University of Virginia Library

THE COMPOSITION AND PRINTING OF MIDDLETON'S A MAD WORLD, MY MASTERS [*]
by
Gerald J. Eberle

In a paper read before the Bibliographical Evidence group at the meeting of the Modern Language Association in Chicago in 1945 I made the point that careful bibliographical analysis of the text of Thomas Middleton's A Mad World, My Masters,[1] 1608, along with a study of Middleton's scribal habits, might afford a more generally satisfactory explanation of the aberrations in the quarto than the previously held theory of revision.

The publication since then of F. T. Bowers' article, "An Examination of the Method of Proof Correction in Lear,"[2] has led me to believe that the analysis of the composition and printing of A Mad World might be interesting in its own right, apart from its application to the editorial problem; for the first five sheets of the play were printed with three skeletons used in a pattern somewhat different from that in Lear. To complicate matters a bit, no variant formes in

these sheets appear in the nine copies of the first edition which I have collated.

Incidentally, the two compositors of the first five sheets were accomplished workmen, one of whom set almost all the type for these first sheets—all but D1-4 and E3v-4. The point is significant because apparently a single compositor was not pressed for time in setting all the type for sheets A, B, and C.

Some of the phenomena to be detailed here are sufficiently similar to those in Lear to recommend comparison, and sufficiently different to prove puzzling and perhaps enlightening. To facilitate comparison, I have used in the following table the same system of indicating running-title transfers that was used by Bowers.

Several oddities deserve special comment, if only as a warning to future students of running-titles used as bibliographical evidence. As indicated in the above table two slightly different titles are used on A4v. Printing starts with title IIIa, found in copies of the play in the Victoria and Albert Museum, the Bodleian, the Folger, the Huntington, and the New York Public Library.

After a considerable number of copies had been printed the broken initial A was replaced by a better one, but the rest of the running-title was not changed. The title in its improved form, IIIb, appears on A4v in the copies of the play in the British Museum, Harvard, the Library of Congress, and the Boston Public Library. Thereafter title IIIb appears in all copies on B3v and C3v, after which it is rejected.

Meanwhile what is apparently the rejected initial A from IIIa re-appears in a new title on B4v, labeled now IXa, which is used again on C4v and then set aside temporarily. It appears again on E4v but its broken initial A is replaced almost immediately, quite by chance by what is apparently the same initial A which had replaced it in title III. Title IXa is found only in the British Museum copy.

The notable fact, it seems to me, is that these pages are invariant except for the changed letter in the running-titles. If these pages were proofed, as their general excellence suggests, what sort of proofing before printing might allow the broken running-title to escape notice? That there was ample time for proofing in the first five sheets of the play is obvious when we study the probable order of the formes through the press.

The cumulative effect of admittedly inconclusive evidence leads me to believe that the outer formes of both sheet A and sheet B preceded their inner formes through the press. If inner B had been imposed and run before outer B, we should be obliged to account for the delay occasioned by the use in inner B of running-titles from both formes of A. This demands the assumption that sheet A was completed at the end of a working day, in which event both skeletons should have been ready for normal transfer to their respective quarters in the formes of B.

Instead we find new running-titles used in outer B, one of them with the previously discarded broken A. Apparently outer B was imposed before either skeleton of A was ready for transfer. Then while outer B printed white paper, the running-titles and furniture from both formes of A were used in inner B.

Similarly, if inner A had been first run, its running-titles would presumably have been first ready for transfer to inner B, in which case normal transfer of running-titles would have placed title IV on B3v instead of on B1v, and title V on B4 instead of on B2. But if outer A was first run, first rinsed, and first stripped, normal transfer would place title II on B4 and title IIIb on B3v, which is precisely where they are used. This conjecture, unfortunately, leaves the puzzle of why title I from A2v was not used. I suspect that by the time the workman was ready to impose B1v and B2 he had available the running-titles from the two adjacent type pages of inner A, which he preferred to use rather than the single remaining running-title from outer A plus another odd one.

