A garrulous old woman, trying to fix the age of a girl she nursed, reaches for the one calendar event she can never forget: the night the ground moved. “‘Tis since the earthquake now eleven years,” the Nurse says in the third scene of the first act, and with that throwaway line she hands four centuries of scholars a riddle she never meant to pose. If the tremor she remembers is a real one, and if we can name its year, we can count backward and pin the composition of the most performed tragedy in the language to a single season. The trouble is that no document tells us when Shakespeare set down the first line of the work, the Nurse is the least reliable witness imaginable, and the earthquake she recalls may be any of several that shook England or Italy in the preceding decades.

That gap between certainty and the desire for it is the whole subject here. The standard account, the one a quick search returns, gives a confident answer with a clean number attached, usually 1594 or 1595 or 1597, and moves on as if the matter were settled. It is not settled. What this page shows is the actual machinery editors use when the record falls silent: how a date gets triangulated from internal allusion, from the grain of the verse, from a printer’s title page, and from a single sentence in a 1598 commonplace book. The verdict most modern editions reach, that the tragedy was composed in roughly 1595, is not a fact recovered from an archive. It is a reconstruction built from contested fragments, and the reconstruction is more interesting than the number it produces.

Why we cannot simply look the date up

No playbill, no diary entry, no letter, and no payment record survives that names the moment Shakespeare began or finished this work. That absence is not unusual. For most of the early modern stage, composition dates are inferred rather than documented, because plays were commercial property of the acting companies, written fast, revised in rehearsal, and rarely treated as literary objects worth dating at the point of creation. The author did not sign and date a manuscript for posterity. The manuscript itself is lost, as are the manuscripts of every work in the canon.

What is a terminus ante quem?

A terminus ante quem is the latest possible date by which something must already have existed, fixed by external evidence that mentions or reproduces it. For this tragedy the boundary is firm. The work had to be finished before it could be printed in 1597, and before Francis Meres could list it among Shakespeare’s tragedies in 1598. Everything earlier than that wall is open territory; everything later is ruled out for the moment of first composition.

The dating problem therefore divides cleanly into two questions that get confused in popular accounts. The first is when the text was written. The second is when it was first acted, first printed, or first noticed in surviving records. These are not the same. A drama could be composed and staged a year or more before any trace of it entered the documentary record, and the surviving traces are themselves accidents of preservation. The 1597 quarto exists because a stationer chose to print it and a copy happened to survive four hundred years of fire, damp, and neglect. Other early performances and printings have left no shadow at all.

Editors work the problem from both ends. They establish the latest the work can be, the terminus ante quem, from the external mentions. They then push the earliest plausible point, the terminus post quem, from the source material the dramatist had to have read and from the company history embedded in the printed text. Between those walls they narrow the field with the least documentary but most pervasive evidence of all, the style of the verse itself. The result is a window, not a pinpoint, and the honest editions present it as a window.

How wide is the dating window editors accept?

The outer bounds run from about 1591 at the earliest, the year the discredited earthquake reading would require, to 1597, the year of the first printing. Almost no serious editor takes the early bound seriously, so the working range collapses to roughly 1594 through 1596, with 1595 the single year cited most often. The width of that range is itself a finding: it tells a reader exactly how much is known and how much is inferred.

Brian Gibbons, editing the second Arden series in 1980, lands on 1595 and ties the work to the cluster of lyrical pieces around it. Jill Levenson, in the Oxford single edition of 2000, treats the question with more caution, surveying the same evidence and settling on a window in the middle of the decade rather than a fixed year. G. Blakemore Evans, for the New Cambridge edition, reaches a similar mid-decade placement. Rene Weis, who edited the third Arden series in 2012, weighs the textual and stylistic signals and arrives near the same ground. The disagreement among them is real but narrow, and tracing where they part company is the fastest route into how the dating actually works.

What is the earliest the play could be?

The absolute floor is set by the source material, and it sits uselessly far back. The plot descends from Arthur Brooke’s long narrative poem of 1562, The Tragicall Historye of Romeus and Juliet, with a parallel prose telling in William Painter’s Palace of Pleasure of 1567. The dramatist plainly read Brooke, lifting the structure, the doomed timetable, and many of the names directly from it, so the work cannot predate 1562. That bound is true and worthless for narrowing, because it leaves more than thirty years open.

