³ÉÈËVRÊÓƵ

One aspect of Brian Vickers's objection to Holger Syme's review is undeniably true: if Syme Tweeted his response to each page as he read it then he was making judgments "before he could have taken in the argument of each chapter". This is poor scholarly practice and should be deplored. It is impossible not to form opinions as one reads, of course, but a professional critic should wait to at least finish the book and consider it as a whole before beginning to formulate a response for others to read. To fail in this basic scholarly practice of reflection makes Syme's review worthless, no matter what the merits or flaws of Vickers's book. No-one who cares about scholarship should encourage hasty readings circulated to the public before the reader has even finished the book.

Other distinguished scholars have previously taken issue with Prof. Vickers¡¯s methodologies. For instance, in Shakespeare, A Lover¡¯s Complaint, and John Davies of Hereford (2007), Prof. Vickers argues for Davies, and not Shakespeare, as a source for the Roman poet Martial¡¯s presences found in the work; however, Andrew Zurcher, in Shakespeare and Law (2010), disagrees with Prof. Vickers¡¯s view, finding Shakespeare¡¯s authorship for A Lover¡¯s Complaint ¡°strong¡±. He cites MacDonald P. Jackson¡¯s review of Vickers¡¯s book in the Review of English Studies as support for his view that ¡°¡­certain words in A Lover¡¯s Complaint may point to Davies¡¯s authorship, but a different set of words might point to a different author (even Shakespeare, whose name appears on the first printed edition of the poem)...the value of Vickers¡¯s stylistic evidence is dubious because arbitrarily selected¡± (2010:290). Moreover, recent archival findings brought to the fore on, yes, Twitter, seem to suggest that Shakespeare experienced his own, direct engagement with Martial via a common law manual of the day, Lambarde¡¯s Eirenarcha ¨C - and that he documented his outrage at this ¡®Martial¡¯ law with a reference in King Lear: http://bit.ly/1OdP26u

Not having any kind of academic expertise in Shakespeare, I'll just point out that "my book has been peer-reviewed by excellent scholars" does not, in fact, say anything. Peer review is not fail safe, and to say that a critic is wrong because "experienced scholars" have peer-reviewed the book is a mere argument of authority; sophistry, in fact.

Holger, you can't have it both ways. If you accept the principle that it is important to take time to reflect on something before giving a view on it--as you seem to ("the tweets were for the most part a second pass")--then there is no reason to use Twitter at all. Twitter is inherently a non-reflective, off-the-cuff medium for sound-bite anti-intellectualism. If you agree that it is worth taking time to reflect (to "complete each chapter before" Tweeting, as you say you did) then it would be even better to take more time and get an even broader perspective. That is, to read the whole book before making public comments on it. I'm surprised that one scholar should have to try to persuade another of this fundamental principle of intellectual labour. I am quite prepared to entertain the possibility that you found factual errors in Vickers's book. I really did pay your Tweets the compliment that you denied Vickers's book and read all of them in one go (in the Storified form) before deciding to comment on them. They contain their own factual errors about the printing process that I'd be happy to point out. This does not mean that in general you are wrong and Vickers is right. My point here is not about the printing of Q1 King Lear--this just isn't a good place to discuss that--but about how we engage with one another's work as responsible commentators. If gaining attention in the media was your purpose in using Twitter to comment on Vickers's book, then you succeeded. But if you wanted to be taken seriously as a mature thinker who reflects before he gives his verdict, you did not. Gabriel Egan

