University of Virginia Library

The Prentice Hand in the Tragedies of the Shakespeare First Folio: Compositor E
by
Charlton Hinman

THE ANALYSIS OF VARIOUS LARGE BODIES OF hitherto unexploited bibliographical evidence in the Shakespeare First Folio is producing many surprising results. It is showing, for one thing, that much of the textual work of the past—much, indeed, that is now being done—proceeds from false premises. But on the positive side this analysis is making possible the reconstruction of the printing-house history of the volume in a fullness and with a precision that would have seemed impossible only a few years ago. In all of its major outlines, moreover, and also in many of its details, this reconstruction can be made with an almost absolute certainty. The evidence upon which it is based is for the most part physical and unequivocal, and leads to conclusions of which the validity is not subject to debate. Points are of course reached where probability must replace fact and speculation must replace demonstration; but inference can now be exercised within, and governed by, a very comprehensive structure of demonstrable truth.

From the standpoint of Shakespearian textual study, perhaps the most important general fact recently discovered about the First Folio is that it was throughout set into type, not by successive pages, but by formes. Copy was therefore always "cast off" (often long in advance of actual setting and printing); and casting off commonly entailed various space-adjustment practices, some of which affected the text itself. The formes for a given quire were usually set by two compositors working simultaneously, one on each page of the forme in hand; for only thus could typesetting stay abreast of presswork, keeping the single press that printed the book more or less continuously supplied with Folio material. A preliminary report of these facts, together with certain

observations about their textual implications, has already been made.[1] The present article is concerned with a more particular aspect of the printing of the Folio, though with one that involves some of the most important of Shakespeare's plays. The primary aim in what follows is to describe some of the irregularities that characterized the printing of the Tragedies, and to suggest that they all stem from the activities—and the limitations—of the very inexpert compositor who not only set Titus Andronicus, Romeo and Juliet, and certain pages of Troilus and Cressida but who also had a significant hand in Hamlet, King Lear, and Othello.

That Titus and Romeo were not set by the regular Folio compositors, A and B, has been known for some time (see Note 8 below); but the most important facts about the third compositor of the Tragedies have not, I believe, been pointed out; nor has it been recognized that Compositor E, as I have designated him, is in part responsible for the Folio texts of Hamlet, Lear, and Othello.[2] His work in these major tragedies has in fact been confused with that of Compositor B—to B's considerable prejudice in our eyes, and to our own further confusion in trying to determine B's characteristics both here and elsewhere in the book.

The salient facts about Compositor E—both what he did and how well or ill he did it—have so immediate a bearing upon various editions of particular plays now being prepared, and also upon certain more general textual studies now in progress, that it seems worthwhile to offer this preliminary statement at once.[3] The full story, with all of its supporting evidence, must be reserved for the comprehensive account of the printing and proof reading of the First Folio upon which I am now engaged.[4] What will be said in the present article about the

order of formes followed in the Tragedies, about proof reading, and about various other matters is essentially factual. The most interesting thing about Compositor E, however, is inferential. Hence nothing further is claimed for a good part of the "argument" of this essay than that it appears to follow necessarily from the observed facts.

I. Titus Andronicus and Romeo and Juliet

Scores, even hundreds, of individual types become distinctive through injury as work on the First Folio goes forward. The precise point at which a given type suffered injury can frequently be observed, and the damaged type then traced in its subsequent appearances through the book. When large numbers of such types are catalogued and studied, very definite patterns of recurrence become apparent, and from these it is possible to determine many things about the printing of the book. One of these things is the exact order in which the various formes that make up the First Folio were set.[5]

Throughout the Comedies and most of the Histories the progression was, on the whole, very regular. Setting usually began with the inner forme for the inner sheet; then came the outer forme for the same sheet, then the inner forme for the middle sheet—and so on. The sequence[6] was thus 3v:4, 3:4v, 2:5v, 2:5v, 1v:6, 1:6v and this order was followed so often, despite occasional deviations from it, that it can clearly be regarded as the normal order of formes within a quire. Normal too is progress from quire A to quire B, then from B to C—and so on, in regular alphabetic sequence. Here again exceptions must be made. Special circumstances sometimes produced irregularity in the order of the successive quires of the Comedies and Histories. Quire X was followed, not by quires Y and Z, but by quires a and b—though these were then followed by Y and Z. Other departures from strict alphabetic order are also found; but this regular order was nevertheless the one normally followed. Never, moreover, was the setting for a new

quire begun before all six formes of the preceding quire were in type. Hence it is very surprising to find that the printing of the Tragedies was characterized by what at first appears to be extreme irregularity. Not merely does the progression both of formes and of quires seem abnormal, but it is distinctly exceptional for one quire to be finished before another, and usually one far removed from the first in the alphabetic sequence, is begun. Only very late, when quire tt is reached, is there a return to normal; and Antony and Cleopatra and Cymbeline are the only two plays in this section of the book that were printed with the same regularity that we find generally in the earlier sections.

