The first quarto of Romeo and Juliet, printed in 1597 and dismissed for generations as a bad quarto, has been revalued by a newer account of its origins: the performance text view, which holds that this short, theatrically vivid book records the tragedy as the playwright’s company actually played it, cut for the stage and carrying the marks of performance in its directions, its casting economy, and its shape. The verdict argued here is graded, because the claim itself is graded. In its strong form, that the 1597 volume transcribes a performance, the claim is unproven and probably unprovable. In its moderate form, that a performed, abridged version of the tragedy stands behind the printed book, the claim is probable, better supported than any rival single explanation, and strong enough to change how the volume should be used. In its weakest form, that the book is worth reading as evidence of early staging, the claim is beyond serious dispute. Everything in the pages that follow works toward those three grades: what the performance text claim would certify if true, the evidence tried in its favour, the evidence tried against it, the counter-case built by cautious textual critics, and the practical consequences for the three kinds of reader who handle the early texts, the student of the words, the editor who must choose between them, and the director who must put bodies on a stage.

The starting point is the book itself, and its portrait belongs to a companion study of the first quarto of 1597 and the ‘bad’ quarto label it has carried ever since: a slim volume of some 2,230 lines against the second quarto’s roughly 3,040, printed without the author’s name, its title page advertising the tragedy as often played publicly by the Lord Hunsdon’s servants. The single inference examined here begins where that portrait ends. Granted the short text exists, granted its directions are unusually alive and its cuts unusually purposeful, what follows? The older scholarship answered with a story of theft: actors or a shorthand writer carried the play out of the playhouse in corrupted form, and the printer John Danter set the stolen goods in type. The newer scholarship answers with a story of use: the book descends from the play in performance, and its defects are the scars of the stage rather than the crimes of a pirate. Between those two stories lies a genuine evidentiary problem, because a book cannot testify about its own ancestry without cross-examination. Theatrical colour can be manufactured in a print shop as well as inherited from a playhouse. Cuts can be made by a compositor short of paper as well as by a company short of time. The task is to find the features that discriminate, the details a print shop would have had no reason to invent and a pirate no ability to supply, and to weigh the features that point the other way with equal honesty.

What the Performance Text Claim Would Actually Certify

Precision about the claim comes first, because much of the quarrel has been conducted between scholars defending different propositions under one label. To call the 1597 quarto a performance text can mean at least three things. It can mean that the printed words record a specific performance, that somewhere behind the book stands a single occasion on which actors spoke and someone, by shorthand, by memory, or by a transcript made in the theatre, captured what was said and done. It can mean that the copy behind the book was prepared for performance, that the words descend from a playhouse document, a promptbook or an abridged set of parts, whether or not any particular night’s playing stands behind them. Or it can mean only that the text as printed reflects the play in its performed shape, shortened and arranged as playing required, whatever the route by which that shape reached the printer. The three meanings carry very different evidentiary weights. The first is a claim about an event. The second is a claim about a document. The third is a claim about a shape. A sceptic can grant the third, doubt the second, and deny the first without contradiction, and much of the best recent work on the short quartos occupies exactly that layered position.

If Q1 is a performance text, what exactly has been proved?

Nothing about a single night, and a great deal about a version. The defensible core of the claim is that the 1597 text embodies the tragedy as a company performed it, abridged to playing length and marked by staging practice, so that its evidence concerns the performed play. Proof of one event would require a witness chain no document supplies.

The distinction matters because the evidence available is almost entirely internal. No playhouse manuscript of the tragedy survives. No actor’s part, no plot scenario, no promptbook annotation connects the printed page of 1597 to a stage. What survives is the book’s own behaviour: its length, its omissions, its directions, its casting demands, its characteristic errors. Each feature must be asked the same discriminating question. Is this feature more likely given performance descent than given the alternatives, a compositor abridging at the case, a reporter reconstructing from memory, a shorthand writer taking dictation in the yard, an early authorial draft reaching print by a side door? Features that every hypothesis predicts equally well prove nothing. Features that one hypothesis predicts and the others strain to explain are where the case is made or lost. The performance text case, properly built, rests on a small number of such discriminating features, and its opponents have done the subject a service by burning away the rest. Title page boasts are not evidence, since the formula of public performance was advertising copy across the trade. Shortness alone is not evidence, since printers abridged for paper and reporters forgot for human reasons. What remains after the burning is narrower and stronger: the pattern of the cuts, the eyewitness character of particular directions, the naming of a real actor at a real entrance, and the fit between the text’s demands and a small company’s resources.

A further precision concerns the word record. A record implies capture, a mechanism that fixes an event as it happens. The Elizabethan theatre had no such mechanism in routine use, and the candidate mechanisms each leave different fingerprints. Shorthand, of the kind Timothy Bright patented in 1588 and John Willis described in 1602, could in principle fix words at speed, but a shorthand transcript predicts a particular error profile, phonetic confusion and collapsed connectives, and struggles to explain how stage business visible only to the eye entered the text in descriptive form. Memory, the mechanism at the heart of the memorial reconstruction debate, predicts a different profile again: fidelity near the reporter’s own lines and scenes, paraphrase and transposition elsewhere. A playhouse manuscript predicts a third: clean copy for what it contains, directions in the imperative house style, and no corruption of the verse it preserves. The 1597 book shows traces that have been claimed for all three profiles, which is why the verdict must be graded rather than absolute, and why the argument in what follows keeps the questions of shape, document, and event separate even when the evidence for them arrives in the same paragraph.

A Shorter Play: Length, Cuts, and the Logic of the Stage

The brute fact is length. By the usual counts the second quarto of 1599 runs to roughly 3,040 lines and the first quarto to roughly 2,230, a difference of some 800 lines, about a quarter of the longer text. Line counts shift a little between editions and methods, but the proportion is stable and the proportion is what argues. A quarter of a play does not fall out by accident in transmission of the kind that corrupts individual words and phrases. Memory fails locally, a speech paraphrased here, a half-line dropped there; it does not reliably remove exactly the detachable material and reliably preserve the load-bearing scenes. The omissions in the short text are patterned. What goes, overwhelmingly, is expansion: lyric elaboration, rhetorical set pieces, repeated wooing, prolonged lament. What stays, almost without exception, is the plot spine, the sequence of encounters and decisions that carries the story from the street brawl to the tomb. A text damaged at random would show random damage. A text shortened for playing shows exactly this profile, because playing imposes a budget that reading does not: the budget announced in the Prologue itself, “the two hours’ traffic of our stage” (Prologue, line 12).

Why do the cuts in Q1 look like the work of a playing company?

Because they fall where playing feels cost and keep what playing cannot spare. Lyric expansion, repeated lament, and rhetorical display contract sharply, while entrances, exits, fights, letters, potions, and deaths survive nearly intact. Random corruption does not sort a text by stage usefulness. Abridgement for performance sorts it in precisely that way, scene by scene.

The playing-time arithmetic is the spine of the modern case, stated most influentially by Lukas Erne in Shakespeare as Literary Dramatist (2003). At plausible rates of early modern delivery, the longer text runs well past three hours; estimates vary with the assumed pace, but no realistic pace brings 3,040 lines inside two hours. The shorter text, at the same rates, lands close to the Prologue’s advertised span, a little over two hours with music, fights, and the jig of stage business that line counts do not measure. Erne’s inference was that the long texts of the period were written for readers as much as for spectators, and that performance meant abridgement as a matter of routine: the company bought a big play and played a smaller one. On that account the short quarto is not an aberration requiring a criminal explanation. It is the normal fate of a long play, caught in print. The inference has a corollary that cuts the other way against the old piracy story. A pirate steals what has value, and what had value in 1597 was the play the town was talking about, the performed phenomenon, applause and all. A transcript of the performed version, however obtained, is the likeliest commercial object; a transcript of an unperformed long draft is an odd thing for any thief to possess.

The cuts repay closer inspection at the level of the individual scene, because their intelligence is the argument. The second Chorus speech, the sonnet that opens the second act and carries no plot, disappears entire. Extended lamentation in the fourth act, where grief repeats itself in formal patterns, contracts hard. The lovers’ parting at dawn in Act 3 Scene 5 keeps its debate about the lark and the nightingale, the emotional hinge of the scene, and sheds lines around it. Juliet’s potion soliloquy in Act 4 Scene 3, a solo structure of mounting fears, survives in shortened form, its architecture intact and its decorations thinned. Again and again the principle is the same: the scene’s function is preserved, its duration is taxed. That is the signature of a cutter who knows what each scene is for, the knowledge of a company that has played the piece and learned where an audience’s attention is spent and where it is merely stored. A compositor shortening copy to fit a paper stock does not work scene by scene with a sense of function; the compositor works gathering by gathering, and the losses land where the paper runs out, not where the rhetoric swells.

Against this, the counter-case enters a fair objection that must be given its weight here rather than later. A skilled abridger need not be a company insider. London in the 1590s was full of professional writers who could cut a play intelligently for any purchaser, and a printer with a long manuscript and a short purse had motive to commission exactly such work. The pattern of the cuts proves that the cutting was intelligent and theatrically literate. It does not by itself prove whose hand held the pen, and the distance between “cut as a company would cut” and “cut by the company” is precisely the distance the graded verdict refuses to jump. What the pattern does exclude is the crudest version of the old story, in which the short text is the debris of incompetent theft. Debris does not sort itself by stage usefulness.

