Every citation of Romeo and Juliet act and scene divisions carries a quiet falsehood. A student who writes “Act 3, Scene 1” for the street fight in which Mercutio dies is using a grid that no Elizabethan reader ever saw, assembled across more than a century by Folio compositors, an Augustan editor working from a fourth-hand folio, a Cambridge schoolmaster with a theory of the cleared stage, and a succession of nineteenth-century printers who turned their combined work into the universal map. The act and scene divisions of Romeo and Juliet are not the author’s marks on his manuscript. They are an editorial inheritance, laid over a play written for continuous performance, and the history of how that grid was built, piece by piece and often by accident, is the subject of this article.

The verdict can be stated at once, because the evidence is unusually plain. The two early quartos of the play, printed in 1597 and 1599, carry no act divisions and no scene numbers at all. The First Folio of 1623 adds act headings irregularly, heads most later acts in Latin, marks a scattering of scene openings, and then lets its own scheme lapse for long stretches. Nicholas Rowe, editing Shakespeare in 1709, supplied the first complete and consistent scheme of acts and scenes, complete with the invented place headings that still frame the play in classrooms. Edward Capell and Edmond Malone refined the counting in the eighteenth century, and the Globe edition of 1864 froze the result into the citation system the whole world uses. What Shakespeare supplied was not a numbered grid but a sequence of stage pictures separated by cleared stages, and beneath that sequence a story whose joints happen, in most places, to fall where later editors wanted their breaks. The divisions deserve respect as a finding aid and scepticism as a claim about authorship. They are the best map of the play ever drawn, and the map is not the territory the playwright walked.
Who Divided Romeo and Juliet Into Acts and Scenes? The Short Answer
The honest answer to the question in the title is a list of tradesmen and editors rather than a single name. The compositors who set the First Folio in type for the printing house of William and Isaac Jaggard made the first move, breaking the inherited quarto text into acts at four points and heading three of those acts in a way that has confused bibliographers since. Rowe made the second and decisive move in 1709, dividing the whole play scene by scene and printing for each play in his six volumes a list of scenes with the places where they unfold. The eighteenth-century editors who followed him, Lewis Theobald, Sir Thomas Hanmer, William Warburton, Samuel Johnson, Edward Capell, George Steevens, and Edmond Malone, disputed individual breaks and renumbered individual scenes, but none dismantled Rowe’s frame. The nineteenth century then industrialized it. The Globe edition prepared by William George Clark and William Aldis Wright in 1864 printed the divisions in the form that school editions, concordances, and examination papers would copy for the next hundred years, and modern scholarly editions, from Brian Gibbons’s Arden text of 1980 to René Weis’s Arden text of 2012, still print substantially the grid that Rowe sketched and Malone settled.
Three features of that answer matter more than the names. The first is lateness. The complete scheme postdates the play’s composition by well over a century, and even its earliest layer, the Folio’s partial act headings, postdates the author’s death by seven years. The second is anonymity at the crucial point. Nobody knows which workman in the Jaggard printing house decided where the Folio’s act headings would fall, or whether the decision was made by a scribe preparing copy, by the editors John Heminge and Henry Condell, or by compositors fitting type to pages. The third is inertia. Once a grid exists, scholarship builds on it: essays cite it, directors block by it, students memorize it, and the cost of renumbering comes to seem higher than the cost of keeping a scheme everyone privately knows to be editorial. The divisions survived because they were useful, and they were useful because everyone else used them.
How many act headings does the First Folio actually print for Romeo and Juliet?
The Folio heads the second, third, fourth, and fifth acts with Latin act headings, each followed by a heading for an opening scene, while the first act begins with no heading at all once the Prologue is spoken, so the scheme presupposes the fifth act rather than announcing it.
That single detail, the missing first heading, is the whole history in miniature. A scheme designed from scratch would head all five acts. A scheme improvised in a printing house, by men working from a manuscript or printed copy that itself had no divisions, begins wherever the first break happened to be noticed and never looks back. The Folio’s divisions for this play are exactly that: noticed breaks, recorded at the ends of gatherings and the tops of pages, rather than a design imposed on the whole. Later editors, confronted with four headings and a gap, did not ask whether the author had wanted five acts. They supplied the missing heading, numbered the scenes the Folio had left unnumbered, and in doing so converted an accident of the printing house into what looked like an author’s architecture. The conversion was so complete that the question this article answers, who divided the play, now strikes many readers as odd. The divisions feel given. They were made.
The Unmarked Quartos: A River of Print Without Banks
The first quarto of 1597, the short early text begun in John Danter’s printing house and finished in Edward Allde’s, prints the tragedy as one unbroken sequence. Its title page promises “An Excellent conceited Tragedie of Romeo and Juliet, As it hath been often (with great applause) plaid publiquely”, and the text that follows keeps that theatrical promise to the letter: entrance follows entrance, speech follows speech, and the only large structural signs on the page are the two Chorus sonnets and the stage directions that clear and refill the stage. A reader of the 1597 quarto who wanted to find the balcony scene had to know the story, not a number. The text runs from the street brawl to the tomb vault without a single act heading, scene heading, or numeral of any structural kind.
The second quarto of 1599, the long text printed by Thomas Creed for Cuthbert Burby and the copy-text from which nearly every modern edition descends, is equally unmarked. Its title page advertises a play “Newly corrected, augmented, and amended”, and the augmentation is real, hundreds of lines fuller than its predecessor, but the corrections did not extend to structure. The 1599 quarto prints the same unbroken river: no act divisions, no scene divisions, no location headings, nothing between the Prologue and the Prince’s closing couplet except speech headings, stage directions, and the second Chorus. The third quarto of 1609 and the undated fourth quarto, usually assigned to about 1622, descend from the second and inherit its unmarked state. When the Folio’s printers took up the play in 1623, the copy in front of them, a quarto of that line, offered them no divisions whatsoever to copy. Every division in the Folio text is therefore an addition made in or for the printing house, not a feature transmitted from the author’s papers through the quarto tradition.
This matters because the quarto state is the closest surviving witness to what the author’s company possessed. Play manuscripts submitted to printers were working documents, and the absence of divisions in both quartos, in a short reported text and in a long text close to the author’s draft, is strong evidence that the play as written and first played had no numbered articulation. Shakespeare’s foul papers, on the usual account of what the second quarto was set from, contained entrances and exits, speeches, and the two sonnet Choruses. They did not contain the words “Act 2”. The playwright’s own structural thinking, whatever it was, expressed itself in the rhythm of filled and emptied stages, in the alternation of street and house and orchard and cell and tomb, and in the great scenic rhymes that critics such as Emrys Jones have mapped: the feast answering the tomb, the orchard answering the Capulet monument. None of that thinking required a numeral, and no numeral records it.
A detail worth pausing on is the Choruses. The Prologue sonnet that opens the play and the second Chorus sonnet that opens what is now called Act 2 are the only passages in the early texts that stand outside the dramatic action and address the audience about the whole. Later editors made much of the second Chorus, treating its position as a fossil of an act break, evidence that the author thought in acts after all. The inference is weaker than it looks. A Chorus in the manner of the old interlude tradition, or of the play’s own source handling, marks a hinge in the story, the turn from “death-marked love” promised to “the fearful passage of their death-marked love” narrated (Prologue, lines 9 and 13 use the same phrase of the lovers’ course), but a hinge is not an interval. Stories have hinges wherever a teller pauses to summarize, and the sonnet Chorus summarizes precisely because the action does not pause: the marriage plot is already moving underneath it. The quarto page, which prints the Chorus without ceremony between the street scene and the orchard scene, understands the Chorus better than the numbered page does. It is a breath inside the action, not a curtain drawn across it.
The Five-Act Idea: Horace, Terence, and the Grammar-School Inheritance
If the divisions did not come from the play’s own manuscripts, where did the expectation of five acts come from, and why did editors reach for it so confidently? The answer lies in a classical rule that every grammar-school boy of the period absorbed with his Latin. Horace, in the Ars Poetica, laid down that a play should neither fall short of nor exceed five acts: the Latin tag runs “neve minor neu sit quinto productior actu”, let the fable be neither shorter nor longer than the fifth act (Ars Poetica, lines 189 to 190). Roman practice seemed to confirm the precept. The comedies of Terence circulated in editions furnished with act divisions by late antique and Renaissance editors, and Seneca’s tragedies, the prestige model for serious drama, were printed in five acts. Commentators such as Donatus and Evanthius, whose remarks on Terence prefaced the school editions, explained the fivefold shape as the natural articulation of a plot into beginning, growth, crisis, reversal, and resolution. A well-made play, the classroom tradition taught, disclosed five movements the way a body discloses its joints.
T. W. Baldwin built the most influential version of this inheritance into a claim about the playwright himself. In William Shakspere’s Five Act Structure (1947), Baldwin argued that the five-act scheme was not a printer’s afterthought but the working architecture of the Elizabethan professional stage, learned from Terentian commentary at grammar school and applied by dramatists as a matter of course. On Baldwin’s account, the absence of printed divisions in the quartos proves nothing about composition, because the divisions were so thoroughly understood that nobody needed to print them: actors knew the act structure the way musicians know the bar lines, and the manuscripts reflected a shared professional grammar rather than a literate apparatus.
The Baldwin case deserves respect, and its strongest evidence is the broad shape of the surviving plays, this one included. Romeo and Juliet does fall into five large movements with almost suspicious neatness. A first movement establishes the feud, the lovers, and the meeting at the feast. A second carries the orchard declaration, the Friar’s consent, and the secret marriage. A third turns on the street fight, Mercutio’s death, the banishment, and the catastrophic interview in the Capulet house. A fourth winds the potion plot: the forced match with Paris, the Friar’s device, the feigned death, the discovery of the body. A fifth runs the Mantua news, the apothecary, and the tomb. Editors did not have to force the play into five acts; the fit was there to be found, and finding it felt like recovery rather than invention. That felt recovery is precisely why the grid hardened so fast once print supplied it. Each generation of editors experienced the divisions as recognition, and recognition is the hardest kind of editorial act to repeal.