The change in procedure which produced these irregularities took place somewhat before the Histories had been finished. Quires t-x, containing the end of Richard III and the whole of Henry VIII, were not set immediately after quire s but only much later, after all of Macbeth and part of Hamlet had been printed.[7] Next set, after the completion of quire s in the Histories, were certain formes for quire aa—and work on the Tragedies was under way. But quire aa (containing Coriolanus pages) was still unfinished when one forme of quire dd (two pages of Titus Andronicus) was set. And presently quire kk (part of Julius Caesar) was composed; then another forme of quire dd, then two formes of quire ll (more of Caesar) and another forme of dd (Titus again). Four formes of quire bb (Coriolanus) were immediately followed by certain formes of quire mm (Macbeth) and of quire ee (Titus and Romeo). Later, after nn1:6v (Macbeth and Hamlet) and then ff2:5v (Romeo), came quire t (Richard III and Henry VIII); but other formes for Romeo and for pre-cancellation Troilus were worked before quires t-x (and hence, at last, the Histories) were finished. The final Romeo forme (2gg1v:2) was set between two formes of quire oo, for Hamlet. Quire Gg (Timon of Athens) was begun immediately after rr1:6v (King Lear), but a start was made on quire ss (shared by Lear and Othello) before Timon had been completed.

This is surprising indeed, and at first seems not merely irregular but chaotic. Yet a certain method in this madness begins to emerge as soon as the order in which the successive formes were set is scrutinized in the light of information as to which compositors set them.

The greater part of the First Folio was set by two compositors known as A and B. But Titus Andronicus and Romeo and Juliet are evidently the work of someone else—of the man to be referred to hereafter as Compositor E.[8] The work of A and B can almost always be identified quite positively by means of very simple spelling tests. E's spelling, however, although in some respects more like B's than like A's, is characterized by a wild mixture of many of the alternative spellings (notably do-doe and go-goe) that are most useful in distinguishing A and B. Thus pages set by E usually show an olio of certain spellings that is rarely found in the work of either A or B; and such a mixture occurs throughout Titus and Romeo. In these two plays, therefore, spellings alone may be taken as a sufficient basis for the identification of Compositor E—although, as will soon be seen, there are also other ways of determining what he set, and in later plays he can in fact be distinguished from A or B independently of spelling tests. In any event there can be no doubt that Compositor E set Titus and Romeo—or, in bibliographical terms, part of quire cc and all of quires dd, ee, ff, and 2gg (the last being the two-leaf quire at the end of Romeo that immediately follows quire ff).

In the light of this fact a re-examination of the order of formes in the Tragedies proves highly instructive. The entire sequence cannot be considered here, but a representative part of it will at once make clear the essential nature of the whole. The following sequence, indeed, is thoroughly typical of the progression from the first forme of quire aa to the third forme of quire pp—immediately after which a very significant change, to be described in due course, took place. Beginning, then, with the first forme of quire mm, work on the Tragedies proceeded thus (the upper-case letter above each page-designator showing which compositor set that page):

The formes set by Compositor E have been underscored for the sake of clarity. It will be seen at once that if these formes are omitted the progress is absolutely regular in quires mm-nn and t-v. The six formes of mm were set in what we have already seen to be the normal order within a given quire, and mm was followed by quire nn where precisely the same order of formes is found. At this point—with the composition, that is, of quire nn—further work on the Tragedies by A and B was temporarily suspended while these compositors set the three quires (t-x) that finally brought the Histories to completion. The order of formes within quires t and v (and also x, which followed) is again, however, absolutely normal. And it may be added that when quires t-x had been finished a return was made to quire oo, the next quire in the alphabetic sequence of the Tragedies, where again the normal order of formes within the quire was followed, as it was also in succeeding quires.

