University of Virginia Library

Spacing Evidence

Considerable aid has been obtained from observations of the compositors' habits of spacing commas at the end of lines of text, and inside short lines. There can be no doubt that the four compositors had different spacing habits which were habitual and distinctive: these practices usually confirm Hinman's compositor identifications. Even a facsimile of the Folio shows that some commas inside lines are followed by spaces, and that some are followed closely by the first letter of the next word: illustration of this is not necessary.[12] If there were any reason to doubt this, perhaps on the ground that different sorts were irregularly centered on the body of the type, the instances of inked spaces and quads which I have noticed in the Folio would confirm that internal spacing in short lines is a real and not an imaginary phenomenon. There is an inked space after a comma in a short line at Jn. 1970 and with commas in long lines at AWW 428 and Jn. 2207, and before a colon in a short line at TN 1324. Other interesting examples occur at Wiv. 2600, TN 1771 and R2 1074 where only one space of the two which must have been used has inked.[13] A wider space can be seen at the end of 1H6 1724.

The proportions of spaced commas to non-spaced commas have been tabulated in the appendices.[14] Compositor B was the most regular in preferring to use spaces, and non-spaced commas are relatively infrequent in his pages. They sometimes occur, as will shortly be seen, in special circumstances. Compositor C also prefers to space commas but in this, as in many other practices, he is not consistent; his pages show many more commas without spaces than appear in B's pages. Usually but not always the relative proportions are sufficient to distinguish B from C but of course it is rare that spacing is the only evidence available to decide between them. Compositors A and D prefer not to space their commas, D's habit being more strongly marked. However, it is worthy of notice here that the spacing practice of A in the histories and WT is more like that of D in its preference not to space commas, and that A in the comedies apart from WT is sometimes inexplicably inconsistent in his practice from page to page.

Although the counts of spaced commas amply confirm this general description of the compositors' habits, this evidence is sometimes of limited use. The number of commas, although they are usually frequent, sometimes varies considerably from page to page. Many pages have few short lines and so the count of spaced commas provides small guide to compositor identification. Consequently, on a particular page it may be impractical to rely on this evidence to distinguish between A and D who did not favour internal spacing, and B and C who did. Further, sometimes there are anomalies for which it is difficult to account. It may readily be supposed that a compositor who had almost exhausted the supply of spaces in his type-case when justifying passages of prose might be obliged to relinquish his preference for spacing in short lines. This might explain why sometimes there is a surprisingly small number of spaces in some pages set by B or C.[15] Or, on the other hand, a compositor such as A or D who preferred close spacing might in a similar situation be led to use more spaces than usual in short

lines after he had become accustomed to setting them in justified lines. This might explain why compositor A in A5v, which is mainly justified prose (with wide internal spacing after punctuation marks) has 20 spaced commas in short lines to 3 non-spaced.[16] It might also happen that the supply of spaces was depleted when non-Folio matter was set; irregularities of spacing practices might occur as such material was distributed back into the case which was being used for the Folio. When C's share of quires O, P, Q and R are studied in the order of composition, he is seen to depart from his preference for spaces after commas for sequences of pages in an order which might be explained in this way. Fortunately there is usually other evidence to show compositor C in these quires, despite the inconsistent evidence of the spaces.