Yet the inference from fit to authorship fails on the bibliographical evidence, and it fails in a way Baldwin’s argument never answered. If the five-act scheme had been the universal professional grammar he proposed, the printers of the quartos, professional men serving professional buyers, might be expected to show it somewhere, in a heading, a numeral, a rule of type, and they never do, in this play or in most plays of the 1590s printed from theatrical copy. More decisively, when the Folio does divide plays, its divisions are demonstrably erratic across the volume: some plays are fully divided, some partly, some wrongly, with act headings misplaced in ways no authorial scheme would produce. The pattern across the 1623 collection is the pattern of a printing house working through copy of mixed character at speed, not of a company archive whose manuscripts all carried the same professional grammar. Baldwin proved that educated men of the period knew the five-act rule. He did not prove that this playwright wrote this play to it, still less that any surviving number corresponds to an authorial intention. The distinction between knowing a convention and inscribing it is the distinction the whole subject turns on.
Continuous Staging: The Two Hours’ Traffic of the Stage
The strongest witness to how the play was built is the play’s own front door. The Prologue promises the audience “the two hours’ traffic of our stage” (Prologue, line 12), a phrase that imagines the performance as a single spell of traffic, roughly two hours of continuous stage business in an afternoon’s daylight. The public amphitheatres for which the tragedy was written in the middle 1590s, the Theatre and then the Curtain in Shoreditch, played in the open air to a standing crowd, and the economics of that house shaped its dramaturgy. Daylight could not be paused and resumed at will; a crowd that had paid its penny expected momentum; and the repertory system, in which a company performed a different play each afternoon, favoured scripts that could be mounted with the company’s fixed stock of properties and personnel. The staging conditions surveyed in the study of Romeo and Juliet on the Elizabethan Stage point the same way from every angle: the play was engineered to run.
Continuous performance did not mean featureless performance. The Elizabethan stage had its own punctuation, and it was written in bodies rather than numerals. A scene ended, in the practical grammar of the playhouse, when the stage emptied: the exit of the last speaker cleared the boards, and the next entrance established a new place and time without scenery, simply by the convention that an empty stage resets the world. Within a sequence, Shakespeare keeps the stage populated across what modern editions print as scene boundaries, or empties it for only a breath, so that the action pours forward. Consider the run of the present Act 2. The Chorus sonnet ends, Romeo enters alone in the orchard, Mercutio and Benvolio call after him and depart, and the orchard scene begins, all without a natural stopping place; the editions print two numbered units (2.1 and 2.2) where the stage experienced a continuous night of pursuit and discovery. Or consider Act 3, where the street fight, the Prince’s judgement, the cell scene, the lovers’ parting, and the Capulet confrontation follow in a single tightening spiral. The divisions that a modern reader sees between 3.1 and 3.5 mark cleared stages, genuinely, but the clears are moments in a continuous acceleration, and the play’s famous compression, five days crushed out of the nine months of its source poem, depends on the audience feeling no seams. The grid, by teaching readers to expect a curtain where the playwright built a hinge, subtly falsifies the play’s tempo.
The act interval itself, the pause with music between acts that later became normal in London theatres, belongs to a different house and a later moment. Indoor hall playhouses such as Blackfriars, where candlelight required periodic trimming and an audience of gallants expected music, developed act breaks with consort music between them; when the King’s Men acquired Blackfriars in 1608 and began playing winters indoors, the practice migrated into their repertory and eventually into print. Andrew Gurr’s work on the Shakespearean stage holds that the amphitheatre companies played without act intervals through the 1590s and that the fivefold performance shape is a hall practice retrojected onto amphitheatre plays. Gary Taylor, arguing from patterns in the Folio texts, has proposed that some act divisions reflect genuine performance breaks introduced at the Globe itself in the years after about 1609. The dispute matters for plays revised or printed late; for a tragedy composed in the middle 1590s and already twice printed by 1599, the interval case collapses. Whatever the Folio’s headings record, they cannot record how Romeo and Juliet was played at the Theatre, because that play in that house ran its two hours unbroken, the lovers married, exiled, and buried inside a single afternoon.
How long would the play run if nobody stopped for an interval?
At the pace of Elizabethan delivery, the second quarto’s roughly three thousand lines play in about two hours, the span the Prologue promises for the traffic of the stage, so the first audiences saw a single unbroken sitting, with the feast, the killings, and the tomb arriving inside one act of attention.
That arithmetic has a critical consequence. A play built to be absorbed in one sitting organizes memory differently from a play built in five instalments. The audience of the undivided performance carried the feast in fresh recollection when the tomb arrived roughly ninety minutes later, and the rhymes between the two, the torches, the music, the watching parents, the bodies laid out in a rich interior, registered as echoes within a single experience. A reader working through the numbered grid across several evenings experiences the same rhymes as architecture, planned and symmetric, rather than as echo. Neither experience is false, but they are different plays in the mind, and the difference was created not by the author but by the history of his editors. The divisions taught the play’s audiences to consume it in chapters. The author had composed it as a flood.
The Folio Compromise: Four Headings and a Lapse
The First Folio of 1623, the collected edition prepared by the author’s fellow actors John Heminge and Henry Condell and printed in the Jaggard shop, is where the divisions begin, and its manner of beginning them tells the truth about their origin. The volume as a whole shows a new ambition: Shakespeare presented as a classic author, in a large format, his plays sorted into comedies, histories, and tragedies, many of them furnished with the act divisions that classical editions of Terence had taught readers to expect. The ambition was applied unevenly. Romeo and Juliet, set in type from a copy in the line of the third and fourth quartos, acquired act headings for the openings of its second, third, fourth, and fifth acts, each styled in Latin in the period’s typographic fashion, “Actus Secundus”, “Actus Tertius”, and so on, with an accompanying heading for the opening scene. The first act received none, the Prologue standing where a heading would go. Within the acts, scene headings appear at some scene openings and not at others, with no consistent principle a reader can reconstruct, until the attempt is effectively abandoned and long runs of cleared-stage transitions pass unmarked.
Bibliographers have proposed several agencies for this patchwork. The copy may have been prepared by a scribe who began marking divisions, in the manner of the transcripts Ralph Crane made for other plays in the volume, and stopped. The compositors, Charlton Hinman’s famous analysis of the Folio’s printing distinguishes the workmen by their spelling habits, may have divided where their copy quires divided. The editors of the volume may have issued a general instruction, divide into acts, that each workman executed according to his energy. What no serious account proposes is that the scheme descends from the author, because the scheme’s defects are of the kind authorship excludes: a missing first act heading, scene markings that begin and lapse, breaks that follow the convenience of the page rather than the joints of the action. The Folio’s divisions are best understood as the first, incomplete attempt to dress a playhouse script in the clothes of a classical text. A full study of Romeo and Juliet in the First Folio (1623) traces how the Folio text itself was assembled from the quarto line; the divisions belong to the same assembly, and carry the same mixed authority.
Two later folios, the second of 1632 and the third and fourth of 1663 and 1685, inherited the Folio scheme and its gaps, since each was printed from its predecessor. By 1685 the English reading public had possessed, for more than sixty years, a Shakespeare whose Romeo and Juliet was partly divided, and nobody had completed the work. The completion waited for a new kind of editor and a new kind of edition: the commercial, corrected, modernized Shakespeare of the eighteenth century, sold not as a monument but as a book for readers. When that editor arrived, he treated the Folio’s ragged scheme the way his century treated all inherited untidiness, as an invitation to regularize.
Rowe Regularizes: A Bookseller’s Editor Draws the Whole Grid
Nicholas Rowe’s edition of 1709, “The Works of Mr. William Shakespear” in six volumes from the Tonson publishing house, is the hinge of the whole history. Rowe, a successful playwright turned editor, worked not from the quartos, which his age despised as stolen and corrupt, but from the Fourth Folio, and he brought to the text the instincts of a man of the theatre and a man of letters combined. He supplied lists of dramatis personae. He modernized spelling and punctuation for contemporary readers. And he divided the plays throughout into acts and scenes, completing for every play the scheme the First Folio had applied by fits and starts. For Romeo and Juliet the result was the first full articulation the play had ever carried in print: five acts, each broken into numbered scenes wherever the stage cleared, with a prefatory list assigning each scene its place, a street in Verona, a hall in Capulet’s house, the orchard, Friar Laurence’s cell, Mantua, the churchyard, so that a reader could hold the geography of the tragedy in one view.
The place headings deserve a paragraph of their own, because they show what regularization really does. No early text of the play names the locations of its scenes; the dialogue supplies what little geography it needs, “What, shall this speech be spoke for our excuse, or shall we on without apology?” belongs to a stage that is wherever the speakers stand (1.4, in the street, as the editions say). Rowe’s locations came from inference, from stage tradition, and from the Augustan demand that a printed play read like a classical text with its scene fixed by rule. Once printed, they acquired the authority of the text they merely glossed. Generations of readers have pictured the Capulet orchard and the Mantuan street through Rowe’s list, and editors who discarded his exact words kept his act of naming: every modern edition still heads its scenes with places, and the famous balcony, a word the play never uses for the window at which Juliet appears, entered the play’s furniture through the same editorial and theatrical tradition of naming what the script leaves unnamed. The divisions, that is, did not merely count the play. They furnished it.