The source therefore functions only as a loose terminus post quem, the earliest possible point, and the real work of dating happens far above that floor. What the source does establish, by contrast with the finished drama, is how much the author compressed and transformed: Brooke spreads the action across roughly nine months, while the tragedy crushes it into a handful of days, and Brooke moralises where the later writer dramatises. Those changes belong to the analysis of the play’s construction rather than its date, but they explain why the 1562 floor cannot help. A writer could have read Brooke at any point across three decades, so the source fixes only that the work is Elizabethan, not which Elizabethan year produced it. The narrowing must come from style and external mention, not from the raw plot.

The lines that carry the evidence

Three passages do most of the work, and reading them closely matters more than reciting the conclusion. The first is the Nurse’s reminiscence in the third scene of the first act. The second is the shared verse the lovers speak at their first meeting, the fourteen lines that form a complete sonnet. The third is not in the body of the text at all but on the title page of the 1597 quarto, where a printer recorded the company that had staged the work. Each passage points somewhere, and the directions do not perfectly agree.

The Nurse and the earthquake

The Nurse is trying to establish that Juliet will turn fourteen at Lammas-tide, and she does it the way an old servant would, by anchoring the girl’s age to domestic catastrophe. She remembers weaning the child, she remembers a fall the toddler took, and she remembers that the whole sequence sits eleven years after the earth shook. Following the Arden third series text edited by Weis, the reckoning runs that it has been eleven years since the earthquake, and that on that occasion the child was already weaned. The detail is pure characterization. A precise mind would give a year; the Nurse gives a tremor, a sore nipple smeared with wormwood, and a husband’s bawdy joke. The earthquake is there to make her vivid, not to date the manuscript.

Yet the line is too specific to ignore. A reader in the 1590s would have heard “earthquake” and, plausibly, thought of a particular one within living memory. England is not a seismic country, so a felt tremor was a marked event, the kind people fixed their private chronologies against exactly as the Nurse does. The most notable English shock of the period struck on the sixth of April in 1580, rattling London hard enough to dislodge masonry and frighten congregations. If that is the tremor the Nurse means, and if her eleven years are meant literally, the scene was written around 1591. That arithmetic is the entire foundation of the early date, and it is a shaky one, as the closer reading below will show.

What the line actually does in the scene is establish a texture, not a chronology. The Nurse anchors the girl’s age to a private calendar of bodily and domestic events: a weaning, a fall, a husband’s coarse joke about the child landing on her face, the wormwood on the breast. The earthquake belongs to that register of remembered shocks, a vivid marker in an old servant’s mental almanac rather than a public datum offered to the audience. Read as characterisation, the tremor is perfect; read as a clue planted for posterity, it is wildly out of key with the speech around it. The dramatist who wrote this scene was not encoding a composition date. He was building a garrulous, sensual, unreliable old woman whose memory works by association, and the earthquake is one association among several. Mistaking that texture for a cipher is the original sin of the early-dating tradition.

The shared sonnet at the ball

When the young Montague first speaks to the Capulet daughter at the feast, the two of them build a sonnet together. He opens by casting her hand as a holy shrine and his lips as two blushing pilgrims; she answers that palm to palm is holy palmers’ kiss; the exchange runs its three quatrains and seals with a couplet and a kiss. Fourteen lines, a perfect Shakespearean sonnet split between two speakers, embedded inside a scene of dialogue without announcing itself. The placement at the first meeting is deliberate, and the form is the point.

This is dating evidence of a different kind. The conceit, the patterned rhyme, the willingness to stop the dramatic action and let two characters spin a closed lyric form together, all belong to a specific phase of the author’s practice. Sonnets were a London craze in the early and middle 1590s, set off by the posthumous publication of Philip Sidney’s sequence in 1591 and sustained by a flood of imitators across the next few years. A dramatist building a sonnet into the spine of a love scene was working with the most fashionable verse form of the moment, and that fashion has a datable peak. The verse, in other words, smells of its decade.

