University of Virginia Library

The promise that compositorial analysis has held out for editors of Shakespeare—that it could significantly affect how editors treat accidentals, lineation, and emendations, as well as indirectly reveal the nature of the printer's copy—has not yet been fulfilled because of the incompleteness of our knowledge about the compositors of the Shakespeare First Folio. Compositorial analysis of the First Folio has been primarily concerned with establishing the number of compositors and identifying their shares of the text.[1] These tasks have been the necessary prerequisite for analysis of the amount and kind of textual corruption that the compositors introduced. Now that recent studies of the Comedies section of the Folio have revealed "who set what,"[2] we can begin to examine the quality of each compositor's work, in that section at least.

The Comedies section is especially important since the most authoritative text for ten of the fourteen plays in this section is the Folio, and five of the six compositors who worked on the Folio set pages in the Comedies. Compositor B was the most important compositor, setting 127 of the 301 pages. Dr. Alice Walker roughly described his habits over twenty years ago,[3] but a thorough analysis should wait until his share in the rest of the Folio is determined. Compositor F was the least important compositor, setting only 22 pages in the first five

plays, and dropping out completely after MM. Compositor A began his work in the Folio with WT, and continued through the Histories and Tragedies. Compositor C and D may have worked only in the Comedies, with Compositor C setting 96 pages and Compositor D 41 pages. Although Compositors C and D are not so important as Compositor B, I have chosen to focus on them because we know nothing about the quality of their work; moreover, methods and generalizations developed while working on these more manageable compositors can then be adapted to the others. Also, by working with these two compositors, I can contrast their very different styles and offer a meaningful comparative analysis. For instance, the fact that Compositor D transposes words once in 25 pages becomes more significant when compared with the fact that Compositor C transposes words once every two pages. I have limited this study to substantive changes because this is the most significant compositorial evidence for editors.

Both Compositors C and D set a number of pages in the four comedies where extant quartos were their copy (Ado, LLL, MND, and MV) as well as pages from Err, MM, AYL, Shr, and AWW. Compositor C also worked in Tmp, TGV, and Wiv. Thus, one can establish qualitative evidence (a list of changes the compositors made from the extant copy which reveals their habits), and apply it to the plays where the Folio is the only or best text. Compositor C was the more careless, in most ways, of the two compositors. In the four plays where his copy is extant, he set 267 corrupt substantives,[4] an average of one in every 21 lines; whereas Compositor D set 93 corrupt substantives, an average of one in every 34 lines (or in other words, Compositor C made one more error per column than Compositor D). These figures suggest that one should examine Compositor C's pages with more care and adopt emendations more readily, but it does not tell an editor what to look for. A mass figure or an average, like the ones above, can be misleading. Hence I have classified the substantive errors according to the traditional categories of substitutions, omissions, interpolations, and transpositions.

               

	COMPOSITOR C					COMPOSITOR D
	Ado	LLL	MND	MV	Total	Ado	LLL	MND	MV	total
pages	13.5	11	9	8.5		5	6	5	9.5
Transpositions	4	4	3	8	19	0	0	0	1	1
Substitutions	55	58	31	30	174	10	19	15	26	70
Omissions	30	12	6	10	58	1	8	1	3	13
Interpolations	7	3	4	2	16	4	2	0	3	9
total	96	77	44	50	267	15	29	16	33	93

These categories allow us to spot some of the specific habits of each compositor, and thus to perceive more clearly their basic differences in method.

TRANSPOSITIONS generally involve two monosyllabic words, and they usually do not change the meaning of a passage because the word order of English is often loose. The transpositions by Compositors C and D are typical: they do not radically alter the meaning, nor do they disrupt the meter in a verse line. Usually transpositions are the result of a memorial error. The compositor who frequently transposes words is probably trying to carry too much copy in his head.

The frequency of transpositions in these four comedies reveals most dramatically how Compositors C and D differed in their type-setting techniques. Compositor D transposed words only once in 25-½ pages: shall it for it shall (P1, 803). Compositor C transposed words nineteen times in forty-two pages.

TRANSPOSITIONS BY COMPOSITOR C

                                       

			Quarto	Folio
Ado	I4	302	you do	doe you
	16v	968	of us	us of
		986[*]	make but	but make
	K3v	1747	are you	you are
LLL	L3	476[*]	was Sampson	Sampson was
	L4	665	I will	would I
		674	shall we	we shall
	M3v	2108[*]	but vouchsafe	vouchsafe but
MND	N3v	700	we can	can you
	N5v	1170	all are	are all
	O1v	1736[*]	patcht a	a patch'd
MV	O4v	219[*]	is it	it is
	P1v	946	are you	you are
	P6v	2223	take then	then take
		2261	so taken	taken so
	Q1	2318[*]	shalt thou	thou shalt
		2321	home with me	with me home
	Q1v	2477	it in	in it
	Q2v	2699	even but	but ev'n

Almost all of Compositor C's changes make sense. Most of his transpositions involved a verb: nine of them are between a modal verb and a pronoun, and four more are about the placement of the adverb before or after the verb. Thirteen of his transpositions are in verse lines but, despite the possible change of emphasis in the line of

verse, his transpositions are nearly impossible to detect without referring to his copy. A few of his transpositions are obviously errors, as on O4v, 219, and P6v, 2223. Apparently Compositor C would read a line and then set it from memory, thereby occasionally changing the word order while retaining the sense of the line. Compositor D, on the other hand, kept checking his copy and rarely misread the word order.