Also apparent is the fact that the formes set by Compositor E themselves show normal order. No two of them are immediately consecutive within the larger sequence; but omit the material set by A and B and the progression in what remains is again completely regular, always from inner to outer forme and from inner to middle to outer sheet; and quire ff follows quire ee in the proper alphabetic order. Thus what we seem to find is a very regular progression of formes within two independent but mixed sequences. The appearance is in some respects misleading, as will be seen; but regularity—method of some sort—is patent.

It should also be noticed that the first three formes of quire mm were set mostly by Compositor A, Compositor B having no part whatever in mm3:4v; and that B alone, on the other hand, set the first four formes of quire nn and also one forme (v2v:5) of quire v. Although A and B together set most of the 24 formes of quires mm-nn and t-v, each compositor is sometimes found setting one or more formes by himself. This, too, is typical of the whole Coriolanus-Hamlet progression; and it is important in showing that each of the two regular Folio compositors was fairly often required somewhere else than at work on Shakespeare.[9]

Both A and B, moreover, were sometimes absent, and many of the formes set by E can only have been put into type at such times. For Compositor E always worked from one or the other of the two type-cases that were used in the printing of the First Folio; and formes set by E are often found between formes that were jointly produced by A and B using the only cases of First Folio text type anywhere available in Jaggard's shop. See, for example, ee1:6v. The immediately preceding forme, mm1v:6, was set by A and B working together, one at each of the two cases; and neither of these cases can have been available to E, of course, while A and B were using both of them. Probably the two pages of mm1v:6 were not finished at exactly the same time, and one case was doubtless vacated, as it were, somewhat before the other. But the two pages of mm1:6v, the forme immediately after ee1:6v, were also set by A and B together. Thus there can be little doubt that both A and B were absent while most, at least, of ee1:6v was set; and that they returned at approximately the same time to work together on mm1:6v. And precisely the same considerations apply to ff2v:5, 2:5v, 1v:6, and 1:6v. Five of the eight Titus-Romeo formes in the present sequence appear—and can be shown to have been set—between formes jointly worked by A and B. Form ee1v:6, moreover, although it was not set from the case used by Compositor A to set mm4v, was set from the same case as that from which A set his next page, mm2v.[10] Hence A (as well as B, who set nothing between column a of mm4 and column b of mm5 in the present sequence) must have been absent during the setting of at least part of ee1v:6; and similar evidence shows that B as well as A must have been elsewhere when ff3v was set. Thus only ff3:4v—only one of the eight Titus-Romeo formes in our sequence—can have been set throughout simultaneously with any of the material in this sequence that was set by A and B or either of them.

This point is emphasized in order to show that the setting done by the third compositor of the Tragedies was not done independently of the work done by A and B and simultaneously with it. The third compositor never worked from a third type-case; and the same skeleton that is used for what is set by A and B is always used with E's formes as well.[11] Both A-B formes and E formes therefore quite certainly belong in a single sequence, are integral parts of one operation—in which, however, a forme set by E is always a kind of substitute for an A-B forme and is never set except in the absence of at least one of the regular compositors. Thus Compositor E set only what may properly be called the intercalary formes of the sequence; and the otherwise orderly progression of the regular formes is larded with intercalary material set by a substitute compositor when one or both of the regular compositors had to be elsewhere.

In this is the essential puzzle. All would be understandable enough if only it could be shown that Compositor E set certain plays while A and B worked independently on others. Thus we should indeed have quite separate sequences of formes, each following a normal order and each produced pari passu with the other in a two-part operation obviously designed to hasten the composition of the book. Yet, as we have seen, it is perfectly certain that we do not have two really separate sequences. And why should Compositor E, although acting only as a substitute for A or B or both of them, always work on different pages altogether, always set other quires than those being set by the regular compositors? There is, I believe, only one explanation that accounts both for the irregular order of formes described above and for certain other phenomena found later in the book: Compositor E was incapable of setting manuscript material acceptably. It was indeed desired, not so much to speed up work on the Tragedies as to avoid at least some of the delay caused by the absences of one or both of the regular compositors who were so often required temporarily elsewhere. Compositor E kept work on the Shakespeare folio going forward. His activities, however, were necessarily confined to printed copy; and hence formes for Titus and Romeo (set from printed quartos) are

found scattered among formes for Coriolanus, Caesar, Macbeth, and Henry VIII (all set from manuscript copy).