The craftsmanship of the embedded sonnet sharpens the point. The form is not decoration laid over the scene but the scene’s actual structure: the two speakers divide the quatrains between them, trade the religious conceit of pilgrim and saint back and forth, and resolve the argument of the form in the same instant they resolve the dramatic moment, the couplet sealed by the kiss. A writer reaches for that device only when the sonnet feels alive and prestigious enough to bear the weight of a play’s emotional centre. By the closing years of the decade the form was curdling into self-parody, and the same author would later have his lovers and wits mock the conceited sonnet rather than build a tragedy’s hinge on it. The confident, unironic use here belongs to the years when the fashion was ascendant, which is the middle of the 1590s. The lovers’ shared lines therefore do more than charm; they timestamp.

The two choruses that frame the opening

The shared lyric at the feast is not the only sonnet in the early scenes, and the others reinforce the same dating signal. The Prologue is a complete sonnet, fourteen lines spoken by a Chorus that lays out the quarrel, the lovers, and the ending in advance. A second sonnet opens the second act, another Chorus passage summing up the swing from the old infatuation to the new. Three sonnets, then, bracket and punctuate the first movement of the drama: the Prologue that frames it, the lovers’ shared lines that ignite it, and the second Chorus that turns it.

That triple use is a strong stylistic marker. A dramatist who reaches for the sonnet not once but three times in the opening stretch, deploying it as a structural device rather than a single ornament, was working at the centre of the sonnet fashion, not at its edges. The second Chorus is often cut in performance and is absent or altered in some early texts, which has its own bearing on the printing history, but its presence in the composition confirms how thoroughly the form pervades the design. The bracketing choruses, taken with the shared lyric, make the sonnet the signature gesture of the work’s first act. The signature belongs to the mid-1590s, when a poet could expect an audience to recognise the form and feel its prestige rather than its fatigue.

The title page no reader was meant to analyze

The 1597 quarto announces itself as an excellent conceited tragedy of Romeo and Juliet, performed often and with great applause in public by the servants of the Lord of Hunsdon. That company name is a gift. The troupe Shakespeare belonged to was the Lord Chamberlain’s Men, but it carried the name Hunsdon’s Men during a specific interval. Henry Carey, the first Lord Hunsdon, was Lord Chamberlain until his death in July 1596, and his company held the prestigious title through him. When he died, his son George Carey inherited the barony but not, at first, the chamberlainship, so the players were known simply as Hunsdon’s Men until George secured the office in the spring of 1597 and the older title returned.

The window in which the troupe was called Hunsdon’s Men therefore runs from the summer of 1596 to roughly March 1597. The title page records performances under that name, which means the work was being staged in public, to applause, during those months. That does not fix the composition date, since a drama already a year or two old could still be playing in 1596. What it does is corroborate that the text existed and held the boards before the first printing, and it quietly rules out any claim that the work was new in 1597.

The InsightCrunch dating ledger

The cleanest way to reach a verdict is to treat each clue as a piece of testimony and grade it. The InsightCrunch dating ledger does exactly that: it takes the five signals editors actually rely on, the earthquake line, the sonnet vogue, the stylistic kinship with neighbouring works, the Meres list, and the 1597 quarto, and scores each as strong, weak, or contested, then asks where the strong evidence converges. The exercise forces honesty. It separates the clue that feels decisive but crumbles under pressure from the dull clue that quietly does the heavy lifting.

Clue one: the earthquake line, contested and probably weak

The earthquake reckoning is the most cited and the least trustworthy of the five. Its appeal is obvious. It offers a number, eleven, and an event, a tremor, and the promise that if both are real and literal the manuscript dates itself. Sidney Thomas, writing in 1949, pressed exactly that case, identifying the 1580 London shock as the Nurse’s referent and arguing for composition around 1591. The argument has a tidy logic and has never quite died.

It fails on several fronts at once. The first is dramatic. The Nurse is a comic figure whose entire mode is rambling, associative, and unreliable; her speech in this scene loops, repeats, and has to be cut off by Lady Capulet. Treating her arithmetic as a coded almanac mistakes the kind of speaker she is. The second is historical. Earthquakes felt in England or reported from Italy were not rare enough to be unique anchors. A tremor shook parts of England in 1580, but smaller shocks and widely reported Continental quakes give other candidates, and a Verona setting invites an Italian referent rather than a London one. The 1570 Ferrara earthquake, devastating and famous across Europe, would push the count to 1581, which no one accepts. The third objection is the simplest. Even granting the 1580 shock, the eleven years need not be exact, and an author writing in 1595 could easily put an approximate fifteen-year-old memory into the mouth of an old woman without intending a precise sum. The clue points everywhere, which is to say nowhere.