Dear Holger Yes indeed your factual claim that I most strongly disagreed with is your Tweet "35. Setting by forme doesn't use less type. It's faster". From practical experience of teaching students the art of letter-press printing I believe the opposite to be true: setting by formes does indeed use less type (because one can commence printing after setting four quarto pages instead of seven) but it is slower (because of the time lost in casting-off the copy and in adjusting later for errors in casting-off). Because it cuts down on the amount of type that is set into pages but not yet usable (because the forme is not yet complete) and because it allows the same type to be used on both sides of the sheet, setting by formes enables a printshop to do more work with the same amount of type (which I take to be a reasonable practical definition of "using less type"). I imagine that you could defend your claim that setting by formes is "faster" by saying that you mean that this method gets the type ready for the press sooner (after only four pages instead of seven). But to this I would respond that such a definition of "faster" is based on the false assumption that the press would otherwise be standing idle until the type is ready, when in fact the printer would occupy his press with the work of machining sheets for other books at the same time rather than leaving it idle. This avoidance of press-idle time negates any supposed benefit of rapidity conferred by the method of setting by formes instead of seriatim. I haven't yet read Vickers's book on Q1 King Lear and (to repeat for clarity) I do not mean by this that he is in general right and that you are wrong; but from your account of your disagreements with him these technical matters about printing would seem to be close to the heart of the dispute. Regards Gabriel Egan (Centre for Textual Studies, De Montfort University)

Dear Gabriel, thank you for that detailed response! You're quite right that the brevity of expression encouraged by Twitter turned that statement into something more general and unqualified than it would otherwise have been. But let me clarify what I meant, and why I don't quite agree with your characterization of the differences between the two type setting methods. I don't know if there are any documented examples of commercially printed books from the early modern period using the same type on both sides of the sheet -- if that happened, it must have been a real anomaly. What this would mean in practice, after all, is that the compositor would wait until the outer forme (or inner, but let's just assume the outer would _always_ be imposed first) was printed, would then distribute the type, and only after that would begin setting the inner forme. I can imagine scenarios in which such a way of working might somehow have made sense, but not many -- instead, the typical process (as I know you know) would have involved a compositor working on the inner forme while the outer was being printed; the outer wouldn't be distributed until after that, which is why reused type typically shows up at some point in the subsequent sheet, but not normally (to my knowledge) on both sides of the same sheet. There are, however, real economies of type-use with setting by formes, not because you can use the same type on both sides of one sheet, but because you can distribute type sooner. If setting seriatim, you'll need to finish eight pages of a quarto before printing can commence. Therefore you will start setting the next sheet while both formes of the previous sheet are still standing. If you're setting by formes, you can distribute the outer forme of sheet A before starting on sheet B (and while the inner of sheet A is in press). So a printer can get away with less type when setting by formes than when setting seriatim -- though exactly what the flexibility gained is really depends on the book and the process. I suppose it could be as much as 4 pages of type. The time savings are more significant, I'd say, at least as long as two compositors are employed -- they can set up both formes at the same time. In theory at least that's twice as fast as when setting seriatim. In practice, you're right (of course) that the press(es) wouldn't be sitting idle, and that such efficiencies of time may not have been fully realized. But I don't think we have historical evidence that setting by formes slowed down the process either -- otherwise, I find it hard to see why any printer with a reasonable amount of type at hand would ever have had his compositors work that way. All of those are general considerations; in the case of _King Lear_, they need to be put in the context of Nicholas Okes' printing house. In Okes' case, we know that he wasn't working on any other books at the same time as _Lear_ (or at least none survive; he may of course have been printing ephemeral jobs that have gone without a trace). We also know, thanks to Blayney, that the amount of type he had on hand exceeded the demands of the _Lear_ MS -- where it fell short was in specific sorts (and almost certainly in blanks). But as I argued in the tweets, and as I would maintain in this comment section, shortages in specific sorts may be alleviated when setting by formes, but can't necessarily be eliminated. It's certainly true that if you don't have enough capital italic Es to set up 8 pages without using alternative sorts, you'll be better served by choosing the method that would allow you to distribute sooner rather than later. But shortages will occur no matter what. Okes may have decided that the difficulty of the MS made setting by formes impracticable or too difficult, and that the aesthetic advantage of being able to use the correct sort more often wasn't worth it. But I think it's important to understand that his decision to set the text seriatim was not so foolish that it had his compositors running out of type -- they ran out of specific sorts that would have been in short supply in most houses not used to printing a lot of plays. (And in the context of the tweet you cite, it's also worth noting that Vickers seems to think the type was under greater stress when used in seriatim setting -- an assumption for which there is no evidence at all. If anything, the higher rate of re-use that setting by formes enables could be argued to put greater stress in individual pieces of type. no?) This is a very helpful and, I think, productive discussion -- thank you! Holger

Sorry-- for "eight" read "seven" in my third paragraph. Obviously.