A refinement of the time argument closes the section, because the two hours of the Prologue are not made of lines alone. A performance spends time on what no line count measures: the fights, choreographed and rehearsed; the ball’s dancing; the music that covers scene changes and fills the tomb’s silences; the processions of the final tableau; and, by the custom of the period’s stage, the jig that followed even a tragedy and sent the audience home. A text cut to the advertised span had to leave room for all of it, which means the spoken text needed to run shorter than the full two hours, and the short quarto’s margin fits that requirement with room to spare while the long text bursts it before a single dance is counted. The refinement matters because it reverses the usual direction of sceptical attack. Defenders of an unabridged stage life for the long texts sometimes plead that delivery was faster, crowds thinner, business brisker in some unrecoverable Elizabethan manner. Each plea spends from the same account, and the music, the fights, and the dancing are not conjectural economies but documented fixtures of the afternoon. The long text’s three-plus hours stand before a note is played. On the stage the play actually inhabited, only one of the two books fits the afternoon, and fitness for the afternoon is the definition of a performance text with which this article began.

Directions That Watch the Play Happen

If the cuts supply the case’s skeleton, the stage directions supply its flesh. The 1597 quarto is famous among the early texts for directions that do not read like instructions copied from an author’s draft but like observations of action performed. An author’s foul papers typically direct the future: enter, exit, a noise within. A promptbook directs the present company in the same imperative economy. The short quarto, at its most distinctive, describes: action narrated in a vivid present, with a specificity about bodies and their relations that serves no compositor’s purpose and answers an eyewitness’s memory. The centerpiece examples have been quoted in every account of the text, and they earn their fame. In the street fight of Act 3 Scene 1, where the longer text contents itself with the fact of Mercutio’s wounding, the short text specifies the mechanics of the blow, Tybalt’s thrust delivered under Romeo’s intervening arm as Romeo steps between the fighters, the very arrangement of bodies that makes the death an accident of peacemaking rather than a fair hit. In the parting scene, the direction brings the lovers on together at the window, fixing the staging of the dawn separation in a single image. At the catastrophe, the direction for Juliet’s death narrates the act itself, the dagger, the fall, in a sequence that reads like something seen rather than something planned.

The InsightCrunch Q1 stage record, set out below, catalogues the features of the 1597 text on which the performance case principally rests, with the limit that attaches to each. A fuller treatment of stage directions in the early texts of the tragedy places these examples in the wider grammar of Elizabethan direction writing; the concern here is narrower, what these particular directions can and cannot certify about one book’s descent.

Feature of the 1597 text	What the book shows	What it suggests	The limit on the evidence
The window direction at the parting (Act 3 Scene 5)	Romeo and Juliet are brought on together at the window for the dawn scene	A staging observed and remembered, the lovers framed aloft for the aubade	A literary copyist could also imagine the window the scene requires
The fight mechanics (Act 3 Scene 1)	The direction specifies Tybalt’s thrust under Romeo’s arm as Romeo intervenes	Eyewitness knowledge of how the wound was staged	Descriptive directions also occur in some authorial and scribal copies
Juliet’s death direction (Act 5 Scene 3)	The stabbing and the fall are narrated as action seen	A reporter or adapter recording business, not merely dialogue	The action is dictated by the words spoken around it
An actor named at an entrance (Act 4 Scene 5)	Peter’s entrance names Will Kempe, the company’s clown	Playhouse knowledge inside the copy; the print knows who played the part	One name could travel with a cast list or a part, not only with a performance
The missing second Chorus	The Act 2 Chorus sonnet is absent entire	Cutting by function: pure exposition is expendable in playing	A printer short of copy space could cut the same detachable item
Compression of the Queen Mab speech (Act 1 Scene 4)	The catalogue of dreams is cut to roughly half its longer-text length	A set piece taxed for pace while its entrance and exit survive	An abridger working from paper could judge the speech equally detachable
Romeo’s Mantua dream and final speeches shortened	Reverie and valediction contract while decisions and actions stand	Duration taxed where the plot is stationary	Literary taste, not only playing time, dislikes stationary rhetoric
A smaller company requirement	The cast of named and functional roles doubles down to a reduced company	A version fitted to fewer performers, as touring or a lean staging required	Doubling estimates depend on assumptions about boys, hired men, and doubling taboos
The title page claim of public playing	The tragedy is advertised as often played with applause by the Lord Hunsdon’s servants	The book trades on a performance reputation	The formula is trade advertising and appears on books of many ancestries

Read as a set, the directions have a quality that is hard to fake wholesale: they are interested in relations between bodies. Where does Romeo stand when the thrust lands. Where are the lovers framed when the lark debate begins. How does the final tableau assemble itself. This is the interest of the playhouse, whose currency is disposition in space, rather than the interest of the study, whose currency is wording. Editors have felt the force of the point in practice for a century: editions based on the second quarto repeatedly adopt the short text’s directions, in brackets or without ceremony, because the directions work, because they solve staging problems the longer text leaves open. Brian Gibbons, in the Arden edition of 1980, set the pattern that later editors have largely kept, a second-quarto text informed at point after point by first-quarto staging intelligence. Practice of that kind is a quiet verdict, delivered by the scholars with most to lose from error: whatever the short text is, it behaves like a text that has been on a stage.

What do the most distinctive Q1 directions actually describe?

Relations between bodies at the moments that decide the story. The thrust that lands under an intervening arm, the lovers framed together at a window for their parting, the heroine’s fall narrated as seen action: each direction fixes disposition in space, the playhouse’s working knowledge, rather than the wording a reader’s copy would preserve.

The honest limit must be stated in the same breath. Descriptive directions are not unique to reported texts, and the early modern book trade knew how to dress a page theatrically. A direction can narrate action because the copyist saw it, because the copyist imagined it vividly, or because the copyist inferred it from the dialogue it accompanies, and the printed page does not label which. The direction for Juliet’s death, for instance, describes exactly what her final words announce; an intelligent compositor or a scribe preparing fair copy could supply it without leaving the print shop. The fight direction is harder to dissolve in that way, since the spoken text nowhere states that the thrust passes under Romeo’s arm; the detail is surplus to the dialogue, and surplus staging knowledge is the strongest currency the performance case trades in. But surplus is a matter of degree, and degree is what a graded verdict is for. The directions raise the probability of performance descent sharply. They do not, singly or together, close the question of the route, and the route is where the counter-case lives.

Named Actors, Doubled Roles, and a Company That Fits the Text

One direction in the book has carried more argumentative weight than any other single line of the text, and it is not a direction about the lovers at all. In the musicians’ scene that follows the discovery of Juliet’s apparent death, the entrance of the servant Peter is given in the 1597 print under the name of the man who played him: Will Kempe, the celebrated clown of the Chamberlain’s Men, the company’s star comic and the successor in that line to Richard Tarlton’s legend. Print shops do not invent such details, and authors drafting at a desk rarely write them, since a name in a direction is useless to a reader and specific to a company at a moment in its life. The name’s presence means that somewhere in the ancestry of the copy stood a document or a memory that knew the casting, playhouse knowledge in the strictest sense. The performance case does not need to claim more for the name than that. It needs only to observe that the fact is of the playhouse’s kind, not the study’s, and that every hypothesis about the book’s descent must find room for it. A reporter who had seen the play could carry the name away in memory. A set of actors’ parts could carry it. A promptbook could carry it. What strains is the hypothesis of clean literary copy, a manuscript of the author’s long version somehow shrinking in a print shop, because literary copy does not know who played Peter.

Why does one actor’s name in a direction matter so much?

Because casting knowledge is playhouse knowledge, and its presence inside printed copy requires a playhouse source somewhere in the ancestry. A desk manuscript has no reason to record that the company’s clown took the servant’s part; a performed version cannot help recording it. The name does not identify the route, but it fixes the territory the route must cross.

The casting argument extends beyond the single name. The short text’s demands on personnel are measurably lighter than the long text’s. Roles that the longer version keeps separate can be doubled more aggressively in the shorter, scenes are arranged so that the same performer can clear one part and take another, and the whole can be mounted by a company of a size that fits the reduced troupes companies sent on the road when plague closed the London houses or when a text travelled to provincial stages. Casting analysis of the kind associated with T. J. King’s work on the documents of performance has made a method of such counting, and its application to the two quartos runs one way: the short text fits a leaner company the way a key fits a smaller lock. The fit supports the moderate claim directly. A version prepared for playing by fewer hands is a version prepared for playing, and a company cutting its big literary play down to a touring shape would produce precisely the object the 1597 book appears to be, shorter, leaner, its directions warm with use.

Two cautions belong to this section, both entered by the sceptics and both just. Doubling estimates are models, not documents: they rest on assumptions about how many boys a company could field, which roles could not double for reasons of dignity or quick change, and how hired men were deployed, and different assumptions move the numbers. And a lean text need not imply a touring company; a London company protecting its two hours against a long script produces the same leanness at home. The cautions limit what the casting evidence identifies, the occasion, without dissolving what it establishes, the orientation. The text behaves like a thing made to be played by a company counting its people. That is the moderate claim again, and it is the claim the evidence keeps returning to from every direction: cuts, directions, casting, length. Convergence of independent features on one explanation is how historical arguments earn probability, and the performance case, disciplined to its moderate form, has that convergence.

The Rehabilitation: From Bad Quarto to Playing Version

The term bad quarto was Alfred W. Pollard’s coinage, launched in 1909, and for half a century it functioned as a verdict rather than a description: texts so labelled were presumed stolen, garbled, and beneath serious attention except as evidence in the detective story of their own corruption. W. W. Greg, the commanding bibliographer of the next generation, gave the presumption its classic mechanism in the theory of memorial reconstruction, actors retailing their play from memory to a printer for ready money. Within that settlement the 1597 Romeo and Juliet required no individual study; its badness was deducible from its class. Harry R. Hoppe’s monograph of 1948, the first full bibliographical examination of the book, complicated the deduction without overturning it, treating the text as derivative and surreptitious while documenting features the simple piracy story explained awkwardly. The settlement held because it was tidy. It explained shortness by theft, corruption by memory, and theatrical colour by the thief’s presence in the theatre, and it left editors free to build on the second quarto undisturbed.