In the pages of King John which Hinman assigns to compositor C (a1-2v, a5-b4) there is a large number of commas at the ends of short lines which are preceded by spaces. There are 25 on a1v alone. These commas are more widely spaced than the spaced terminal commas which occasionally occur on pages set by the other compositors, and they are quite characteristic of C. Inked spaces before terminal commas can be seen at TGV 531 and MND 1166. As such spaced terminal commas are absent from a3-4v which Hinman assigned to B or alternatively to C, his first attribution is the more likely to be correct. The number of such commas varies considerably from page to page, usually on account of long lines or because there were not many terminal commas (C did not space every comma at the end of a short line), but occasionally justified lines which extend to the full width of the column show widely spaced commas at the end of the line. As these are not often found in prose passages set by other compositors, they may be used with spaced terminal commas in short lines to support attribution of pages to compositor C. There are occasional instances of spaced terminal commas in pages which were clearly set by other compositors (e.g., at l. 1576 on E2, set by B), but, as the tables below show, they are undoubtedly characteristic of C. When, therefore, they have been observed on pages attributed to other compositors, I have taken this as a warning that such pages need especially close examination. I have not allowed this evidence to outweigh the testimony of spellings and the other evidence which will be discussed shortly. In fact, spacing evidence is consistent overall with the spellings and compositor identifications of Hinman and Miss Walker, and this shows that it gives good evidence for compositor identification in the Folio.

Some tables will illustrate the different compositorial spacing habits, and the nature of the evidence which I have been able to apply to doubtful pages. In Jn., which was set by C and B, the distribution of internal and terminal spaced commas in an equal number of pages set from case y is:

Spacing in 'King John'[17]

                         

COMPOSITOR C				COMPOSITOR B
Page	Terminal spaced,	Medial spaced,	Non-spaced,	Page	Terminal spaced,	Medial spaced,	Non-spaced,
a1	5	24	10	a3		48	1
1v	25:1j	32	27	3v		40	4
2	25	31	14	4		42	4
2v	23	24	18	4v		42	5
5	12	51	14	b1	1	47	6
5v	13:1j	47	15	1v		65	3
6	11	51	18	2		50	4
6v	21	42	15	2v		61	3
b3v	15	22	13	3		53	7
4	27	38	7	4v		51	4
	177:2	362	151		1	499	41

A glance at this table shows that although both B and C are alike in preferring to insert spaces after commas, B's habit is much stronger than C's and he rarely uses a space with a terminal comma.

A group of A's pages from the early comedies, excluding pages with large amounts of prose, and anomalous pages such as B3v and 4v shows how his habit differs from C's.

Compositor A's Spacing[18]

                         

COMPOSITOR A				COMPOSITOR C
Page	Term. #,	Med. #,	Non sp.	Page	Term. #,	Med. #,	Non sp.
A1v	1	10	49	A2	1	61	6
B1		19	40	2v	6	35	10
1v	1	17	58	B2	5	54	12
5		20	30	2v	2	43	15
5v		19	48	3	1	52	11
C4v		20	40	6	4	21	3
5	3	6	49	6v	6	33	6
5v		12	31	C2v	12	25	21
6	2	15	43	3	7	39	18
6v		14	39	3v	8:1	38	23
	7	152	427		52:1	401	125

Here, plainly, A does not prefer to use spaces after commas, and his habit is the converse of C's. The spaced terminal commas in A's pages are not found in such relative profusion in later A pages, and few of them have the wide spacing that is so characteristic of C.

A selection of pages set from case z, including those that Hinman used to prove that there was a fourth compositor in the Folio comedies,[19] illustrates the spacing practice of compositor D. Here, however, I must admit pages with large numbers of long lines, in order to make up a sufficiently large selection.[20] Contrasting pages set by C show also how prose affected his usual practice of spacing commas: it demonstrates that nevertheless he can be identified by spaced terminal commas.

Compositor D's Spacing

                                 

COMPOSITOR D				COMPOSITOR C
Page	Term. #,	Med. #,	Non sp.	Page	Term. #,	Med. #,	Non sp.
*K2v		2	12	*K1	10	11	8
* 5v	1j	1	43	* 1v	9	23	27
* 6		0	18	*L1	6	29	25
L2		5	17	1v	14	19	18
* 4v		4	20	3v	10:1	17	12
* 5		9	45	4	10	25	10
5v		8	29	*N2	6	13	20
6		5	24	3	15	19	25
* 6v		4	12	3v	11	38	27
N4		9	39	5v	16	34	31
* 4v		10	27	*O1	6	16	9
5	1j	9	31	* 1v	7	25	30
6		6	75	* 2	1	25	14
6v		7	48	2v	11	36	46
	2j	79	440	132:1	330	302

Compositor D's preference for close spacing is stronger than A's, and he is more consistent. Sometimes, however, it is not possible by this evidence alone to distinguish them in pages with many long lines.