Some of the most striking evidence upon which this suggestion is based is encountered only after the point in quire pp at which the alternation between regular and intercalary formes ceases, another method of working is adopted, and E is found in an important new role. This will be described in due course, and the whole of the evidence as to E's chief limitation can then be reviewed. First, however, certain facts about E's earlier activities should be noticed.

(1) The Folio text of Titus Andronicus contains one scene (III.ii) not found in the quarto. This scene (about 85 lines, falling partly in dd3v and partly in dd4) must have been set from manuscript copy and was undoubtedly set by Compositor E. Once, therefore, E did set from manuscript. Now it cannot be determined, in the absence of the copy itself, precisely how well or ill Compositor E managed to reproduce the manuscript text of Titus III.ii—or, of course, how long it took him. But Professor H. T. Price, now preparing the New Variorum edition of Titus, has assured me that this scene is certainly one of the most corrupt passages in the play. Which, considering the rest of Folio Titus, seems to be saying a good deal.[12]

(2) Whatever E's capabilities with respect to manuscript copy, it is clear that he could not set even printed copy well. This is almost immediately apparent to anyone who will examine Folio Titus and Romeo with half a proof reader's eye. They are full of errors, and of errors not found in the printed quartos from which they were set. The liberties which Compositor B was wont to take with copy may create more serious problems for an editor than some of the more obvious bungling of Compositor E. But B's shortcomings are not due to inexperience or lack of mechanical skill. E's work, on the other hand, usually seems to reflect both. It inevitably suggests, not the swift though careless expert, but the tyro. Hence it is especially interesting to note that E was apparently expected to make a great many errors. In striking contrast with much of the material set by A and B in the Tragedies, practically everything that Compositor E set, not only in Titus and

Romeo but elsewhere, was proof read. E's deficiencies were evidently well known to his superiors, and all that he did was subject to more or less systematic review. He was certainly not, in this respect at least, on the same footing in Jaggard's printing house as Compositors A and B.[13]

(3) As we trace the whole progress of the printing of the Tragedies from Coriolanus to Hamlet we observe that Compositor E's activities gradually change. He begins by setting a page (dd3v) from B's cases without any of the immediately prior distribution of wrought-off material that is normally the first step in setting a new page. Then, more regularly, he distributes s1 and sets his second page (dd4). A little later we find that he is usually distributing considerably more material than he sets, and hence that the regular compositor who presently returns to the case at which E has been working finds it already well supplied with type and is able to proceed at once, without preliminary distribution, to the setting of his next page. E is being used for extra distribution—to the advantage of the regular compositors, now one and then the other.[14] Soon after this, moreover, there is a further development. Compositor A set column b of mm4 and then all of mm3:4v (see p. 8). But when he finished setting mm4 from what may be called the x case, he proceeded to set mm3 from the y case, then moved back to the x case to set mm4v. This looks very much as if E were here doing A's distributing for him—so that A could devote his whole time to setting.[15] There is no sudden and striking increase in foul case in the Tragedies: Compositor E could obviously distribute better than he could set. This, too, his superiors appear to have recognized and exploited. In any event it is patent that here, in

B's absence, E did not set any part of quire mm. And this would be strange indeed if E were a fully qualified typesetter and could have functioned in B's place by setting one page of forme mm3:4v—strange, that is, unless E was unable to set manuscript copy with sufficient speed and accuracy to warrant his being employed on it.

II. Hamlet, Lear, and Othello

The complete sequence in which intercalary formes are encountered among regular formes is a long one, embracing six complete plays and parts of three others—a total of some 89 formes. Compositors A and B set Coriolanus, Caesar, Macbeth, the end of Richard III, Henry VIII, and about half of Hamlet. And E set Titus, Romeo, and part of Troilus. Only twice in the whole Coriolanus-Hamlet progression do two intercalary formes appear consecutively. And such formes usually appear, not only singly, but between small groups of regular formes. Thus E's work on Titus and Romeo got on only slowly. Eventually, however, he completed the last forme for quire ff and is presently found at work on a quire containing both the final pages of Romeo and the early ones of Troilus and Cressida.