The untidying began when scholars started asking what the old story actually rested on. Paul Werstine’s examination of a century of bad quarto narratives, published in 1998, showed that the memorial reconstruction hypothesis had been built by analogy and assertion rather than by documentary evidence: no record survives of any actor selling any play from memory, and the narratives Greg and his successors constructed for individual texts routinely contradicted the physical evidence of the books themselves. Laurie E. Maguire’s Shakespearean Suspect Texts, in 1996, applied the diagnostic tests rigorously across the suspect corpus and found the supposed symptoms of memorial reporting distributed in ways the theory could not predict, present where no reporting was alleged and absent where it was certain. Kathleen O. Irace, in Reforming the “Bad” Quartos (1994), pressed the positive alternative, that several of the short texts are best understood as adaptations, deliberate reworkings of plays for performance conditions, made by or for the companies that owned them. The ground had shifted before Erne built on it. What Erne’s Shakespeare as Literary Dramatist added in 2003 was the complementary half of the argument: if the long texts are literary in conception, written with readers in view and exceeding any plausible playing time, then short theatrical versions are not the scandal of the system but its ordinary product, and the first quarto of the tragedy becomes the type specimen of a play as played. The rehabilitation since has been cumulative rather than revolutionary. The short text is edited, taught, and staged from; its directions inform productions; its cuts are studied as evidence of early modern playing practice of the kind surveyed in accounts of Romeo and Juliet on the Elizabethan stage. A text once mined only for its corruptions is consulted for its knowledge.

What changed when the performance view displaced the piracy verdict?

The burden of proof moved. Under the old settlement, the short text had to justify every feature against a presumption of theft. Under the newer account the presumption reverses for the patterned features, the cuts, the directions, the cast economy, and corruption, not theatricality, is what requires a mechanism. The book stopped being a crime scene and became a document.

The rehabilitation has a risk that its friends should name. A revalued text tempts its advocates to overclaim, to move from performance-shaped to performance-recorded without noticing the step, and to treat every vivid direction as a photograph. The scholars who built the newer view at its best, Erne included, kept the discipline of grades: his argument concerns versions and their destinations, literary against theatrical, and licenses conclusions about the performed shape of the play rather than about any night at the Theatre or the Curtain. Where popular summary has run ahead of that discipline, the cautious critics have been right to pull it back, and the pullback defines the counter-case to which the argument now turns.

The Counter-Case: Reporters, Print Shops, and Corruption Abridgement Cannot Explain

The performance case explains the short text’s virtues. Its vices are another matter, and the counter-case begins by insisting that any theory of the book must explain both. The 1597 text is not merely shorter and more theatrical than the 1599 text; at line level it is frequently worse: verse flattened into prose rhythms, metre broken where the longer text scans, phrases paraphrased into near-equivalents, lines anticipated from later scenes or recalled from earlier ones, connective tissue replaced by approximation. A company abridging its own playbook cuts; it does not, in the ordinary course of business, garble the lines it decides to keep, because the kept lines are the ones its actors must speak that afternoon. Corruption of the retained text is the fingerprint the performance case cannot produce from its own premises. Something in the transmission degraded the wording while preserving, even enhancing, the theatrical shape. That combination, shape sound and wording damaged, is the central datum of the whole problem, and every adequate theory is a theory of that combination.

The cautious critics have built their case on exactly this ground. Maguire’s contribution was to test the traditional diagnostics rather than recite them: anticipation, recollection, and paraphrase, the textbook symptoms of memorial reporting, turn out on inspection to be neither necessary nor sufficient markers, appearing in texts with no plausible reporting history and failing to pattern by role or scene in the ways the classic theory predicted. The conclusion she drew was austere and has aged well: the mechanisms proposed for the suspect texts exceed the evidence, and certainty about any of them is unavailable. Werstine’s parallel demonstration, that the narratives of piracy and reconstruction were constructed without documentary foundation and maintained by repetition, cut in the same direction. Between them, the two sceptics did not prove the performance case. They cleared its ground, by showing that the incumbent explanation was a story told with confidence it had not earned. But their scepticism is even-handed, and it is the part of the counter-case that advocates of the new view quote least: the same evidential discipline that dissolves Greg’s certainty dissolves any certainty that the book records performance. Theatrical colour, the sceptics remind the enthusiasts, is exactly what a reporter brings back from a theatre; it is also what a competent hack adds to a manuscript to make it saleable; the observation “this reads like the stage” is compatible with nearly every route by which a play reached print in the 1590s.

Why can abridgement alone not explain the state of the kept lines?

Because cutting and garbling are different operations with different agents. Abridgement removes whole speeches and leaves the survivors word-perfect, since the survivors must be spoken. The short text’s survivors are paraphrased, transposed, and metrically broken. Some agent in the transmission worked from imperfect knowledge of the wording while commanding the shape, and abridgement has no such agent in its cast.

The print shop deserves its paragraph as a suspect, because early modern printing was not passive reproduction. Compositors set from copy they sometimes could not read, corrected in the press room by guess, and abridged when paper or copy ran short; publishers commissioned adaptations and modernisations; scribes prepared copy with a reader’s comfort in view. The 1597 volume’s own production was irregular, divided between printers, unentered in the Stationers’ Register, anonymous on its title page where the 1599 volume would advertise itself as newly corrected, augmented, and amended against it. A sceptic is entitled to ask how much of the book’s theatricality a theatrical milieu outside the playhouse could generate. Playwrights for hire moved between stage and page; the men who might adapt a manuscript for Danter knew the stage’s conventions as intimately as any actor, and a direction describing a fight’s mechanics could come from a professional imagination stocked by attendance rather than participation. None of this is provable either, and that is the point. The counter-case at its strongest does not install a rival certainty. It establishes underdetermination: the printed features underdetermine the route, and any verdict stronger than the evidence is advocacy dressed as bibliography.

One more alternative must be kept in view, because it explains part of the evidence better than either rival. The short text may descend in places from an earlier state of the play itself, a draft the author later expanded into the long version, so that some of its differences are not losses but survivals, the play before revision rather than the play after cutting. The hypothesis has attracted scholars impressed by passages where the shorter reading seems prior, rough in the way of first thoughts rather than corrupt in the way of bad copies. It cannot explain the actor’s name, which belongs to a cast and not a draft, nor the pattern of theatrical cuts, which falls on a performed shape. But as a partial account of particular passages it remains alive, and its existence is a further reason the verdict comes in grades: the book is a confluence, and confluences do not have single sources.

Two Questions, Not One: Performance Origin and Memorial Origin

The brief for this article posed the complication directly: are performance origin and memorial origin mutually exclusive? They are not, and seeing why they are not dissolves much of the quarrel’s false geometry. The two phrases answer different questions. Performance origin concerns what the text reflects: the play in the shape playing gave it. Memorial origin concerns how the copy travelled: through the memory of someone who had played or watched it. A text can satisfy both descriptions at once, and indeed the classic reconstruction story, at its most coherent, always implied both: actors remembering a play remember the performed version, its cuts, its business, its running order, because that is the version their bodies know. The performed shape is what memory holds best; wording is what it holds worst. Read that sentence again and notice that it predicts the central datum exactly, theatrical shape preserved and enhanced, verbal texture degraded. The conjunction of the two hypotheses is not a compromise pasted together to end an argument. It is the only account under which the book’s combined profile, vivid directions, intelligent cuts, damaged verse, stops being a paradox.

Conversely, a performance text need not be a memorial one. Copy could reach a printer from the playhouse by documentary routes: a promptbook or a transcript of it, actors’ parts assembled into a whole, a hired scribe’s copy of the company’s abridged book. A text travelling by those routes would be a performance text of the strongest documentary kind, and its wording should be correspondingly clean, cleaner than the 1597 volume in fact is. The documentary routes are the ones the state of the wording argues against; the memorial route is the one the state of the shape argues for. Holding the two questions apart converts a shouting match into a matrix, and the matrix has a most probable cell: performance-shaped, memorially transmitted, with the print shop’s handling and possibly an earlier draft state contributing at the margins. That cell is where the verdict of this article finally rests, and it gives the neighbouring debate its proper size: the memorial reconstruction controversy concerns the transmission mechanism, which is a genuine and open question, while the performance question concerns the reflected shape, which the evidence settles much more nearly. A scholar may deny that any actor reconstructed anything and still read the short quarto as the performed play’s portrait. A scholar may affirm reconstruction and deny that the portrait is faithful. The positions are independent, and the literature’s habit of treating them as a single switch has cost the subject decades.

Could a promptbook stand behind Q1, and what would that settle?

It would settle the route in favour of documents and against memory, and it would predict clean wording in the lines the book keeps. The prediction fails: kept lines arrive paraphrased and transposed often enough that a playhouse manuscript, copied whole, is the least comfortable fit. Performance descent survives the failure; the documentary route does not.

A third route deserves naming because it dissolves the apparent competition between the two named ones: dictation. Suppose the performed version reached the printer because someone who carried it in memory spoke it aloud to a scribe, an actor or a pair of actors reciting the play as played, a hired rememberer performing his stock for a stationer’s money. Dictation is memorial in its source and documentary in its product, and it predicts a distinctive mixture in the result: the fluency of oral reconstruction, passages flowing where the reciter’s confidence flows, and the characteristic lesions where memory under dictation pressure smooths, substitutes, and leaps. It also predicts the directions’ presence with unusual economy, since a reciter describing the play as performed supplies business along with lines, narrating the action to make the transcript intelligible, exactly the descriptive texture the 1597 directions show. No document certifies that dictation occurred for this book, and the graded verdict files the route where it files the other mechanisms, under probable carriers of an established shape. But the route’s availability matters for the logic of the debate: the scholarly tradition long treated performance text and memorial text as exclusive verdicts between which a jury must choose, when the transmission’s plausible mechanics are hybrids almost by definition. The matrix, not the courtroom, is the right instrument, and in the matrix the 1597 book’s most probable cell has been stable throughout this article: performed in shape, remembered in carriage, handled in print by men who added their own noise at the margins.

