The text of Romeo and Juliet that the seventeenth century read, quoted, and handed down was not the text of the famous quartos of 1597 and 1599 alone. It was the text as two later reprints carried it forward: the third quarto of 1609, a page-for-page descendant of the 1599 edition, and the undated fourth quarto that followed it. Between them, these unglamorous books did the work on which everything else depends. The third quarto became the copy from which the First Folio was printed, so that every Folio reading, and every edition that trusted the Folio for three hundred years, rests on a 1609 reprint. The fourth quarto carried the play to the edge of the Folio moment and, in some of its copies, put the author’s name on the play’s title page for the first time in the quarto line. The verdict of this study is that the reprints are not an appendix to the transmission but its load-bearing middle: remove them, and the road from the playwright’s papers to the collected edition of 1623 has a gap exactly where the traffic was heaviest.

That verdict reverses the usual telling, in which the story of the early text stops at the contest between the short quarto of 1597 and the full quarto of 1599 and resumes, briefly, at the Folio. The reprints correct that picture in three ways. They show how a printed text actually travels through a trade: not by fresh encounters with manuscripts, but by compositors setting type from the previous printed page, correcting a little, corrupting a little, and passing the accumulated result on. They show what the play was worth as property, since a publisher bought the rights in 1607 and kept the title in print for thirty years. And they settle questions the famous quartos cannot settle about themselves, including which book the Folio’s compositors really had open on their desks. The third and fourth quartos reward the attention they have rarely been given, and the pages below take them in order: the books themselves, the property deal that produced the third, the drift each reprint introduced, the annotated copy that carried the third into the Folio, the undated fourth quarto’s contested year and divided title page, and the market that the whole reprint sequence measures.

What Are the Third and Fourth Quartos of Romeo and Juliet?

The third quarto is a book of ninety-two pages, printed in London in 1609 for the publisher John Smethwick and sold at his shop in Saint Dunstan’s Churchyard in Fleet Street, under the Dial. Its title page offers “The Most Excellent and Lamentable Tragedie, of Romeo and Juliet,” describes the play as acted “by the Kings Maiesties Seruants at the Globe,” and repeats the proud formula of its copy: “Newly corrected, augmented, and amended.” The printer’s name appears nowhere on the page; the printing is identified as John Windet’s by the ornaments and type his shop used, a standard attribution in the bibliographical record. The fourth quarto is a slimmer book of eighty-eight pages, undated, printed for the same publisher, sold at the same shop, and bearing the same title and the same claim to correction, with Smethwick’s device, a punning image of a smew, set between the claim and the imprint. No printer is named on its page either, and no year: the two facts that bibliographers most want from it are exactly the facts it withholds.

Both books are reprints in the strict sense. Neither was set from a manuscript. The third was set from a printed copy of the second quarto of 1599, and the fourth from a printed copy of the third, and the descent can be proved from the inside, because each book inherits its copy’s accidents along with its words: the readings that only make sense as misreadings of the earlier page, the punctuation habits, the very shape of famous passages on the page. The Cambridge editors, William George Clark and William Aldis Wright, put the chain in one sentence in their preface to the play: the 1609 book “was printed from Q2,” the undated book “was printed from Q3 and the Quarto of 1637 was printed from it.” That sentence, reached by collation in the nineteenth century, has never been displaced. What later scholarship added was everything around it: who owned the copy, what the reprints changed, which one the Folio used, and when the undated book was really printed.

A reprint, understood properly, is not a transparent window on its copy. It is a fresh act of typesetting performed by workmen reading a printed page, often at speed, in a shop where the copy was money and time was money too. The compositors of 1609 read the 1599 text and set it again, and in the setting they normalized spellings that caught their eye, expanded or contracted words to fill their lines, misread letters, corrected what looked wrong, and occasionally improved what they touched by accident or intelligence. The fourth quarto’s workmen did the same to the 1609 page a decade and more later. The result is a text in slow motion: nearly the same at any two adjacent points, measurably different across the whole journey. The study of the reprints is the study of that motion, and of the uses a trade makes of a text it owns.

How big are the third and fourth quartos?

The third quarto runs to ninety-two pages, matching its copy page for page, and the fourth contracts to eighty-eight, the same text set a little more tightly. Paper was a book’s chief cost, so the sizes are prices: both reprints were built to sell at the established figure for a play quarto.

Extent is a trade fact before it is a textual one: a reprint that held its copy’s size held its price, and one that saved four pages saved money on every copy in the edition. The numbers frame both books as commercial objects built to a known market, not as monuments, and the reprint that kept the tragedy at the size and price of a cheap unbound book kept it in the hands of the readers who had made the earlier editions sell.

The imprints of the two books teach the trade’s grammar to any reader who pauses over them. “Printed for Iohn Smethwicke” names the man who financed the edition and owned the copy, not the man who set the type: in the imprint language of the period, the publisher follows “for,” and the printer, when named at all, precedes him. That is why Windet’s name is absent from the third quarto’s page and why the fourth quarto’s silence about its printer is the rule rather than a mystery. The shops are recovered by the other evidence, ornaments, devices, and type, which is why the bibliography of these books is a science of physical clues. The device deserves its own word: the smew, a small duck, carrying a scroll, is a rebus on the publisher’s name, Smethwick’s visual signature set into every copy, branding the book as his in an age before title pages carried much else of identity. When the fourth quarto’s two title states are compared, the device stands in both, the publisher’s mark constant while the author’s name comes and goes above it: the hierarchy of the page in 1622 still put the man who sold the book ahead of the man who wrote it, and the variant states catch that hierarchy in the act of inverting.

Why do the reprints matter if they only copy earlier books?

They matter because copying is how the text survived, and because these copies were used. The 1609 reprint became printer’s copy for the First Folio, and the undated reprint was corrected against the first quarto and passed its text to the fifth quarto of 1637. Used books are evidence twice over.

A witness that later craftsmen worked from shows what those craftsmen thought worth doing, and editors who skip the reprints misjudge both the Folio, which cannot be weighed until its copy is identified, and the play’s standing in its own century, which the reprint rhythm measures better than any compliment. The unglamorous books carry the history the glamorous ones merely begin.

How Did Smethwick Get the Play? The Rights Transfers of 1607

The third quarto exists because of a property transaction, and the transaction survives in the Stationers’ Register. The second quarto had been published in 1599 by Cuthbert Burby, printed for him by Thomas Creede. On 22 January 1607 Burby transferred the right in the tragedy to Nicholas Ling, and on 19 November of the same year Ling transferred it onward to John Smethwick, in an entry in the Stationers’ Company’s Liber C that moved sixteen titles at once. Smethwick was building a list. The same Ling transfer carried other proven Shakespeare properties, Hamlet among them, and Smethwick worked the portfolio steadily: he issued a Hamlet quarto in 1611 from the same acquisition, and the tragedy of the Veronese lovers in 1609, each title a known seller whose earlier editions had already proved the demand. The reprint history of the play begins, therefore, not in a printing house but in a rights market, where the tragedy was an asset with a track record.

That origin explains the third quarto’s character. Smethwick had bought a text, not a manuscript: what the transfer conveyed was the right to print the play, and the copy his printer worked from was the printed second quarto, the best text the trade possessed. There was no occasion to seek the author’s papers, which in any case belonged to the acting company rather than to any stationer, and no mechanism by which a publisher of 1609 would have obtained them. The reprint was the rational product: the owned text, freshly set, attractively titled, and offered to a market that had absorbed two editions already. When the Folio’s syndicate came to print the play fourteen years later, it faced the same arithmetic and reached the same shelf, taking a copy of Smethwick’s 1609 book as its copy. The rights deal of 1607 thus stands behind both the third quarto and the Folio text, a single purchase propagating through the century’s two most important printings of the play.

The geography of the transaction is worth a glance, because the reprint history has an address. Saint Dunstan’s Churchyard, in Fleet Street under the Dial, where Smethwick’s shop stood, was one of the trade’s established bookselling quarters, a few minutes’ walk from Saint Paul’s Churchyard, the industry’s centre. A buyer who wanted the tragedy in 1609, in the early 1620s, or in 1637 went to the same shop sign for it, across a span in which the play’s author died, its company changed its king, and its text entered the great folio. Continuity of address is a small fact with a large implication: the play was not rediscovered by successive generations of stationers. It was kept, by one house, in one place, as a standing item of stock, and the reprint history’s steadiness is the steadiness of a shop that never took the title off its list.

The demand the deal priced was real, and other records catch it. Sir John Harington, the courtier and translator, compiled lists around 1609 and 1610 of the play quartos in his own collection, and the tragedy appears among the holdings of a gentleman reader who collected drama as literature. A manuscript commonplace book of the early 1620s, copying extracts to remember, takes one of its eleven Shakespearean extracts from this play, alongside passages from Richard II, Hamlet, and Othello. These are the readers the reprints served: not the playhouse crowd, who needed no book, but the growing company of buyers who wanted the famous plays on a shelf. A text reprinted in 1609, again about 1622, and again in 1637 was being bought by such readers at a steady pace for forty years, and the reprint sequence is the hardest evidence of the play’s literary afterlife before the Restoration that the record offers.