The evidence by which it is possible to show precisely how much of Troilus was in type before work on this play was discontinued does not concern us here. It need only be said that a beginning was made on a regular six-leaf quire containing the end of Romeo and the beginning of Troilus but that it was decided to stop work on Troilus long before this quire had been finished and at about the time A and B completed the setting of Henry VIII and resumed work on quire oo for Hamlet. In consequence of this decision Compositor E prepared two formes for a one-sheet quire (present 2gg; pages 73-76) and Romeo and Juliet was finished independently of Troilus.[16] Forme 2gg1v:2, the second of the two formes for quire 2gg, is thus the forme with which E's work for Romeo came to an end. And it is also the last forme set by E that appears between two regular formes set by A and B. A new development now takes place. Its nature will at once be suggested by another short sequence. The progression of formes from 001v:6 through pp2:5v was as follows:

Through pp3:4v this sequence is clearly of a piece with the one given on page 8 above. A and B together ultimately set three of the four regular formes, but A was required elsewhere before he had set all of 001, and B therefore finished 001 and set all of the next forme, pp3v:4. E, as always, set the intercalary forme; and the next forme was set jointly by A and B. But pp4v is Compositor A's last page. He reappears briefly, after even the preliminaries had been finished, to share one final task with Compositor B: except for two pages earlier set by E and preserved, the present Folio text of Troilus was set by A and B working together once more. But Compositor A had nothing to do with any page of Hamlet after pp4v—nor with any of the following Tragedies right through Cymbeline. And A's place as B's partner was taken by Compositor E, beginning with pp5.

This is made perfectly clear by various kinds of evidence. Types show conclusively that two different men, working from different cases, set the different pages of pp2v:5 and pp2:5v—as well as of many later formes. Spellings indicate that neither of these was A, that one of them was B, and that the other uses the same mixture of certain A and B spellings that have been noticed in E's earlier work in Titus and Romeo. And above all, perhaps, the quality of the setting done by this man continues to reflect the same inexpertness so obvious in the earlier plays. Hence the following facts about the seven formes listed above are of interest:

• (a) there are no press-variants in any of the pages of the formes that were set by either A or B;
• (b) both pages of E's forme 2gg1v:2 are press-variant;
• (c) both pp2v:5 and 2:5v are press-variant but the variants are in each case confined to but one page of the forme: pp5 and 5v, the two pages set by the compositor who here replaces A as B's partner, show numerous press-corrections.

It may be added that most of the later work of this man was also proof read. Titus and Romeo, it has been noticed, are remarkable for the large number of their formes that were proofed and corrected. Only in Hamlet, Lear, and Othello do we find similar concentrations of press-variants.

Thus types, spellings, quality of setting, an unusual amount of press-correction—all point one way: Compositor E set pp5 and 5v; and similar evidence shows that he also set many subsequent pages in quires pp-ss. Nor is this surprising, viewed in the light of the whole course that the printing of the Tragedies had followed up to this point. With the setting of 2gg1v:2 Compositor E had finished all of the material on which he had so far been working: Titus was in print, the last forme

of Romeo was about to be printed, and work on Troilus had been indefinitely postponed. E had demonstrated that, although what he set might require more proof reading than the work of A or B, he could be useful; and A was needed for other tasks. Compositor E was therefore put to work with B and on what was for him fresh material— Hamlet to begin with—and A went elsewhere. In due course, to be sure, E also disappears. Quires Gg-hh and tt-bbb (Timon, most of Othello, and all of Antony and Cymbeline) were set wholly by B. Compositor B did not work alone, however, on the last formes of quire pp nor on quires qq-ss; and the following pages of the Folio texts of Hamlet, Lear, and Othello were set by Compositor E:

Certain peculiarities in the later tragedies of the Folio—in the plays set after A's departure, that is—throw additional light on E's capabilities and limitations. Three of these may be mentioned.

(1) The Folio text of King Lear, as has long been known, was set from a quarto that had been extensively "corrected" through collation with a manuscript version of the play. The copy must have been considerably more difficult than that used for Titus or Romeo. Hence it may be very significant that Compositor E did not set certain pages of Lear that would ordinarily have fallen into his stint.[17]

When two compositors work together on successive quires of the First Folio, the compositor who sets the pages for the second half of one quire ($4-6v) also sets those for the first half of the next ('$1-3v). Thus we find, for example:

This is the almost invariable practice throughout the book, and it is the obvious consequence of a simple division of large blocks of the copy: Compositor A held the copy for a continuous sequence of pages from the middle of quire t to the middle of quire v (t4-v3v); B held the copy for the preceding pages of quire t and for the pages from the middle of quire v to the middle of the following quire. Nor is this way of dividing copy ever permanently upset by other demands upon the workmen in question. One compositor often sets both pages of one or more formes for a quire for which both nevertheless do some setting: Compositor A may be temporarily absent, and his block of copy is at such times turned over to B so that the necessary formes (each containing, of course, one page from each half of the quire) can be kept going to press. When A returns, however, he resumes work on his block of copy. The result of this practice—and the evidence shows a remarkable constancy in it—is that the same compositor does not set both early and late pages for the same quire unless he is working alone and therefore setting some pages for the absent partner whose block of the copy he temporarily holds. When both compositors are present we simply do not find, normally, either one of them setting both early and late pages of the quire in hand.