Rowe’s scene breaks themselves followed the cleared-stage convention: a new scene begins when the stage empties. The rule sounds mechanical, and in Rowe’s hands it mostly was, but it carried an interpretive payload, because not every cleared stage is the same kind of joint. Where the playwright empties the stage for a breath between two halves of one continuous movement, Rowe’s rule cuts a numbered boundary, and the boundary, once numbered, invites readers to treat the units as comparable blocks. The twenty-four scenes of the modern grid are not twenty-four equal things. Some are vast public movements, the street fight of 3.1, the tomb sequence of 5.3; some are transitions of a dozen lines. The grid flatters them into a false symmetry, and directors and teachers have been fighting or exploiting that symmetry ever since. Still, fairness to Rowe is necessary. He worked quickly, for money, from a corrupt base text, and the amazing fact is not that his scheme contains accidents but that it has proved durable enough for three centuries of scholarship to be conducted inside it. The longer history of editing Romeo and Juliet from Rowe to Malone gives the careers their full span; within the history of the divisions, Rowe is simply the author, the one figure who can be said to have divided the play as a deliberate and complete act.
The Eighteenth Century Refines the Count
Rowe’s successors inherited the grid and quarreled within it, and their quarrels refined the numbering toward the form now universal. Lewis Theobald, whose 1733 edition established collation of the early texts as the editor’s duty, corrected readings against the quartos and folios but kept the Augustan frame. Sir Thomas Hanmer’s handsome Oxford edition of 1743 to 1744 and William Warburton’s edition of 1747 adjusted scene divisions at the margins, working over a scheme neither thought of questioning. Samuel Johnson’s edition of 1765, the most intellectually distinguished of the century, defended the playwright in its famous Preface against the charge of violating the unities of time and place, and treated the machinery of acts and scenes as a convenience of printers rather than a law of drama: Johnson understood, earlier than most, that the numbers were furniture. His edition none the less printed them, because by 1765 a Shakespeare without numbers was unmarketable.
The century’s real technician of the divisions was Edward Capell. Capell’s edition of 1767 to 1768, the fruit of decades of collation of the early copies, applied the cleared-stage rule with a rigour Rowe had not attempted, counting a scene strictly by the emptying of the stage and recording, in his companion volumes of notes, the state of the divisions in every early edition he handled. Where Rowe had divided by instinct, Capell divided by doctrine, and some of the small peculiarities of the modern numbering, the exact points at which scenes begin in the later acts, descend from Capell’s arithmetic. George Steevens, first as Johnson’s collaborator and then in his own editions, brought the quartos into the editorial mainstream and with them a new sensitivity: the early texts’ lack of division could now be seen in the apparatus, even as the page above it grew ever more regular. Edmond Malone’s edition of 1790, the most scholarly of the century, collated systematically, dated the plays, and settled a text whose divisions, refined from Rowe through Capell, became the standard that the great variorum edition of 1821, completed from Malone’s papers by James Boswell the younger, transmitted to the nineteenth century as settled fact.
What does a cleared stage signal in the numbering editors built?
A cleared stage signals a full stop in the practical grammar of the early playhouse: every speaker has exited, the boards stand empty, place and time may reset without scenery, and so eighteenth-century editors, Capell most strictly, counted each refilling of the emptied stage as a new scene and numbered the play by counting its emptyings.
That doctrine, so simple to state, conceals the deepest assumption of the whole system. Counting emptyings presumes that the playwright structured his play as a chain of discrete blocks, each sealed by an exit. But the Elizabethan stage also knew the linked transition, in which one group departs as another enters, the stage never quite empty, the action dovetailing. Where the text offers such dovetails, editors must either split a continuous movement at its nearest emptying or merge across one, and different editors chose differently, which is why editions still differ at a handful of boundaries even while agreeing on the total shape. The doctrine also presumes that emptiness means the same at every point: a cleared stage before the lovers’ parting (the close of 3.5) and a cleared stage before a servants’ interlude (4.4, the wedding preparations played against the feigned corpse upstairs) are counted as identical units of structure, though one is the emotional summit of the tragedy and the other a held breath of dramatic irony. The numbering cannot tell the difference. Readers must, and the grid, by making the two boundaries look alike, quietly discourages them from trying.
From Malone to the Globe Edition: How the Grid Hardened Into Fact
The nineteenth century did not redraw the divisions; it did something more powerful, which was to build the entire infrastructure of Shakespeare study on top of them. The variorum tradition culminating in the Boswell-Malone edition of 1821 printed the divisions with a full record of editorial opinion beneath them, so that a reader could watch Rowe, Capell, and Malone align into consensus. Charles Knight’s pictorial edition, John Payne Collier’s editions, Alexander Dyce’s scholarly text, and Howard Staunton’s all transmitted the same articulation. Then, in 1864, Macmillan published the Globe edition, prepared by Clark and Wright from the Cambridge Shakespeare then in progress, in a single portable volume priced for the general reader. The Globe text became the reference Shakespeare of the English-speaking world. Its act, scene, and line references were adopted by concordances, by the great lexicons, by examination boards, and by the lecture room, until “3.1” ceased to be an editorial address and became simply where Mercutio dies, a fact of the play like his name.
The hardening deserves a moment’s scrutiny, because it happened almost without argument. Nobody in the nineteenth century staged a campaign for Rowe’s numbering; nobody needed to. The grid spread the way standard gauges spread, by the convenience of interoperability. A scholar in Boston and a schoolmaster in Bombay could correspond about the orchard scene only if both called it 2.2, and so both did, and the agreement of the network masqueraded as the discovery of a fact. Meanwhile the stage itself, the nineteenth-century stage of Henry Irving and his contemporaries, cut and rearranged the play freely under cover of the same numbers, restoring scenes, transposing them, building the balcony into architecture, treating the printed grid as a menu rather than a map. Print fixed what performance kept fluid, and the fixity of print, being the fixity scholars cite, won the memory of the culture.
One more layer completed the system in the twentieth century. Charlton Hinman’s Norton Facsimile of the First Folio (1968) and the through line numbering devised for it gave scholars a second, rival address system keyed to the Folio’s own columns, and modern critical editions now commonly cite both: the act, scene, and line of the edited text and, in textual studies, the through line numbers of 1623. The coexistence of the two systems is the healthiest feature of the whole apparatus, because it advertises what the act and scene numbers alone conceal: that every address is an address within an edition, chosen by editors, and that the play itself, the sequence of entrances and speeches, sits beneath all its addresses unchanged.
The InsightCrunch Division Timeline
The table below, the InsightCrunch division timeline, compresses the argument of this article into a single reference: each early witness and landmark edition, the state of the divisions it carries, and what it contributed to the grid readers use. Read down the middle column and the whole thesis appears at a glance: nothing, then something partial, then everything, laid on in three centuries of afterthought.
| Witness or edition | State of the divisions | Contribution to the grid |
|---|---|---|
| Second quarto copy behind the 1599 text | No act or scene divisions | The author’s draft tradition transmits an unmarked sequence |
| First quarto, 1597 | No divisions; two Chorus sonnets only | Shows the play in performance shape, unbroken |
| Second quarto, 1599 | No divisions | The copy-text of modern editions carries no grid to inherit |
| Third quarto, 1609, and fourth quarto, about 1622 | No divisions | The Folio’s quarto copy still unmarked |
| First Folio, 1623 | Latin act headings for Acts 2 to 5; first act unheaded; scene headings sporadic, then lapsing | The first divisions of any kind, added in the printing house |
| Second, third, and fourth Folios, 1632 to 1685 | Folio scheme inherited with its gaps | Sixty years of partly divided Shakespeare |
| Rowe, 1709 | Complete acts and scenes throughout; list of scenes with invented place headings | The full grid and its geography, drawn for the first time |
| Theobald, 1733; Hanmer, 1743 to 1744; Warburton, 1747 | Rowe’s frame retained; readings collated; breaks adjusted | Collation enters editing without disturbing the frame |
| Johnson, 1765 | Frame retained, defended as convenience | The numbers understood as furniture, and printed anyway |
| Capell, 1767 to 1768 | Scenes recounted strictly by the cleared-stage rule | Several modern boundaries fixed by Capell’s doctrine |
| Steevens and Malone, 1773 to 1790 | Quartos collated; text and numbering settled | The standard scheme in its mature form |
| Boswell-Malone variorum, 1821 | Divisions printed with the record of editorial opinion | Consensus displayed as authority |
| Globe edition, 1864 | Rowe-Malone grid in a cheap one-volume text | The citation standard for the English-speaking world |
| Hinman facsimile, 1968 | Through line numbering added for the Folio | A rival address system that exposes the grid as editorial |
| Gibbons, Arden, 1980; Levenson, Oxford, 2000; Weis, Arden, 2012 | Grid retained with textual notes on its origins | Modern editions keep the map and confess its making |
The pattern the timeline exposes is the pattern of the whole editorial tradition in miniature. Authority flows backward: the most finished and regular state of the text is treated as the play’s natural form, and the earlier, messier states are treated as its corruption, when the documentary order is exactly the reverse. The unmarked quartos are the play in its working clothes. The numbered grid is the play in the dress uniform that three centuries of editors tailored for it. Both are worth studying; only one of them was worn in 1595.
Does the Five-Act Shape Fit? The Play’s Own Joints
The complication every honest account must face is the fit. If the five-act scheme were an alien frame forced on the play, its history would be a simple story of imposition, but the scheme is not alien, and the story is not simple. The tragedy does articulate into five movements, as the summary earlier in this article showed, and the movements hinge at points any structural analysis would choose: the feast, the marriage, the killings and banishment, the potion, the tomb. Brooke’s narrative poem, the play’s chief source, already falls into phases, and the dramatist’s compression preserved the phases while quickening them. An editor seeking act breaks in 1709 did not scatter them at random; he laid them along grooves the story itself had cut. Rowe’s breaks feel inevitable because, at the level of the acts, they nearly are.