Q3 (1609): A Reprint of Q2, Page for Page

John Windet’s compositors set the third quarto from a copy of the second, and the proof of the descent lies in the details no independent text would share. Where the 1599 book errs distinctively, the 1609 book follows it into error; where the 1599 book’s spelling is idiosyncratic, the 1609 book echoes the idiosyncrasy too often for coincidence; where the 1599 book dislocates a famous passage, the dislocation passes through untouched. The rose speech of the balcony scene is the cleanest exhibit. In the second quarto the sentence arrives with its catalogue of body parts dislocated and its great question unpunctuated, reading “word” where memory insists on “name”: “That which we call a rose, by any other word would smell as sweet” (2.2). The third quarto reprinted that sentence entire, dislocation, bare question, and “word” together, handing the whole unsettled state of the passage into the copy from which the Folio would be set. No compositor working from a manuscript produces that pattern. Only a compositor working from the 1599 page does.

The title page makes the same point in the trade’s own language, provided it is read as a document and not as an advertisement. The 1609 page keeps the second quarto’s boast, “Newly corrected, augmented, and amended,” a formula that by then described nothing anyone had done to the text in 1609; it was inherited copy, reprinted with the rest of the page, because the page itself was part of what Smethwick had bought. What the 1609 stationer actually changed is more telling than what he copied. The acting company is updated: where 1599 named the Lord Chamberlain’s servants, 1609 names the King’s Majesty’s servants at the Globe, the company’s style since 1603 and its famous address. The book is dressed for the market of its own year, its company current, its claim traditional. One omission in that dress deserves emphasis, because the next quarto turns on it. For all its updating, the 1609 title page still does not name the play’s author. The tragedy had been in print for twelve years, through three editions, and no quarto title page had yet said who wrote it.

Windet’s shop produced a competent trade reprint, neither careful nor careless beyond the norms of its kind. The Cambridge editors’ judgment on its text has stood for a century and a half: printed from the second quarto, it “differs by a few corrections, and more frequently by additional errors.” The proportions matter. A reprint’s corrections are usually small and obvious, the mending of turned letters, broken words, and nonsense that any attentive workman spots; its additional errors are the steady tax of resetting, the misread minim, the skipped or doubled word, the spelling modernized in one place and fossilized in another. Brian Gibbons, editing the play for the Arden Shakespeare, and Jill Levenson, editing it for the Oxford Shakespeare, both treat the 1609 book exactly so: a derivative witness whose agreements with the second quarto prove nothing, because agreement is what reprinting means, and whose disagreements are evidence only of the printing house. That disciplined contempt is correct as far as it goes. What it can miss is the book’s second career, for the derivative witness of 1609 became, in 1622 and 1623, the most influential copy of the play in the world.

Did the third quarto correct the second quarto’s mistakes?

It corrected some and added others, in the proportions the Cambridge editors named: a few corrections against more frequent additional errors. Neither class required a manuscript, and neither shows one at work. The 1609 text is the 1599 text after ten years and one resetting, which is what its place in the chain required.

The corrections are the kind a printing house makes unaided, clearing obvious nonsense, repairing broken words, tidying pointing where the copy’s pointing baffled the compositor. The additions are the kind a printing house cannot help making, the small misreadings and substitutions that accumulate whenever type is reset from a printed page. The 1609 book never restores a passage its copy lacks, and never repairs a famous difficulty with authority, and those two silences are the surest signs that no second source stood open beside the 1599 page.

How Q3 Was Made: Windet’s Shop and the Setting of a Reprint

A reprint begins as a physical task, and the task’s shape explains the text’s. Smethwick’s copy of the 1599 edition went to John Windet’s printing house as the exemplar, a printed book to be read by workmen and converted, sheet by sheet, into standing type. Windet was a veteran London printer whose shop turned out a heavy list of books across four decades; his name is attached to the third quarto not by its imprint, which names only the publisher, but by the ornaments and typographic habits bibliographers use to assign anonymous printing, an attribution the standard catalogues record without dispute. Two years after the edition appeared Windet was dead, and his shop’s materials passed to his successor William Stansby, a succession that matters to the fourth quarto’s story, because the undated book’s types belong to the same printing milieu and could be dated only by following that milieu’s dated work.

The shop’s first job was arithmetic. Before a line could be set, the copy had to be cast off: counted and divided so that each portion of text could be assigned to a compositor and fitted to its pages, in a book printed, as quartos were, by sheets that were set and machined out of reading order. For a reprint the arithmetic is eased by the copy itself. The 1609 book matches the 1599 book’s extent exactly, ninety-two pages against ninety-two, and the match is the fingerprint of page-for-page setting, the compositors following their exemplar’s pagination so that each page of new type answers to a page of the old book. Page-for-page setting is the reprint trade’s natural method: it keeps the cast-off copy honest, it lets two workmen share a book without confusion, and it makes the finished edition a physical echo of its copy before a word is compared. The echo is also a trap for the unwary reader, since a page-for-page reprint looks more faithful than it is; the words on each page correspond, while the small accidents of setting, the turned letter, the modernized spelling, the substituted word, accumulate beneath the correspondence.

Correction in such a shop was a press-side activity, not a study discipline. A proof reader, where the work had one, read the setting against the copy for gross faults; compositors corrected what they noticed as they worked; and where a sheet was corrected while printing ran, the edition itself divides into states, early copies carrying the uncorrected reading and later ones the correction, so that no two copies of a hand-press book are guaranteed identical in every word. The third quarto’s text must be imagined with that grain in it: not a single fixed state but a small family of states, differing at the level of the letter and the stop. Editors collate copies for exactly this reason, and the reprint history the collations build is a history of states as well as editions. What no stage of the process involved was comparison with anything beyond the copy. No manuscript stood in Windet’s shop, no author walked through it, and the authority the finished book could claim was the authority of a faithful act of copying, performed at trade speed by men whose skill was real and whose instructions were the page in front of them.

That last point deserves its weight, because it explains the third quarto’s peculiar double character, derivative and influential at once. The shop did its work well by the trade’s standards: the edition is complete, orderly, and close to its copy, close enough that identifying its copy is a matter of proof rather than impression. But trade standards measure fidelity to the exemplar, not to the lost manuscript behind the exemplar, and a perfect reprint of 1599 would still stand one remove from the papers of 1595. The third quarto is a very good copy of a book. Its long afterlife, as the Folio’s exemplar and the fourth quarto’s copy, would convert that goodness into something like authority, as later craftsmen treated the convenient page in front of them as the text of the play. The conversion of convenience into authority is the reprint trade’s deepest habit, and this play’s history is its cleanest demonstration.

How Reprint Errors Accumulate: The Drift Between Q2 and Q3

The mechanism of drift deserves a section of its own, because it explains not only these two books but every derivative text in the play’s history. A compositor setting from printed copy reads a phrase, holds it in memory, and sets it; the errors of that process are not random in the way that matters. They cluster at the points where the copy is hard: unfamiliar spellings get modernized, similar letters get exchanged, short function words get substituted for their neighbours, and verse lines that the copy prints as prose, for economy, are very hard to restore to verse by eye. The second quarto had set the Queen Mab speech as prose to save space, a famous economy at 1.4, and the 1609 compositors, following their page, kept it prose. The reprint transmits the saving as faithfully as the words. What a reprint inherits, therefore, is not only its copy’s text but its copy’s compromises, and each generation of the chain freezes the previous generation’s shortcuts into the appearance of authority.

Accumulation has a direction as well as a rate. Errors that make a kind of sense survive proofreading; errors that produce nonsense attract correction, and the correction is a guess. Across two reprints the text therefore drifts toward the plausible: readings get smoother, spellings get more familiar, punctuation gets more helpful and less faithful. The rose passage again supplies the miniature. The second quarto’s unpunctuated “Whats in a name” troubled no one in 1609, and the third quarto passed it on bare; only in the fourth quarto does a compositor supply the pointing that turns it into the modern question, “What’s in a name?” The improvement is real and the improvement is the problem: every such kindness moves the received text one step from the copy and presents the step as the text itself. Multiply the kindness by two reprints and a Folio, and the distance between the play as printed in 1599 and the play as read in 1700 becomes a distance no single editor chose, but thousands of small mercies composed.

The drift also flows through the feature of the 1599 text that most embarrasses its champions, the dawn passage printed twice. The second quarto carries Romeo’s dawn quatrain at the end of the balcony scene and its twin at the opening of the Friar’s cell scene, an unresolved doublet that betrays the working papers behind the 1599 printing. A reprint has no machinery for noticing such a thing. Windet’s men set what stood on their copy’s pages, and the doublet passed into 1609 entire, and from 1609 toward the Folio, its duplication invisible to a trade process that checks words against copy, not sense against structure. The lesson generalizes: a reprint chain transmits problems with the same fidelity as solutions, and the critic who would weigh any later witness must first reconstruct the chain, because the chain, not the witness, determines what its agreements and disagreements mean. How compositors work, and what their habits can and cannot prove, is the subject of a companion study of compositor analysis and the printing house, whose findings the reprint history presupposes at every step.

Set against the drift, the corrections stand out as the chain’s counter-current, and they are the reason the reprints cannot be dismissed as mere decay. The 1609 compositors repaired some of the 1599 book’s faults as they went, and the fourth quarto’s printer, as a later section shows, went further and consulted the first quarto to mend the third. The chain is thus not a photocopier but a series of workshops, each adding noise and each removing a little. The net direction, across 1599 to 1637, is what the collations show: a text recognizably constant in substance, slowly modernizing in dress, accumulating small faults at a rate the occasional corrections never quite match. Editors who base their texts on the second quarto are not being nostalgic; they are walking back up the chain to the point nearest the manuscript, before the drift began.