Quire qq is triply anomalous in this respect. What we find is this:

Compositor E set the late pages of quire pp, B the early. Hence we should expect E to set the early pages of quire qq (1-3v) and B the late pages (qq4-6v); but the reverse is true. Yet not wholly so: E set the earlier page (qq2v) and B the later page (qq5) of the first forme after qq3:4v, which had been set wholly by E. After qq2v:5, however, another switch was made and E set the late pages (qq5v, 6, and 6v), B the early pages (qq2, 1v, and 1), of the last three formes. This can only mean that the copy for quire qq was most irregularly divided—or else, and more probably, that there were highly irregular exchanges between E and B of normally divided copy. Which of these alternatives better accounts for the observed facts is irrelevant. The result is in either case the same: certain pages that would normally have been the responsibility of one compositor were in fact set by the other. Since each page set anomalously by one compositor implies a corresponding page set anomalously by his partner, the irregularities are reciprocal and can be described in terms of what either compositor did, and did not, set. The departures from normal may therefore be listed as follows:

• (a) B set qq3v, instead of qq4, at the beginning of the quire;
• (b) B set qq5, instead of qq2v, and
• (c) B then set qq2 instead of qq5v.

Compositor B also set qq1v and 1 instead of qq6 and 6v; but this irregularity is probably only a consequence of the shift that gave him the copy for qq2, the early page in the forme that immediately preceded qq1v:6 and 1:6v. That no further abnormal shifts were made is indicated by his proceeding quite regularly from the early pages of quire qq (2, 1v, and 1) to the late pages of quire rr.

What, then, does all this mean? That normal practice was not followed is a matter of demonstrable fact; but why should the copy for certain pages of quire qq (one and a half pages of Hamlet and 10 pages of Lear) be subject to such special treatment? Why, more specifically, should the copy for qq3v, 5, and 2—all of which are pages in King Lear —be so manipulated as to fall to B rather than to his less skillful partner? We cannot be sure. One explanation, however, immediately suggests itself. The copy for Lear cannot have been easy, and parts of it must have been more or less excessively annotated. Was the copy for these three pages too hard for Compositor E—or at least difficult enough to make advisable the anomalous exchanges that have been noticed? The inference is consistent with other indications of E's inability to deal efficiently with all kinds of copy; and no other explanation seems to accord so well with all that we know both of Folio Lear and of the compositors who set it.[18]

It may be added, finally, that a similar manipulation of copy is found in quire ss (shared by Lear and Othello) —where ss3, again a page of Lear, was set by Compositor B, but set from copy that would normally have fallen to Compositor E. These irregularities in quires qq and ss, in each case involving the copy for Lear, are without parallel elsewhere in the Tragedies.[19]

(2) Work on Troilus as the immediate successor to Romeo and Juliet was only well begun when it was abruptly discontinued; but the gap thus left between Romeo and Caesar (long since printed) was not quickly filled. It was eventually filled, though very imperfectly, by Timon of Athens. The printing of Timon (and with it a reset version of the last half-page of Romeo) thus appears to mark the final abandonment of the original plan for Troilus. In any event, Timon (quires Gg and hh) was not begun until after quire rr had been set—and set

jointly by E and B. The new E-B partnership had been established and was working. It had completed quire pp, produced quires qq and rr, and was presently to turn out quire ss. But Timon of Athens, which falls between quires rr and ss, was set wholly by Compositor B. Compositor E was at hand during part of the interval, for he appears to have done the distributing that immediately preceded B's setting, first of Gg3, then of Gg4v; and he clearly set part of quire ss while B was at work on the later pages of Timon: forme ss3v:4 is in fact an intercalary forme, set wholly by E, between the fourth and fifth formes of quire hh; and the second ss forme, also wholly by E, falls between the last hh forme and an ss forme set jointly by B and E. The partners, in short, were both available; but one of them had no share whatever in the setting of Timon of Athens. Again, therefore, we find a highly irregular method of working, but also one that is eminently reasonable if Compositor E was unable to deal competently with the manuscript copy from which Folio Timon must have been set.