The scene level tells a different story, and it is at the scene level that the imposition bites. The act breaks coincide with the story’s great pauses; the scene numbers slice the movements into units of wildly unequal weight and pretend the slices are comparable. Worse, the numbered unit trains the reader to expect each scene to behave like a chapter, with its own completeness, when the play’s characteristic unit is the linked sequence: the orchard flowing out of the street search, the wedding flowing into the brawl’s aftermath, the parting at dawn flowing into the forced betrothal. Emrys Jones, in Scenic Form in Shakespeare (1971), made the foundational case that the playwright’s true structural unit was the scene conceived as a shaped stage picture, often paired or patterned against its neighbours, and that reading by numbered blocks obscures the patterns that reading by scenic form reveals. The two-part rhythms Jones traced, preparation and fulfilment, question and answer, public scene answered by private scene, cross the numbered boundaries constantly. The grid is not wrong about where its breaks are; the stage does clear at those points. It is wrong about what the breaks mean. They are joints in a limb, not walls between rooms, and a limb is misread when its joints are treated as separations.
The interval question sharpens the point for the acts themselves. Modern performance almost universally breaks the play once, usually after the marriage or after the banishment, which means the five-act grid is not even used as five: the theatre takes from the editors a single permission to pause and discards the other three pauses the scheme pretends to offer. Directors who have experimented with playing the tragedy entire, in the Elizabethan manner, regularly report the same discovery, that the second half’s acceleration becomes terrifying when the audience is given no recovery from the street fight onward, and that Juliet’s isolation reads as a continuous closing of doors rather than a sequence of episodes. The play’s design, in other words, is legible in both arrangements, but its pressure is legible only in one. A shape that must be suspended for the work’s full effect to register is a shape the work does not contain. The five acts are a reading technology, excellent for finding passages, poor for feeling them, and the mature verdict holds both halves of that sentence without flinching.
Where does the modern numbering put Mercutio’s death, and who fixed it there?
Mercutio dies in the scene modern editions number 3.1, the opening scene of the third act, struck under Romeo’s arm in the street fight; the position was fixed by Rowe’s scene counting in 1709, refined by Capell, and ratified by the Globe edition, though no early text numbers the scene or marks an act anywhere near it.
Notice what the numbering does to that death. As the opening of an act, 3.1 presents itself as a fresh start, the beginning of the tragedy’s second half, and generations of criticism have duly treated the play as a structure balanced on this hinge, comedy before, tragedy after. The hinge reading is powerful and partly true: the play’s weather does change at the killings, and the change is permanent. But the numbered frame exaggerates the balance into symmetry. In continuous performance the street fight arrives not as a new beginning but as the detonation of everything the feast and the marriage loaded into it, Tybalt’s challenge issued in 2.4, the insult festering through the wedding hour, the heat of the day insisted on by Benvolio’s opening speech: “The day is hot, the Capels are abroad, And if we meet we shall not scape a brawl” (3.1, lines 1 to 3). The scene’s power is cumulative, not inaugural. Calling it 3.1 teaches the memory to file it as a commencement; playing it unbroken teaches the body to receive it as a consequence. The difference between those two lessons is the difference the whole history of the divisions finally comes to.
The Scholars Disagree: Architecture, Accident, or Performance?
Three positions divide the scholarship, and the article’s verdict requires choosing among them rather than blending them into mush. The first is Baldwin’s: the five-act structure was the professional dramatist’s inherited architecture, so the printed divisions, however late, recover a shape the author composed to. The second, associated with James Hirsh’s The Structure of Shakespearean Scenes (1981) and with the bibliographical scepticism of W. W. Greg’s generation carried further by later textual critics, holds that the playwright composed in scenes, in continuous sequences of stage action, and that act divisions in the early printed texts are the work of scribes, book-keepers, and printing houses, carrying no authorial authority at all. The third, developed by Gary Taylor and others from patterns of division in the Folio, proposes a middle origin: that some act divisions, in some plays, register genuine performance practice, the intervals with music that London theatres adopted in the early seventeenth century, so that the divisions are theatrical rather than authorial, evidence about staging rather than about composition.
Adjudicated for this play, the second position wins, with a qualification borrowed from the third. The documentary chain is decisive against Baldwin: two quartos without divisions, a Folio scheme that is partial, Latin-headed, missing its first act, and demonstrably irregular, and a complete scheme that waits until 1709. Architecture that has to wait 114 years for its first complete printing is not architecture recovered; it is architecture assigned. The qualification concerns what the assigned architecture caught. Editors did not impose the five breaks on a formless text; they found a text whose cleared stages clustered at real joints, the ends of the feast sequence, the marriage sequence, the catastrophe sequence, the potion sequence, and so the imposed frame and the story’s own articulation coincide at act level to a degree they do not for every play. Taylor’s performance intervals, meanwhile, are chronologically possible for the Folio’s preparation but impossible for the play’s first life: an interval practice of the Blackfriars era cannot describe performances at the Theatre in 1595, and the Folio’s headings, whatever they record, record at most a revival practice of the author’s old age or after his death. The divisions of Romeo and Juliet are therefore triply removed from the writing desk: assigned by editors, ratified by the market, and at best distantly connected, through the Folio, to performance habits the play’s first performances did not have.
Greg’s own role in this history is double, and the doubleness is instructive. As the leading bibliographer of his generation, Greg established the methods by which the Folio’s printing, its copy, and its compositors could be studied, and those methods are what allow the patchwork character of the Folio’s divisions to be demonstrated rather than asserted. Yet Greg the editor also worked comfortably inside the inherited grid, as every editor must, because an edition without the standard addresses cannot be used alongside any other scholarship. The grid’s survival owes less to anyone’s belief in it than to this professional interlock: each scholar’s private scepticism is outweighed by the public convenience of the shared map. Recognizing the interlock is not cynicism. It is the precondition for using the map intelligently, citing the numbers while knowing their provenance, as a sailor uses a chart whose coastline was surveyed by dead men with imperfect instruments and is none the less, for most purposes, right.
One further scholarly voice belongs in the adjudication because it reframes what a division is. Tiffany Stern’s research on early modern rehearsal and the making of plays has shown how much of a script’s life happened in parts, in actors’ sides containing only their own lines and cues, in plots posted backstage that summarized the sequence of scenes for the company. The backstage plot, the document by which a performance was actually coordinated, divided the play into its entrances, scene by scene in the practical sense, and survives for a handful of plays of the period. No plot survives for Romeo and Juliet, but the genre of the document matters here: the company’s own articulation of the play was a running list of stage events, closer to the quartos’ unbroken sequence than to Rowe’s grid, keyed to who enters and what property is wanted. The play’s first real divisions, in the only form its first performers used, were a stage manager’s checklist. Print replaced the checklist with a classicizing grid, and literary history has been reading the grid ever since as if it were the checklist’s meaning. It is not. It is its tombstone and its monument in one.
Prologues, Choruses, and Other Things the Grid Cannot Count
A scheme is best tested at its edges, and the divisions of this play have instructive failures at theirs. The Prologue stands outside the system entirely: fourteen lines of sonnet that no edition numbers as a scene, prefixed to an act that the Folio forgot to head. Is the Prologue part of Act 1? Editions answer by typography rather than argument, printing it on its own footing and citing it by its own name, an admission that the grid has an unnumbered antechamber. The second Chorus is the mirror case. It is, in substance, the same kind of sonnet, and it performs the same summarizing office, yet the grid must put it somewhere, and where it puts it has wandered: heading the second act in most modern editions, folded into the close of the first in a few older ones. A text that fit its frame would not leave its two most formal passages straddling the frame’s joints.
Why does Act 2 open with a Chorus instead of a character in modern editions?
Act 2 opens with a Chorus in modern editions because the second Chorus sonnet stands in the early texts between the street scene of the masque’s departure and Romeo’s return to the Capulet orchard, and editors, counting that cleared stage as the seam between acts, placed the sonnet at the head of the new act as its prologue in miniature.
The wandering of the second Chorus exposes the deeper truth that the act break itself, the most consequential division in the tragedy, the one modern theatres often use for the interval, rests on nothing firmer than an editor’s reading of a sonnet’s position. Brooke’s poem pauses nowhere here; the quartos pause nowhere here; the sonnet looks backward over the feast and forward over the marriage plot in a single gesture, “Now old desire doth in his deathbed lie, And young affection gapes to be his heir” (Chorus 2, lines 1 to 2), and its gesture is Janus-faced precisely because the action at that point is continuous growth, not recommencement. To head an act with it is to mistake a summary for a threshold. Yet the mistake, once made by Rowe and ratified by everyone since, has structured real experiences of the play for three centuries, and an honest textual history must record both facts: the break is editorial, and the break is now part of the play’s received form. Received form is a genuine category. It is simply not the same category as authorial form, and the confusion of the two is the error this article exists to correct.
Smaller edges tell the same story. The wedding scene, in which the Friar marries the lovers, is the shortest numbered scene in the tragedy relative to its consequence, a few dozen lines between the orchard consent and the street catastrophe: “So smile the heavens upon this holy act, That after-hours with sorrow chide us not” (2.6, lines 1 to 2). The grid gives it a full scene number, 2.6, equal in dignity to the vast tomb sequence 5.3 that runs nearly three hundred lines, and the equality is a pure artefact of counting emptyings. Or take the discovery sequence of Act 4, where the Nurse finds Juliet feigned dead (4.5), a scene of lamentation whose lines the early texts arrange without the regularity the editors impose, some speeches in the quartos being laid out as prose or shared verse that the grid silently regularizes. At every edge the pattern repeats: the play’s own articulation is rhythmic and proportional, and the grid’s is arithmetic. Arithmetic is easier to cite. Rhythm is what the author built.