Q3 in Detail: What Collation Finds in the 1609 Text

Collation, the word-by-word comparison of the 1609 text against the 1599 text, produces a profile distinctive enough to recognize across the whole family of derivative editions. The agreements are massive and meaningless, the necessary result of reprinting. The disagreements sort into kinds, and the kinds tell the story. The largest kind is accidental: spellings and punctuation, the surface of the page, where the 1609 compositors dressed the text in the habits of their own decade, regularizing some of the older book’s forms, varying others without pattern, and repointing sentences where the copy’s rhetorical pointing no longer matched a Jacobean compositor’s sense of a pause. Accidentals are the reprint’s weather: pervasive, unarguable, and almost never significant for meaning, except that an editor who prints the second quarto’s accidentals is choosing the layer nearest the manuscript, and must therefore walk past the 1609 layer to get them.

The second kind is the small substantive traffic of resetting: a word misread for its neighbour, a short word dropped or doubled, a speech’s wording smoothed where the copy was crabbed. These are the errors the Cambridge verdict counts, more frequent than the corrections, and their distribution is the drift’s signature: they thicken where the copy is difficult and thin where the copy runs clear, exactly as errors distribute when a text is being copied by eye rather than composed from sense. The third kind is the corrections, compositorial or a corrector’s, where nonsense in the 1599 page has been mended into sense in 1609. A correction of that kind proves attention, and attention of a strictly local kind: the mended reading is the obvious reading, the one any competent workman supplies, and it never extends to information the copy does not contain. Where the 1599 text is corrupt beyond a workman’s guessing, the 1609 text transmits the corruption without a tremor, and those silences, collated across the book, are the strongest internal proof that Windet’s men had the printed page and nothing else.

Two inherited features deserve naming because the Folio would inherit them again from this book. The first is the standing doublet of the dawn passage, Romeo’s quatrain at the close of the balcony scene answered by its twin at the opening of the Friar’s cell, which the 1609 compositors set twice as their copy printed it, neither noticing nor empowered to notice that the play had been given the same lines twice. The second is the pair of Chorus speeches, which the third quarto transmits entire, the sonnet Prologue and the second sonnet that opens the action after the feast. Both features matter downstream: the Folio’s printers, working from this book, reproduced the doublet’s substance and lost the second Chorus, and the difference between those two fates is a map of where the Folio’s text was passive and where its preparation intervened. The reprint itself intervened nowhere; it was the perfect conduit, which is why its readings, where the Folio repeats them, carry no independent weight, and why its errors, where the Folio repeats them, identify the Folio’s copy beyond argument. A conduit is known by what flows through it unchanged.

Modern editions handle the 1609 book with a consistency that amounts to a verdict on it. The apparatus of a scholarly edition records the third quarto’s variants where they matter, which is chiefly where the Folio agrees with them, since that agreement is the Folio’s pedigree; the text is never taken from 1609 where 1599 is available, and no reading is admitted on the third quarto’s sole authority. Gibbons’s Arden edition and Levenson’s Oxford edition, approaching the play from different editorial generations, agree in this handling because the bibliography leaves them no choice. The 1609 book’s scholarly function is diagnostic rather than constitutive: it tells the editor where the Folio came from, and it tells the historian how the play read in the year of the King’s Men’s maturity at the Globe. Those are large services for a book that contributed almost nothing to the text it carried, and they are the services the rest of this study keeps calling on.

The Annotated Q3: How the 1609 Reprint Became the Folio’s Copy

In 1623 the syndicate that published the First Folio needed copy for every play in the collection, and for this tragedy it used the book the trade had ready: a copy of the third quarto. The fact is established by the same internal evidence that establishes every link in the chain. The Folio text follows the 1609 reprint through its distinctive errors, errors the Folio compositors could not have found in the 1599 book because the 1599 book does not contain them; where the third quarto drifted from the second, the Folio drifts with it. W. W. Greg, in his bibliographical history of the Folio, placed the tragedy’s copy in the line of the third quarto, and no subsequent investigation has moved it. The Folio text of the play is therefore a reprint of a reprint of the 1599 edition, at two removes from the “good” quarto and three from the manuscript behind it, and every judgment about the Folio’s readings has to begin from that address in the chain.

But the copy the Folio’s printers used was not a clean copy of the 1609 book, and here the reprint history produces its most consequential episode. The Folio departs from the third quarto at points where mere typesetting does not explain the departure: stage directions appear that the 1609 page lacks, fuller and more theatrical than a compositor invents, and readings shift in ways that suggest someone had prepared the exemplar before it reached the printing house. Barbara Mowat and Paul Werstine, introducing the Folger Shakespeare text, state the inference plainly: the Folio “reprinted the Third Quarto, but its printer’s copy must have been annotated by someone, because the Folio departs from the Third Quarto in ways that seem beyond the capacities of mere typesetters.” Someone, that is, marked up a copy of Smethwick’s 1609 reprint, adding directions and touching readings, and Isaac Jaggard’s compositors set the Folio from the marked book. The anonymous annotator is the last craftsman in the chain whose work can still be detected, and the third quarto is the page he worked on.

Two adjudications follow. The first concerns the fact of annotation, and there the verdict goes with Greg and the Folger editors: the pattern of departures, above all the added stage directions, exceeds compositor improvisation, and an annotated exemplar is the economical explanation. The second concerns the annotator’s source, and there the record supports less than has sometimes been claimed. A playhouse manuscript, even the promptbook, has been proposed as the authority behind the marks; the marks themselves do not prove it. Nothing the annotator added requires a manuscript: the directions could come from theatrical knowledge, the corrected readings from collation against another printed text, and the most dramatic departure of all, the Folio’s loss of the second Chorus speech, “Now old desire doth in his deathbed lie” (2.0.1), is as easily an accident or a cut in the exemplar as a promptbook’s cancellation. Jill Levenson’s Oxford edition accordingly prints the descent without inflating it: the Folio derives from the third quarto, its exemplar annotated, its independent authority nil. The annotated copy is best understood as a working book in a working trade, prepared by a hand that knew the stage, not as a lost channel to the author. What the annotator used the third quarto for is treated at full length in the companion study of Romeo and Juliet in the First Folio; what matters for the reprint history is the identity of the page annotated, and the page was 1609’s.

The consequences for editors are exact. Because the Folio’s copy is identified, its departures from that copy can be listed exhaustively, and a witness whose departures can all be listed and priced has no residual authority: it cannot surprise. Modern editions therefore take their text from the second quarto, the subject of the companion study of the second quarto of 1599, the ‘good’ quarto, and admit Folio readings only where the 1599 text is manifestly wrong and the Folio’s variant manifestly right, a door through which very little passes. Three centuries of readers met the play in the Folio’s dress, through editions descended from Rowe, who printed from the late quartos and the Folio line; the reprint history is what allows a modern reader to see that dress as a costume put on the text in 1623, over a 1609 reprint, over the 1599 book that stands closest to the playwright’s hand.

Why does the Folio text differ from the second quarto if it descends from it?

It differs because descent ran through an intermediary and a preparation. The third quarto had already drifted from the second in spelling, pointing, and a scatter of readings; the Folio inherited that drift entire and added its own compositorial layer and the annotator’s marks. Shared descent is not sameness, only kinship.

Two texts can share a line of descent and stand measurably apart at every point where the chain’s workshops touched them. The Folio’s differences from the 1599 book are therefore not evidence of a second source behind the Folio; they are the sum of the chain’s small events, priced link by link. That is why collation, which reconstructs the chain, must precede judgment, which weighs the witnesses: a Folio reading that merely repeats the third quarto’s error is one witness speaking twice, not two witnesses agreeing.

A Marked Book at the Press: How the Annotated Exemplar Was Used

The annotated copy of the third quarto deserves a section of its own, because it is the most worked-on book in the play’s early history and the least visible. Picture the object: a copy of Smethwick’s 1609 edition, its margins and interlines carrying a second hand’s additions, stage directions entered where the print had none, readings touched, passages marked. Such a book was a tool, prepared for a printing house, and its preparation converted a trade reprint into copy for a collected edition. Whoever prepared it worked through the whole play with the printed text as the base and theatrical knowledge as the supplement, and the Folio’s departures from the 1609 page are the fossil record of that work. The directions the Folio adds have the flavour of the playhouse rather than the study, businesslike notes of who enters, who dies where, what the stage must show, and they sit in the Folio text to this day, reprinted by editors who take their base text from 1599 and their staging from a marked copy of 1609 by way of 1623.

At Jaggard’s shop the marked book was copy in the technical sense: the compositors set from it, casting off its pages into the Folio’s columns and following its text as compositors follow copy, with their own habits superimposed. Charlton Hinman’s great study of the Folio’s printing identified the workmen and their stints across the volume, and for this play the setting proceeded from the annotated quarto as copy, its readings passing through the compositors’ fingers with the usual small toll of error and normalization. The important structural fact is the double mediation. Between the reader of 1623 and the text of 1599 stand two acts of copying, the annotator’s pen on the 1609 page and the compositors’ type from the annotated page, each act adding its own character: the annotator’s theatrical, the compositors’ mechanical. An editor pricing a Folio variant has to ask which act produced it, and the answer is usually recoverable, because the two acts err in different ways. Directions and deliberate-looking corrections point to the pen; misreadings and spelling habits point to the type.

The marked book itself does not survive, and its loss is worth a moment’s regret, because it would settle the argument about the annotator’s source at a glance. A promptbook’s descendant would look one way, a reader’s collation another, and the hypotheses that now compete on inference would compete on evidence. In its absence the reprint historian works with probabilities, and the probabilities favour modesty: an annotator with the theatre in his head, preparing a popular play’s reprint for a prestige volume, in a trade where annotated quartos demonstrably served as Folio copy for other plays as well. W. W. Greg allowed the mechanism across the Folio, and the Folger editors’ formulation for this play, printer’s copy annotated by someone, keeps the claim at its provable size. What the marked book proves, even lost, is the third quarto’s centrality: when the most ambitious publishing venture of the age needed the tragedy, the book it reached for, marked up, and set was the reprint of 1609.