Most editors discount it for these reasons, and the ledger marks it contested, leaning weak. Gibbons rejects the early date outright. Levenson treats the earthquake as a red herring that has misled the unwary. The line earns its place in the conversation only because it keeps resurfacing, not because it carries weight. A separate study in this series traces the tremor argument in full, walking through the 1580 reading and its rivals; the dedicated piece on the Nurse’s earthquake dating clue lays out why the count cannot bear the load placed on it, and the companion study of the 1580 tremor debate follows the seismic record itself.

Clue two: the sonnet vogue, strong by accumulation

The strongest stylistic clue is the work’s saturation in sonnet practice. The lovers’ shared fourteen lines at the feast are the showpiece, but they are not isolated. The Prologue is a sonnet. A second chorus, opening the second act, is another. The verse throughout favours end-stopped rhyme, antithesis, oxymoron, and the conceited compliment, the whole apparatus of the fashionable lyric. This is not background noise. It is a dramatist writing at the height of a craze and building the craze into the architecture of the drama.

That craze is datable. The publication of Sidney’s sequence in 1591 detonated a London fashion for sonnet sequences that ran hard through 1592, 1593, 1594, and 1595, with the major sequences of Daniel, Constable, Drayton, and others crowding the years 1592 to 1595. By the later 1590s the form was tipping toward parody and exhaustion, and the author himself would mock its excesses elsewhere. A work this committed to the sonnet, treating it as a living and prestigious mode rather than a tired one, sits most comfortably in the middle of the decade, when the fashion was at full pressure. The ledger marks this clue strong, with the caveat that style dates a sensibility rather than a calendar year. A fuller account of the verse fashion appears in the study of the 1590s sonnet vogue and the play’s verse, which sets the lovers’ shared sonnet against the printed sequences it echoes.

The named sequences pin the peak with some precision. Sidney’s Astrophil and Stella reached print in 1591 and set the model; Samuel Daniel’s Delia and Henry Constable’s Diana followed in 1592; Thomas Lodge’s Phillis appeared in 1593; Michael Drayton’s sequence in 1594; and Edmund Spenser’s Amoretti in 1595. That run is the vogue, and it crests in the very years the other evidence favours. The dramatist’s own sugared sonnets were circulating in manuscript among private friends by 1598, when Meres notes them, which confirms that the writer was steeped in the form across the middle of the decade. A tragedy that opens with a sonnet, brackets its first movement with a second one, and seals its lovers’ meeting in a third was made by a poet living inside that fashion at its height. Place the composition in the years of Daniel, Constable, Drayton, and Spenser and the saturation needs no special explanation; place it in 1591 or in 1599 and the dense, unironic sonneteering looks either premature or stale.

Clue three: the lyric-period kinship, strong and corroborating

Editors group the canon into stylistic phases, and this tragedy falls squarely into what is usually called the lyrical period of the mid-1590s. The works clustered there share a signature: heavy rhyme, formal patterning, conceited imagery, and a delight in set-piece eloquence. A Midsummer Night’s Dream, Richard the Second, and Love’s Labour’s Lost belong to the same moment, and the family resemblance is not impressionistic. The comedy of the woods burlesques a Pyramus and Thisbe love-tragedy that is the broad farce twin of the very plot under discussion here, as if the author worked the same raw material in two registers at once. The history play in rhyme and the love-comedy of courtly wit share the tragedy’s verbal fingerprints.

This kinship is the quiet workhorse of the ledger. It cannot name a year, but it places the work among neighbours whose own dates cluster in 1594 to 1596, and it does so by a measure, verbal style, that is hard to fake and pervasive across thousands of lines. Where the earthquake line is one sentence that can be read a dozen ways, the lyric-period signature is everywhere in the text and points consistently to the same span. The ledger marks it strong. Style is the least documentary evidence and, paradoxically, the most reliable, because it cannot be selectively quoted to prove a thesis; it either runs through the whole work or it does not, and here it runs through everything.

Clue four: the Meres list, strong as a hard boundary

Francis Meres published Palladis Tamia, a commonplace book of comparisons between classical and modern writers, in 1598. In a famous passage praising Shakespeare, he lists the comedies and tragedies then known, naming Romeo and Juliet among the tragic works. The entry is a fixed point. By 1598 the work existed, was attributed to Shakespeare by a literary observer, and was reputable enough to cite. Meres sets the terminus ante quem in stone.