How Modern Editions Present the Divisions
A modern scholarly edition of the play is a negotiation between the grid its readers expect and the documents its editor has sworn to represent. Brian Gibbons’s Arden edition of 1980, based on the second quarto, prints the standard divisions and records in its textual apparatus where the Folio’s headings stand and where they are absent, so that the reader who consults the notes can watch the grid being laid over the copy-text in the act of using it. Jill Levenson’s Oxford edition of 2000, the most fully documented modern text, takes the same double approach to its logical end: the reading page carries the familiar numbers, while the introduction and commentary reconstruct the early texts’ unmarked state and the editorial centuries that numbered it. Jill Levenson and the Oxford Edition receives its own study in this series; for the history of the divisions, Levenson’s edition is the model of the honest compromise, a grid worn openly as a grid. René Weis’s Arden edition of 2012 continues the practice into the present generation of scholarship, retaining the standard articulation while treating the play’s early textual state with a fidelity that earlier centuries, armed with the same grid, rarely matched.
The compromise has rules, and knowing them converts a passive reader into an active one. Divisions that stand in the edition’s copy-text or in the Folio are printed as part of the text; divisions supplied by the editor are conventionally recorded in the apparatus and, in some editions, flagged in the text itself by brackets or by notes at the scene head. Location headings are universally editorial, however immemorial they feel: when an edition heads the orchard scene “Capulet’s orchard”, the words are Rowe’s legacy, not the quarto’s. Line numbering within scenes, the third element of the modern citation, is likewise the edition’s own count, which is why the same speech can carry slightly different line addresses in different editions and why scrupulous citations name the edition used. Students assembling the evidence of this article for a seminar presentation, laying the InsightCrunch division timeline out row by row as slides through the ReportMedic PPT Viewer, are doing in small what Levenson does in large: keeping the map and its makers in the same field of view.
Performance texts stand at the opposite pole. Acting editions and production scripts renumber freely, cutting scenes, merging them, restoring the Chorus to the interval, because the stage answers to the evening rather than to the archive. The freedom is legitimate and even Elizabethan in spirit, and it throws the scholarly fixity into relief: the grid is a library convention, and the library and the theatre, which shared the play for its first century without any numbers at all, have simply agreed to remember it differently. Neither memory cancels the other. The historian’s task is to keep the two accounts distinct, and the reader’s privilege is to move between them, now studying the tragedy as the numbered classic it became, now imagining it as the unbroken afternoon it was.
The Citation Culture: How Numbers Taught Readers to Remember
It would be a mistake to treat the grid as a mere finding aid with no effect on interpretation, because finding aids shape what is found. A criticism organized by numbered scenes favours the scene as the unit of thought: essays are written on 2.2, on 3.1, on 5.3, and the units accumulate small orthodoxies, the balcony scene as lyric idyll, the street fight as hinge, the tomb as machine of ironies. Each orthodoxy contains truth, and each is partly the creature of its frame. The balcony scene, isolated as 2.2 and anthologized as a jewel, loses in isolation the quality it has in sequence, of an answer improvised under pursuit, with Mercutio’s mockery still ringing from the street below the wall: Romeo has climbed into the orchard straight out of his friends’ bawdy conjuring, and the scene’s holiness is asserted against that noise, in the hearing of the audience if not of Juliet. The grid cannot hold that context, because the context is precisely the boundary the grid draws. Teachers fight the effect with throughlines and scene-linking exercises; productions fight it by overlapping transitions; but the printed page, where most readers meet the play, keeps issuing its quiet instruction to read in blocks.
Citation also disciplines memory across the culture. “2.2” and “3.1” function as the play’s postal districts, known to people who have never opened a scholarly edition, reproduced in film credits, programme notes, quiz questions, and the titles of artworks. A reference system that reaches so far beyond its makers has become a cultural fact in its own right, and the historian of the divisions must grant it the same respect granted to any other part of the play’s reception. The numbers are wrong as a claim about 1595 and indispensable as a fact about every year since 1864. Scholarship lives inside that doubleness without embarrassment. What it cannot afford is amnesia: the moment the grid’s origin is forgotten, an editorial convenience hardens into a superstition about the author’s intentions, and arguments begin to be built, as they demonstrably have been built, on the supposed significance of act structures that the author never numbered and scene counts he never made. The prophylactic is the history itself, which is why an article-length account of who divided the play, long overdue in the reference literature, is not pedantry but hygiene.
Other Systems the Editors Rejected: Entrances, French Scenes, and Lines
The grid that won was not the only grid available, and the rejected alternatives clarify what the winning system assumes. One alternative counts by entrances. French classical practice, codified in the theatre of Corneille and Racine and analyzed by their commentators, divided a play into scenes at every entrance or exit of a significant character, whether or not the stage emptied, so that a scene was a configuration of persons rather than a filling of space. Under French counting, Romeo and Juliet would splinter into far more than twenty-four units: the orchard scene alone would divide at the Nurse’s call within, at Juliet’s exits and re-entries at the window, at each shift of the speaking group. English editors knew the French system and declined it, partly from insularity and partly from practical sense, since the cleared-stage rule could be applied mechanically to any old text by a compositor, while French counting demands a judgement at every door. The choice had lasting consequences. A configuration-based count would have made the play’s scenes into studies of grouping, foregrounding who is on stage with whom, which happens to be one of the tragedy’s real structural languages, the lovers perpetually reconfigured against parents, friends, friar, and prince. The space-based count the editors chose foregrounds instead the rhythm of emptying and filling. Both are defensible instruments. Only one became invisible through familiarity, and its invisibility is the thing to correct.
A second alternative is the line itself. The Globe edition’s innovation, completed by its successors, was to number lines within each scene, creating the three-part address that every citation now uses. Before that standardization, references ran by page and signature, by quotation, or by the rough geography of “toward the end of the fourth act”, and scholarship was correspondingly harder to verify. The line numbers completed what the scene numbers began: the total discretization of the play, its conversion into a fully addressable object in which any ten words can be summoned by a coordinate. Bibliographers of a later generation added the Folio’s through line numbers, and textual scholars cite those when the Folio itself is the subject, a standing reminder that the edited grid and the historical document are different objects. No one proposes abolishing the coordinates. The proposal of this article is narrower and, in the end, more useful: that the coordinates be taught with their birthdays attached, so that a student citing 3.1 knows the address was assigned in the age of Queen Anne, settled in the age of Malone, and universalized in the age of steam printing, and reads the scene the more attentively for knowing that its frame is younger than the steam engine’s grandfather.
What the Author’s Own Articulation Actually Looks Like
Strip the numbers away and the play does not dissolve into formlessness; it reveals a different and more interesting articulation, built from devices the early texts do transmit. The first device is the Chorus pair, two sonnets framing the love story’s launch and its turn toward marriage, prose of the gods in verse of the strictest form. The second is the cleared stage used as punctuation of variable force: a full stop after the catastrophe of the street fight, when the stage empties behind the Prince’s judgement, “Let Romeo hence in haste, Else, when he is found, that hour is his last” (3.1, lines 195 to 196), and a mere comma at the dovetailed transitions of the orchard night. The third is scenic rhyme, the patterning Jones taught readers to see: feast answered by tomb, the Friar’s cell answered by the apothecary’s shop, two poor men’s rooms where the plot’s poisons and remedies are dispensed, the orchard’s vertical longing, window above and lover below, answered by the monument’s vertical horror, Juliet above on her bier and Romeo below among the dead. The fourth is the alternation of public and private pressure, great street and chamber scenes succeeding one another so that the feud and the marriage breathe against each other. None of these devices needs a numeral. All of them are damaged, in different degrees, when the numbered grid trains attention on block boundaries instead of on pattern.
What marks a new scene in the second quarto if nothing on its page is numbered?
In the second quarto a new scene is marked only by the clearing of the stage and the next entrance direction, so that the reader follows a chain of entrances and exits, with place and time supplied by the dialogue alone, and the modern twenty-four scenes exist as twenty-four unlabelled emptyings and refillings.
This is the sense in which the editors’ work was genuinely editorial rather than merely clerical. The emptyings were there, in the author’s stage directions and the logic of his entrances, and the editors counted something real. What they added was everything that makes a count into a system: the numbers themselves, the act framework the numbers hang on, the location headings that fix in print what the dialogue left fluid, and the presumption, absorbed by every reader since, that the counted units are the play’s anatomy rather than one legitimate dissection of it. Anatomy is the right metaphor, and it cuts both ways. A dissection reveals structure that the living body hides; it also stills the motion that was the life. The undivided quarto is the living body, hard to cite and impossible to lecture from. The numbered edition is the anatomy atlas, indispensable, exact, and dead on the page unless the reader restores the pulse. The verdict of this article is that scholarship needs the atlas and should never mistake it for the body, and that the whole history, from the Jaggard shop’s four Latin headings to the Globe edition’s universal coordinates, is the story of the atlas being drawn.
The Grid on Stage: What Directors Do With the Numbers
The theatre’s dealings with the divisional inheritance are the practical test of its authority, and the theatre’s verdict is unambiguous: the numbers are a convenience to be used and discarded nightly. Nineteenth-century actor-managers, working from acting editions that preserved the grid in print, rearranged the play behind the proscenium with a freedom no editor would have dared on the page, transposing scenes, cutting the Rosaline material to a whisper, expanding the tomb sequence into a scenic apotheosis. The twentieth century’s historically minded directors went further and made the grid itself a directorial decision: to play with one interval or two, to restore the Choruses to their quarto positions as hinges rather than thresholds, to run the street fight and the parting and the betrothal crisis of Act 3 as the single escalating sequence the quartos describe. Each such decision is an implicit textual criticism. A director who refuses the conventional interval after Act 2 is asserting, with the company’s bodies as the evidence, that the break is Rowe’s and not the play’s; a director who restores a pause before the tomb is asserting the contrary, that three centuries of received form have themselves become a kind of authority. Both are right about different objects, the document and the tradition, and the confusion of those objects is, once again, the error the history exists to dispel.