Q4: The Undated Quarto and the Argument Over Its Year

The fourth quarto refuses to say when it was printed, and the refusal has generated a longer argument than many dated books enjoy. Its title page gives the title, the company at the Globe, the inherited boast, Smethwick’s device, and the imprint “Printed for Iohn Smethwicke,” with no printer’s name and no year. Before its date could be argued, its place in the sequence had to be fixed, and that was the nineteenth century’s contribution. Charles Knight could still speak of a quarto of 1607, and the British Museum’s catalogue conjecturally dated its copy of the undated book to that year, which would put the supposed fourth edition before the dated third. The Cambridge editors closed that possibility with internal evidence: collation proves the undated book was printed from the 1609 quarto, and that the dated quarto of 1637 was printed from the undated one. A book printed from a book of 1609 cannot be of 1607. The undated quarto is fourth, and its year, whatever it is, falls after 1609 and before 1637.

Within that window the proposals spread widely, and grading them is a tidy exercise in bibliographical evidence. The British Museum’s 1607 conjecture fails on descent and can be discarded. Datings that cluster the book near its copy, in the years just after 1609, rest on nothing firmer than a feeling that reprints follow quickly; the trade’s rhythm for this play, a decade and more between editions, argues the opposite. The date the catalogues settled on, and the English Short Title Catalogue still carries, is a bracketed 1622, a judgment of approximate contemporaneity with the Folio’s production rather than a demonstrated year. Against the bracket stands the one dedicated investigation: R. Carter Hailey’s typographical analysis of the book, which traced its types and their progressive damage through dated work from the same printing milieu and fixed its printing in 1623. The Folger Shakespeare’s editors print Hailey’s conclusion: the fourth quarto “has been dated by Carter Hailey as having been printed in 1623.”

The verdict this study reaches is that Hailey’s 1623 is the best-supported single year, with the catalogues’ 1622 understood as the same moment described less exactly: the book belongs to the Folio years, 1622 to 1623, and typographical evidence of the kind Hailey marshalled is the strongest currency in which an undated book’s year can be paid. The grading matters because the date changes the book’s meaning. A quarto of 1611 would be a Jacobean reprint keeping a favourite warm; a quarto of 1623 is something stranger and more interesting, a trade reprint issued in the very season when the great collected Folio was in the press, offering the same play in the old cheap format to the old market while the syndicate dressed it for posterity a few streets away. And it raises the question the Folio’s copy answers from the other side: when Jaggard’s men needed the tragedy, they took the 1609 book, annotated, not the new quarto, because copy is chosen from what a shop holds and has prepared, and the newest reprint had no part in the Folio’s making. The fourth quarto ran in parallel to the Folio, not into it, and its line continued on its own to 1637.

The season the dating assigns deserves one more look, because 1622 and 1623 were not ordinary years in the play’s existence. The Folio was in production across both, its sheets passing through Jaggard’s shop in a long campaign that reached the tragedies in its later stages, and the syndicate behind it had gathered the copy, settled the rights, and set the volume’s order while Smethwick’s new quarto of the same play was being prepared elsewhere in the same small trade district. London printing was a village: the same paper merchants, the same typefounders, the same carriers served both ventures, and the workmen of one shop drank with the workmen of the other. In that village the fourth quarto is not a rival to the Folio but a symptom of the same moment, the moment when the play’s value stood so high that two formats of it could be manufactured at once without either spoiling the other’s market. The cheap quarto did not fear the folio, because their buyers were different men; the folio did not notice the quarto, because its copy had been settled before the quarto existed. Hailey’s date, by putting the books in the same season, converts the fourth quarto from a footnote in the Folio’s prehistory into the Folio’s contemporary portrait of the trade that made them both.

Was the fourth quarto printed before or after the First Folio?

The two books are twins of the same season, and their order within it is not certainly known. Hailey’s typographical dating puts the quarto in 1623, the Folio’s own year; the catalogues’ 1622 puts it just before. Neither book used the other, and printers did not wait for one another.

What the evidence settles is the independence of the two descents: the Folio was set from an annotated copy of the third quarto, and the fourth quarto from an unannotated one, each reaching 1623 by its own path from 1609. In the same months, the tragedy existed in London in a new cheap quarto and in the great folio volume, the twin issues of a single property at the height of its value.

The Ghost Quarto of 1607 and the Question of Lost Editions

Every long bibliography grows a ghost, and this one’s is a quarto of 1607 that never existed. Charles Knight, the industrious Victorian editor, mentioned a quarto of the play in that year, and the British Museum’s catalogue, faced with an undated copy of the fourth quarto, the anonymous one now pressmarked C. 34. k. 56, conjecturally dated it 1607 and so gave the ghost a shelf to haunt. The conjecture had a certain surface logic: the anonymous title page looked primitive beside the dated 1609 book, and a cataloguer reaching for a year could imagine the plain page the earlier. Collation dissolved it. The undated book’s text descends from the 1609 quarto page by page, inheriting readings that did not exist before 1609 was printed, and a book cannot be set from a copy that postdates it. The quarto of 1607 is the British Museum’s guess, Knight repeated the guess, and the guess is the whole of the evidence. Bibliographies are full of such entries at their edges, and the moral generalizes beyond this play: a date in a catalogue is a claim, and a claim’s value is the evidence beneath it, which here was a cataloguer’s hunch about a title page’s look.

The ghost earns its section because killing it raises the honest question it travestied: could an edition of the play have existed and vanished entirely? The survival of early playbooks is poor enough to allow the thought. The first quarto of this play survives in a handful of copies; some editions of other plays survive in one copy, or in fragments, or are known only from a Register entry or a bookseller’s list. Peter Blayney’s studies of the playbook trade estimate editions in the hundreds of copies, five hundred to a thousand being the plausible range for a play quarto, and of such runs, sold cheap, read hard, and bound late if at all, the loss of every copy of an edition is a physical possibility no bibliographer can exclude in principle. Register evidence is the control: Smethwick’s title to the play passes through the two transfers of 1607 and no entry records a further edition, and no contemporary list, Harington’s included, names an edition the chain does not know. The verdict is therefore disciplined: a lost edition between 1609 and the early 1620s cannot be disproved, and there is no reason to posit one. The chain as collated is complete at every joint, each book’s copy identified in the previous book, and a hypothetical edition would have to hide in a joint that shows no gap.

The exercise matters for the dating argument, because it clarifies what kind of question the undated quarto poses. Its year is not lost in the way a lost edition is lost; the book survives, its types survive in fixed states of damage, and the evidence for its year exists and can be weighed, which is what Hailey’s investigation did. The difference between the ghost and the problem is the difference between speculation and bibliography. Knight’s 1607 had nothing beneath it but a look; Hailey’s 1623 has the progressive damage of identifiable type, checked against dated books. The reprint history contains both kinds of claim, and learning to tell them apart is most of what the discipline of the subject consists in.

Q4’s Two Title Pages: The Author’s Name Appears, and Disappears

The undated quarto survives with two forms of title page, and the difference between them is the most eloquent square inch in the play’s print history. In one form the page runs exactly as the 1609 page had run, title, company, boast, device, imprint, and no author. In the other form a line has been inserted after the Globe line: “Written by W. Shakespeare.” Everything else stands. The copies differ in the title only, as the catalogue of the 1916 Shakespeare exhibition at the New York Public Library put it, one form with the name inserted, the other anonymous; the exhibition’s copy of the anonymous form was one of three then known. With that inserted line, the author’s name appears on a title page of this play in the quarto line for the first time, in some copies of an undated book, twenty years and more after the play was written and several years after its author died.

The bibliographical question, which form came first, produced a clean disagreement, and the disagreement is instructive because both sides argued from the same copies. J. O. Halliwell-Phillipps held that the named form came first: copies with the name were published, and the name was then omitted in copies subsequently issued, a sequence Edward Dowden’s Arden edition of 1900 reports with Halliwell-Phillipps’s authority attached. The exhibition catalogue of 1916 judged the opposite: the anonymous form is “probably the earlier,” presenting it in its entries as the fourth edition’s earlier title and the named form as the later state. Each ordering implies a story. If the name came first and was dropped, the insertion looks like a false start, a publisher testing the author’s brand and retreating to the familiar anonymous page. If the anonymous form came first and the name was added, the sequence shows a seller discovering, mid-issue, that the name had become the selling point, and stopping the press to put it in.

The adjudication this study reaches is that the order cannot be proved from the surviving copies, and that the scholarly value of the two forms does not depend on solving it. Survival counts are too small and too accidental to sequence issues; the anonymous form’s three known copies in 1916 measure collecting luck as much as print runs. What the two forms establish jointly is firmer and more interesting than either sequence: in the early 1620s, at the end of the quarto line’s long anonymity, the name of Shakespeare had become a thing a title page might carry or might not, inserted by a trade decision rather than assumed as a fact of the book. The fifth quarto of 1637 settled the question the way the market settles questions, by keeping the name: its page reads “Written by W. Shakespeare” as a matter of course, above the imprint of R. Young, printing for the same Smethwick who had issued the play anonymously for thirty years. The author’s name conquered the title page not by proclamation but by a variant state in an undated reprint, and any account of the play’s early fame that begins the name’s reign in 1609 has misread which book introduced it.

