures may be of interest: a comma was replaced by a semicolon 34 times, by a colon 28, by a full point 32, by a question mark 29, by an exclamation mark once; a semicolon was replaced by a comma 10 times, by a colon 9, by a full point 15; a colon was replaced by a comma 14 times, by a semicolon 28, by a full point 30, by a question mark 5; a full point was never reduced to a semicolon but it was replaced by a comma 8 times, by a colon 6, by a question mark 19, by an exclamation mark once (two full points were accidentally omitted at the ends of speeches and two necessary ones inserted); a question mark was replaced by a comma 5 times, by a semicolon once, by a colon 3 times, and by a full point 53 (one question mark was added and another deleted). B capitalized the word following a mid-line colon 5 times when it was not capitalized in Q1 and he reduced such capitals in Q1 4 times, although some of these changes were doubtless independent of punctuation. A capital follows a comma twice, a semicolon 3 times, and a question mark 7 where Q1 gives lower-case letters. Finally, B added parentheses 6 times, omitted them once, and correctly repositioned a misplaced bracket. Of all the relevant changes, 102 resulted in lighter punctuation and 538 in heavier. Such an analysis of punctuation changes shows the need for much freer treatment of Folio punctuation generally (did compositor A set many other commas like that in Macbeth II.ii.64 — "Making the Greene one, Red."?)

It may eventually be possible to see some system in B's capitalization, but any attempt to impose a classification on the limited number of examples provided by Q2 is predestined to failure — there are too many exceptions. Of the 253 changes made (apart from those required by repunctuation or the substitution of upper- for lower-case letters at the beginning of verse lines), 193 represent words capitalized and 60 represent words reduced to lower case. Words capitalized by B against his copy more than twice are: Casket 10, Conscience 3, Ducats 3, Gentlemen 3, Grace 5, Iudge 3, Law 8, Letter 10, Loue 3, Mistresse 3, Mother 3, Moone 4, Musicke 4, Ring 26, and State 3. Yet we must beware of concluding that B's preference for some of these forms was necessarily a very strong one, for he also gave iudge 5, and mistris 1. A similar inconsistency is seen with ile 3 and Ile 2. More significant perhaps are father 10, although Father also occurs once, master 11, sonne (i.e. son) 3, and summer 2 together with winter 1.

It has been claimed that B, in the Folio at least, regularly centred his entry directions. However of the 46 entries in Q2, only 29 are

centred with some care and the remainder are roughly centred. There are moreover 4 pages on which one entry is carefully centred and another not. These figures, and in particular the variant positioning of entries on the same pages, would seem to indicate a measure of indifference on B's part.

Compositor B apparently also had a preference, though again not a hard and fast one, for setting exit directions full out to the right-hand margin. Of the 41 exits in Q2, 35 are set in this manner and only 6 are indented. Again in two cases indented and non-indented forms appear on the same page.

Apart from entries and exits there are only 5 other directions in Q2. The first of these is set full out (Hee reads. — not in Q1); the second is indented (He opens the Letter. — Q1 open the letter. indented); and the third is carefully centred (Musicke playes. — Q1 play Musique. set to right). Of the 2 remaining directions, one (Iessica aboue. — really an entrance — after II.vi.25) is carefully centred, as in Q1, and the other (after III.ii.62) is centred to the page, again as in Q1, but not to the verse measure.

With regard to speech prefixes there are two main points to be considered: whether compositor B abbreviated or expanded the speech prefixes of his copy and whether in either case his changes were made for the sake of uniformity. I assume that substantive changes in these forms may be revisional. Excluding omissions and insertions, B altered the speech prefixes of his copy 230 times (48 times in justified lines). Of these changes, 17 were confined to spelling only and did not affect the length of the prefix. Of the remaining 213 changes, however, 162 resulted in abbreviation of the copy-text form and 51 in expansion of it. Compositor B must therefore be considered to have had a pronounced preference for the shorter forms. The preference is in fact stronger than would appear for the Salanio and Salarino prefixes had to be expanded to at least the fifth letter, Solanio being deleted altogether from Q2. Other expansions, like that of An. to Ant., Bas. to Bass., Ies. to Iess., Clo. and Clown. to Clown., seem to have been made from a desire for some degree of uniformity.

Compositor B's practice in expanding or abbreviating and, Master, Lord and so on was, as we might expect, variable. He abbreviated and to & 8 times and expanded & to and 6 times and to If once. In 9 of these changes justification could have been a factor. He twice abbreviated Maister to M. and twice expanded L. to Lord. The evidence is insufficient to justify generalisation.