The Queen Mab Test: One Speech Tried Both Ways

General arguments about patterns gain credibility when a single passage is tried in detail, and the dispatch for this article assigns the Queen Mab speech as evidence at passage length, reserving the microscopic collation for its companion study. Mercutio’s set piece in Act 1 Scene 4 is the tragedy’s most detachable famous speech: it advances no plot, introduces a fairy mythology the play never uses again, and exists to display a verbal virtuosity that delays the lovers’ meeting. Q2 gives the speech its full catalogue, the coach drawn by atomies, the dreams visited on lovers, courtiers, lawyers, ladies, parsons, and soldiers, a crescendo of social satire inside a fantasy. The short text keeps the speech’s frame and its opening movement, “She gallops night by night / Through lovers’ brains, and then they dream of love” (Act 1 Scene 4), and then contracts the catalogue to roughly half its length, reaching Romeo’s interruption and Mercutio’s admission of talking of nothing at a pace the long version declines to attempt.

Try the passage for the performance case first. A company playing the tragedy learns within a week of performances which passages hold a yard and which tax it, and a virtuoso aria that halts the action before the ball, the play’s first great spectacle, is the natural candidate for the knife. The cut as made respects the speech’s function with professional precision: Mercutio’s fantastical dominance of the scene survives, the dream logic survives, the exit line survives; what goes is enumeration, the middle of the catalogue where one dream more or less changes nothing. That is abridgement by a hand thinking in minutes and attention, and it is continuous with the pattern across the whole book. Now try the passage against the case. A memory-based reporter predicts similar damage for a different reason: catalogues are what memory sheds, lists being the least memorable form of verse, and a paraphrasing transmission would thin the speech whether or not any company ever cut it. A literary abridger working for a printer predicts the same cut again, since the speech’s detachability is visible on the page to any reader. Three hypotheses, one observation. The Mab passage therefore confirms the book’s patterned shortening without discriminating among its agents, and it is used in this article as corroboration, not as proof. The discriminating evidence remains the surplus staging knowledge and the playhouse name, features only one family of hypotheses predicts. A discipline that knows which of its exhibits are corroborative and which are decisive is the difference between the newer scholarship at its best and the advocacy that has sometimes borrowed its conclusions.

One consequence of the Mab cut deserves separate notice because it shows the performed version interpreting the play, not merely shortening it. In the long text, the catalogue’s full crescendo makes Mercutio’s virtuosity the scene’s centre of gravity: the speech swells until the character seems to command the play’s imagination, and Romeo’s interruption arrives as relief from a display that has outgrown its dramatic errand. In the short text the balance shifts. With the catalogue halved, the speech remains a brilliant obstruction but ceases to be a rival sun; the scene’s weight stays on the movement toward the ball, and Mercutio registers as the hero’s dazzling companion rather than as a competing protagonist. Companies and critics have long remarked that Mercutio threatens to walk away with the tragedy in performance, a problem the author himself solves only by killing the character at the midpoint. The performed version, on the evidence of this cut, had begun solving it earlier and more quietly, by rationing the display. An abridgement that interprets while it shortens is evidence of a specifically theatrical intelligence at work on the text, the intelligence of rehearsal rather than of the desk, and it is the kind of observation that converts waverers: a pirate abridging for print has no stake in the performed balance of a scene, while a company living inside that balance has nothing else.

Fitted for the Road: Touring, Doubling, and the Economics of the Short Text

The casting argument sketched earlier deserves its full form, because it connects the book to the material life of a 1590s company in a way no purely literary hypothesis does. A playing company was a business with a payroll. The Chamberlain’s Men of the middle 1590s comprised a core of sharers, a changing strength of hired men, and a small number of boys for the women’s parts, and every text the company mounted had to be cast from that finite pool, with the same faces returning in new costumes across an afternoon. The long text of the tragedy, played entire, presses hard against that pool: crowd scenes in the first and third acts, a ball full of named and unnamed Capulet guests, musicians, watchmen, citizens, servants in two households, an apothecary, a page for Paris and a page’s worth of attendants, besides the dozen principals. Doubling can absorb much of it, but doubling has friction: a performer cannot exit as one character and enter as another in the same breath, cannot double two figures the audience must distinguish in a shared scene, and cannot be asked to carry two heavy parts whose scenes interleave. The short text eases every one of these pressures. Business shrinks, supernumerary colour thins, scenes settle into sequences a reduced company can staff, and the whole becomes mountable by a troupe noticeably smaller than the long version wants.

The touring hypothesis gives that leanness a destination. Companies travelled when London playing was interrupted, above all in plague years, and a travelling party was a skeleton crew, the sharers and a few hired men, playing in guildhalls and inn yards to audiences who expected the famous London pieces in portable form. Records of provincial performance survive in quantity, and the repertory’s greatest hits plainly travelled. A version of the tragedy cut to fit a dozen or so performers, running near two hours, keeping every famous scene and shedding the rhetorical surplus, is precisely the instrument such travel required. The hypothesis cannot be proved from the book alone, and its friends have sometimes stated it with a confidence the records do not fund. But it integrates facts the rival stories leave scattered: the leanness, the length, the preserved spectacle and shed rhetoric, the directions warm with observed business. It also explains the book’s existence without villains. A company asset, worked hard on the road, is the kind of text that leaks into print in a decade when companies guarded their stock unevenly and stationers paid for copy with a famous title attached.

A subtler economic point supports the same reading. The long text and the short text are not competitors in this account but complements, the one the author’s full conception and the company’s capital asset in manuscript, the other the asset in daily use. Erne’s literary dramatist argument generalises the point across the period’s drama: playwrights writing for companies that doubled as publishers’ suppliers had reason to produce texts exceeding the stage’s capacity, and companies had routine machinery, the book-keeper’s pencil, the adapter’s hand, for bringing those texts within playing compass. On that understanding the 1597 quarto is evidence less of a crime than of a workflow, the ordinary conversion of literature into repertory, and its survival in print is the accident that lets the conversion be studied at all. The workflow account also predicts something checkable across the wider corpus, that other long plays of the period should show similar short theatrical counterparts, and several do, which is why the Romeo case has become the exemplar of a general phenomenon rather than a local mystery.

How the Comparison Is Actually Worked

The argument conducted in this article rests on a method any reader can repeat, and describing the method is part of the article’s evidence, because the performance case is finally a claim about patterns visible only in systematic comparison. The work begins with the two books side by side, the 1597 text and the 1599 text, collated scene by scene and line by line, with every divergence sorted into kinds: omissions, additions, transpositions, paraphrases, direction differences, speech assignment differences. Sorted so, the material stops being an impression of badness or vividness and becomes a distribution, and distributions are what hypotheses must answer to. The omissions cluster in rhetorical expansion; the paraphrases cluster in connective and catalogue verse; the direction differences cluster at moments of physical decision. Each cluster was predicted by one account and is an embarrassment to the others, and the discipline of the method is that the clusters, not the anecdotes, carry the verdict. Students entering the field repeat the exercise in miniature, often presenting the two texts of a single scene in parallel columns or slides; laying the parting scene’s directions side by side in a presentation built with a tool such as the ReportMedic PPT Viewer makes the distribution visible at a glance, the short text staging what the long text only says.

The method has a second stage, external comparison, which disciplines the first. Features claimed as signatures of performance descent must be checked against texts whose ancestry is known: authorial drafts that survive for other plays, licensed manuscripts, promptbooks with theatrical annotation, and printed texts set from each kind of copy. The check is sobering in both directions. Descriptive directions, the performance case’s prize exhibit, do occur in some authorial papers, though more rarely and with a different texture; paraphrase and transposition occur in memorial texts and also in careless scribal copying; actor names in directions are the rarest feature of all across every comparison class except documents of the playhouse itself. It is the rarity distribution of the Kempe entrance, measured against the external corpus, that converts it from anecdote to argument. Method talk of this kind belongs inside the article rather than in a footnote, because the graded verdict is a method’s output: probability assigned cluster by cluster, exhibit by exhibit, with the confidence of each grade tracking the discriminating power of the evidence beneath it.

The Surviving Playhouse Documents: What a Part and a Plot Can Teach

The external comparison stage of the method deserves its exhibits named, because the period’s playhouse paperwork survives in fragments eloquent enough to show what performance-derived paper looks like, and the 1597 book’s features can be measured against the genuine articles. Begin with the actor’s part. An Elizabethan performer learned a role from a scroll containing his own speeches and nothing else, each speech preceded by a short cue of the words that summoned it, so that the part is the play experienced from inside one body: entrances known, partners’ lines present only as triggers, the whole drama reduced to a thread of responses. One famous survival shows the format entire, Edward Alleyn’s part for the title role of Robert Greene’s Orlando Furioso, preserved at Dulwich College, a working document annotated and corrected in use. A text assembled from such parts, or reconstructed from the memories parts had drilled, would inherit their perspective exactly: the reporting actor’s own scenes held with cue-locked fidelity, the remainder held as gist. The fidelity-gradient claims made for the short quartos, disputed in detail for the tragedy, are claims that transmission passed through documents or memories of this shape, and the parts explain why that shape is the natural one for any playhouse leakage to take.