Does the fourth quarto name Shakespeare on its title page?

Some copies do and some do not, which is the whole point of the book’s divided survival. The edition was issued with its title page in two forms: anonymous in one, and in the other carrying the line “Written by W. Shakespeare” after the Globe, the author’s first title-page appearance in this play’s history.

A buyer in the early 1620s might therefore carry home either the old anonymous tragedy or the newly branded one, in sheets otherwise identical, and neither buyer would know the other form existed. The variant is a reminder that early modern books were not uniform objects, and that a claim about what an edition says must always be checked against more than one surviving copy.

What Did Q4 Change? A Reprint That Consulted Q1

A reprint that merely reprinted would be a simpler object than the fourth quarto turns out to be. Its copy was the third quarto, and the descent shows in the inherited errors and readings that bind the two books; but the workmen or the corrector who prepared its copy did not follow the 1609 page blindly. The Cambridge tradition noticed the fact early: the undated book’s text is, in the old editors’ phrase, more correct than the text of 1609. The Folger editors specify the method: the fourth quarto’s printer “appears to have consulted the First Quarto for some corrections and word choices.” The despised short text of 1597, displaced in 1599 and treated ever since as the tradition’s embarrassment, was on the desk in the printing house of the early 1620s, and readings were taken from it to mend the reprint of the reprint of its rival.

The gesture deserves a moment’s wonder, because it is the chain’s only backward step. Everywhere else the transmission flows downhill from print to print, each book taking the last as its authority; here a craftsman checked the current page against the oldest witness and preferred the oldest witness at points. The corrections are small, word choices and mendings rather than restorations, and no passage suppressed since 1599 returns; the first quarto was consulted, not adopted. But the consultation proves that the printing house, not only the study, practised a primitive collation, comparing witnesses and choosing between them, a generation before the Folio and a century before the editors made the method famous. How the two early quartos differ, and what an editor does when they disagree, is worked through passage by passage in the companion study of Q1 vs Q2 and what the differences reveal; the fourth quarto’s shop-floor version of the same exercise is its modest claim on the history of editing.

The detection of the consultation is itself a small lesson in method, worth spelling out because it shows how reprint history is proved rather than asserted. The fourth quarto agrees with the third in the mass of readings, as a reprint must; the significant cases are the minority where the fourth departs from the third and agrees with the first quarto against both the third and the second. A departure of that shape cannot come from the copy, since the copy reads otherwise; it cannot easily be a compositor’s guess, since the guess would have to land repeatedly on the first quarto’s exact words; and it cannot come from the second quarto, which the departure also contradicts. The remaining explanation is consultation: a copy of the 1597 book open in the shop, used as a corrector’s check. Each individual case is small, and scepticism about any one of them is healthy; the pattern across cases is what convicts. Bibliographical proof is cumulative in exactly this way, and the fourth quarto’s quiet consultation of the first is as firmly established as anything in the chain, by the same logic of shared and unshared readings that establishes the chain itself. There is even a fitness in the detail: the book that first printed the play, dismissed by the 1599 title page and by three centuries of editors after it, was still earning its keep in a printing house in the 1620s, consulted by workmen who cared about the words and had no theory to defend.

The rest of the fourth quarto’s work is the usual reprint mixture at slightly better odds. Its pointing is tidier than its copy’s: the rose sentence’s question receives in this book the pointing that makes it a question on the page, “What’s in a name?” dressed almost as modern readers know it, though the disputed word itself passes through unchanged, as reprints pass cruxes rather than judge them. Its eighty-eight pages against its copy’s ninety-two show a shop setting economically, a slightly tighter page buying the same text at less paper. And its errors are its own, a fresh scatter laid over the inherited drift, so that the fifth quarto of 1637, set from it, inherits two reprints’ worth of accident beneath the 1599 text. By 1637 the chain ran Q2, Q3, Q4, Q5, four settings of type between the reader and the manuscript’s nearest printed heir, and the fifth quarto, the last of the play printed before the theatres closed, is the chain’s natural terminus: Smethwick’s property, still selling, its text still recognizably the 1599 text, its title page now carrying the author’s name as the first fact of the book.

The Reprints as Stage Evidence: Sundry Times at the Globe

The reprints testify about the stage as well as the page, and their testimony is the kind historians of performance rely on precisely because it is incidental. Each title page in the chain repeats the formula of performance: the play “hath beene sundrie times publiquely Acted,” by the Lord Chamberlain’s servants in 1599 and by the King’s Majesty’s servants at the Globe in 1609 and after. Title pages lie about texts often enough, as the correction boast shows, but a performance claim in 1609, renewed in the early 1620s and again in 1637, is a different kind of statement: it advertises a play the buyer could go and see, in a city where the claim was checkable any afternoon. The Yale edition’s history of the play drew the inference plainly a century ago: the fact that the tragedy was reprinted four times in quarto, with the evidence of its title pages, supports the belief that it held the stage until the theatres closed in 1642. The reprint sequence and the repertory sequence corroborate one another, two independent traces of the same continuous popularity.

The updated company style on the 1609 page does finer work still. When Smethwick’s compositor set “the Kings Maiesties Seruants at the Globe” in place of the 1599 book’s Chamberlain’s servants, he was not merely modernizing an imprint; he was recording which company owned the play in performance and where it played, the King’s Men at the Globe on the Bankside, the house the tragedy had helped to fill since 1599. A reprint’s title page, reset from its copy, had every reason of laziness to repeat the old style, and the change shows the shop attending to the play’s living circumstances. Bibliography and theatre history meet in that line: the book that would carry the text to the Folio also certifies, in passing, that the play it carried was a current repertory piece of the kingdom’s leading company, not a revivalist’s reconstruction. The fourth quarto repeats the Globe style unchanged, and the fifth repeats it in 1637, by which time the formula had become the play’s permanent stage address in print, outliving several rebuildings of the playhouse itself. The reprints are thus the play’s continuous programme as well as its text, and the stage historian who ignores them loses the cheapest run of performance evidence the period provides.

One refinement keeps the evidence honest. A title page’s performance claim is a publisher’s sentence, renewed at each reprint by a tradesman whose interest was to present the play as a living repertory piece, and a claim repeated in 1637 may owe something to the copy it was set from as well as to the Globe’s stage in that year. The corroboration works because the claim does not stand alone: the reprint rhythm itself, edition following edition at a working play’s pace, is independent testimony that cannot be faked by a formula, and the two together, the words on the page and the money behind them, make the case neither could make separately. Evidence in this subject is nearly always of that compound kind, a statement priced by an investment, and the reprint title pages are its model instance: believed, rightly, because Smethwick kept paying to repeat them.

After the Folio: Q5 and the Reprint’s Long Shadow

The chain’s last link was forged in 1637, when R. Young printed the fifth quarto for Smethwick, setting it from the fourth quarto as the fourth had been set from the third. The 1637 book closes the quarto line with a kind of symmetry. Its text is the chain at full length, the 1599 play after three reprints, sound in substance and drifted in detail, the accumulated accidentals of three shops upon it. Its title page completes the journey the fourth quarto’s variant began: “Written by W. Shakespeare” now stands as a standard line of the page, and the book is sold, openly and entirely, as the work of its author, forty years after the first quarto offered the same play with no name at all. Between the anonymous pamphlet of 1597 and the branded book of 1637 lies the whole early history of Shakespeare as a selling name, and the reprint line, not the Folio, is where that history can be watched happening in stages.

Meanwhile the Folio line ran its own descent, and the two lines together governed what the next three centuries read. The Second Folio of 1632 reprinted the First with corrections, some intelligent and many not, and the third and fourth folios followed, the text passing from folio to folio as it had passed from quarto to quarto, with the same drift and the same occasional mending. When Nicholas Rowe came to edit the works in 1709, the first editor in the modern sense, he printed from the late folio line, which is to say from the end of a chain whose quarto links he did not consult. The play the eighteenth century read was therefore the play of the annotated 1609 reprint at several further removes, its Folio dress further tailored by folio compositors and by Rowe’s own regularizing hand. The scholarly recovery of the second quarto in the nineteenth century, when the Cambridge editors collated the quartos and named the 1599 book “unquestionably our best authority,” was precisely the walking back of that chain, link by link, past the folios, past the annotated copy, past the 1609 reprint, to the book printed from the author’s papers. The reprint history is thus not a detour from the main story of the text. It is the map of the territory the main story crossed, twice, in opposite directions.

There is a final symmetry worth drawing. The quarto line ends in 1637 because its world ended: the theatres closed in 1642, the trade that fed on new plays and steady reprints was broken by war, and when playing resumed at the Restoration the play returned in altered forms, through adaptations that belong to another history. The 1637 quarto is therefore a terminus in fact as well as in collation, the last state of the text that the first age of the play produced. What survived the interruption was the chain itself, the physical books in libraries, from which the nineteenth century could reconstruct the descent and the twentieth could build its editions. The reprints were the ark as well as the road: had the trade’s copies perished entire, the Folio line alone would have carried the play, and the modern text, which rests on a quarto the reprints frame and explain, would stand on poorer evidence than it does.

What the Reprint Trade Shows About the Play’s Market

Count the editions and a rhythm emerges: 1597, 1599, 1609, the early 1620s, 1637. After the first correction, the play settled into a reprint every ten to fourteen years, a pace sustained for four decades under two publishers, Burby for the first two quartos and Smethwick for the rest. Stationers did not reprint from sentiment. A reprint was ordered when stock ran low and demand was expected to continue, and each new edition of this tragedy represents a fresh investment of paper, press time, and shop space in a title already a generation old. Most plays of the period were printed once and never again; the stock of unsold or unrepeatable titles is the silent majority of the drama catalogue. A play reprinted four times before the Civil War, and carried into the period’s great collected edition besides, was a property of unusual durability, and the reprint sequence is the documentary proof of the durability that the play’s stage reputation merely asserts.