The clue is strong but limited. It tells us the latest the work can be, not the earliest, and it sits a year behind the 1597 printing, so it adds confirmation rather than precision. Its real value is corroborative weight: an independent witness, with no stake in a printing, names the work as established by the end of the decade. The ledger marks it strong as a boundary, neutral as a pinpoint. The companion study of Meres’ 1598 list and the play’s early fame examines the entry and what it reveals about the work’s reputation within a few years of composition.

Clue five: the 1597 quarto, strong for performance, silent on composition

The first printing carries two pieces of datable freight. The date 1597 on the title page sets a printing boundary slightly earlier than Meres. The company name, the servants of the Lord of Hunsdon, narrows the performance window to the months between mid-1596 and early 1597, as the company history above establishes. Together these confirm that the work was written, rehearsed, staged repeatedly, and successful enough to attract a stationer, all before the spring of 1597.

What the quarto cannot do is date composition, and the reason is structural. The text is short, garbled in places, and widely held to be a reported or memorially reconstructed version rather than an authorial copy, the kind of edition long labelled a bad quarto. A reported text reaches print after performances, sometimes well after, so the 1597 printing is a floor on staging, not on writing. The work could have been composed a year or more before the Hunsdon’s Men performances the title page advertises. The ledger marks the quarto strong for performance and printing, silent on composition. The textual character of this edition is examined in detail in the study of the first quarto of 1597 and why it is called a bad quarto.

Where the strong evidence converges

Lay the five graded clues side by side and a pattern emerges. The one clue that promises an exact year, the earthquake, is the one that collapses. The two clues that set hard boundaries, Meres and the quarto, agree that the work was finished, staged, and famous by 1596 to 1598 but say nothing about how much earlier it began. The two clues that actually narrow the field, the sonnet vogue and the lyric-period kinship, both point to the middle of the decade, with the fashion for sonnets at full pressure and the neighbouring works clustering in 1594 to 1596.

The intersection of the strong evidence is a window of roughly 1594 to 1596. Within it, 1595 is the point of best fit, far enough into the sonnet craze for the form to feel current, close enough to the 1596 Hunsdon’s Men performances that no awkward gap opens between writing and staging, and consistent with the cluster of lyrical works that share its verbal skin. The InsightCrunch dating verdict is therefore a composition around 1595, stated as a centre of probability rather than a recovered fact, with 1594 and 1596 as the honest margins of error. That verdict is not a discovery. It is the place where four independent lines of weak-to-strong evidence cross, and the crossing point is what the confident single numbers in the thin pages quietly stand on without showing their work.

Reading the date out of the verse

The claim that style is the surest clock deserves to be made concrete, because there is a long technical tradition behind it, and the methods are more rigorous than the word style suggests. Victorian scholars built a whole apparatus for ordering the canon by counting features of the verse, and that apparatus, refined by modern statistics, places this tragedy firmly in the middle of the decade. The metrical evidence is the part of the case least visible to a casual reader and most decisive to a specialist.

What are verse tests?

Verse tests are tallies of measurable features of dramatic poetry that change across a writer’s career: the proportion of rhymed lines to blank verse, the share of lines that run on into the next without a pause, the frequency of feminine endings where a line closes on an unstressed extra syllable, and the placement of light and weak stresses at line ends. A writer’s early verse and late verse differ systematically on these counts, so the tallies can sort works into a sequence even when no document dates them.

The New Shakspere Society, founded in 1873 under Frederick Furnivall, made these counts the engine of its chronology, and F. G. Fleay tabulated the canon by them. The pattern they found is consistent and intuitive once stated. Early dramatic verse leans heavily on rhyme, keeps most lines end-stopped with a clear pause at the close, and uses few feminine endings, producing a regular, song-like movement. Late verse abandons rhyme almost entirely, runs sentences across line breaks so that the metre nearly dissolves into speech, and piles up feminine and weak endings, producing the loose, conversational rhythm of the final romances. Between those poles the works fall into an order, and the order tracks the independently known dates wherever those dates exist.