The plot, the second survival, is the part’s complement, the play experienced from the tiring-house wall. Companies prepared a scene-by-scene outline of entrances, properties, and personnel, posted backstage so that the performance could be marshalled by the book-keeper and the gatherers. The handful of surviving plots, among them the plot associated with The Seven Deadly Sins, list performers by name against roles and entrances, because a plot that named parts instead of men would fail at its only job. Here, in undisputed playhouse paper, is the ancestry of the tragedy’s most famous anomaly: a document class in which entrances carry actors’ names as a matter of function. When the 1597 print names Will Kempe at Peter’s entrance, it shows a feature the plot tradition generates naturally and the literary tradition does not generate at all. The point is not that a plot stands behind the book. The point is that the name belongs to a paperwork ecosystem, parts, plots, allowed books, whose business was performance, and any ancestry proposed for the copy must pass through that ecosystem or explain how the feature immigrated.

The third exhibit is an archive rather than a document: the papers of Philip Henslowe, preserved at Dulwich, with accounts of payments to playwrights, inventories of costumes and properties, and the record of a repertory run as a business. Henslowe served the Admiral’s Men, rivals to the Chamberlain’s, and no line of his diary touches the tragedy directly. Its value here is structural. It proves that the theatrical world generated dense paperwork around performance and almost none of it was meant to survive; what remains is the fraction chance preserved from one company’s landlord. For the Chamberlain’s Men, the tragedy’s own company, no equivalent archive exists, and the allowed book of the play, the licensed manuscript that was the company’s capital, is lost with the rest. The documentary desert around the 1597 quarto is therefore not an anomaly requiring suspicious explanation but the normal condition of the evidence for any play of the decade, and the graded method, patterns weighted, claims sized to what patterns can bear, is the only method the desert permits. Measured against the genuine paperwork, the short text’s exhibits rank themselves: the actor’s name aligns with the plots, the fidelity profile aligns with the parts, the abridged shape aligns with the repertory economics the diary documents. Alignment is not identity, and the grades already stated stand. But the book’s features are at home in the playhouse’s paper world in a way they are at home nowhere else.

What the Second Quarto’s Title Page Concedes

When the second quarto appeared in 1599, printed by Thomas Creede for Cuthbert Burby, its title page made a claim that functions in the argument as hostile testimony, all the more valuable for being hostile. The new volume advertised the tragedy as “newly corrected, augmented, and amended”, a formula announcing to buyers that the edition in circulation was defective in wording, deficient in quantity, and wrong in particulars, and that the remedy was now on sale. The formula is evidence of a limited and precise kind. It certifies that the trade regarded the 1597 book as a corrupt and incomplete version of something fuller, which no one disputes. It certifies nothing about routes: corruption and incompleteness are the symptoms every hypothesis claims to explain. Yet one word of the formula deserves dwelling on. Augmented speaks to quantity, to a play grown longer in its proper form, and it sits comfortably with the performance account, under which the short book is the fuller text’s playing abridgement and the 1599 volume restores what playing removed. Under the piracy account, augmentation must mean the restoration of lines the pirates dropped, which is equally sayable. The title page, like the 1597 page before it, advertises rather than testifies, and both pages together establish only the framing fact of the whole inquiry: contemporaries who knew the play in the theatre recognised the short book as the performed phenomenon’s imperfect relative, and paid for the distinction. The grading of that imperfection, theft or adaptation, memory or manuscript, is work the title pages leave entirely to the evidence inside the books.

The Adjudication: Erne Against the Sceptics, Point by Point

The run’s standard requires a genuine disagreement surfaced and adjudicated, and the materials are now on the table. Set the principals in their strongest forms. Erne’s position: the long texts of the period exceed plausible playing time by design, their author writing with readers in view; performance required abridgement as routine; the short quartos, the 1597 tragedy among them, are the surviving witnesses of that routine, best described as performance texts, and their study is the study of the plays as early audiences actually encountered them. The sceptical position, in the complementary forms given it by Maguire and Werstine: the narratives by which short texts have been explained, piracy, memorial reconstruction, shorthand theft, are documentary fictions maintained by repetition; the diagnostic features invoked to identify transmission mechanisms do not behave as the theories require; and the honest description of the evidence is that the routes by which these books reached print are unknown and, on surviving evidence, unknowable in detail. Between the principals stands the editorial tradition represented by Gibbons, declining to decide what it does not need, building its text on the second quarto while quietly importing the first quarto’s staging intelligence wherever the longer book is silent or clumsy.

Adjudicate point by point. On playing time, Erne wins outright: the arithmetic of the long text against the Prologue’s two hours is not answerable by appeals to faster delivery, and no sceptic has produced the stopwatch that saves an unabridged 3,040-line performance inside the advertised span. On the presumption of theft, the sceptics win outright: the piracy narrative for this book has no documentary foundation, and Werstine’s demonstration that the tradition argued by analogy from text to text stands unrebutted. On mechanism, the sceptics win against both older parties: Maguire’s tests disable confident diagnosis of memorial reconstruction, and the same rigour disables confident assertion of promptbook descent, which the state of the kept wording independently discourages. On the positive characterisation of the short text, Erne wins the point that matters and loses the ornament: the patterned cuts, the surplus staging knowledge, the playhouse name, and the casting economy converge on performance descent as the best available explanation of the book’s shape, while nothing in his argument, or anyone’s, certifies the book as a record of performance in the event sense. The editorial tradition, judged as a fourth party, turns out to have practised the correct adjudication for decades before the theorists stated it: trust the long text for the words, trust the short text, cautiously and with attribution, for the staging. Theory has caught up with the editing room.

Notice what the adjudication does to the original question. Asked whether the 1597 quarto is a performance text, the literature’s shouting suggested a coin to flip. Graded, the answer is a distribution: the shape claim is established to the standard history can reach, the document claim is open between memorial and mixed transmission, the event claim fails for want of any possible evidence, and the practical claim, that the book is a primary source for early staging, is as secure as anything in the field. Scholars who continue to fight over the label are usually fighting over which grade the label should be allowed to hide. Stating the grades ends the fight without splitting the difference, because each grade is argued, not averaged.

What a Reader, an Editor, and a Director May Each Take from Q1

A verdict that changes nothing is a decoration, so the consequences come now, one for each of the book’s three professional readers. The general principle distributes the early texts by competence: the long text governs the words, the short text informs the staging, and neither is permitted to annex the other’s jurisdiction without special argument in the particular case.

For the common reader, the short quarto offers what the long one structurally cannot: the tragedy’s performed rhythm. Reading the 1597 text is the closest a page can come to the experience of the first audiences, a story that moves at the speed of event, in which the machinery of courtship, feud, marriage, exile, and tomb succeeds itself without the long text’s contemplative pauses. The reader who knows only the short text knows a swifter and in some ways crueller play, one whose haste is itself an interpretation, the lovers crushed not by meditation but by momentum. The licence has a boundary, and the boundary is quotation: lines existing in differing forms in the two texts must be quoted from the long text, whose wording stands nearer the author’s papers, and the short text’s variants are evidence about performance, not alternatives for citation. A reader may learn from the first quarto how the play went. A reader may not cite it for what the play says where the witnesses differ, except to make the performance point itself.

For the editor, the verdict ratifies existing best practice and sharpens its rationale. Copy-text remains the second quarto, on grounds untouched by anything argued here: derivation from authorial papers, fuller and cleaner wording, the 1599 volume’s own advertisement of correction. The first quarto enters the edition in three licensed capacities. Its stage directions are admissible evidence of early staging and may be adopted, marked as adopted, where the copy-text is silent, as editors since Gibbons have adopted the window direction for the parting and the mechanics of the fatal thrust. Its substantive variants belong in the collation complete, because a performance-shaped witness records choices made under theatrical pressure, and some of those choices may preserve readings the long text’s own transmission disturbed. Its cuts are evidence in the introduction and the commentary, the primary exhibit for the play’s early life in the theatre. What the editor may not do is promote the short text’s wording against the copy-text on the ground of its theatricality: theatrical descent certifies shape, not readings, and an abridgement’s paraphrase has no authority against the fuller witness it abridges.

For the director, the licence is the widest of the three, and the most practically valuable. The short quarto is a worked example of the play cut by professionals to professional ends: which speeches halve without bleeding, which scenes fuse, where spectacle must be kept and rhetoric may go. A company facing the long text’s three hours inherits a set of solutions tested, on the performance account, in the play’s own first years. The business is a second inheritance: the parting staged at the window, the fight choreographed around an intervening body, the catastrophe arranged as a sequence of visible acts. A director may take all of it as evidence of what the first stage did, while remaining free, as every generation of the stage has remained free, to find other solutions. The one use the verdict forbids is citation of the short text as an authenticity machine, a claim that a production reproducing its cuts and business thereby reproduces 1597. The book certifies a version’s existence, not an evening’s contents, and a director who understands the grades gains a resource; a director who ignores them gains only a slogan.

Where does Q1 help each of its readers most?

In three different currencies: pace for the reader, business for the director, and a control on both for the editor. The short text teaches how the tragedy moved, shows how its decisive moments were disposed in space, and lets an edition distinguish the play’s literary fullness from its performed shape without confusing them.

Underlying all three licences runs a single change of regard, the rehabilitation’s real content. The 1597 book was for generations an embarrassment to be explained away, a corruption between the reader and the play. On the graded verdict it is a witness of a kind the long text cannot be: evidence that the tragedy had a performed life distinct from its literary one, and a partial transcript of that life’s shape. The cliché dismissed the quarto because it was not the play entire. The performance view restores it because the play entire was never what the first audiences saw, and the book preserves the dimensions of what they did see.