The ownership pattern sharpens the point. The tragedy was valuable enough to be traded twice in a single year, 1607, passing from Burby to Ling to Smethwick inside eleven months, and Smethwick, the most deliberate accumulator of Shakespeare properties in the trade, then held it for the remaining thirty years of the quarto line, reissuing it three times and never letting the right lapse. The Folio syndicate, assembling the collected works in 1622 and 1623, had to work around such holdings: Smethwick was not of the syndicate, and the Folio’s use of his 1609 edition as copy is the visible trace of the negotiation between the old quarto trade and the new collected venture, the syndicate printing from the book the rights had made standard. When the fourth quarto appeared from Smethwick’s shop in the same season, the two publishing models stood side by side in the same street: the cheap single play, sold at the Dial in Saint Dunstan’s Churchyard to the readers who had always bought it, and the great folio, sold by subscription to gentlemen. The play was large enough to fill both markets at once.

Demand of that steadiness leaves other fossils. Harington’s lists show the quartos on a courtier’s shelves; the commonplace book of the early 1620s shows a reader copying the play’s lines for private use, the extract trade that print both served and stimulated; the stage kept the play in repertory, as the title pages’ repeated claim of performance at the Globe attests in the trade’s own terms. Reading, watching, and collecting fed one another, and the reprints are where the three meet in evidence. The reprint history thus does double duty for the play’s biography: it is the record of the text’s descent, and it is the record of the audience that made the descent worth a stationer’s money. A criticism that knows the play only as a script for the stage, or only as a poem for the study, misses the third thing it demonstrably was: a book that sold, in the same small format, to two generations of buyers who never saw a performance.

The economics beneath the rhythm can be sketched from the trade’s general shape, which Peter Blayney’s work on playbook publication has done more than any other to clarify. A play quarto was a small speculation: an edition of some hundreds of copies, commonly estimated between five hundred and a thousand for a title of this kind, printed on a few dozen sheets of paper per copy and sold unbound for pence. The publisher’s profit lay in selling the run through before the title cooled, and his risk was bounded by the modest scale. Against that shape, the tragedy’s record is eloquent. Five quarto editions and a folio printing in forty years mean the speculation was repeated six times by professionals who could count, and the intervals between editions, a decade and more, mean each run sold at the pace of a steady backlist title rather than a sensation. Sensations burn faster and are not reprinted at decade intervals for forty years. The play was not a craze. It was a staple, and staples are what publishing houses are built on.

Smethwick’s handling of the title fits the staple pattern exactly: no experiments with format, no enlargement, no illustration, the same small book at the same shop address under the same device, refreshed as stock required. Even the title page’s one novelty in forty years, the author’s name appearing in the fourth quarto’s variant state, is a staple-keeper’s novelty, a marginal adjustment to the packaging of a known product, tried in some copies and standardized by 1637 when its value was clear. Read as trade documents, the reprints describe a mature market in classic drama a generation before the trade had a word for classics: the play had become the kind of property a stationer bequeathed, as Smethwick’s long hold and the transfers of 1607 show it being treated, like a rent-paying asset in the list.

Why did the same publisher issue the play four times?

Strictly, Smethwick issued it three times after buying the right in 1607, and the reason is the arithmetic of the backlist. Each reprint cost him paper and presswork against a demand the earlier editions had already proved, the safest merchandise a stationer owned. He reprinted when the stock ran out.

The copy had been bought as part of a portfolio of proven titles, and a known play by a known company, requiring no new copy and no advertising beyond its title page, carried its risk already retired. The intervals between his editions, 1609, the early 1620s, and 1637, are simply the measure of how long each printing took to sell.

The Surviving Copies: What the Census Shows

The books themselves survive in the small numbers typical of cheap print, and the census of their survival is the last body of evidence the reprint history offers. Greg’s bibliography recorded the copies known to it library by library, and the great collections, the Folger Shakespeare Library above all, the British Library, the Bodleian, and the Huntington, hold the bulk of what remains. Of the fourth quarto’s anonymous title page, the 1916 exhibition catalogue knew three copies; the British Museum’s, now the British Library’s, is the copy pressmarked C. 34. k. 56 whose conjectural dating to 1607 launched the ghost edition. Scarcity of that order is not evidence of a small original edition so much as testimony to a hard-used one: play quartos were bought to be read, carried, lent, and read again, and a book can be a commercial success and a bibliographical rarity by the same biography, its copies consumed by the use that proved its popularity. The paradox runs through the whole chain, whose most successful books are among its rarest survivors.

Survival is also biased in ways the historian must correct for. Copies endured where institutions or wealthy collectors preserved them, and the copies that survive are disproportionately the copies that were bound early, owned gently, and kept dry: the Folger’s shelves represent the play’s most fortunate readers, not its typical ones. The annotated exemplar of the third quarto, the hardest-worked copy of the play in the century, does not survive at all, because working books are used up. The census therefore shows the reprint history in a distorting mirror, its monuments preserved and its tools discarded, and the argument of this study has been built with that distortion in view: the books we can inspect are the chain’s relics, and the books that did the chain’s work are known by their effects on other books. That is the normal condition of print history, and the reprints of this play, scarce, variant, and eloquent in their effects, are its textbook case.

Q3 and Q4 by Their Labels: How Bibliographers Name the Reprints

Even the names of these books are modern impositions, and knowing it disciplines the argument. No early reader called anything Q3: the sigla are the coinage of the bibliographical tradition, systematized when Pollard and Greg ordered the Shakespeare quartos into numbered sequences, and the numbers smuggle a tidiness the trade never felt. “Third quarto” suggests a series planned as a series; the reality was a publisher reprinting when stock ran out, in a trade where the previous edition was copy and the next was not contemplated. The labels are indispensable and misleading in the same breath, indispensable because collation needs short names for its witnesses, misleading because they make a genealogy look like a publishing programme. This study has used the sigla throughout, as all scholarship does, but the narrative beneath them should be remembered in its untidy form: not a sequence designed, but a property maintained.

The labels also fix the questions editors ask. To call the 1609 book Q3 is to define it by its distance from the manuscript’s line, third in descent, and to predetermine its treatment as a witness whose agreements are cheap. To call the 1599 book the “good” quarto and the 1597 book the “bad” is to import Pollard’s courtroom into the printing house. The reprints suffer most from the naming habit, because a book called a reprint is a book excused from attention, its interest presumed exhausted by its copy. The record refutes the presumption at every point this study has examined: the copy for the age’s greatest book was a reprint, the play’s author was first named by a reprint’s variant, the chain’s only act of collation before the editors was a reprint’s, and the property history of the play is a reprint history entire. Bibliographers will go on saying Q3 and Q4, and should; the correction needed is not in the labels but in the reflex that hears them as a dismissal.

The Chain in Scholarship: Collation from Mommsen to the Modern Apparatus

The chain this study describes was not always known, and the manner of its discovery is part of its meaning. The nineteenth century began with the Folio line enthroned and the quartos imperfectly distinguished; the undated book wandered the catalogues misdated, and the third quarto’s role as the Folio’s copy was a suspicion before it was a proof. The proof came from collation, and its pioneer for this play was Tycho Mommsen, whose 1859 edition printed the first and second quartos on facing pages with a collation of the remaining quartos and the folios in its notes: laborious, incomplete for the fourth and fifth quartos by the Cambridge editors’ later account, and foundational, because it put the witnesses side by side where their agreements could be counted. The Cambridge Shakespeare of Clark and Wright then drew the conclusions the collation supported, identifying each book’s copy from internal evidence and stating the chain in the sentences this study has quoted. Collation is slow virtue, and the reprint history is its monument: every link in the table below rests on some scholar’s counted agreements in error.

The twentieth century professionalized the chain. W. W. Greg’s bibliography of the English printed drama catalogued the quartos with their surviving copies and gave them the STC numbers by which they are still cited, the third quarto as 22324 and the fourth as 22325a, the lettered suffix itself a small fossil of the bibliographical trouble the undated book caused. Greg’s Folio studies fixed the Folio’s copy in the third quarto’s line, and Hinman’s compositor analysis showed the Folio’s text in the act of being made from its annotated exemplar. Later bibliographers, Hailey with the undated book’s types, refined the dates. The modern critical editions inherit the whole structure: Levenson’s Oxford text and Gibbons’s Arden text print their genealogies in their introductions as settled matter, and the Folger edition’s introduction compresses the chain into a paragraph for general readers, reprint, annotated copy, and all. What was once a research frontier is now the frame every student meets, which is the usual fate of bibliography that succeeds.

One gap in the record should be marked, because this study has had to work around it and honesty requires the marking. The reprints have never had an editor of their own in the way the first quarto has, no modern edition presents the third or fourth quarto entire with a full introduction, and the scholarship on them lives inside other books’ introductions and articles: the Folio’s bibliography, the second quarto’s editions, the dating literature. Their variants are in the apparatus of the standard editions, their title pages in the catalogues, their dates in the bibliographical journals, but the books themselves, as books, wait for their historian. The present study is an attempt to state what that historian would find: that the least regarded volumes in the play’s transmission are the ones through which the transmission actually ran.