Schools stage the same negotiation in miniature. The examination syllabus, built on the grid, asks candidates to write on numbered scenes, and candidates oblige, learning the play as a cabinet of labelled drawers. The label is a genuine aid to memory and reference, and no reformer proposes teaching the play from an unnumbered facsimile. But the label carries its smuggled lesson, that the drawers were built by the cabinet’s maker, and the lesson survives into adult reading, where it does its quiet damage in programme notes and reviews that attribute to the playwright the architecture of his editors. Against that damage the remedy is not abolition but annotation: a single sentence in any programme or edition, stating that the divisions are editorial and the play was written to run unbroken, restores the true proportion of the thing at a cost of twelve seconds. That so few programmes print the sentence is a measure of how completely the grid has naturalized itself. Naturalization, in textual matters as in others, is what power looks like when it has finished its work.
The Twentieth Century Looks Behind the Grid
For most of the grid’s history, its users could not easily inspect what lay beneath it, because the early texts were rare books in great libraries. The twentieth century changed the conditions of knowledge, and the change slowly altered the status of the divisions from fact to exhibit. The bibliographical revolution associated with A. W. Pollard, W. W. Greg, and R. B. McKerrow taught scholars to treat every early printed book as a manufactured object with a recoverable biography, and under that discipline the Folio’s ragged headings stopped looking like the author’s architecture and started looking like evidence about a printing house. Hinman’s mechanical collation and his analysis of compositor stints completed the demystification for this play’s Folio text: once a reader knows which workman set which pages and how the copy reached him, the Latin act headings take their place among the other accidents of manufacture, datable, explicable, and entirely human.
A second movement pressed further. Leah S. Marcus’s Unediting the Renaissance (1996) gave a generation of critics the word and the method for reading early texts against their edited descendants, arguing that the smoothing work of editors, of which regularization of divisions is a textbook instance, systematically effaces the local and historical character of the works it transmits. Randall McLeod’s essays, written under shifting pseudonyms in a typographic style that enacts their argument, displayed the material page of the early editions against the abstractions of the edited text, and for the divisions the display is devastating in its simplicity: the quarto page has no numbers on it, and everything the number claims must therefore be argued, not assumed. The Oxford Complete Works of Stanley Wells and Gary Taylor (1986), though it retained the standard articulation for practical citation, embodied the new scepticism in its whole editorial policy, printing texts that advertise their documentary bases and, in the famous case of King Lear, refusing to merge two early states into a single conflated reading text. The lesson of that refusal travelled: an edited text is a construction, and its furniture, divisions included, belongs to the construction.
The third change was technological, and it is still completing itself. Photographic facsimiles, and then digital facsimiles of the quartos and the Folio held by the great libraries, put the unmarked pages in front of any reader with a connection, so that the primary evidence of this article, the bare quarto page running from the brawl to the tomb without a numeral, can be checked in minutes by a student who will never enter a rare book room. The effect on the divisions has been paradoxical and healthy. The grid has not weakened; citation still requires it, and digital editions encode it more thoroughly than print ever did. But the grid’s invisibility has ended. A reader can now hold the number and the unnumbered page in view at once, can see 3.1 as a label pasted on a passage that flows unbroken out of the wedding scene in the 1599 quarto, and the pasted label, seen as pasted, loses its false antiquity while keeping its use. That double vision is exactly the condition this article has argued for throughout: the map honoured as a map, the territory restored to its own weather. The twentieth century did not repeal Rowe. It taught his readers to see his hand, and seeing the hand is enough.
The Verdict: Cite the Numbers, Doubt the Numbers
The history is now complete enough to judge. The act and scene divisions of Romeo and Juliet were not written by William Shakespeare, whose surviving texts, in both their quarto states, carry none. They were begun, irregularly and anonymously, by the makers of the First Folio in 1623, who headed four acts and let the scheme lapse; completed by Nicholas Rowe in 1709, who numbered every scene and invented the play’s printed geography; refined by the eighteenth-century editors, above all by Edward Capell’s doctrine of the cleared stage and Edmond Malone’s settling scholarship; and universalized by the Globe edition of 1864, whose addresses the whole citing world adopted. The five-act frame they hang on descends from Horace’s rule and the Terentian commentary tradition, and it fits the tragedy’s story because the story’s own phases, inherited from Brooke and sharpened in compression, fall obligingly into five movements. The fit is real, and it is not authorship. It is the luck of a good story meeting a ready-made frame, and the editors who exploited the luck deserve credit for the exploitation rather than the credit, so often assigned by default, for the design.
How, then, should the divisions be regarded? As the most successful editorial intervention in the play’s history: a finding system of genius, a teaching instrument of proven power, and a falsification only when it is mistaken for the thing it indexes. The scholar’s duty is double and slightly paradoxical, to use the grid with complete fluency and to remember, at the moment of use, that it is a grid. Cite 3.1 for Mercutio’s death, because no other address will find the passage for a reader; and know that in citing it one cites Rowe’s arithmetic and Capell’s doctrine and the Globe edition’s reach, laid over a moment that the first audience experienced as the unbroken consequence of a wedding an hour old. The play survives both descriptions. It was built, by a writer working for a daylight stage and a standing crowd, to be taken in one breath, and it has been read, for three hundred years, in numbered chapters. The chapters are excellent. The breath was the author’s.
What follows the verdict in this article is application rather than argument: a walk through the counted grid, scene by scene, and a pair of worked cases from the editor’s desk, testing the general history against the particular boundaries where it does its work. Readers who keep the verdict in mind through those applications will find the numbers changing character under their eyes, from facts of the play into entries in a ledger of editorial decisions, each with its date and its maker’s name, which is the condition in which the whole apparatus deserves to be used.
Walking the Counted Grid: Twenty-Four Scenes Under the Lens
The surest way to see what the numbering does is to walk the play with it in hand, act by act, watching where the count falls easily and where it strains. The first act counts cleanly. Five scenes, five full changes of place, street to house to street to hall to street, each sealed by a general exit, and the feast at the centre blazing with the lovers’ first exchange, the shared sonnet that begins “If I profane with my unworthiest hand” (1.5, line 93). Rowe’s arithmetic and the play’s geography coincide here almost perfectly, because the opening movement is genuinely episodic: the playwright is establishing a city, and he establishes it in blocks. If the whole play behaved like its first act, the grid would be a discovery rather than an imposition, and much of the confidence editors have always felt in their scheme comes from this obliging first act.
The second act is where the counting starts to editorialize. Six scenes are counted, but they are not six journeys; they are one night. The Chorus dismisses the old desire and ushers in the new, Romeo scales the orchard wall, is conjured and mocked from the street, speaks with Juliet at the window through the long central scene, goes at dawn to the Friar’s cell, endures the street banter and the Nurse’s embassy at noon, and is married in the cell by afternoon. The cleared stages are real, the stage does empty between the orchard and the cell and the street, but the emptiness is the emptiness of a held breath inside a single accelerating design, night deepening into morning into noon into the wedding hour. The grid numbers the breaths. It cannot number the acceleration, which is the scene-sequence’s actual content, and readers who meet the act as six labelled boxes consistently underrate its speed. Levenson’s commentary is alert to the continuity, printing the numbers while reminding the reader that the action pours.
The third act counts five scenes and contains the play’s most consequential editorial illusion, treated earlier in this article: the presentation of the street fight as an act’s opening, a fresh movement, when the sequence makes it a detonation. The act then runs the Prince’s judgement, the cell scene of Romeo’s despair, “Art thou a man? Thy form cries out thou art” (3.3, line 119), the dawn parting, and the Capulet confrontation, in a tightening spiral that the numbering presents as four further boxes. Observe the strain at the act’s end: the confrontation scene (3.5) contains within one numbered unit the lovers’ parting at the window, the mother’s announcement of the Paris match, the father’s storm, and the Nurse’s betrayal, four movements that a configuration-based count would separate and that even cleared-stage counting might split, were the stage ever to empty. It does not empty, and so the grid files one of the play’s most violent accelerations under a single number, while granting the twelve-line wedding preparations of 4.4 a number of their own. The arithmetic is faithful to its rule and faithless to the play’s proportions, which is the definition of a system doing exactly what it was built to do and no more.
The fourth act counts five scenes around the feigned death, and here the grid’s weaknesses and the playwright’s dovetailing meet head-on. The Friar’s proposal (4.1), Juliet’s consent and soliloquy (4.3), “Farewell! God knows when we shall meet again” (4.3, line 14), the household’s wedding bustle (4.4), and the discovery of the body (4.5) follow in story-time across some forty hours, yet in stage time they pour without pause, the wedding preparations played for savage irony against the corpse upstairs, the discovery turning bustle to lamentation in a breath. Rowe counts five; the audience experiences one long closing of the trap. The fifth act counts three, the fewest, for the reason given in the questions below: catastrophe does not stop to change the scene. Mantua, the cell, the churchyard, and the churchyard scene alone, 5.3, holds Paris’s death, Romeo’s great final speech over the body, “Thus with a kiss I die” (5.3, line 120), Juliet’s awakening and suicide, the watch, the Friar’s narrative, and the reconciliation of the houses. In the 1599 quarto all of it is one unbroken press of entrances upon a stage that never clears until the Prince dismisses the survivors. The number 5.3, pasted over that press, is the grid in its purest form: accurate as an address, silent as an account. Walking the count in this way, act by act, is the discipline this article recommends to every reader who cites it. The numbers are good servants of reference. They are poor masters of attention, and attention, in the end, is what the play was written to command.