Shorthand on Trial: The One Mechanism That Claims to Fix an Event

Of all the transmission mechanisms proposed for the short text, shorthand alone would make it a performance record in the strong sense, a transcript fixed during performance itself, and it therefore deserves a trial of its own. The technology existed. Timothy Bright published his Characterie, a system of stenography, in 1588, dedicating it to Queen Elizabeth; Peter Bales offered a rival method in The Writing Schoolmaster of 1590; John Willis set out a fuller art of stenography in 1602. Sermons were demonstrably taken down by shorthand in church, and the practice’s migration to the playhouse was alleged early and often: Thomas Heywood complained in print that plays had been stolen by shorthand note-takers, and the figure of the spectator scribbling in the yard became a stock joke. If a shorthand writer worked through performances of the tragedy at the Theatre, collating several afternoons into a composite copy, the resulting manuscript would be a genuine performance record, and its sale to a printer would explain the book entire.

Set the hypothesis against its predictions. A shorthand transcript fixes the order of an event: what was spoken, in the sequence spoken, including errors of the afternoon, an actor’s skip or stumble fossilised in type. The 1597 book does not behave like an event transcript. Its errors are of the settling kind, paraphrase that smooths, recollection that imports lines from elsewhere in the play, anticipation that reaches forward to material an audience has not yet heard at the point of error. Anticipation is the diagnostic embarrassment: a stenographer cannot record in Act 2 a phrase that belongs to Act 4 unless the afternoon itself contained it, while a rememberer, saturated with the whole play, produces anticipation as a matter of course. The shorthand hypothesis also struggles with the book’s most theatrical feature, its descriptive directions: stenography captures words, and the business directions would require a second observer noting action in parallel prose, an apparatus for which no evidence exists. Finally, the system’s known speeds and the compression of its symbols make the capture of rapid dialogue in a noisy yard a heroic assumption, one the sermon parallel does not rescue, since the preacher’s pace and the actor’s pace are different instruments. The verdict on shorthand is not impossible but unsupported, with its strongest predictions contradicted at the anticipation points. Its historical importance lies elsewhere: it was the mechanism the strong claim needed, and its failure is why the strong claim fails with it.

The sermon parallel, shorthand’s best supporting evidence, repays a closer look because it shows both the technology’s reach and its limits in a single comparison. Sermons were taken down in church by practised hands, and printed sermon texts derived from such notes circulated with the preachers’ complaints attached, complaints that specify the failure modes: words misheard, arguments collapsed, the note-taker’s own phrasing substituted where attention lapsed. A sermon, however, is the most favourable capture conditions imaginable, one voice, a stationary speaker, a predictable structure of heads and proofs, an audience seated in silence, and a text the speaker has often substantially composed beforehand. The playhouse inverts every condition: many voices in rapid exchange, movement and noise, verse delivered at playing speed, speeches overlapping action that itself carries meaning. That shorthand succeeded, imperfectly, under the first set of conditions is evidence it could be attempted under the second; that even sermons emerged scarred by substitution and collapse is evidence of what the attempt would yield. Transferred to the tragedy, the parallel predicts a text whose wording is a stenographer’s reconstruction as much as a record, which is to say it predicts the 1597 book’s texture by a mechanism whose first requirement, verbatim capture, it simultaneously refutes. Shorthand thus survives the trial in a reduced office: a possible auxiliary in the transmission, a rememberer’s aid or a spectator’s supplement, rather than the recording instrument of the strong claim. The strong claim needed a machine. The evidence keeps supplying a memory.

The Verse Damage, Mapped by Kind

Return to the counter-case’s central datum with a taxonomy in hand, because the kinds of damage discriminate among mechanisms better than the fact of damage does. Four kinds recur through the collation. Paraphrase substitutes a near-equivalent for the copy-text’s phrase, sense preserved and music lost, the signature of a mind holding meaning without wording. Transposition reorders elements within a line or speech, the signature of recall reassembling a structure whose parts are remembered better than their sequence. Anticipation and recollection import material across scene boundaries, the signature of total familiarity, a transmitter who knows the whole play and cannot keep its compartments sealed. Metrical collapse flattens verse toward prose rhythm, the signature of wording recalled at the level of gist and reconstituted under pressure. Now set beside the taxonomy the damage an abridger does. Abridgement deletes. It does not paraphrase what it keeps, does not transpose, does not import, does not flatten, because its raw material lies open on the desk in correct form. The four kinds of damage are unanimous testimony against desk transmission of the kept text, and they are equally unanimous in favour of a transmitter whose medium was memory, whether an actor’s, a regular spectator’s, or a hired rememberer’s. That is as far as the taxonomy reaches, and its limit deserves emphasis: memory of what? The performed version, on the argument of the preceding sections, since the shape memory preserved is the abridged shape, the cuts already in place, the business already fixed. A rememberer recalling an unperformed long draft could not reproduce cuts the draft did not contain. The damage map and the shape map, drawn independently, point to the same cell of the matrix: a performed abridgement, carried to the printer through memory.

One refinement keeps the refinement honest. Memory in the period was a trained professional instrument, and actors’ memories were stocked by aural learning from dictation and by the part system, in which a performer received only his own lines with cues. A reconstruction drawing on several actors’ parts and memories would show fidelity gradients, scenes crisp where a reporting actor stood centre and haze where none did, and claims of exactly such gradients have been advanced and disputed for this text without resolution. The unresolved state of that finer question is recorded here without embarrassment, because the graded verdict does not need it. Whether one rememberer or several carried the performed version to Danter’s shop, the version carried was the performed one. The mechanism’s personnel remain unknown; the mechanism’s object is identified.

The Sister Witnesses: What the Other Short Quartos Teach

No verdict on one book should ignore the family it was assigned to, because the old scholarship’s deepest error was taxonomic: bad quarto named a drawer into which unlike objects were swept, and the drawer’s label did the arguing. Set the 1597 tragedy beside its supposed siblings and apply the same instruments, length, cut pattern, direction character, wording damage, and the family resemblance dissolves into separate cases that illuminate one another by contrast. The comparison also disciplines the performance claim: if the features graded as performance signatures here appear in books where performance descent is implausible, the grades must fall; if they pattern only where performance descent is independently plausible, the grades hold.

The 1600 quarto of Henry V is the nearest relative in profile. It prints roughly half the matter of the long text preserved in the Folio, sheds or shrinks the Chorus speeches that frame the action, compresses the rhetoric of kingship, and keeps the campaign’s narrative spine and its famous scenes in working order. Scholars who reject every performance argument for the tragedy must still explain a martial play halved with its spectacle intact and its meditation taxed, the profile of a version fitted to playing resources and playing time. The parallel strengthens the present verdict by showing its pattern is not a one-book anomaly requiring a one-book criminal. The 1603 first quarto of Hamlet cuts the other way in instructive fashion. It too is drastically shorter than the long texts, but its divergences include reordered scenes, a soliloquy displaced to an earlier position, names altered, Polonius appearing as Corambis, and a compression that scholars have read as damage in the transmission at least as much as design in the cutting. The Hamlet quarto warns against overreading design into every divergence: some short texts are abridgements, some are casualties, and several are both in layers no analysis fully separates. Applied home, the warning trims the Romeo verdict’s confidence about particular passages while leaving the convergent pattern, cuts, directions, casting, length, intact, since the tragedy’s divergences do not include the Hamlet quarto’s structural disorder.

The first quarto of Richard III, printed in the same year as the tragedy, is the comparison that breaks the old taxonomy outright. It is a long, substantially sound text, the foundation of every edition of its play, and whatever its copy’s exact nature, no account of it as a pirate’s garbled theft has ever commanded assent. A class that contains both that book and the short Merry Wives quarto of 1602, a text compressed and disordered to a degree the tragedy never approaches, is not a class but a shelf. Maguire’s corpus-wide survey demonstrated the point statistically: the features supposed to define the class refuse to stay inside it. The consequence for the present question is clarifying. Nothing follows about the 1597 tragedy from membership in a set that does not exist. Everything follows from its individual profile, which is why this article has argued from profile throughout, and why its verdict, unlike the old label, would survive the complete abandonment of the term bad quarto. Indeed the verdict implies the abandonment. A text best described as a performed abridgement imperfectly transmitted has no natural home under a heading meaning stolen and corrupt, and the heading’s retirement, already far advanced in the scholarship, is one of the performance view’s cleaner victories.

The Graded Verdict

The question in the title can now be answered at the length it deserves, claim by claim. That the 1597 quarto embodies the tragedy in a performed shape, abridged to playing length by cuts that sort by stage usefulness, furnished with directions carrying surplus staging knowledge, requiring a company measurably smaller than the long text’s, and running to the Prologue’s advertised two hours: this is established to the standard historical evidence reaches, and no rival account explains the same distribution without multiplying its machinery. That the copy reaching the printer came through memory, the memory of participants or habitués carrying the performed version in their heads and setting it down with the damage memory inflicts on wording while sparing shape: this is probable, the best fit for the combined evidence, held against the sceptics’ correct insistence that the personnel and the occasion are beyond recovery. That playhouse documents, a promptbook or assembled parts, also fed the transmission at some point, contributing the actor’s name and perhaps the sharpest directions: possible, compatible with the memorial route rather than exclusive of it, and unprovable on present evidence. That the book records a performance, any single afternoon at the Theatre fixed as it happened: not established, contradicted in its strongest shorthand form by the anticipation evidence, and to be retired from serious statement of the case. That the book is a worthless corruption, a text to be quarantined from the study of the play: refuted, by the convergence of its evidence and by a century of editorial practice that never stopped using its staging intelligence even while repeating the quarantine in prefaces.