Where the tragedy sits by these counts

On every one of these measures the tragedy reads early-to-middle, not late. Its proportion of rhyme is among the highest of any work in the tragic vein, a direct consequence of the sonnet saturation already described: the Prologue, the second chorus, the shared lyric at the feast, and dozens of rhymed couplets scattered through the dialogue push the rhyme count far above what the mature tragedies show. Its lines are predominantly end-stopped, moving in the regular, closed units of the lyrical phase rather than the run-on speech of the later style. Feminine endings are sparse compared with the later works. By the Victorian tallies the drama lands squarely with the lyrical group of the mid-1590s and nowhere near the looser verse of the great tragedies that followed.

This is the technical backbone of the stylistic argument. Where a single allusion can be argued over for pages, the rhyme proportion is a number derived from thousands of lines, and it cannot be massaged to suit a theory. It places the work among its mid-decade siblings by a measure no one can selectively quote. The verse counts do for the date what the sonnet vogue does for the fashion: they anchor the impression of an early-middle style in something countable.

The limits of the counts

The method has genuine weaknesses, and the careful editions acknowledge them. Rhyme proportion responds to genre and occasion as well as to date; a work consciously experimenting with lyric form, as this one is, will show inflated rhyme for reasons of design rather than chronology, so the very feature that places it early is partly a deliberate choice rather than a pure index of when it was written. The Victorian counters sometimes treated their tallies as more precise than the data could support, ordering works to the year on differences that fall within the noise. And the tests cannot separate composition from revision, since a writer who returned to an old script would smear the counts.

Modern stylometry, working with function-word frequencies and computational authorship tools, has revisited the question with finer instruments, and the broad result has held: the work groups with the mid-1590s plays and stands well apart from the later tragedies. The refinement has tightened the placement rather than overturned it. The lesson for dating is that the verse tests, taken with their limits in view, corroborate the sonnet vogue and the lyric kinship rather than competing with them. Three independent stylistic measures, the saturation in sonnet form, the family resemblance to the dream-comedy and the rhymed history, and the metrical counts of rhyme and line ending, all point to the same span. That triple agreement is why the stylistic case carries so much weight in the ledger, and why the editors who lean on it are not being soft. They are reading a clock with three faces, all telling the same hour.

The critical conversation

The editors who have wrestled this question into their introductions do not all reach the same year, and the texture of their disagreement is worth setting out, because it shows that even a near-consensus is built from arguable steps rather than handed down whole.

Gibbons and the case for 1595

Brian Gibbons, in the second Arden series of 1980, commits to 1595 with comparatively little hedging. His reasoning leans hardest on the stylistic kinship, reading the tragedy as a sibling of A Midsummer Night’s Dream and Richard the Second, works whose own dates sit in 1595, and treating the shared lyrical manner as decisive company-keeping. For Gibbons the verse settles the matter: a writer producing this density of rhyme and conceited patterning was working in his lyrical phase, and that phase has a known midpoint. He sets the earthquake reading aside as a distraction and lets the style carry the date.

The strength of his position is that it rests on the most pervasive evidence rather than a single line. Its vulnerability is that stylistic dating runs in circles when the neighbouring works are themselves dated partly by style; the cluster coheres, but it floats slightly free of any external anchor. Gibbons knows this and braces the stylistic argument against the external boundaries of Meres and the quarto, which is why his 1595 has held up.

Levenson and the cautious window

Jill Levenson, editing the single-volume Oxford text in 2000, declines to commit to one year and presents a window instead. She surveys the full apparatus, the earthquake line, the company history on the quarto title page, the Meres entry, and the stylistic placement, and concludes that the evidence supports composition in the middle of the decade without licensing a sharper claim. Her treatment is the most scrupulous about the difference between what the evidence establishes and what convenience would prefer, and she is sharp on the earthquake clue, which she regards as having led earlier scholars astray.

The disagreement between Gibbons and Levenson is instructive precisely because it is small. Both place the work in the mid-1590s; both discount the early date; both rely chiefly on style and the external boundaries. They part only on how much precision the evidence will bear. Gibbons judges that the stylistic cluster justifies naming 1595; Levenson judges that the same evidence justifies only a range. The adjudication between them turns on a question of method rather than fact: how confident a single year may be when no document supplies it. The honest answer sides with Levenson on principle and with Gibbons in practice, naming 1595 as the likeliest point while flagging that the margins are real.

