This list suggests several things about the process of revision. The corrections are numerous but in these instances limited to a few words or a single word. They tend to be clustered in certain scenes and involving certain interlocutors. The first cluster is in 1.2, when Ovid Senior is lecturing his son on the evils of poetry, abetted by Lupus and Tucca. In 3.1, Jonson seems to have paid particular attention to the speeches of Horace: of the ten alterations entered in this scene, eight are from Horace’s utterances and two from Crispinus. Three corrections in 3.2 are also in Horace’s speeches. In 3.4, where the correcting is especially heavy, some fourteen out of seventeen changes are from Tucca’s speeches; some changes are minor, but several alter the tone of Tucca’s abrupt humour, as in the shifting of ‘Leueret’ to ‘ferret’, ‘Caprichio’ to ‘PANTALABVS’, ‘Paunch’ to ‘Stiffe toe’, ‘Twentie i’the hundred' to ‘shifter’, ‘Death of Pluto’ to ‘life of PLVTO’, and ‘Rascall’ to ‘stiffe-toe’. The changes in 4.3 also are in speeches belonging to Tucca, and in 4.7 his ‘my Prophet' is changed to ‘my noble prophet’ and ‘Noble Horace’ to ‘little fat HORACE’. Some scenes are untouched, including 3.3, 4.1, and 4.2; others, including 1.3, 4.4, 4.5, 4.6, 4.8, 5.1, and 5.2, are only sparsely visited. The largely untouched scenes in act 4 are devoted especially to the presence at court of the ladies and Caesar. Revision is marked in the long final scene, many of the changes being in the speeches of Horace, of Tucca, and of Crispinus when he forced to vomit up his Marston-like neologisms. Some changes alter terms for money, as with ‘sesterces’ for Q‘s ‘Drachmes’ at 3.4.148, ‘a sesterce’ for ‘six pence’ at 4.7.6, and ‘a drachme’ for ‘Twopence’ at 5.3.158. (The ‘Drachmes’ at 3.4.22 and 302, the ‘Sesterties’ at 3.4.49, and the ‘Drachme’ at 4.3.137 in Q are not substantively changed in F1; generally, the changes are away from terms for English money.) A few terms sound as though they have been reworded for considerations of religious sensitivity, as when Horace’s ‘Redeemer’ is changed to ‘releeuer’ at 3.2.1 and when Tibullus renames the ‘Via sacra’ as ‘holy street’ at 4.3.14. (This is of course the street on which Horace encounters the unnamed bore in his Satires, 1.9.) Jonson’s close attention to the utterances of Horace and Tucca is what we might expect, given Horace’s identification with Jonson himself (in this play and in Dekker’s Satiromastix) and Tucca’s role as proxy in Jonson’s spat with Dekker and Marston.

Proof correction of F1 took place during the printing of the book, beginning in late 1615. Stop-press correction and resetting were heavy in F1 as a whole, heavier indeed than as enumerated by H&S. Jonson may well have assisted the press corrector in this labour, especially in the early plays, including Poetaster. According to David Gants’s tabulation (1999, 43), Jonson’s interventions were especially numerous in quire 2A of Poetaster, with substantial corrections also in quires Z, 2E, 2F, and 2G, followed by 2D, and with no Jonsonian interventions (and only one by the press corrector) in 2B and 2C. Throughout, according to Gants, Jonson’s own interventions substantially outnumber those of the press corrector. Cain (1995), 285 comes to a different conclusion, implicitly crediting the press corrector with numerous changes that Gants would regard as authorial. The difference would appear to be that Gants is counting as Jonson’s any literary, ‘stylistic or indifferent alteration of punctuation with which a printing-house corrector would not bother’ (42); and this is the more plausible argument on its face. The upshot of this debate is to reaffirm H&S’s position, in the early plays at least, that most of the corrections are authorial. Many pages exist in uncorrected form, evidently because the pressmen machined the early states over a period of time before running the sheets through the press again, presumably in the interests of efficiency. Even so, on occasion a run of perhaps 200 to 300 sheets would be printed out of a total production of 750 copies. Sometimes a forme might be printed while the corrector was still at work on that forme, and in some instances the initial proofing may have missed some items that were later caught (Gants, 40). The result is that press corrections are found in three states, variously distributed in existing copies of F1. See F1 Collation for a detailed listing.