At the Boundaries: Two Worked Cases From the Editor’s Desk
Abstract doctrine becomes concrete at the boundaries editors actually fight over, and two cases from this play show the whole system in operation. The first is the junction of the third and fourth acts. In the quartos, the Capulet confrontation closes with Juliet alone, resolved on the Friar, and the next entrance brings Paris to the Friar’s cell to arrange the wedding: two scenes in the modern count, 3.5 and 4.1, in two places, on two days of the story’s clock. Rowe needed an act break somewhere in the tragedy’s second half, and the Folio’s heading for the fourth act stood at just this seam, the one place in the later play where the early printing offered him a ready-made division. The break he ratified there has shaped interpretation ever since: it casts the potion plot as a fresh movement, a fourth act with its own machinery, and it invites the interval that many productions take at this point. Yet the seam is a fold in the story rather than a joint in the action. Juliet leaves her father’s house for the cell in a single continuous resolve; her soliloquy before drinking (4.3) is already implicit in her exit line of 3.5, and the catastrophe’s tempo depends on no recovery being granted between the threat of the Paris match and the device that answers it. Editors keep the break, because the Folio offers it and three centuries endorse it, but the best commentaries warn the reader that the play’s own pulse does not pause there. A boundary can be documentary and misleading at once; this one is the type specimen.
The second case is smaller and, for that reason, purer: the division of the fourth act’s discovery sequence. The wedding preparations (4.4) run barely a dozen lines of bustle, the Capulet household at dawn, music ordered, logs fetched, the Nurse sent to wake the bride. The discovery (4.5) follows without a cleared stage in any meaningful dramatic sense, the Nurse passing from her errand to the bedchamber within the same unbroken stage event, and yet the count separates the two, because the household exits toward the feast as the Nurse proceeds to the chamber and the place, by the editors’ convention, has changed. Capell’s doctrine ratified the split; some modern editors have noted the strain, a scene number granted to a transition shorter than many speeches, while the vast press of 5.3, with its five deaths and a civic reconciliation, counts as one. The pair of cases, the act break that folds a continuous resolve and the scene break that severs a continuous minute, shows the grid’s arithmetic operating exactly as designed and exactly as the play does not. No editor is at fault: the rule must be applied uniformly or it is nothing. The fault lies only in the reader who supposes that uniform arithmetic yields uniform units, that 4.4 and 5.3 are the same kind of thing because both are scenes. They are the same kind of thing the way a comma and a chapter are the same kind of thing, both marks on a page, and the history of the divisions is, at its most practical, the history of learning to tell such marks apart.
One further boundary deserves record because it concerns the play’s first moments. The street brawl that opens the action is preceded, in every witness, by the Prologue, and the editions must decide whether the brawl is the first scene of the first act or whether the act, unheaded in the Folio, begins in some sense with the Chorus. The universal solution, brawl as 1.1 with the Prologue unnumbered before it, is Rowe’s tidiness again, and it conceals a genuine oddity: the most formally patterned passage in the play, a perfect sonnet spoken by a figure outside the fiction, frames an act that its first printers forgot to declare. Scholars citing “the Prologue” rather than any number perform, without noticing, the one act of citation in the whole system that remains documentary. The exception proves the argument of this history better than any rule could.
The worked cases converge on a single practical lesson about how the apparatus of a modern edition should be read. Every boundary in the grid is a decision, and decisions have dates, makers, and reasons, all recoverable: the act break before the cell scene descends from the Folio’s heading at that seam, ratified by Rowe because his scheme needed a hinge where the Folio offered one; the severing of the discovery from the wedding bustle descends from Capell’s doctrine, applied with a rigour its author would have defended as fidelity; the unnumbered Prologue descends from nobody’s decision at all, a survival the system never digested. Levenson’s Oxford edition shows the mature way to handle such an inheritance, printing the standard numbers for use while keeping their provenance visible in the textual notes, so that the reader is never more than a glance away from the knowledge that a given boundary is Folio, Rowe, Capell, or the settling hand of Malone. An edition that hides that provenance teaches a falsehood by silence; an edition that displays it turns its own apparatus into a short history of editing. The grid, handled this way, stops being a superstition and becomes what it should always have been, a dated and signed instrument, like a chart that carries its surveyor’s name and the year of the survey in the margin for all to see.
Frequently Asked Questions
Q: What is the InsightCrunch division timeline, and what pattern does it expose?
The InsightCrunch division timeline is the table in this article that lists every early witness and landmark edition of the play, the state of its act and scene divisions, and its contribution to the modern grid. The pattern it exposes is a three-century gradient: the quartos of 1597 and 1599 carry no divisions at all, the First Folio of 1623 adds four Latin act headings and a scattering of scene headings before lapsing, Rowe completes the whole scheme in 1709 with invented place headings, Capell and Malone refine the counting, and the Globe edition of 1864 freezes the result into the universal citation system. Read in sequence, the rows show authority flowing backward, with the most finished and regular state of the text mistaken for its original form, when the documents prove the opposite: the unmarked texts are the earliest, and every number is an accretion.
Q: Why does the First Folio print no heading before Act 1?
The First Folio prints no act heading before the first act because its divisions were improvised during manufacture rather than designed in advance. The compositors, working from quarto copy that carried no divisions, began marking acts only at the first great seam they reached, the opening of the second act, and headed the third, fourth, and fifth acts as they came to them. A scheme drawn up beforehand would have headed all five acts uniformly, as every later edition does. The gap at the head of the play is therefore the clearest internal evidence that the Folio’s articulation is a printing-house afterthought, the record of breaks noticed in passing, and not the transmission of an authorial plan. Rowe silently repaired the omission in 1709 by supplying the missing framework, and the repair was so successful that most readers never learn there was anything to repair.
Q: What did Nicholas Rowe add to the play’s scene divisions?
Rowe added the complete scene division of the play, together with the geography that still frames it. Editing from the Fourth Folio in 1709, he broke the whole text into numbered scenes wherever the stage cleared, applied the five-act framework uniformly from the first act onward, and prefixed a list of the scenes with the place of each: the street, the Capulet house and orchard, Friar Laurence’s cell, Mantua, the churchyard. No early text names these places scene by scene; Rowe inferred them from the dialogue and from stage tradition. His other additions, dramatis personae lists and modernized accidentals, have been revised away by later editors, but his divisions and his locations proved permanent, surviving in essence in every modern edition, because they answered a reader’s need that no successor scheme has answered better.
Q: How did Edward Capell change the way the play’s scenes were counted?
Capell changed the counting from instinct to doctrine. Rowe had divided where a scene seemed to end; Capell, in his edition of 1767 to 1768, applied the cleared-stage rule with systematic rigour, treating every complete emptying of the stage as the boundary of a scene and recording the divisional state of the early editions in his notes with unprecedented care. The effect was to fix several boundaries at the points where modern editions still place them and to give the numbering a defensible rationale that Rowe’s practice lacked. Capell’s example also established the expectation that an editor should be able to show, from the early copies, why every break stands where it does. Where the stage dovetails, one group entering as another leaves, his doctrine still forces a judgement, which is why a handful of boundaries continue to differ slightly between editions.
Q: Why is the balcony scene numbered 2.2 in every modern edition?
The balcony scene is numbered 2.2 because editors count the cleared stages of the second act in order: the Chorus stands unnumbered at the act’s head, Romeo’s solitary entrance and his friends’ search form the first counted scene, and his entry into the Capulet orchard, with the stage reset to night and a single figure, begins the second. The number therefore records nothing about the scene’s content, its fame, or its author’s intention; it records its position in a sequence of emptyings first counted by Rowe. In the second quarto the passage follows the street search without any mark of division, and the two units the editions separate, Romeo mocked by Mercutio and Romeo answering Juliet’s window, play as one continuous night movement. The universal agreement on “2.2” is an agreement among editors, transmitted through the Globe edition’s citation standard, not a fact transmitted from the 1590s.
Q: Did the first audiences watch the play in five acts or in one continuous sweep?
The first audiences watched it in one continuous sweep. The play was written for the open-air amphitheatres of the middle 1590s, where performance ran by daylight without intervals, and its own Prologue promises “the two hours’ traffic of our stage”, a single spell of stage business. Act intervals with music belonged to the indoor hall playhouses, and their migration into the public repertory came after the King’s Men acquired Blackfriars in 1608, more than a decade after this tragedy was new. The five-act pattern of the printed texts therefore cannot describe the experience of the audiences who first saw the lovers married and buried inside one afternoon. Modern productions that insert a single interval are compromising between the play’s original tempo and modern habit, and productions that play it entire are recovering, in this one respect, the conditions of its first life.
Q: What classical rule demanded that a play have five acts?
The rule is Horace’s, from the Ars Poetica: a fable should be neither shorter nor longer than five acts, “neve minor neu sit quinto productior actu” (lines 189 to 190). Renaissance readers met the rule reinforced from every side. Terence’s comedies circulated in school editions furnished with five-act divisions and with the commentary tradition of Donatus and Evanthius explaining the fivefold articulation of a plot; Seneca’s tragedies, the prestige model for serious drama, were printed in five acts; and Italian critics of the sixteenth century hardened the precept into law. A grammar-school education saturated the period’s dramatists and editors alike in this framework, which is why the five-act expectation was available, fully formed, when printers and editors later looked at an undivided play and saw something unfinished. The rule describes an ideal of classical construction. It does not describe the documents of this play’s early life.
Q: Why are some of the Folio’s act divisions for this play missing or misplaced?
They are missing or misplaced because they were added during the manufacture of a large book under commercial pressure, by several hands working from undivided copy. The First Folio’s Romeo and Juliet was set from a quarto in the line of the third and fourth quartos, neither of which carries divisions, so every heading had to be invented at the point of setting. The invention was begun and not finished: the second through fifth acts received Latin headings, the first act none, and scene headings were supplied at some openings and then abandoned. Bibliographers attribute the patchwork variously to a scribe who prepared copy incompletely, to compositors dividing where their stints divided, and to the absence of any single supervising instruction. The irregularity is itself the evidence. An authorial scheme transmitted through copy does not lapse; a scheme improvised in a printing house does, and this one demonstrably did.