It seems likely, then, that the order of formes through the press thus far was outer A, inner A, outer B, inner B; and the pattern of running-title transfers

could be indicated as follows: set X, set Y, set Z, set XY, or three different sets plus a fourth composed of parts of the first two.

This last point may seem unduly labored; the reason for the insistence lies in the odd repetition of that same pattern in sheets C and D, though of course the running-titles used in inner and outer C are not newly set.

The outer forme of C uses the titles from outer B, and is presumably first printed, to be followed by inner C, which uses three titles from inner B plus one new one. The new title can be plausibly accounted for if we assume that imposition of outer C was begun after outer B had been rinsed and stripped. But when the imposition of inner C was begun, the skeleton of inner B was not quite ready for transfer. Overlooking the set-aside running-title from A2v, the compositor set a new one and imposed C1v. By the time he had completed that task the inner forme of B had been rinsed and stripped and its running-titles were available for normal transfer.

The situation in sheet D is similar. Outer D, like outer B, uses in effect a new skeleton, though two running-titles previously set aside from A2v and B1v form part of the skeleton. Composition is again far ahead of presswork; the outer forme of D is imposed with a new skeleton, and the inner forme of D is being imposed shortly before the outer forme of C has been rinsed. Rejecting the broken title IXa, the compositor imposes three quarters of the inner forme of D, making two normal transfers from C3 to D4 and from C1 to D2, and one unusual transfer from C2v to D3v instead of to D1v. By this time the inner forme of C has been rinsed and is available. The normal transfer of running-title from C1v to D1v completes the imposition of the inner forme of D.

It will be noted that inner D might have been imposed entirely with the skeleton from outer C, had the broken running-title not been rejected. It remains a fact, nevertheless, that the inner forme of D uses running-titles from both formes of C, thus completing the cycle of four sets as in sheets A and B with the same pattern, X, Y, Z, XY.

Laboring the point may be wasted effort, especially when one cannot explain exactly why the pattern exists; the point is made because it suggests that sheets B and D were completed at the end of a working day and immediately calls attention to a somewhat similar pattern in Lear.[3]

In any case, composition was well ahead of presswork, the new compositor at work in D is not needed to relieve any pressure, and the compositor evidently had sufficient type for sixteen standing pages. As a corollary it seems clear that the resources of the shop were fully utilized in the printing of these sheets.

If both formes of D marked the end of a working day, both skeletons might have been available for the imposition of either forme of E. As a matter of fact, both formes of E use running-titles from D. In sheet E, however, it is

rather difficult to determine which forme was first imposed and printed because the outer forme of E uses two running-titles from inner D and in addition a single running-title from each of two other formes, outer C and inner C.

The problem is not insoluble, however, if we can correctly interpret the following facts:

• 1. Title X from C1v is used on D1v and E2v.
• 2. Title V from C2 is used on E1.
• 3. Titles IIIb and II from the other two pages of inner C were probably distributed before outer E was machined, for we note that the initial A from title IIIb is used to replace the rejected A in the previously set aside title IXa that now appears on E4v.
• 4. The use of title X on E2v and VIII on E3 is the normal pattern of running-title transfers from an inner to an outer forme; and the only sensible reason why the other two titles from inner D were not used must be that the other quarters, E1 and E4v, were already imposed.

In short, the compositor went back to earlier set-aside running-titles because neither skeleton of D was available, but by the time he had imposed E1 and E4v, side-by-side on the stone, the skeletons of D had been rinsed and were available. The needed two running-titles were taken from the quarters of inner D which are appropriate to normal transfer from inner to outer formes, D1v to E2v and D4 to E3.

Finally the entire skeleton of the outer forme of D was transferred in normal order to the inner forme of E. And that concluded the printing of the first five sheets of the play.

With sheet F a wholly new situation arises. All running-titles from the first five sheets are discarded; they are replaced in sheet F by four new running-titles, one of which has a comma instead of a period after the word "Masters." These four are used in both formes of F and G and in the inner forme of H. Four other running-titles, one of which has a badly broken upper case W in the word "World", were used in outer H, and three of these in the final three pages of the play, I1-2.