Dear Holger Much of what you characterize as normal printing practice seems to me rather speculative. It is certainly the case that the printshop could proceed on a book in the orderly fashion you describe, with "a compositor working on the inner forme while the outer was being printed; the outer wouldn't be distributed until after that" and with the work proceeding alphabetically through the gatherings. But the work did not have to follow such a regular order, and setting by formes gave the printers the freedom to print sheets, and the individual sides of the sheets, in any order they felt like. The ability this conferred to work on several books at once, and so regulate the work-flow of the whole printshop to maximize efficiency (rather than finishing any one particular book), may well have been the major attraction of printing by formes. Type recurrence evidence cannot give us certainty about the order of printing, because well over 99% of the type recurrence is invisible to us. We can detect the recurrence of broken pieces of type, but not the recurrence of the (far more common) perfect pieces: they were meant to be indistinguishable, one from another, and indeed they are. The thus-limited type recurrence evidence we uncover may turn out to be consistent with the orderly practice you describe, but it can never prove that practice. Obviously, a piece of type can't be in two places at once so its appearance on both sides of a sheet would prove setting by formes rather than seriatim. (Except of course recurrence on page 4v of a quarto: that's still technically possible with seriatim setting.) But beyond that we can only use inductive not deductive reasoning. The famous illustration of deduction is "I see a black swan, therefore not all swans are white", while induction can only take us as far as "That's the 100th white swan I've seen, I expect that all swans are white". Likewise, no amount of type-recurrence evidence consistent with orderly progression through a book, alphabetically by sheet doing first one side then the other, can prove that this was the order in which it was done. Type recurrence evidence--which is the major reason to believe in the orderliness that you assume--is further hindered by scholars failing to agree on what is damaged type and what is simply poor inking. Around half of the type-recurrence in Q1 King Lear that was claimed by Peter Blayney was rejected by subsequent scholars looking at the same book. The other main category of evidence that is consistent with but does not prove orderliness is the reuse of skeleton formes and Blayney himself was instrumental in showing how misleading that evidence could be. Regards Gabriel Egan

I'm must confess, Gabriel, that I'm not entirely sure which side of the case you're arguing now. Speaking in _general_ terms, you are, of course, right that type recurrence can't conclusively prove the order in which any given book was set -- we've known this since McKenzie's "Printers of the Mind" essay. But by the same token, if type recurrence is highly problematic evidence, our ability to distinguish between the two types of setting is also severely compromised. We've moved rather far from your original objection, though I will note that the argument you propose now, that setting by formes allowed for maximum flexibility in distributing work (which I don't disagree with), has little to do with the claim that setting by formes uses less type. That said, I don't know of any examples of a printing house operating in quite the way you suggest, setting one forme of a sheet before abandoning that project in favour of another, only to come back to perfect that sheet at a later point. It's not inconceivable that this happened, but the potential complications strike me as more significant than the advantages. Can you name some instances where we know this type of division of labour occurred? (That printing was frequently split, among multiple printers and/or across multiple book projects within the same printing house is obviously an uncontroversial assertion at this point -- but that such splits occurred in the middle of a sheet seems a more striking claim to me, and one I'm not sure I've come across before.) In a similar vein, I'll reiterate that I haven't really seen many arguments that distribution of type in the middle of a sheet was common -- and I would like to see some examples of "a piece of type ... appear[ing] on both sides of a sheet." Again, my point is not that this an impossible scenario (it obviously isn't), but that it seems highly unusual to me, somewhat at odds with the commonly accepted accounts of early modern printing house practices, and at first glance improbable as a widespread habit. I'd be happy to be convinced otherwise, but given what you've said about your views of the unreliability of the available evidence, I'm not sure what the standard of proof I'm supposed to apply might be. Lastly, I'll note that although you're right that two reviews (Hammond and Werstine) of Blayney's book cast doubt on his identification of a large number of type recurrences, neither of them questioned that the evidence he had assembled proved conclusively that _Lear_ was set seriatim; Hammond concluded that the account of type distribution Blayney offered was generally sound _despite_ his skepticism concerning between 33% and 69% of Blayney's counts of individual type recurrences. That strikes me as an important addendum to your point.