(3) Compositor E finally disappears from the Folio picture after quire ss. It is a little surprising that he had no hand in quires tt-vv, into which the greater part of Othello falls. It is not surprising, however, that Antony and Cymbeline were set wholly by B. Like Timon, the last two plays in the Tragedies can only have been set from manuscript copy.

Compositor B set ss1:6v, but most of quire ss, including all of the first six pages of Othello, was E's work. After setting ss6, however, E is seen no more. His disappearance thus coincides very closely with the end of the first act of Othello, only the last 19 lines of which fall in ss6v and so were set by B. It is therefore tempting to suppose that the copy for all but Act I of Othello was, as parts of the Lear copy seem to have been, beyond E's powers—perhaps because more heavily annotated than the copy for Act I. Possibly this suggestion is worth testing; but if so it must certainly be tested by other evidence than what is provided by the bibliographical peculiarities of the First Folio alone—where there is little to show that Compositor E was not simply required elsewhere after setting ss6, and did not again become available for work on the Shakespeare Folio until after Othello had been completed. Yet it may be mentioned that, whereas E's last four pages are not press-variant, an unusually large number of the remaining pages of Othello, all set by B, are so. Why this should be can hardly be said out of hand; but it is at least consistent with the hypothesis that the copy for Act I was relatively easy, while the copy for the rest of the play was more difficult and considered likely to give rise to serious error.

III. Summary: Compositer E

That the same compositor who set Titus and Romeo and pre-cancellation Troilus also set certain pages of Hamlet, Lear, and Othello cannot seriously be doubted. That he was very inexpert is also demonstrable. But that this man, Compositor E, was incapable of setting manuscript material acceptably and was therefore allowed to set printed copy only—this, of course, is inferential. It cannot be absolutely proved. Hence the way of prudence would be to eschew, as possibly not fully established, the proposition that Folio Hamlet and Othello were, like Folio Lear, set from annotated quartos, and to suggest no more than that Compositor E was not ordinarily employed on manuscript copy. Nevertheless—and without any wish to commit myself overenthusiastically to folly—I feel bound to say that the inference that E could not competently set manuscript material seems to me inescapable. It is eminently consistent with everything that has so far been observed about the printing of the Tragedies, and it explains a whole complex of phenomena, large and small, that are otherwise inexplicable.

The strange but evidently methodical switching back and forth between regular and intercalary formes in the earlier part of the Tragedies is simply incredible unless Compositor E was for some reason obliged to set only certain texts (all of which were in fact set from printed copy) and was not permitted to work on any of the material (all of it in fact set from manuscript copy) upon which A and B were engaged, even though he acted only as a substitute for one or both of them in this part of the book. And the inference is strengthened by evidence that E did not function as a fully qualified substitute at times when his doing so would have had manifest advantages. It is not merely that we should expect a qualified substitute simply to continue, in the absence of the regulars, from the point where they left off. There are many formes throughout the earlier Tragedies that were set wholly by A or by B alone. Compositor E may here be found distributing type for the benefit of the regular compositor who remains at work on the Shakespeare folio, a practice that would have speeded up operations slightly; but maximum speed-up is never attempted: E never serves as a full partner, never sets type. And the same thing happens after A's disappearance. E now works as B's partner on the end of Hamlet, on Lear, and on the beginning of Othello; but joint setting by E and B is interrupted while B alone sets Timon, just as he later set all of Antony and Cymbeline.

Thus the evidence as a whole—and it has a remarkable coherence throughout the Tragedies—seems almost to force the conclusion that

Compositor E was not regarded as competent to deal with manuscript copy; and that he was never, after setting III.ii of Titus, permitted to do so again.[20]

Compositor E's precise status in the Jaggard establishment is not of very great importance; but that he was an apprentice is suggested by many considerations. Whatever his capabilities with respect to manuscript copy, he appears to have been less expert than we can reasonably suppose any full-fledged typesetter of the time would have been, and it is clear that his superiors were perfectly familiar with his limitations and considered it necessary to proof read most, at least, of what he set. He was not supplied with type-cases and working space of his own, and he is occasionally found doing some of the more routine work of the regular Folio compositors. Compositor E may not, of course, have been an indentured apprentice of Jaggard's house; yet there is certainly some warrant for referring to his work for the First Folio as that of a prentice hand.

Notes