Apparently the outer forme of H was imposed and printed before the inner forme, a second skeleton being introduced to speed up completion of the job. It is likely that the order of the formes through the press was inner F, outer F, inner G, outer G, outer H, inner H. There is obviously nothing to prevent simultaneous imposition of both formes of I. In fact half-sheet imposition seems likely, the running-titles from outer H being transferred as follows: H1 to I1, H2v to I1v, and H3 to I2.

Signatures F-I differ from the earlier sheets of the play in another important respect—composition. The two efficient compositors of sheets A-E were replaced by one compositor who is not an accomplished workman. His work is easily distinguished from that of the other two by his spelling, by his handling

of contractions and apostrophes, and by his sometimes stupid blunders, which are noticeable chiefly in sheet H. Evidently, sheets F and G underwent some sort of correction, for we find only three variants in these two sheets, two on F1v and one on G3v. When we find numerous errors throughout sheet H, the comparative absence of errors in F and G strongly suggests that F and G underwent some sort of proofing earlier.

Sheet H is truly amazing, exhibiting as it does both formes variant in wildly erratic fashion. For instance we find in outer H in the uncorrected state these flaws: on H1 seven stupid blunders; on H2v a generally good page with one nonsense line at the bottom; on H3 a faulty speech-tag, An.c later changed to Anc.; on H4v a good invariant page.

In inner H the situation is much the same; on H1v there are five stupid blunders limited to the upper half of the page; signature H2 is invariant and good, and shows signs of having been set or extensively corrected by the compositor of the greater part of the first five sheets; on H3v one word is wrong and one speech-tag is changed from Short. to Short.R.; signature H4 exists in three states. The wholly uncorrected state,[4] found only in the British Museum copy, has the incorrect catchword Short. All other copies properly read Mast Pan. The third from the last line of text in this copy reads in part "ſhall I intreat a cuteſy?" All other copies properly add the missing r in "curteſy" and omit one of the l's in "ſhall", probably to avoid the necessity of spacing out the line anew; it was certainly not lack of room in the line. The British Museum copy uses Short. as a speech-tag five times. The partially corrected states remove all the errors listed above except the use of Short., which is hardly an error. The twice corrected states change Short. to Short.R. twice near the top of the page, but leave it unexpanded the other three times it occurs. Here surely is haphazard correction!

To complete the discussion of composition we may add that the three pages of sheet I exist in two states. Signature I1 is uncorrected in the Dyce copy of the play in the Victoria and Albert Museum. The variants in I1v and I2 seem to me to be the result not of correction strictly speaking but of an accident in the course of printing. In those copies of the play that do not lack the FINIS the upper part of signatures I1v and I2 near their common margin show signs of badly justified lines with type slipping from side to side. I assume that the type in this area pied locally and necessitated re-adjustment. In those copies that lack the FINIS these badly justified lines have been tightened up, but in those areas mistakes have crept in; for instance the i from "ſeriouſly" slips down into the next line with this result:

When we add to all this evidence of careless work the significant fact that the badly damaged running-titles in the last sheets of the book were not replaced, as were those in the first sheets, we get a picture the cumulative force of which demonstrates beyond a doubt that the two bibliographically distinguishable sections of A Mad World were composed and printed in quite different ways.

Sheets A-E were carefully composed and printed with all the resources of Ballard's shop directed toward turning out a reasonably good piece of work. Then for some reason the printing of A Mad World was stopped. The distribution of running-titles as outlined in step number three above suggests that the decision to set aside this play was reached before sheet E was in the press, but after its type had been set at least in part. When the printing was resumed, a wholly different attitude prevailed. Composition was done by one man in the main, probably an apprentice, certainly not a good compositor.

For, as I demonstrated before the Bibliographical Evidence group of the Modern Language Association in Chicago, the entire play shows signs of having been set from Middleton's holograph manuscript, and Middleton's handwriting is very easy to read, unlikely to prove troublesome to a good compositor. Even the printing did not make use of all the facilities of Ballard's shop, presumably because his accomplished workmen were busy with some other job.

This entire analysis undoubtedly introduces more questions than answers, but it may prove helpful to some other student with a similar bibliographical problem.