Q: When did act and scene numbers become the standard way to cite the play?
Act and scene numbers became the standard through a long ratification rather than a single act of adoption. Rowe’s complete scheme of 1709 gave readers a citable grid; the eighteenth-century editors refined it; the Boswell-Malone variorum of 1821 printed it with the record of editorial consensus; and the Globe edition of 1864, cheap, portable, and everywhere adopted, turned the refined grid into the common reference system of the English-speaking world. Concordances, lexicons, and examinations then built on the Globe addresses, and the network effect completed the standardization: everyone cited the numbers because everyone else did. Through line numbering, devised for the Folio in the twentieth century, added a second, documentary system for specialists, but the act, scene, and line address remains the public standard, roughly two and a half centuries after Rowe first drew the whole map.
Q: What is a French scene, and does Romeo and Juliet use that system?
A French scene is a unit of the classical French theatre counted not by the clearing of the stage but by the configuration of characters: a new scene begins whenever a significant character enters or leaves, even if others remain. Under French counting, a play’s scenes multiply, since every change in the speaking group starts a fresh count. Romeo and Juliet, as edited in English, does not use that system; its twenty-four scenes are counted by the English cleared-stage rule that Rowe applied and Capell codified. The contrast is instructive because the French method would foreground one of this play’s real structural languages, the perpetual regrouping of the lovers against parents, friends, the Friar, and the Prince, while the English method foregrounds the rhythm of emptying and filling. The choice between the systems was a choice about what a scene is, and it was made by editors, not by the playwright.
Q: Why does Act 5 have only three scenes while Act 1 has five?
The uneven counts arise because scenes are counted by cleared stages, and the play’s late action clears the stage far less often than its early action. The first act crosses Verona repeatedly, street, house, street, hall, street, emptying and refilling at every move, and so accumulates five counted scenes. The fifth act drives forward in three long movements: Romeo in Mantua receiving the false news and buying the poison, the Friar discovering the miscarried letter, and the great tomb sequence, which holds the stage continuously through the deaths of Paris, Romeo, and Juliet and the reconciliation over their bodies. The playwright’s tempo, accelerating and widening, produces the arithmetic; the editors merely recorded it. The imbalance is thus one of the grid’s more honest features, a place where the counting registers a real property of the writing, the way catastrophe in this play refuses to stop and change the scene.
Q: How does the five-act grid change where readers locate the play’s turning point?
The grid invites readers to locate the turning point at a numbered boundary, and the invitation has largely been accepted: Mercutio’s death in 3.1, the first scene of the middle act, is conventionally taught as the hinge on which the tragedy turns from comedy toward catastrophe. The reading is powerful, and the play’s weather does change permanently at the killings. But the frame exaggerates the symmetry. In the undivided sequence, the street fight is not a fresh start but a detonation, prepared by Tybalt’s challenge in 2.4, by the secret wedding an hour before it in stage time, and by the heat Benvolio names in the scene’s opening lines. The catastrophe reads as consequence, not commencement. The grid, by presenting 3.1 as an act’s opening, teaches the memory to file it as a beginning and so flatters a structural symmetry the continuous play does not quite perform. Knowing the frame’s origin restores the more disturbing, and more accurate, account.
Q: What evidence shows the play was written for continuous staging?
Four kinds of evidence converge. The early texts themselves are the first: both quartos run from the brawl to the tomb without act or scene marks, transmitting a sequence rather than a structure of blocks. The Prologue is the second, promising two hours’ traffic of the stage, a single span. The third is the known practice of the amphitheatre playhouses of the 1590s, where daylight performance admitted no intervals, the hall practice of act music being a later import. The fourth is internal and dramaturgical: the play’s compression, five days crushed from its source’s nine months, depends on momentum, with scenes dovetailing, one group entering as another leaves, and consequences arriving before recovery is possible, as the street fight follows the wedding within the same stage hour. Together the documents, the venue, and the writing’s own tempo describe a play built to run unbroken, whatever its editors later made of it.
Q: Why did eighteenth-century editors insist a scene must end when the stage empties?
They insisted because the cleared stage was the only scene boundary the early texts visibly offered, and their editorial creed required boundaries to be found in the documents rather than invented. A scene in a Shakespeare play has no authorial mark; what the quartos and Folio do show, hundreds of times, is the pattern of entrances and exits, and the emptying of the stage is the strongest recurrent signal in that pattern. Rowe adopted the empty stage as his practical rule, and Capell, the century’s most rigorous collator, elevated it into doctrine because it alone could be applied consistently and defended from the copies. The insistence also suited the period’s theatrical sense: in the proscenium playhouse of the 1700s, a scene visibly ended with a change of the painted shutters on an emptied stage. The doctrine, in short, married the documentary evidence to the stage practice of the editors’ own century, and the marriage has lasted.
Q: Why does the Prologue stand outside the act and scene numbering?
The Prologue stands outside the numbering because the system has never known where to put it. It is not a scene: no character of the action speaks it, the stage picture is a single Chorus figure, and it precedes the first counted emptying of the boards. The early texts print it without ceremony at the head of the play, and the Folio, which failed to head the first act at all, left it doubly unframed. Editors have answered by typography, printing the Prologue on its own footing and citing it by name rather than by number, an honest admission that the grid has an antechamber it cannot number. Its partner, the second Chorus, suffered the opposite fate and was absorbed into the frame as the head of Act 2. The pair, identical in kind, divided by editorial convenience, are the neatest demonstration in the play that the numbering is a service performed on the text, not a property found in it.
Q: Why is the tomb scene called Act 5 Scene 3 if the Folio never numbers scenes?
It is called 5.3 because the numbering counts the cleared stages of the fifth act as Rowe and his successors regularized them: the Mantua scene, in which Romeo hears of Juliet’s death and buys the poison, is the first; Friar John’s return to the Friar’s cell with the undelivered letter is the second; and the churchyard sequence, beginning with Paris and his Page and continuing unbroken through the deaths and the reconciliation, is the third. The Folio, which gave the act a Latin heading and then abandoned its scene markings, supplies the act’s bare existence and none of its arithmetic. The number is therefore a small fossil of the whole history: an Augustan count, refined by Capell’s doctrine, settled by Malone, and universalized by the Globe edition, attached to a passage whose early texts present it simply as what happens after Friar John leaves the stage.
Q: What did Samuel Johnson conclude about the act divisions in Shakespeare’s plays?
Johnson concluded, in effect, that they were printers’ furniture rather than laws of the drama, and his 1765 edition embodies the conclusion: it prints the inherited divisions, because readers required them, while its Preface dismantles the classical theory that had given the divisions their prestige. Johnson argued that Shakespeare’s plays hold their audiences without observing the unities of time and place, that the spectator’s imagination moves from Verona to Mantua without needing the machinery of acts to certify the journey, and that the supposed rules were critics’ abstractions rather than conditions of theatrical pleasure. On the divisions themselves his practice was conservative and his theory quietly subversive, a combination typical of the man. Later scholarship has ratified the theory: the numbers are conveniences of the book trade, valuable for reference, empty as evidence about how the plays were conceived.
Q: How did Edmond Malone’s edition settle the numbering that editions still follow?
Malone’s edition of 1790 settled the numbering by the weight of its scholarship. Malone collated the early texts systematically, established the chronology of the plays, and produced a text whose every feature, divisions included, carried the authority of the most learned editorial labour the century could show. Where Rowe had sketched and Capell had doctored individual boundaries, Malone’s settlement presented the scheme as a finished system, and the great variorum of 1821, completed from his papers by James Boswell the younger, printed that system surrounded by the record of two centuries of editorial opinion, converting agreement into apparent fact. The Globe edition then took the Malone-line numbering to the world. Modern editions that differ from it at a boundary must argue the difference in their apparatus, which is the practical meaning of a settlement: the burden of proof has rested, since 1790, on anyone who would move a number.
Q: How do modern editors mark divisions they have added to the text?
Modern editors mark their additions through the apparatus and, in the most scrupulous editions, through signals in the text itself. The reading page normally carries the standard act and scene numbers without interruption, since the edition must remain usable alongside all other scholarship, but the textual notes record which divisions stand in the copy-text or the Folio and which are editorial, following the practice that Brian Gibbons’s Arden edition of 1980 and Jill Levenson’s Oxford edition of 2000 exemplify. Location headings are understood by convention to be editorial throughout, however ancient they feel. Line numbering is each edition’s own count, which is why precise citation names the edition. The honest modern edition thus performs a double service: it gives the reader the universal map, and it keeps, in small print at the foot of the page, the documentary truth that the map is drawn, not found.
Q: What is lost when the play is read as five sealed acts instead of a continuous action?
What is lost is the play’s pressure. Read as five sealed acts, the tragedy becomes a symmetrical structure, exposition, complication, crisis, consequence, catastrophe, admired the way a building is admired. Experienced as a continuous action, it becomes a mechanism that will not stop: the feast generates the marriage, the marriage the killings, the killings the banishment, the banishment the potion, the potion the tomb, each stage arriving before the last has been absorbed, exactly as the lovers’ own haste drives them. The first audiences, held in a single two-hour sitting, felt the doors close in sequence without relief, and the closing is the meaning: a world in which everything happens slightly too fast for rescue. The numbered grid, by inserting its pauses and dignifying its blocks, teaches a slower and more contemplative play than the one written. Nothing in this verdict requires abandoning the grid, which remains the indispensable map. It requires only reading with the map’s legend in view: drawn by editors, over a country the author crossed in one journey.