Were spacing the sole evidence by which compositors could be identified, it would not be very helpful even though there would be little difficulty in distinguishing C from B when C used his characteristic spaced terminal commas. Taken with other peculiarities, however, it is useful evidence, especially when discriminating spellings are scanty or ambiguous. Page H2v in Err. has 2 'do', 4 'doe' and 3 'here' spellings: one of the 'here' is a rhyme and another occurs in a long line. The spacing, however, shows that this page was set by C and not

by D to whom Hinman assigned it, mainly on account of the 'do' spellings. Another page in Err. with scanty and mixed spellings is H4 of which the spacing is, column a: 6/15/7; column b: 0/11/17.[21] This evidence is entirely consistent with the habits of compositors C and D.

Hinman occasionally referred to A's habit of representing dashes by two or three hyphens. This is not helpful evidence in the comedies, for the only occurrence of this kind of dash before WT is on D4 which Hinman and Miss Walker assign to C; this is supported by spacing evidence (1:1/13/9). Two short dashes are found in A's D3v, but dashes are not good evidence in the comedies, for none occurs in D's pages and both B and C prefer to set a dash by a long rule. Nor can catchwords which give the speech-prefix or the speech-prefix and the first word of the speech, supply clear-cut evidence. Compositors A and B in the histories are distinguished by their almost invariable practices, A for giving the first word of the speech, B for omitting it. But in the comedies, although B's habit is uniform, there are five pages on which A omits the first word of the speech out of 11 with this kind of catchword. C's practice is inconsistent (22 with the speech word, 11 without) but D almost always provides it: out of 19 pages, only on G1 is the speech word omitted from the catchword.

There is another compositorial practice which sometimes helps to identify the compositor of a page. There are occasions on which compositors, obliged to turn over a line of verse which was too long for their measure, and unable or unwilling to set it as a turn-down or turn-up at the right margin of the column, have carried the remaining part of the line on to the next line. Often these have been indented from the left-hand margin. Lines 2658-9 on A's G5v (the only occurrence of this in his pages) is an instance. What is remarkable about this is that it is infrequent in the pages of A and C (with whom D is most likely to be confused) and relatively frequent in the pages of B and D. One might expect that over a large number of pages set by the respective compositors the number of occasions on which they encountered long lines of verse requiring to be turned over would be roughly equal. To WT, however, the figures are:

Indented Turn-overs

     

COMPOSITOR	A	B	C	D
PAGES	1	14	2	20
OCCURRENCES	1	20	2	94

It is therefore more likely than not that any page with more than one such indented turn-over was set by compositor D, and it is very unlikely to have been set by A or C.—This evidence, together with spellings, was employed to re-examine Hinman's compositor attributions in the comedies. Also helpful was the different compositorial treatment of 'will' and 'shall' elisions (we'll, they'le, sheel and so on), especially as knowledge of Ralph Crane's spellings made it possible to allow for the influence on the compositors of copy in his handwriting.[22] There were also contrasting orthographies of 'the' elisions in i'th, o'th' and the like which served to distinguish compositors. The occurrence of this evidence is mentioned when necessary and may be consulted in the appendices.

After the first form of this paper, in my thesis, was finished, Professor Cairncross's discussion of the compositors (largely in the comedies) came to hand; this offered the possibility of a stringent check on the methods and conclusions set out here. Mr. Cairncross is concerned mainly with C and D. To distinguish them he uses, besides the familiar spellings, the evidence of the future elisions (wee'll, they'll, etc.), the layout of run-on stage-directions, and the setting of 'fareyouwell' as one word for compositor C; for D, the future elisions (weele, theyle, etc.), -ie or -aie endings, and 'the dropping of final l, as in cal, fal, wil'. He confirms or changes Hinman's tentative attributions for 66 pages of the Folio up to sig. V. By my analysis, Cairncross is correct for all but 28 pages, although not often on the basis of the evidence he brings forward. It would be tedious to discuss every point of difference page by page but, undoubtedly, it is important to examine the kinds of evidence he has drawn upon and the way in which it has been used: nothing which leads towards certainty in these studies should be neglected.