Evans, Weis, and the settled middle

G. Blakemore Evans, for the New Cambridge edition, and Rene Weis, for the third Arden series of 2012, occupy the ground between the two. Evans places the work in the mid-1590s, weighing the same stylistic and external evidence and resisting the early date, and his caution about the earthquake reckoning matches Levenson’s. Weis, editing more than three decades after Gibbons, inherits a question that has largely stabilised; his apparatus weighs the textual situation of the early printings alongside the stylistic placement and arrives near the same span, with attention to how the quarto evidence bounds the performance history. None of these editors revives the 1591 reading. The drift of two generations of scholarship has been toward consensus, not away from it, and the consensus is a window centred on 1595.

The one genuine fault line that remains, and the disagreement worth adjudicating, is the status of the earthquake line. A minority tradition descending from Sidney Thomas treats it as the strongest internal clue and lets it pull the date back to 1591. The majority treats it as the weakest, a comic flourish that cannot bear arithmetic. The adjudication is not close. The early reading requires the Nurse to be a reliable chronometer, requires the 1580 London shock to be her unambiguous referent despite an Italian setting and competing Continental quakes, and requires the count of eleven to be exact, three improbable conditions stacked on one another. Against that stands a coherent body of stylistic and external evidence pointing to 1595. The earthquake clue loses on every count, and the verdict of the field is right to set it aside.

Chambers and the standard chronology

Behind the modern editors stands the reference chronology that fixed the field’s habits. E. K. Chambers, in his 1930 survey of the documentary facts and problems of the canon, assembled the external records and the verse-test results into the ordering that later scholarship inherited, placing this tragedy in 1594 to 1595. Chambers was a sceptic about the Victorian counters’ precision, trimming their confidence while keeping their broad sequence, and his placement of the work in the mid-decade gave the editors who followed a stable starting point. When Gibbons names 1595 and Levenson states a mid-decade window, both are working within the frame Chambers set, refining rather than overturning it. The continuity from Chambers to the present editions is one reason the date has stabilised: each generation has tested the placement against new evidence and found it holds.

The argument about method

Underneath the small disagreement over the year runs a larger and more interesting quarrel about method, and it is worth surfacing because it shapes every verdict. One camp trusts the countable: verse tests, function-word statistics, the hard external boundaries of printing and citation. The other trusts the contextual: the feel of the style, the fashion the verse draws on, the company history, the relationships among neighbouring works. The two camps are not really opposed, since the best editions use both, but they weight the evidence differently, and the weighting decides how precise a claim each is willing to make.

The adjudication is that neither approach alone suffices and their convergence is the real argument. The countable methods place the drama early-to-middle by metre and bound it by Meres and the quarto, but they cannot name a year. The contextual methods read the sonnet saturation and the company history toward the mid-decade, but they trade in judgement rather than proof. Where both point to the same span, as they do here at 1594 to 1596, the conclusion rests on independent foundations and becomes hard to shake. The lesson the dating case teaches about Shakespearean scholarship generally is that the strongest results are the ones where the statistical and the interpretive evidence cross, and the weakest are the ones, like the earthquake date, that rest on a single clue read in isolation. The field’s near-consensus on roughly 1595 is strong precisely because it is overdetermined, supported by metre and fashion and company and citation at once, and no longer dependent on any single thread.

The early printed and acted life

A composition date means little in isolation; it gains meaning when set against what happened to the work in the years immediately after. The interval between writing and printing, and the company history folded into the title pages, both sharpen the picture and explain why the dating question keeps getting muddled with the printing question.

How long between writing and printing?

If the work was composed around 1595 and the first quarto appeared in 1597, the gap between writing and printing was short, two years at most, and the Hunsdon’s Men performances advertised on the title page fall right in the middle of it. That tight interval is itself a small argument for the mid-decade date. A long delay between composition and the documented performances would open an awkward silence, a span in which a popular work left no trace. The 1595 placement closes that gap and produces a clean sequence: written around 1595, staged to applause in 1596 and early 1597 under the Hunsdon’s Men banner, reported into print in 1597, and named by Meres in 1598.

The first printing was followed by a far better one. The second quarto of 1599 advertised itself as newly corrected, augmented, and amended, and it is roughly half again as long as its predecessor, carrying the fuller text from which modern editions chiefly descend. The relationship between the short 1597 text and the fuller 1599 text has its own large literature, but for dating purposes the lesson is narrow: the existence of a reported short text by 1597 confirms that performances were frequent enough to be memorised and pirated, which fits a work already a year or two into a successful run rather than a fresh script.