Possibly some of Compositor C's transpositions were not accidental; rather, they may have been conscious attempts to make the line clearer. Some of his changes are smoother than the quarto reading; for example, the change on 16v makes the line smoother and clearer: "I remember a pretty jest your daughter told us of" (968). Similarly on K3v the Folio version, "Why then you are no maiden" (1747), is more direct. Still, making a line more regular might also be the result of memorial error—the compositor recalls a simpler phrasing. Some of Compositor C's transpositions may have been intentional, but I think they were simple memorial errors. As we will see, Compositor C was clearly prone to similar kinds of memorial errors.

SUBSTITUTIONS are the exchange of one or more letters or words. Approximately 20% of the substitutions by Compositors C and D are corrections of obviously corrupt readings in their copy. The rest are corruptions that they introduce. Unintentional substitutions are of two types: the first type is orthographic—the omission or deletion of a single letter or the transposition of two letters (literals), or the confusion of common roots like willingly and willfully. The second type changes the whole word, like can for should, often roughly synonymous with the original. In general the first type results from misreading the copy; whereas the second seems to be memorial. (See the chart in Appendix I.)

As with transpositions, Compositor C's substitutions are more frequent and seem to be more often memorial than Compositor D's substitutions, which seem to be more often misreadings of the copy. Compositor C averages a substitution every 32 lines; Compositor D substitutes on an average of once every 48 lines. Occasionally both compositors will substitute a number of words within a few lines (for example, Compositor C on L4 and Compositor D on N5), but generally they do not substitute words in clusters that might suggest illegible copy or compositorial fatigue. Compositor C makes five or more errors of substitution on eighteen of his forty-two pages, while Compositor D makes five or more errors of substitution on only five of his 25-½ pages. The difference does not appear to be patterned (due to the

nature of the copy or position in the quire, for instance); rather Compositor C seems to be more erratic.

Because of the low number of substitutions, it is difficult to see from the above lists if either compositor is prone to particular kinds of errors. The following chart divides their substitutions into some non-exclusive categories.

                 

	Compositor C	Compositor D
number of substitutions	174	68+2[5]
literals	89	38
same root	29	14
final s omitted or added	10	14
contraction, elision, or expansion	20	0
nouns and verbs	76	36
proper nouns	11	1
in prose	59	29

There are, of course, other possible categories, but these reveal some of the striking features of Compositor C's and D's substitutions. Over three-fourths of Compositor D's changes are of the orthographical sort. He had a particular fondness for omitting or adding a final s; nearly 20% of all his substitutions involve a final s. Compositor D consistently misread the word in his copy and composed a word that looks similar but means something different. Line 1415 in MND is a good example: "Whiles I in this affair do thee imply" (Q: apply); or line 509 in MV: "I like not faire teames, and a villaines mind" (Q: terms). This kind of error is easier to spot than are memorial slips and, when we know that Compositor D habitually makes such errors, easier to emend.

Compositor C, on the other hand, is more likely to make substitutions due to memorial error. Proportionately he does not have so many orthographical errors or misreadings as Compositor D; instead he more often forgets the exact wording or spelling. For instance, he occasionally does not recall if a word (especially a name) has been contracted or not; thus he makes such substitutions as upon for on (I4v, 444) and 'Fore for Before (L1, 1012). He also makes another kind of memorial error when he sets a word that he has recently used in a similar context. Compositor C is occasionally influenced by nearby words or phrases; an apparent example of either eye-skip or memorial error is immaculate for maculate (L3, 396), since immaculate appears in line 395 directly above Compositor C's error. A second example of

this kind of slip is the mixing of pronouns on M3v (you for her, 2083, and you for we, 2122), where Boyet is acting as a go-between for the two sets of lovers. Compositor C cannot match Boyet's ability to change the pronoun as he repeats the speech.[6] As one might expect of memorial slips, Compositor C changes the less significant connectives and qualifiers more often than the nouns and verbs.

Some of his changes appear to be deliberate. Many of the changes are roughly synonymous, as looke for see (K3, 546) and doe for use (N3, 589). These changes might be due to spacing or attempts to improve the copy, rather than memorial errors. About a dozen of his substitutions seem to be in response to spacing difficulties.[7] A good example is on crowded I4, where he changes at least to yet:

Another example of substitutions as a result of crowding can be found in lines 349 and 358 at the bottom of L2v. Sometimes Compositor C apparently tries to improve his copy beyond correcting the obvious corruptions. He straightens out proper names in the speech-prefixes (K1, 1256) and in the text (O2, 1957). He 'corrects' at least one malapropism by Dogberry (K3, 1650). Also, he seems more sensitive to meter than Compositor D; he drops (L4, 669) or adds (N2, 253) syllables to regularize iambic pentameter lines.[8] The only comparable sophistication by Compositor D is dombe for dead (K6v, 2531) for the sake of rhyme.