Dear Holger You're right, we're a long way off the main point here. I'll close just by remarking that I see our current state of knowledge about what actually happened in printshops to be more uncertain than you do. Werstine, for example, was not convinced that Blayney had proved that Q1 King Lear was the only book that Okes was working on at the time he printed it: ". . . the possibility remains that 'Lear' and 'Beza' (containing Biblical quotations in the same fount of type used in 'Lear') went through the press together" (Shakespeare Quarterly 36 (1985): 122). The recent history of research into printing has been one of increasing realization that we just don't know what we thought we knew. The great 'discoveries' of the Virginian school of New Bibliography have turned out to be illusory. We now know that we cannot reliably distinguish compositors' stints by spelling and other setting habits, that the work of printing a book need not proceed in a linear fashion from beginning to end, that evidence from type reuse is easily overstated and can be used only inductively, and so on. You ask for a concrete example of interrupted printing, in which sheets printed on just one side were laid aside while something else was got on with. The easiest case to cite is the interrupted printing of the Folio text of Troilus and Cressida. You may say of course that this was a highly exceptional circumstance. I would reply that the Folio is by a long way the most closely studied book of this period and we just don't know how many similar cases there are in books that no-one has attended to so closely. Regards Gabriel

In his Twitter review of _The One King Lear_, Holger Syme notes that Sir Brian Vickers ¡°spends 7 pages battling for the idea that Q1 was set from holograph. A notion with which few to none disagree.¡± Vickers rationalized authorial printer¡¯s copy because his hypotheses rely on the ¡°idea¡±; its history reflects Shakespeare scholarship in general. Introducing their 1983 Two-Text manifesto, _The Division of the Kingdoms_, Wells and Taylor cite P.W.M. Blayney¡¯s ¡°two-volume study (half of which has since appeared) . . . investigating . . . [Q1 and] the great variety of non-literary evidence that points to . . . Shakespeare¡¯s own draft manuscript. It would have been foolish ¨C and unhelpful ¨C to attempt to duplicate Blayney¡¯s research.¡± Blayney refers to ¡°abundant bibliographical evidence throughout [Q1] which shows that the lineation and punctuation of the printed book were continually affected by a printing-house problem which had nothing whatever to do with the lineation, punctuation, legibility, or authority of the copy.¡± He saved the ¡®problem¡¯ for a second volume otherwise reserved for origins and textual (literary) criticism of Q1 and F. Howard-Hill: ¡°Blayney's rigorous analysis . . . concluded that [Q1 was set] from Shakespeare's own foul papers. . . . Because the second volume . . . remains to be published we can only assume that Q1's memorial errors are scribal, or . . . compositorial (1985). Howard-Hill: "The consensus established by Warren, Urkowitz, Taylor, Blayney, and others understands that Q1 prints Shakespeare's early draft¡± (1997). Jowett: "[Q1] used to be regarded as . . . memorially contaminated . . . Opposing [that view] is the impressive authority of Peter Blayney's work . . . . The revision theory would be vulnerable without [his] conclusion that [Q1] is . . . an unusually poor piece of printing. Once the compositor's propensity to err is taken into account, no substantial barrier remains to viewing the copy for Q1 as an authorial draft¡± (2000). Murphy, citing Wells: ¡°¡®[Blayney] concludes that [Q1 copy] was a much altered autograph manuscript containing the . . . near-final text.