The InsightCrunch Reprint Chain

The whole argument of this study stands in one table. The InsightCrunch reprint chain sets the editions in their order of descent, with the copy each was printed from and the role each plays in the transmission, so that a reader can see at a glance why the reprints, not the famous quartos alone, are the road the text travelled.

Edition	Year	Printed from	Printer and publisher	Role in the transmission
Second quarto (Q2)	1599	The author’s working papers, with one stretch set from Q1	Thomas Creede for Cuthbert Burby	The authority: nearest printed heir of the manuscript
Third quarto (Q3)	1609	A printed copy of Q2	John Windet for John Smethwick	The bridge: its annotated copy became the Folio’s copy
Fourth quarto (Q4)	Undated; 1622 to 1623 on the evidence	A printed copy of Q3, corrected in places from Q1	Printer unnamed; for John Smethwick	The parallel line: first quarto page to name the author, in some copies
First Folio (F1)	1623	An annotated copy of Q3	Isaac Jaggard’s shop for the Folio syndicate	The collected text: a reprint of Q3 with an annotator’s marks
Fifth quarto (Q5)	1637	A printed copy of Q4	R. Young for John Smethwick	The terminus: last quarto of the play before the theatres closed

Read down the third column and the chain’s law appears: after 1599, no edition of the play was printed from a manuscript, and every edition was printed from the edition before it, with the single branching at 1609, where the third quarto feeds both the Folio and the fourth quarto. Read down the last column and the reprints’ dignity appears with it: the bridge, the name-giver, and the terminus are all reprints, and the two most famous books in the story, the second quarto and the Folio, are the chain’s ends rather than its substance. Teachers who lay the transmission before a class find the chain easier to hold as a single visual argument than as a narrative; slides assembled through a presentation tool such as the ReportMedic PPT Viewer let the five rows stand side by side while the class traces the descent with its eye, which is the form in which the evidence is most persuasive.

What the Reprint Record Cannot Show

A study built on reprints should end its evidence by marking the record’s edges, and the edges are real. No sales figures survive for any edition of the play: the reprint rhythm proves demand in the way footprints prove a walker, by sequence and recurrence, and the size of any edition is an estimate from the trade’s general practice, not a counted fact about this title. No document names the Folio copy’s annotator, and none records the fourth quarto’s year, its printer, or the order of its title pages; each of those silences has been filled by inference of graded strength, Hailey’s typographical dating at the strong end, the title pages’ sequence at the weak end, where this study has declined to pretend a verdict the evidence will not carry. And no copy survives of the annotated exemplar itself, the single artifact that would convert the Folio’s preparation from inference to inspection.

The limits cut in the study’s favour as much as against it. A record this lacunose would not support the claims usually made for the famous quartos if the reprints were not there to discipline them: it is the chain’s completeness at the joints, each book’s copy identified in the previous book by shared error, that allows the strong conclusions to stand where they stand, on the Folio’s descent, the drift’s mechanism, and the property’s history. Bibliographical knowledge is mostly knowledge of exactly this kind, a skeleton of proved descents fleshed with graded inference, and the reprint history of this play is an unusually complete skeleton. The honest marking of what it cannot show is what entitles the rest to be believed.

The Verdict on the Reprint History

What, finally, does the reprint history contribute, beyond completing a bibliographer’s diagram? Four things, in ascending order. It identifies the Folio’s copy, and thereby converts the Folio from an authority into a measurable witness, whose every departure from a surviving exemplar can be listed and priced; three centuries of editorial confusion about the Folio’s weight dissolve once the 1609 book is named as its source. It demonstrates the mechanism of textual drift in the wild, with named shops, dated books, and a checkable chain, so that the abstract warning that texts corrupt in transmission becomes a documented rate: two reprints between 1599 and 1623, each adding errors faster than its corrections removed them, each freezing the previous shop’s economies into the received text. It restores the play’s commercial biography, the rights traded in 1607, the portfolio logic of Smethwick’s list, the decade rhythm of reissue, without which the play’s fame is a reputation in search of its evidence. And it corrects the small famous error about the author’s name, which entered this play’s books not with the third quarto’s modern dress but in a variant state of the undated fourth, in some copies only, as a trade experiment at the end of the quarto era.

The critics whose work this study has followed disagree at the expected joints, and the adjudications can be gathered. On the undated quarto’s year, the older conjectures fail on descent, the catalogues’ bracketed 1622 stands as a fair approximation, and R. Carter Hailey’s typographical investigation earns its conclusion: 1623, the Folio’s year, with the two books printed in parallel from the same copy by different shops. On the order of the fourth quarto’s title pages, Halliwell-Phillipps’s named-first sequence and the 1916 catalogue’s anonymous-first judgment cancel out on the surviving evidence, and the honest verdict is that the order is unproved while the fact of variation is certain. On the Folio’s annotated copy, the verdict goes with W. W. Greg and the Folger editors on the fact of annotation, and against the stronger promptbook claim on its source: the marks are a craftsman’s, theatrical in flavour, bibliographically unplaceable, and the third quarto needed no manuscript behind its annotator to become the Folio’s copy. On the chain itself, from Clark and Wright’s collation to Brian Gibbons’s and Jill Levenson’s editions, there is no disagreement left to adjudicate: the reprints descend as the table shows, and the only error available is to ignore them.

For the general reader the consequences are concrete, and worth stating without apparatus. The words in a modern edition descend from the 1599 book because editors walked the chain back to its head; the pointing of the play’s most quoted question descends in part from an anonymous compositor in the undated quarto’s shop; the stage directions a company rehearses from descend in part from an anonymous annotator’s pen on a copy of the 1609 reprint. Three unnamed workmen of the reprint trade stand inside every modern experience of the play, and the reprint history is the only account of the text that can introduce them. A reader who knows the chain reads the play differently: not as a monument lowered intact from 1595, but as a living text that survived by being copied, and whose survival left marks that are themselves a kind of authorship, the trade’s.

The cliche knows two quartos and a folio, a bad book, a good book, and a monument. The record knows a working text in a working trade: corrected once, in 1599, and then carried for forty years by reprints that corrected a little, corrupted a little, sold steadily, and at the decisive moment supplied the copy for the monument itself. To study the third and fourth quartos is to watch literature being transmitted by the only machinery that has ever transmitted it, paper, property, and piecework, and to find that the machinery left its fingerprints on every line the later centuries read. The reprint history is not the footnotes to the tragedy’s textual story. It is the story, and the famous quartos are its first two chapters.

Frequently Asked Questions

Q: What is the InsightCrunch reprint chain, and what pattern does it expose?

The InsightCrunch reprint chain is the table in this study that sets the play’s editions in their order of descent: the second quarto of 1599, the third quarto of 1609 printed from it, the undated fourth quarto printed from the third, the First Folio of 1623 set from an annotated copy of the third, and the fifth quarto of 1637 printed from the fourth. The pattern it exposes is that after 1599 the text travelled from print to print, never again from a manuscript, so that every later witness is a reprint whose value is fixed by its place in the chain. Agreements between descendants prove nothing twice; only the head of the chain speaks with the manuscript’s accent.

Q: Who published the third and fourth quartos, and who printed the third?

John Smethwick published both, selling them from his shop in Saint Dunstan’s Churchyard in Fleet Street, under the Dial. The third quarto of 1609 was printed by John Windet, whose name appears nowhere on the title page but whose ornaments and type identify his shop’s work. The fourth quarto names no printer at all, one of several facts it withholds. Smethwick had acquired the right to the play in 1607 and kept it for the rest of the quarto era, issuing the third quarto in 1609, the fourth in the early 1620s, and the fifth in 1637, a thirty-year tenure that makes him the tragedy’s publisher for the whole of its early reprint history.

Q: What did John Smethwick buy in the rights transfers of 1607?

He bought the right to print the play, transferred through two entries in the Stationers’ Register in a single year: Cuthbert Burby, the second quarto’s publisher, transferred the tragedy to Nicholas Ling on 22 January 1607, and Ling transferred it to Smethwick on 19 November, in an entry that moved sixteen titles at once. What Smethwick acquired was not a manuscript and not the acting company’s book, but the trade right in the printed text, together with the market the 1599 edition had made. The purchase explains the third quarto’s nature: a publisher who owns a text reprints it, and Smethwick’s printer worked from the printed second quarto because that was the asset.

Q: How can scholars tell that the third quarto was printed from the second and not from a manuscript?

The proof is the pattern of shared accident. The 1609 book follows the 1599 book into its distinctive errors, echoes its idiosyncratic spellings, and reproduces its dislocations, including the unsettled state of the rose passage in the balcony scene, entire. A text set from a manuscript would differ from the second quarto independently, sometimes agreeing with it against sense and sometimes departing with authority; the third quarto instead departs only as resetting departs, by misreading, modernizing, and small correction. The Cambridge editors stated the conclusion from collation: the 1609 book was printed from the second quarto. Bibliographers call such proof descent by shared error, and it is the same method that fixes every link in the chain.

Q: What errors did the third quarto add to the text it reprinted?

The third quarto added the ordinary tax of a resetting: misread letters and words, small substitutions of one function word for another, spellings modernized in one place and left archaic in the next, and pointing adjusted to a compositor’s sense rather than the copy’s. The Cambridge editors summarized the balance as a few corrections against more frequent additional errors. No single added error is famous, which is itself the point: reprint corruption is anonymous and cumulative, a scatter of small faults rather than a catastrophe, and its danger lies in invisibility. Each fault, once printed, looked like the text, and the fourth quarto and the Folio’s copy inherited the scatter along with the words.

Q: Why do editors call the third quarto the bridge to the First Folio?