Two misunderstandings of the verdict should be headed off before it is quoted. The first treats the performance account as a compliment that raises the short text to parity with the long one, an alternative Romeo and Juliet of equal authority to be chosen by taste. Nothing argued here supports parity. The two books answer different needs with different credentials, and the long text’s authority over the words is not dented by the short text’s authority over the staging; a reader choosing the 1597 version as a reading text chooses the performed dimensions at the price of the author’s phrasing, and should make the choice knowingly. The second misunderstanding treats the verdict as a licence to ransack the short text for authenticity effects, its cuts revived as period practice, its business reproduced as documentation. The grades forbid that confidence. A version’s existence is certified; its every detail is not, and details travelling through memorial transmission arrive with the static of that channel on them. The honest use of the book is probabilistic throughout, its patterns trusted, its particulars corroborated before they are staged or printed as fact.

The verdict also travels, with its grades adjusted, to the family of texts from which this one was artificially separated. Wherever a short early text shows the convergent profile, patterned cuts, surplus staging knowledge, casting economy, playing length, the performance account should be the starting hypothesis and the burden placed on rivals; wherever the profile fails, as it fails in stretches of the Hamlet quarto’s disorder, the account should be withheld without embarrassment. Method, not loyalty, distributes the label. That portability is the final measure of the newer view’s worth. The piracy story it replaced could explain any text by adjusting its criminals; the graded performance account can be wrong about a given book in checkable ways, and a theory that can be wrong is the only kind that can also be right.

So: is Q1 a performance text? Yes, in the sense that survives cross-examination. The first quarto is the performed Romeo and Juliet, abridged, staged, cast, and timed, transmitted to print by the imperfect instruments the period offered, and legible ever since to any reader willing to grade its testimony instead of dismissing its character. The newer view, disciplined by its critics, does not ask the book to be believed. It asks the book to be read, as the record of the only Romeo and Juliet the sixteenth century ever actually saw.

What Would Change the Grades: Standards of Proof in a Documentary Desert

A graded verdict earns trust by stating its own revision conditions, and the exercise is worth a section because it exposes why the grades sit where they do. The field works in a documentary desert: for the tragedy, no manuscript of any kind survives, no company accounts name the text, no contemporary describes its transmission, and the two printed books are the whole of the direct evidence. In such a desert, arguments are built from internal pattern and external analogy, and their grades should track the strength of pattern each claim needs. The shape claim needs only the book itself, whose patterns are inspectable by any reader with the two quartos open, and it accordingly stands highest. The mechanism claim needs the patterns to discriminate among routes that predict overlapping damage, a harder test, and it stands one grade lower. The event claim needs a witness chain the desert does not contain, and it fails. The architecture of the verdict is the architecture of the evidence, and a critic who wants a different verdict must attack the architecture, not the adjectives.

Name the discoveries that would move the grades, since naming them proves the grades are falsifiable rather than temperamental. A promptbook or book-keeper’s copy of the tragedy with abridgement marks matching the short text’s cuts would raise the document claim to certainty and reframe the memorial evidence as secondary contamination. A surviving actor’s part whose wording matched the short text’s distinctive errors in the scenes that part covers would identify a transmission channel and a personnel for the first time. A Stationers’ Company record, a bookseller’s inventory, a lawsuit over copy, any paper trace of the 1597 book’s acquisition by its printers, would convert route speculation into route history overnight. Press analysis of a new order, demonstrating that the short text’s copy arrived at the print shop in cast-off gatherings of differing character, could separate the memorial strata from the documentary ones compositor by compositor. None of these discoveries is expected; the desert has been sifted for a century. But the verdict’s willingness to be overturned by them is the difference between bibliography and belief, and it is the standard to which the old piracy verdict should also have been held. Stated as revision conditions, the old verdict fares worst of all: no imaginable document short of a signed confession from a pirate actor would establish it, and its century of dominance rested on narrative charm rather than on evidence that could, even in principle, have been checked. The performance view, in its graded form, is not merely the better story. It is the first account of the book whose claims are sized to the evidence that exists and exposed to the evidence that might.

Frequently Asked Questions

Q: What separates a performance text from a text meant for readers?

A performance text is shaped by the conditions of playing: a running time near two hours, a cast a working company can field, spectacle kept and rhetoric taxed, directions concerned with bodies in space. A readerly text is shaped by the page, free to expand speeches, multiply reflection, and exceed any afternoon’s capacity. The second quarto of the tragedy, at roughly 3,040 lines, fits the second description; the first quarto, some 800 lines shorter with its cuts sorted by stage usefulness, fits the first. The distinction drawn in Lukas Erne’s Shakespeare as Literary Dramatist is one of destination, not of worth: the same play could exist in both shapes, the long version for the study and the short version for the stage, and the 1597 book’s claim on attention is that it preserves the staged destination’s dimensions.

Q: Which scholar made the strongest modern case for Q1 as a performance text?

Lukas Erne, in Shakespeare as Literary Dramatist (2003), made the case that has reorganised the field. Erne argued that the long texts of the period were written with readers in view and run far beyond plausible playing time, so that performance required abridgement as a routine operation, and the short quartos survive as witnesses of that routine. For the tragedy, the short text lands close to the Prologue’s advertised two hours while the long text exceeds three at any realistic pace, and its cuts fall on lyric expansion rather than plot. Erne’s framing reclassified the first quarto from a pirate’s garble to a playing version, while his own discipline kept the claim at the level of versions and destinations rather than transcripts of particular nights. Earlier work by Kathleen O. Irace on the short quartos as deliberate adaptations prepared the ground his argument built on. Erne extended the case in Shakespeare and the Book Trade (2013), situating the long texts within a print culture that valued substantial playbooks for reading, and the combined argument has shifted the default: the short quarto is now approached first as a version with a destination, and only then as a puzzle of transmission.

Q: How much shorter is Q1 than Q2, and why does the difference matter?

By the usual counts, the first quarto runs to roughly 2,230 lines and the second to roughly 3,040, a difference of some 800 lines, about a quarter of the longer text; counts vary slightly by edition and method, but the proportion is stable. The difference matters because of its size and its distribution. Transmission accidents corrupt at the scale of words and phrases; they do not remove a quarter of a play while sparing its plot entire. The missing quarter is patterned, concentrated in lyric elaboration, rhetorical set pieces, and repeated lament, with the sequence of encounters and decisions intact. Patterned loss at that scale requires an agent with a purpose, and the purpose the pattern fits is playing, a company taxing duration wherever the story is stationary. Size makes the fact; distribution makes the argument.

Q: Why does Q1 name the actor Will Kempe in a stage direction?

In the musicians’ scene of Act 4 Scene 5, the 1597 print introduces the servant Peter under the name of Will Kempe, the Chamberlain’s Men’s celebrated clown. The likeliest explanation is that playhouse knowledge stands in the book’s ancestry: casting is knowledge a desk manuscript has no reason to hold and a performed version cannot help recording, since the name identifies who actually walked on at that entrance. Every theory of the book must find room for the fact. A reporter who had seen the play could carry the name away in memory; a set of actors’ parts or a promptbook could carry it in writing; a compositor inventing theatrical colour has no motive to coin so checkable a detail. The name does not identify the route by itself, but it fixes the territory the route must cross, the playhouse itself.

Q: Does Q1 cut any scene that the plot cannot spare?

Essentially none, and the near-perfect preservation of the load-bearing structure is central evidence for the performance account. Every encounter the story requires survives: the brawl, the ball, the balcony, the secret marriage, the street fight, the banishment, the potion, the tomb. What disappears or contracts is material the plot can spare by definition, the second Chorus sonnet entire, stretches of lyric wooing, the middle of the Queen Mab catalogue, formal lamentation in the fourth act, reverie and valediction in the fifth. A text damaged by forgetfulness or by a printer’s economy would be expected to lose load-bearing and decorative material indifferently, and to lose them in the order the copy or the memory failed. Instead the losses sort by function with a consistency that points to a cutter who understood what each scene was for.

Q: Can a text be a performance record and a memorial reconstruction at the same time?

Yes, because the two descriptions answer different questions. Performance origin concerns what a text reflects, the play in the shape playing gave it; memorial reconstruction concerns how the copy travelled, through the memory of someone who had played or watched it. Actors remembering a play remember the performed version, its cuts, its business, its running order, because that is the version their bodies learned. The conjunction predicts the first quarto’s combined profile with precision: theatrical shape preserved and even enhanced, verbal texture degraded by paraphrase, transposition, and anticipation. The descriptions compete only when they are confused. A promptbook transcript would be a performance text without memorial transmission, and a spectator’s reconstruction of an uncut reading text is barely conceivable. For the 1597 book, the evidence supports the performed shape strongly and the memorial route as its most probable carrier.

Q: What evidence in Q1 tells against a purely theatrical origin?

The state of the lines the book keeps. Abridgement by a company working from its own playbook removes whole speeches and leaves the survivors word-perfect, since the survivors must be spoken that afternoon. The 1597 text’s survivors are frequently paraphrased into near-equivalents, transposed in their elements, flattened in metre, and contaminated by lines anticipated from later scenes or recalled from earlier ones. None of those defects is an abridger’s defect; all of them are a rememberer’s. A purely theatrical origin, meaning descent from a playhouse manuscript of the performed version, therefore fails on the wording it should have protected. The counter-case built by Laurie E. Maguire and Paul Werstine adds the methodological half: the mechanisms traditionally invoked for such books lack documentary foundation, so certainty about the route is unavailable in either direction.

Q: Could Q1’s stage directions have been composed in a print shop rather than a playhouse?