Compositor B invariably italicised all entries, exits and directions,

even though 25 of Q1's directions were partly in roman. Usually, though far from invariably, he also italicised mythological names. He changed Q1's roman to italics 4 times and Q1's italics to roman 3. The use of italics for countries, towns or their inhabitants, and adjectives derived from them, suggests a more fixed preference. In 3 instances B went against Q1 to set such forms in italics, and although he sometimes followed Q1's roman he never departed from Q1's italics. Characters' names when they appear in the text were normally italicised by B although he did twice put such names in roman against Q1's italics. Similar words italicised against his copy are Mounsier, Abram, Daniel, Barrabas and Faucenbridge. Finally, B consistently italicised quotations, letters and so on.

There is little point in discussing the spelling changes introduced by B into the characters' names, but there are several other proper nouns whose spelling differs from that of Q1. Cressed becomes Cressada against both Q1 and the metre (F1 keeps Cressed, but the form in Troilus is usually Cressida or Cressid), Caribdis becomes Charibdis (but Cholchos is altered to Colchos), Daniel is 6 times spelt with a single -l, Fauconbridge is changed to Faucenbridge, Franckford to Frankford, Fraunce of course is normalised to France, Genowa on 3 occasions becomes Genoway, Germanie is altered to Germany, Italie to Italy, Nazarit to Nazarite, Neopolitane to Neapolitane, Palentine to Palatine, Romaine to Romane and Venecian to Venetian.

Although we really have no means of telling when words were altered to justify a line, I have throughout treated full prose lines as if they had in fact been deliberately justified by spelling. To treat all full lines as suspect is the only way of isolating possibly justified words. On this basis, I have counted the spelling changes and other alterations in a number of groups where justification could have been a factor. Omitting instances which overlap, we find that shorter forms are given 269 times and longer (including the insertion of apostrophes in elided words) 96 times. That is to say, B normally justified by deletion, in keeping with his general preference for shorter word forms. Four of B's additions to the text and 5 of his omissions occur in justified lines.

Q1's verse lineation was changed 10 times in Q2. One of these changes (at I.i.6) is insignificant since the division is in effect only a turn-under. More significant is the fact that this particular line ("I am to learne: & such a want-wit sadnes makes/of me,") is really a line and a half, but the extra half-line goes unnoticed and apart from breaking it at "makes" instead of "sadnes" Q2 keeps Q1's arrangement. Twice Q1's division is corrected (at II.v.53 and II.vi.45) but in the

remaining 7 instances it is wrongly changed. Q1's lineation is indicated by the bar:

• 1. II.vi.1-2 This is the pent-house vnder which Lorenzo / desir'd vs to make stand.
• 2. II.vi.45-7 So are you sweete, Euen in the louely garnish of a boy, But come at once,/for the close night Doth play the run-away,
• 3. III.v.87-9 for the poore rude world Hath not her fellow. Lor. Euen such a husband / hast thou of me, As she is for wife.
• 4. IV.i.6-9 I haue heard, Your Grace hath tane great paines To qualifie / his rigorous course: But since he stands obdurate,
• 5. V.i.1 The Moone shines bright. In such a night as this, /
• 6. V.i.112-13 He knowes me as the blinde man knowes The Cucko, / by the bad voyce.

There seems to be little reason for the incorrect division in 1 (the verse measure would have taken Lorenzo) except that our analysis of punctuation changes shows that B preferred a pronounced pause at the end of a line, and the pause in this instance may have seemed to him more natural at which than at Lorenzo. In 2, B correctly divided line 45, but failed to recognise the half-line in 46 and so misdivided 47. The greater attraction of hast to husband than of As to me, combined with B's preference for a pause at the end of the line (given by his new arrangement), is probably responsible for the change in 3, and a similar explanation may apply to the incorrect division found in 4. The analysis of punctuation changes showed that B frequently inserted a comma before an infinitive, so that possibly here he divided line 7 to achieve the same effect, and this of course affected the division of line 8. In 5 the division would appear to have been made simply to give uniformity with the arrangement of Jessica's and Lorenzo's succeeding speeches which for metrical reasons had to be set as in Q1. The most likely explanation for the final change is that the Cucko would not fit B's verse measure, and since he could not end the line on the he carried both words over. Had he mistaken the verse for prose, as the New Shakespeare editors suggest, it is unlikely that he would have capitalized The. Disregarding I.i.6 then, B redivided the verse lines of his copy 9 times, twice correctly but 7 times wrongly. The explanations for these 7 errors are necessarily conjectural but it would seem that two influences could have been at work —B's desire for uniformity and his preference for a reasonably strong pause at the end of a line.