Because the Folio’s printers worked from it. When Isaac Jaggard’s shop set the tragedy for the 1623 collection, its copy was an exemplar of the 1609 reprint, annotated before use, and the Folio text follows the third quarto through errors the second quarto does not contain. Every Folio reading therefore reaches the second quarto only across the 1609 book, and any editor weighing the Folio must first subtract the bridge: the third quarto’s drift, the annotator’s marks, and the Folio compositors’ own layer. The metaphor is exact in one further respect. Traffic crossed the bridge in one direction only; nothing from the Folio flowed back into the quarto line, which continued from the third quarto to the fourth and fifth on its own side of the river.

Q: What marks in the Folio text show that its copy of the third quarto was annotated?

The Folio departs from the 1609 text at points where typesetters do not improvise: stage directions stand in the Folio that no quarto prints, fuller and more theatrical in character, and readings shift in ways that suggest preparation of the copy rather than accident at the press. Barbara Mowat and Paul Werstine, introducing the Folger Shakespeare, draw the inference directly: the Folio’s printer’s copy “must have been annotated by someone,” because its departures exceed what mere typesetters do. W. W. Greg’s bibliographical history of the Folio supports the annotated exemplar. What the marks do not show is the annotator’s source; a playhouse manuscript has been proposed, but theatrical knowledge and collation against print explain the marks without it.

Q: What does the fourth quarto’s title page omit besides a date?

It omits the printer’s name, and in one of its two forms it omits the author’s name as well. The page gives the title, the acting company at the Globe, the inherited claim “Newly corrected, augmented, and amended,” Smethwick’s device, and the imprint “Printed for Iohn Smethwicke,” and there it stops: no year, no printer, no shop address beyond what the imprint implies. The silences are why the book has to be dated by its types and placed by its errors rather than read off its own face. Early modern title pages are advertisements, not title deeds, and the fourth quarto is the extreme case, a book whose bibliography must be reconstructed entirely from evidence it never intended to give.

Q: What dates have scholars proposed for the fourth quarto, and which is best supported?

The proposals span the whole window the book’s descent allows. The British Museum’s catalogue once conjectured 1607, and Charles Knight repeated a quarto of that year, but collation kills the guess: the undated book was printed from the quarto of 1609, so it cannot precede it. Dates clustering just after 1609 rest on assumption rather than evidence. The catalogues long carried a bracketed 1622, and the one dedicated typographical investigation, R. Carter Hailey’s, fixed the printing in 1623. The best-supported answer is therefore the Folio years, 1622 to 1623, with Hailey’s 1623 the most precisely argued single year: the fourth quarto is the First Folio’s contemporary, not its predecessor by a decade.

Q: Who dated the fourth quarto to 1623, and on what kind of evidence?

R. Carter Hailey dated it, on typographical evidence: the analysis of the book’s types, ornaments, and the progressive damage that individual pieces of type accumulate as a printing house uses them. An undated book can be placed among dated books from the same milieu by matching the state of its materials, a letter’s nick or an ornament’s crack appearing in one condition in a book of one year and in a worsened condition in a book of the next. Hailey’s work on the undated Shakespeare quartos applied that method systematically, and for this book produced 1623. The Folger Shakespeare’s editors accept and print the conclusion, which is why the date now appears in modern editions’ textual introductions.

Q: How do the fourth quarto’s two title pages differ?

They differ in a single inserted line. In one form the page is anonymous, running from the Globe line straight to the inherited claim of correction, exactly as the title pages of the play had always run. In the other form the line “Written by W. Shakespeare” stands after the Globe line, before the claim. Everything else, title, company, device, imprint, is the same, and the catalogue of the 1916 New York Shakespeare exhibition states the position exactly: copies of the edition differ in title only. The two forms are states of one edition, not two editions, and a reader who has seen only one form has seen only half of the book’s bibliography.

Q: Which of the fourth quarto’s title pages came first, the one naming the author or the anonymous one?

The order is unproved, and the authorities divide. J. O. Halliwell-Phillipps held that the named form was issued first and the name then omitted in copies subsequently published, a sequence Edward Dowden’s edition reports on his authority. The 1916 exhibition catalogue judged the anonymous form probably the earlier, presenting it as the edition’s earlier title. Survival cannot settle the question: the anonymous form was known in three copies in 1916, and counts that small record the accidents of collecting more than the order of issue. The honest answer is that either sequence is possible, and that the variation itself, the name treated as an optional line, is the fact of lasting interest.

Q: When did the author’s name first appear on a title page of the play?

In the undated fourth quarto, in the copies that carry its named title page, where “Written by W. Shakespeare” stands after the Globe line. The first, second, and third quartos are all anonymous on their title pages, the 1609 book included: for all its modern dress, its updated company, its Globe address, the third quarto still sold the play without naming its author. The name’s arrival therefore belongs to the early 1620s, at the end of the quarto line’s anonymity, and it arrived unstably, present in some copies of the edition and absent in others. The fifth quarto of 1637 prints the name as a matter of course, completing a conquest the fourth quarto’s variant state began.

Q: How did the fourth quarto’s printer improve on the third quarto’s text?

By going behind it. The fourth quarto descends from the third and inherits its errors, but its printer also consulted the first quarto of 1597 and took corrections and word choices from it, as the Folger Shakespeare’s editors note. The old Cambridge judgment that the undated book’s text is more correct than the text of 1609 registers the result: a reprint whose makers compared their copy against an older witness and chose the better reading at points. The improvements are local, word-level mendings rather than restored passages, and the first quarto was consulted, never adopted as copy. Even so, the consultation is the printing house practising collation on its own initiative, a generation before the Folio and long before the method had a name.

Q: What happened to the play’s text in the fifth quarto of 1637?

It completed the chain. The fifth quarto was printed by R. Young for John Smethwick, the tragedy’s publisher for thirty years, and it was set from a printed copy of the fourth quarto, inheriting the accumulated drift of 1599 through 1609 and the early 1620s beneath the author’s name, which its title page now prints as standard. No manuscript and no new authority entered at this last link; the 1637 book is the 1599 text after three reprints, recognizably itself and measurably farther from its source. It was the last quarto of the play printed before the theatres closed in 1642, and after it the text’s future belonged to the Folio line, the collected editions through which the next three centuries read the play.

Q: How often was the play reprinted between 1599 and 1637, and what does the rhythm imply about demand?

Three times: 1609, the early 1620s, and 1637, following the correction of 1599, a reprint roughly every ten to fourteen years across four decades. The rhythm implies the steadiest kind of demand. A stationer reprinted when stock ran low and sales were expected to continue, so each edition certifies that the previous one had sold through and that the publisher judged the market unexhausted. Most plays of the period were printed once and never reprinted; a title renewed at decade intervals under two publishers, traded twice in one year as an asset, and collected into the Folio besides, was a backlist property of the first rank. The reprint rhythm is the hardest quantitative evidence of the play’s early literary standing that survives.

Q: Why did the Folio’s printers use the third quarto rather than the newer fourth?

Because the third quarto was the copy in hand, prepared for the purpose. The Folio’s exemplar of 1609 had been annotated before it reached Jaggard’s shop, and copy preparation, not novelty, governs a printing house’s choice. Chronology also presses: on Hailey’s dating the fourth quarto appeared in 1623 itself, and on the catalogues’ dating in 1622, in either case too close to the Folio’s production to displace an exemplar already marked and in use, and the tragedy stood in the Folio’s tragedies section, set while the volume was still in the press. A newer reprint offered the syndicate nothing it needed: the same text at one more remove, unannotated, owned by a stationer outside the syndicate, whose rights the Folio’s publishers had instead accommodated by using his earlier edition.

Q: Can a reading found only in the third or fourth quarto ever be what Shakespeare wrote?

Almost never as evidence, and never by the reading’s own authority. A reprint’s unique reading is overwhelmingly likely to be its printing house’s work: a misreading, a modernization, or a compositor’s guess at nonsense. Two narrow exceptions keep the answer honest. A reprint’s correction can hit the truth by intelligence, as when the fourth quarto’s printer mended the third quarto from the first; the reading is then right because the first quarto carried it, not because the reprint vouches for it. And the annotated copy behind the Folio might, in principle, transmit a reading from a source now lost; in practice no Folio reading of this play requires that explanation. Editors therefore cite the reprints’ variants in the apparatus and print them only when an older witness stands behind the agreement.

Q: Where can the third and fourth quartos be consulted?

In the research libraries that hold the surviving copies, and in digital facsimile for any reader with a screen. The Folger Shakespeare Library holds copies of both books and presents the third quarto in its digital collections; the British Library’s copy of the fourth quarto, pressmark C. 34. k. 56, is the anonymous copy its catalogue once dated to 1607. Early English Books Online carries the quartos as page images for subscribing institutions, and modern editions print the reprints’ variants in their apparatus even where the books themselves are not reprinted entire. The chain that this study describes can thus be checked link by link against the originals, which is the proper fate of a bibliographical argument.

Q: How should an essay use the InsightCrunch reprint chain to argue that the received text is a transmitted text?

Begin from the third column of the chain, the copy each edition was printed from, and show that after 1599 no edition of the play touched a manuscript. Then take one passage with a documented journey, the rose sentence passing unchanged through 1609 and acquiring its modern pointing only in the undated quarto, and price each link’s contribution to the form in which readers now quote it. Close with the Folio: because its copy was an annotated third quarto, the most authoritative-looking book of the seventeenth century is a reprint of a reprint, and every claim for its readings must be discounted by the chain it stands at the end of. The essay’s thesis writes itself: the received text is not a document but a descent.