Because Compositor C's substitutions are most often either memorial slips or conscious sophistications which preserve the original meaning, they are harder to detect than Compositor D's. Also, Compositor C's copy is harder to reconstruct since he often fails to leave any orthographical clues.

OMISSIONS by both compositors generally seem to be either unconscious errors or attempts to save space, rather than deliberate

improvements. As one might expect, Compositor C omits words more frequently than Compositor D. He also is significantly more careless on some of his pages than others; over half of his omissions occur on eight of his forty-two pages. (See chart in Appendix II.)

Somewhat surprisingly, omissions occur more frequently in verse than in prose. Usually both compositors are sensitive to meter (though Compositor C has the better ear), and one would expect omissions to consist of extra, less important parts of speech, in prose passages. Both compositors usually omit a single word, often a pronoun or preposition. But only 15% of Compositor D's omissions fall into these two categories, compared with 40% for Compositor C. Once again this discrepancy can be explained according to their differing compositorial styles: memorial errors are likely to involve the less important parts of speech. Thus, Compositor D's omissions are often more significant than Compositor C's: for instance, he is slightly more prone to skip a line than Compositor C. Compositor D misses one or more lines three times in his 25-½ pages. Compositor C misses a line or more six times in his 42 pages, but two of them were probably editorial rather than compositorial. His two-line skip on K1v may be due to a reference to the Spanish, "and a Spaniard from the hip upward, no doublet";[9] and the three-line slip on K4v may be due to the four references to God in one line, "they serve God: and write God first, for God defend but God should goe before such villaines." A minor cause of omitting a line is eye-skip. On two occasions Compositor C omits lines because of eye-skip. I4 in the Folio is a classic example:

The second instance is on K3v, where 'do' is repeated at the end of four clauses in a row. Another minor cause for skipping a line is lack of space. Compositor C clearly skips a line because of lack of space on L1, the last page of Ado. Compositor D might have skipped a line on N6v because of lack of space, since the whole page is crowded (see line 1378). Yet curiously a space is left where line 1348+1 should be.

Lack of space is more frequently a cause for omitting a word, rather than a line. Some omissions in verse lines are clearly due to

space problems. On L1, where Compositor C omitted a whole line because of lack of space, he also cut words in lines 2639, 2640, and 2641, rather than have each verse extend onto a second line. Similarly, Compositor D dropped a word on L4v, where he also had omitted whole lines. If he had included 'maister' in line 779, he would have had to use two lines rather than one. Omissions due to spacing problems are few, however; most omissions by both compositors are simply careless errors.

INTERPOLATIONS may be Compositor C's and D's least interesting changes. In most instances they do not add anything significant; rather, they simply make the meaning more explicit by adding an article, preposition, or pronoun. Over a third of Compositor C's interpolations are articles. (See chart in Appendix III.) In contrast to the other kinds of errors, interpolations occur with equal frequency in both compositors' pages. Possibly a compositor prone to memorial errors, like Compositor C, is less likely to interpolate than to substitute or omit. One significant difference in their interpolations is that Compositor D adds words in prose proportionately more often than Compositor C. This difference is probably due to Compositor C's greater sensitivity to meter: only four of Compositor C's additions are in verse lines (M1v, N3, P4, and P5v), and three of them regularize the meter. Two of these verse interpolations emphasize parallel grammatical constructions: 'The sea will ebbe and flow, heaven will show his face' (M1v), and 'a husband / Hast thou of me, and she is for a wife' (P5v).

Compositor C's most significant interpolation is the new speech-prefix on L3v, 512. He attempts to correct the quarto's reading that gives the whole speech (504-525) to an unannounced, and thereafter silent, Queen. However, as Furness notes, the comma after Boyet is puzzling:[10]

Why did Compositor C split the speech between the missing Queen and the Princess, rather than give the whole speech to the Princess?

Compositor D inserts several interpolations more important than Compositor C's. Four of Compositor D's six interpolations in verse lines disrupt the regular meter (L5, 988, N4, O5v, and P1). On the

other two occasions, where he keeps the meter regular, he substantially changes the line. On L2 he has changed the quarto, 'Clymbe ore the house to unlocke the little gate' to 'That were to clymbe ore the house to unlocke the gate.' The change seems pointless; neither his interpolation nor his omission is an improvement over the original, nor does there seem to be some circumstantial reason for the change. The addition of passionate on N6 is a similar mystery. Although interpolations are not as important as the other kinds of changes, they provide a neat final example of how these compositors differ: Compositor D adds words in verse more often than in prose and spoils the regular meter; Compositor C adds words in prose more often than in verse and supports a regular meter.