There has in the past been a great deal of discussion about the extent to which elisions found in Folio texts were Shakespearian or compositorial, and the uncertainty led to further disputation about the quality of elided vowels. It is safe to say however that compositor B, at least in Q2 Merchant, did sometimes elide and sometimes expand the forms in his copy, and that the marking of elisions by an apostrophe had nothing to do with pronunciation. For the most part B's treatment of elision in past participles or the third person singular present indicative simply amounted to the insertion of an apostrophe. Altogether he altered Q1's forms 101 times, but only 17 of the changes represent expansion of Q1 elisions, and of the 84 contracted forms only 20 involve actual changes in the spelling (apart from those where the whole word is altered). Among the participles, B substituted -'d for Q1's -d 58 times; for -t or-'t 4 times; for -ed twice. He replaced -ed by -'de once and by -d 3 times. He only once gave -t for -'d. He also gave entred for enterd, engendered for engendred and, by mistake no doubt, ser'ud for serud. On the other hand, B expanded Q1's -'d or -d to -ed 10 times and to -de 5 times. Among the remaining verb forms, B once gave -st for -'st, but made the reverse change 4 times; and he once expanded -st to -est but twice elided-est to -'st. He also altered call'dst to cald'st and threatenst to threatnest.

Compositor B's effective contractions of other words (the contractions that is which reduce the number of syllables) total 25 (19 of them occurring in verse lines), and his expansions 18 (9 occurring in verse lines). A further 56 changes simply involve the insertion or omission of an apostrophe. Where B altered Q1's youle or weele, he usually put you'l and wee'l; normally he set an apostrophe in here's, that's and there's; on the other hand, he twice omitted an apostrophe given in Q1 to mark the elision of e in a superlative. Some words however B actually altered and these are much more serious. I have already mentioned prethee which was 6 times substituted for pray thee, but we also find ifaith for in faith, farwell or farewell for far(e) you well, ist for is it, on't for of it, Let's for Let vs, ha for haue, Ime for I am, my for mine before a consonant, to for vnto and y'are for you are. Among the expansions are in the for ith, say it for say't, vnto for to

and uppon for on. Nowhere did B use the forms to'th, to'th', by'th', they'are and so on against his copy, although he did insert an apostrophe in th'interiour, t'intrap and th'afternoone.

Finally, B's handling of verse requires special attention, the main point being whether he tried deliberately to improve the metre. First, as we have just remarked, one of the most significant things about B's punctuation was his insertion of 186 commas at the end of verse lines and of 113 within verse lines, indicating a preference for a strong pause at the end of a line. Some though by no means all of his mid-line insertions coincide with the caesura, but there is little to suggest a special attempt to alter the metrical structure. Second, with regard to verse re-lined by B, we should remember that his alteration of Q1's verse lineation resulted in only 2 corrections but 7 errors. But more important than either his new punctuation or line-division is his addition or deletion of syllables in verse lines. In brief, B lengthened the verse lines of his copy 22 times and shortened them 33. Of these changes, one resulted in a line of 8 syllables, 18 in lines of 9, 13 in lines of 10, 16 in lines of 11, and 7 in lines of 12. Only 4 changes could possibly be considered corrections. It is certain that none of B's changes was made with an eye to, or rather an ear for, the metre.

The detailed study of compositors' working practices, bleak by-way though it seemed at first, has now become one of the main highways to Shakespeare's text. But the journey is far from over yet. We also need to know a great deal more about their social and educational background as well as their apprenticeship training. Naturally, the less certain we are of the nature of the copy-text for any play, the less able we are accurately to assess the compositor's role in giving us the printed version. That is why the patient analysis of reprints, where the compositor or compositors can be positively identified and the copy is not seriously disputed, offers the surest method of determining the nature of compositorial corruption. What holds for printed copy may not hold for manuscript, and a compositor's preferences might change. These however are suppositions which I hope others will demonstrate from similar and comparative analyses of other plays.

APPENDIX
Number of times that test spellings occur in Q2:

                                                             

	Word	Unjustified	Justified	Total
A	doe	7	4	11
B	do	91	12	103
A	goe	14	2	16
B	go	34	7	41
A	here	11	4	15
B	heere	58	4	62
A	young	8	5	13
B	yong	7	5	12
A	yeere	--	NA	--
B	yeare	6	NA	6
A	deare	4	NA	4
B	deere	12	NA	12
A	griefe	--	NA	--
B	greefe	2	NA	2
A	deuill	2	NA	2
B	diuell	12	NA	12
A	cousin	--	--	--
B	cosin	1	--	1
A	suddainely	--	--	--
B	sodainely	2	--	2
A	hower	1	NA	1
B	houre	9	NA	9
A	graunt	--	--	--
B	grant	4	--	4
A	chuse	2	--	2
B	choose	29	11	40
	----	----	----
Totals		316	54	370	A -- 64
	----	----	----
					B -- 306

Notes