¡¯ [This idea] has displaced the dominant view [that Q1] is a reported text. . . . When one sets aside this [idea] . . . consequences of the most far-reaching sort emerge¡± (1987). Robert Clare: ¡°Two important textual studies [recently appeared]. . . . P.W.K. Stone argued in favor of revision, but by another hand. . . . Blayney investigated the working conditions and practice of Q's printer . . . . He did not deal with the text of Q itself, or with the textual implications . . . which are promised in a second volume . . . . Given their contention that Q [is] authoritative, it is worth noting that nowhere do the revisionists satisfactorily explain . . . the inferior phraseology and the baffling and inconsistent imperfections of lineation . . . (1997). Vickers: Blayney¡¯s book is ¡°a remarkably detailed study that solves many of the problems . . . . For instance, we now know that [Q1] was not pirated, that its inaccuracies are due not to a shorthand report but to a messy authorial manuscript¡± (2011). Bate: ¡°Blayney proved decisively by means of meticulous and highly technical bibliographic investigation that [Q1] was not a bad text . . . but an authoritative one, almost certainly deriving from Shakespeare¡¯s own holograph. . . . Blayney and a group of other scholars concluded that both Quarto and Folio texts were authentically Shakespearean¡± (2016, _The Spectator_). Blayney believes: ¡°that Q1¡¯s copy was an authorial manuscript; that the adaptation was made by someone other than Shakespeare from the printed Q1 rather than from a playhouse manuscript of any kind; and that F1 was printed from a manuscript (either the adapter¡¯s final draft annotated for promptbook use or a promptbook prepared from it) with the assistance (primarily for punctuation) of a copy of Q2. But while it may be useful for people to know that the author of volume 1 [_The Texts of ¡®King Lear¡¯_] does believe those things, until or unless he explains why he does they should of course allow that he might be wrong¡± (2009). For thirty-four years scholars have cited a non-existent argument as the very basis of their opinions and to counter earlier opinion that Q1 is a reported text. Further research would be ¡°foolish and unhelpful¡±; ¡°we can only assume¡±; ¡°consensus understands¡±; ¡°theory would be vulnerable without impressive authority¡±; ¡°we now know¡±; ¡°proved decisively.¡± The citations affirm that late acceptance of ¡®foul-paper¡¯ Q1 copy appeals to authority, not only in aid of Shakespearean revision, but of Q1/F conflation and (twice over) Sir Brian¡¯s _One Lear_. But alternative, ¡°far-reaching consequences¡± are not eliminated by ¡°settings aside.¡± Vickers learned that Blayney made no ¡®authorial¡¯ case. Its repeated citation is that much the worse for the passing generation, as admitting of nowhere else to turn; Vickers wouldn¡¯t rely on revisionist argument. S.W. Reid: ¡°[Urkowitz¡¯s] discussion of [Q1 copy] is a little less than lucid. There is a slight tendency . . . to rely unquestioningly on traditional concepts of Shakespeare's manuscripts that do not really fit the case. [His] use of certain critics to establish that Q1 represents an early Shakespearean draft cannot help but strike one as a bit facile.¡± Yet Richard Knowles (Variorum _Lear_, forthcoming) had already adopted the same uncertain critic, Madelaine Doran, whose (later disavowed) case consisted in spelling and handwriting comparisons. Vickers had no choice but to reassert Doran¡¯s lackluster 1931 effort. Perhaps Holger Syme can say why he accepts authorial Q1 copy. Howard-Hill: ¡°Stone apparently did not know of Blayney¡¯s conclusion . . . for he maintained . . . that Q1 was a report.¡± The evidence supports Stone (d. 1983); he probably would have asked to see more than a conclusion. _The Textual History of King Lear_ is the best textual study, not only for its unsurpassed analysis; his unbiased, fresh look at the play and his personal manner are exemplary. Gerald E. Downs