Some of them could, and the concession defines the evidence’s limit. A direction describing exactly what the surrounding dialogue announces, such as the heroine’s final fall, could be supplied by an intelligent compositor, scribe, or hired adapter without any playhouse source. Directions are also occasionally descriptive in authorial and scribal copies elsewhere in the period, so vividness alone cannot certify staging. The directions that resist the print-shop explanation are those carrying surplus knowledge the dialogue does not contain: the mechanics of the thrust that kills Mercutio, delivered under Romeo’s intervening arm, and the entrance that names the company’s clown. A print shop could invent theatrical colour in general; it had no access to, and no interest in, the specific dispositions and personnel of the Chamberlain’s Men. The case rests on the surplus, not on the colour.

Q: What can a director legitimately take from Q1?

A worked example of the tragedy cut by professionals to professional ends, and a body of staging evidence from the play’s first theatrical life. The cuts demonstrate which speeches halve without bleeding, the Queen Mab catalogue, the fourth act’s lamentation, Romeo’s reveries, and which scenes must stand entire. The directions preserve dispositions a modern staging can test: the lovers framed together at the window for the dawn parting, the fatal fight choreographed around an intervening body, the catastrophe assembled as visible acts. What a director may not take is an authenticity certificate. The book certifies that a version of this shape was played; it does not certify the contents of any single evening, and a production reproducing its cuts is reviving a version, not re-enacting a documented night at the Theatre.

Q: Why do editors print Q2’s words but borrow Q1’s stage directions?

Because the two quartos have different competences, and editing follows competence. The second quarto descends from authorial papers, runs fuller and cleaner, and was advertised in 1599 as newly corrected, augmented, and amended; it is the proper copy-text for wording. The first quarto descends from the play in performance, and its authority, where it has any, concerns staging rather than phrasing. Where the copy-text is silent about disposition, the short text often speaks with the playhouse’s knowledge, as in the window direction for the parting scene that editions since Brian Gibbons’s Arden text of 1980 have adopted with attribution. Borrowing a direction from a performance witness costs the edition nothing in verbal authority, while borrowing a reading from an abridgement’s paraphrase would cost a great deal. The practice is the graded verdict applied at the desk.

Q: What can a reader learn from Q1 that Q2 cannot teach?

The tragedy’s performed rhythm. The long text teaches the play as meditation, its lovers thinking at length in some of the most celebrated lyric verse in English. The short text teaches the play as event, a machine of encounters accelerating from brawl to tomb at the speed the first audiences experienced. Reading both reveals that tempo is interpretation: in the short version the catastrophe feels less like fate contemplated than like momentum unchecked, the lovers crushed by a story that never pauses long enough for rescue to organise itself. The short text also preserves the proportions of the performed play, which characters and episodes bulked large on stage rather than on the page, the Nurse and Mercutio vivid, the philosophical friar trimmed. No amount of annotation on the long text supplies that knowledge, because the long text does not contain it.

Q: How long would Q1 take to perform, and why does playing time matter to the argument?

At plausible rates of early modern delivery, the short text runs a little over two hours with fights, music, and stage business included, close to the span the Prologue advertises in the line “the two hours’ traffic of our stage”. The long text, at the same rates, runs well past three hours. Playing time matters because it converts an impression, that the short text feels stagey, into a measurement that hypotheses must fit. A version cut to the advertised span by patterned removal of stationary rhetoric behaves like a text prepared under the stage’s central constraint. A reporter stealing a literary manuscript or a printer trimming a long copy has no particular reason to land on the Prologue’s figure, and the fit between the book’s length and the play’s own advertised duration is one of the performance case’s least assailable exhibits.

Q: How does the smaller cast Q1 requires support the performance case?

A playing company is a finite pool of sharers, hired men, and boys, and every text it mounts must be cast from that pool under the frictions of doubling, quick changes, and scenes where two roles meet. The long text presses hard against the pool, with its crowds, ball guests, musicians, watchmen, and double households of servants. The short text eases each pressure: supernumerary colour thins, business shrinks, and the whole becomes mountable by a noticeably smaller company, of the size a touring party or a lean London staging could field. Casting analyses in the tradition of T. J. King’s work on performance documents run consistently one way for the two quartos. A literary manuscript has no reason to fit a smaller company; a version prepared for playing has every reason. The fit supports performance preparation directly, while leaving the occasion, London or the road, open.

Q: What does Q1’s title page claim about performance, and how much weight can the claim bear?

The 1597 title page advertises the tragedy as often played publicly, with great applause, by the right honourable the Lord Hunsdon’s servants, the company that performed it. As evidence of transmission the claim bears little weight, because the formula is trade advertising: performance boasts appear on early editions of many ancestries, including books set from authorial papers, and a title page’s business is to sell copies by invoking the playhouse’s glamour. As evidence of reputation the claim bears more: it certifies that the book traded on a performed phenomenon the town recognised, which is consistent with the copy descending from the performed version in some form. The correct handling is to log the formula as marketing, extract the modest corroboration it offers, and rest nothing on it. The performance case is built from the book’s internal patterns, which no advertising department could have arranged.

Q: Is Q1 a record of one particular performance?

No, and the distinction between a performance text and a performance record is the one the whole argument turns on. A record would fix a single occasion, one afternoon’s words and business captured as they happened, most plausibly by shorthand. The one mechanism that could produce such a record fails its own predictions in the 1597 book: shorthand cannot anticipate in Act 2 a phrase belonging to Act 4, cannot supply descriptive directions for business it only hears, and cannot explain why the text’s errors are those of settled familiarity with the whole play. What the evidence supports is a version, the tragedy as the company performed it across a run, abridged and stabilised, carried into print by transmission that mixed memory with, possibly, playhouse documents. The book is a portrait of the performed play, not a photograph of a performance.

Q: What would count as stronger proof that Q1 is a performance text?

Documents, of kinds the field can specify precisely because their absence defines the argument’s limits. A promptbook or book-keeper’s copy bearing abridgement marks that match the short text’s cuts would convert performance descent from pattern inference to documentary fact. A surviving actor’s part whose wording reproduced the short text’s distinctive errors in the scenes that part covers would identify both a channel and a carrier. A Stationers’ Company record, a lawsuit, a bookseller’s inventory tracing the copy’s acquisition by the 1597 printers would turn route speculation into route history. Failing such finds, press analysis separating the copy’s strata compositor by compositor could sharpen the mechanism grades. The graded verdict is built to be revised by exactly this evidence, which is what distinguishes it from the piracy narrative it replaced, a story no imaginable document short of a signed confession could have established.

Q: How does the performance-text view change the meaning of “bad quarto”?

It dissolves the label for this book. Bad quarto, Alfred W. Pollard’s coinage of 1909, bundled a description, short and corrupt, with a verdict, stolen and worthless, and for half a century the verdict did the work: the 1597 tragedy required no study because its class condemned it. The performance view separates the description from the verdict. The book is indeed shorter and its wording indeed damaged, but the shortness patterns as abridgement for playing and the damage patterns as memorial transmission of the performed version, neither of which implies theft or worthlessness. Corpus-wide testing by Laurie E. Maguire showed the supposed class has no stable features anyway; its members differ from one another more than the label allows. Retiring the term for the tragedy is not politeness. It is the removal of a premise that had been masquerading as a conclusion.

Q: Why did scholars resist the performance-text view of Q1 for so long?

Because the piracy settlement was tidy, prestigious, and useful. Pollard’s label and W. W. Greg’s theory of memorial reconstruction explained shortness by theft, corruption by memory, and theatrical colour by the thief’s presence in the theatre, all in one motion, and the explanation carried the authority of the century’s commanding bibliographers. It also served editors well, licensing undisturbed reliance on the second quarto. Resistance weakened when the settlement’s foundations were audited: Paul Werstine showed in 1998 that no documentary evidence records any actor selling any play from memory, and the narratives had been built by analogy between texts. Kathleen O. Irace’s account of the short quartos as deliberate adaptations, and then Erne’s playing-time argument, supplied positive alternatives the old story could not absorb. What fell was not a fact but a presumption, and presumptions with famous names attached fall slowly.

Q: If Q1 records the performed play, should editors make it the copy-text instead of Q2?

No, because performance descent certifies shape, not wording. A copy-text is chosen for authority over the words, and on wording the second quarto’s claims are untouched by anything in the performance argument: descent from authorial papers, fuller and cleaner text, contemporary advertisement as newly corrected, augmented, and amended. The first quarto’s wording arrives through abridgement and memorial transmission, the two processes least entitled to govern an edition’s readings. What the performance verdict changes is the first quarto’s standing in the edition’s other departments: its stage directions become admissible staging evidence, its variants earn full collation as the record of choices made under theatrical pressure, and its cuts become primary evidence for the play’s early life. Editors since Brian Gibbons have in practice held exactly this position. The verdict ratifies the practice and explains it.

Q: How should the claim that Q1 is a performance text finally be graded?

In parts, because the claim is several claims wearing one coat. That the 1597 quarto embodies the tragedy in its performed shape, abridged to playing length, staged with a company’s knowledge, cast for a smaller troupe, and timed near the Prologue’s two hours: established, on convergent internal evidence no rival explains as economically. That its copy reached print through memory of the performed version, possibly assisted by playhouse documents: probable, the best fit for wording that is shaped like the stage and damaged like recollection. That it transcribes any particular performance: unproven, and contradicted in its shorthand form by the book’s anticipations. That it is a worthless corruption: refuted. The grades are the answer. Stated whole, the first quarto is the performed Romeo and Juliet in imperfect transmission, and that is a distinction no earlier generation of scholarship thought to grant it. The grading is not indecision dressed as nuance; it is the accurate report of an evidence base in which the shape of the book is inspectable, the route of its copy is inferable, and the event behind it is gone beyond recovery, leaving the version, not the evening, as the object the evidence can certify.