Cain (1995), 285 posits that F1 Poetaster was set by formes, even though Stansby’s printing house seems to have had a sufficient supply of type at least much of the time to be able to print by quires, which would allow more time for proofing -- some of it even off-premises, as at Jonson’s residence. Gants and Donovan (1987) argue, on the other hand, that the text pages for most of the composition of F1 were set seriatim: in Gants’s words, ‘Throughout most of the Folio composition of the gatherings seems to have been seriatim, i.e. sequentially by page: sig. 1r, sig. 1v, sig. 2r, sig. 2v, etc.’ Gerritsen (1959), 52-5 estimates that Stansby had thirty-seven folio pages still standing toward the end of the printing of F1. Cain hypothesizes that one compositor set most of quire Z and half of 2A of Poetaster. This differentiation of compositorial stints is, however, based on a few spelling preferences, such as ‘Ile’ (not found after 2B3v) and ‘verie’ and ‘everie’ (not found after 2A6), and Cain admits that the balance between y and ie endings is more evenly balanced after quires A and B. At the same time he wants to point out that, except for words like heavenly, the middle quires from 2C through 2F generally favour ie endings, frequently changing Q’s final y to ie. Is a third compositor to be discerned, Cain wonders, in 2G, with its strong preference for the final y? This is hazardous as a tentative conclusion. What’s more, the data on y and ie do not fall out as Cain proposes. Compositorial preference for y formations, as in changing poetrie to poetry, is to be found on many pages of F1 throughout the play.

The compositorial assignments are difficult to determine, in part because the spellings and punctuation of Q are often followed in F1; after all, the F1 compositors worked with corrected printed copy except for the few additions. Some changes are systematic; many capitals in Q have been reduced to lower-case, even in a word like ‘Helicon’ (1.1.9). Generally roman caps and small caps are employed for proper names like ‘ENVIE’, ‘ROME’, ‘IOVE’, etc., in place of Q’s preference or italics. The speech prefixes have been carefully regularized in four letters, cap and small cap: TVCC. (or, once on p. 341, TVCCA.), OVID., OVID se. or OVID. SE. (when the two are onstage together), OVID iu. or OVID. IV. (similarly), LVSC., LVPV., PYRG., TIBV., CRIS., ALBI., CHLO., MAYD. 1., MAYD 2., IULI., CYTH., PLAV., PROP., HERM., GALL. (or, on p. 313, anomalously, COR. GALL.), ALL., HORA., ARIS., LICT., MINO., DEMET. (twice, anomalously, on p. 306) or (more commonly) DEME., 1. PYR., 2. PYR., TREB., HIST., MECÆ or MECŒ., CAESA. (anomalously on pp. 323-4 in five letters, but representing CÆSA.) or (more often) CAES., VIRG., EQVES 1., and EQUITES. (in plural use). This regularization means that the speech prefixes are not useful in determining compositorial stints. Other changes too seem independent of compositorial preference. Some roman words are italicized, and the reverse. Hyphenation is sometimes introduced into words like ‘ey-strings’ (Induction, 28). Sometimes apostrophes are introduced into words like ‘endowed’, ‘squeezd’, ‘armed’, and ‘scornd’ (Induction, 43, 45, 51, 65). Ile’ is regularized into ‘I’le’ throughout most of the text. A word like ‘iuice’ sometimes becomes ‘juice’. The handling of y/ie endings seems arbitrary: in the Prologue of the Induction, for example, at 73 ‘Allegory’ is changed to ‘allegorie’, whereas two lines later ‘follie’ is changed to ‘folly’. (Granted, in much of the play F1 prefers ie to Q’s y.) Ampersands occurring in the crowded Q printing are changed throughout to ‘and’, as at 1.1.13.

These many changes occurring throughout notwithstanding, compositorial spelling preferences are perhaps discernible, especially in the changes away from the spellings of Q. In the printing of words like ‘me’, ‘we’, ‘be’, ‘she’, and ‘he’, many pages of quires 2A and 2B -- namely all of 2A except 2r and 4v, and, in 2B, 1v, 2r, 2v, 5r, 5v, and 6r -- exhibit a marked preference for ‘mee’, ‘wee’, ‘bee’, ‘shee’, and ‘hee’, changing the Q spelling in this direction some 36 times as compared with 7 instances in the reverse direction. Tentatively, let us identify the compositor here as Compositor A. Some other pages in 2A (2r and 4v) and 2B (3r, 4v, 6v) are indeterminate in their preference for ‘mee’ or ‘me’ spellings. Conversely, 2A4r prefers ‘he’ and me’; 2B3r changes Q’s ‘mee’ and ‘Wee’ to ‘me’ (twice) and ‘We’. In quire 2C, certain pages (notably 2r and 3v) change ‘wee’ to ‘we’ etc. with notable frequency. Tentatively, we can identify this as the work of Compositor B. Meanwhile, 2C1v (with 3 ‘he’ types) and 4r (with 5 ‘mee’ types) were perhaps set by Compositor A. In 2D, the first four formes give us 12 ‘bee’ types against 1 ‘be’, while in 3r-4v and 3v-4r, ‘be’-type spellings predominate 10 to 3. In quire 2E, ‘bee’ types predominate in 2r, 3v, and 6r, while 2v, 3r, and 4v tend to favour ‘he’ etc. In 2F, ‘mee’ spellings dominate on 1r, 2v, 5r, while ‘he’ spellings are favoured on 3r, 4r, and 4v.