Mr. Cairncross distinguishes C from B by C's practice of indenting a two- or three-line stage direction by two or three ems; B centers the following lines after the first.[23] Compositor C is clearly partial to this kind of arrangement and there is nothing new about this. What is new is the extent to which Mr. Cairncross is prepared to take the occurrence of the indented stage direction as a sign of C's hand without the support of other evidence. Also, the value of his paper is seriously

compromised by his reluctance to notice evidence which does not support his conclusions. For example, compositor C's kind of indented layout may be seen in stage directions on A1 and I3 (together with a B-type centered direction)[24] and I3, and in the letter on G2v, which are all B pages; other indented directions occur on A5 (A), B1 (A) where there are 3 indented and 1 centered direction, and H3v (my compositor D). As for compositor C, there are centered directions on C2v, G4, H2 and K4v which he reassigns to C, and on P4v. Mr. Cairncross, observing that C disappears from F after Jn., concludes that 'An important negative result of the new criteria is the elimination of Compositor C from quires d-gg . . . and t-x . . . oo and pp. . . . C's run-on stage directions are also absent.'[25] but this is not supported by C-type indented SD on d1v (Hinman's B?C but B from the evidence of spacing), l3v (B), n1 (A), p2v (A), t6, t6v, v2 (all C?A), x1v (B), x3, x4 (C?A), aa2v (B), aa3v, cc4 Gg2v, Gg3v oo6, xx5, yy2, bbb2v, bbb3, bbb3v (all B). The elisions are in general good evidence of C's hand and so too is 'fareyouwell', although again, Mr. Cairncross omits to mention an instance of this on P2, which he reassigns from A to D.[26]

With the help of such evidence Mr. Cairncross transfers 22 pages in sigs. F, G, H, I and K from compositor A or D to C;[27] in the same pages I find the hands of four compositors. The relative likelihood of Mr. Cairncross's or my attributions being correct should appear more clearly from the following discussion. It needs to be mentioned here however that his 'new' evidence is distributed very sparsely indeed throughout these pages. For instance, in F2, F3v, F6v, G1v and G2, the first 5 pages he discusses, there are elisions in long lines on F2, F3v, G1v, no new evidence at all on F6v, but G2 at last provides a single elision in a short line. On G5v the sole evidence adduced for reassignment to C is 2 F-type elisions. Only on two pages (H3v, I4v) does more than one kind of new evidence support reattribution.

Mr. Cairncross's discussion of the evidence from which pages were confirmed or reassigned to compositor D is seriously inadequate. The elisions, as I found, are good evidence. For D's spellings of -ie and -aie and 'l/ll no statistics at all are given. In fact, when one surveys 5 pages (O5v, O6, T5, V3 and V3v) which he attributes (correctly) to D,

there are 86 -ie/aie to 70 -y/ay spellings and, mysteriously, only 2 -l spellings in long lines to 113 -ll forms.[28] In P1-3v which he assigns to D, the proportion of -ie/aie to -y/ay spellings favours only his attribution of P3v, and in the same 6 pages, there is only one -l spelling, and that is in a long line. On P4 however, there are 5 -l spellings, but this page was unquestionably set by C. Neither of these spelling tests then can on this showing be used to characterise D, and Mr. Cairncross's D evidence reduces to the elisions. It is not surprising, therefore, that he was not able to detect the alternation of C and D in H1-4 which he reassigned to C.—To conclude, I have re-examined all the reattributions discussed in the following pages and I have found no reason to alter any of them in light of Mr. Cairncross's observations.