The size of the gap between the two quartos is itself suggestive. A first quarto running well under the length of the second, missing passages and garbling others, is the fingerprint of a text assembled from memory after the fact rather than from an authorial manuscript. Memorial assembly presupposes a body of performances dense enough to lodge the play in someone’s memory, and that density takes time to build. The short text of 1597 therefore implies a staging history reaching back well before its printing, which is consistent with first performances in 1596 and composition the year before. The very defects of the first quarto, read as evidence, push the writing earlier rather than later, away from 1597 and toward the mid-decade.

Why is there no performance record like other plays have?

Many plays of the period can be dated from Philip Henslowe’s Diary, the account book in which the manager recorded daily receipts for performances at the Rose, chiefly by the Admiral’s Men, from 1592 onward. It is the richest surviving record of the Elizabethan repertory, and for the works it covers it often fixes a first performance to the day. This tragedy is not in it. The absence is not a mystery but a consequence of company affiliation: the drama belonged to the Chamberlain’s Men, whose accounts do not survive in anything like Henslowe’s detail, so there is no equivalent ledger to date it.

That gap in the record explains why this particular work is harder to date than some of its contemporaries. A rival author’s play staged at the Rose might carry a precise Henslowe entry for its premiere, while this tragedy, performed by a company without a surviving account book, leaves only the indirect traces of style, the quarto title page, and the Meres mention. The contrast is instructive. The dating is not vague because the work is mysterious; it is vague because the company that owned it kept records that did not last. Had the Chamberlain’s Men left a Henslowe, the question this page answers by triangulation might be settled by a single line in a ledger. They did not, so triangulation is the only route, and the route leads to a window rather than a date.

What the company name proves and what it does not

The Hunsdon’s Men detail is often overstated. It does not date the writing; it dates a phase of performance to mid-1596 through early 1597. Read carefully, it is a corroborating clue, not a foundational one. Its value is that it independently confirms the work was on the boards and popular before the printing, and it does so through a fact, the patron’s title, that had nothing to do with the drama and could not have been invented to suit a dating theory. That independence is what makes it useful. A clue that no one shaped to prove a point is worth more than a clue, like the earthquake line, that can be bent to support whatever year a reader prefers.

The plague years and the company’s formation

A piece of London theatre history tightens the lower bound more usefully than the source poem does. Plague shut the city playhouses for long stretches between 1592 and 1594, and the closures scattered and reorganised the acting companies. Out of that disruption the Lord Chamberlain’s Men formed in 1594, the troupe Shakespeare belonged to and wrote for thereafter. The title page of the first quarto records the tragedy as performed by that company under its Hunsdon’s Men name, which places its staging within the company’s life, after the 1594 reorganisation.

The inference is modest but real. If the drama was written for the Chamberlain’s Men, composition most plausibly follows the company’s 1594 formation, since the playhouses were closed and the troupes in flux through the preceding two years and offered no stable home for a new script. That consideration pushes the floor up from the useless 1562 source-bound to a working lower margin around 1594, exactly where the stylistic evidence also bottoms out. Two independent lines, the institutional history of the company and the texture of the verse, agree on the same earliest plausible point. The agreement is part of why editors are comfortable closing the early bound at 1594 and treating the 1591 earthquake date as an outlier that the company history cannot accommodate, since a 1591 composition would predate the very troupe whose name the first printing records.

The first performance no one recorded, and there will almost certainly never be one. The earliest documented staging that survives is the applauded public run the 1597 quarto advertises, but that run was clearly not the premiere, since reported texts presuppose a performance history dense enough to be memorised. The premiere lies somewhere in 1595 or 1596, in a playhouse whose records do not survive, before an audience whose response went unrecorded except in the general claim of great applause. The opening night of the most staged tragedy in the language is, fittingly, lost. The dating exercise recovers the season but never the night.

Why the date matters

A confident reader might ask why any of this is worth the trouble. The answer is that the dating restores the work to a working life and a working calendar, and that restoration changes how the tragedy reads. The framing that treats the drama as floating outside time, a possession of all ages and no particular year, is precisely the flattening that obscures what the writer was doing in 1595.

A tragedy made inside a busy career

Placed around 1595, the work sits inside one of the most productive stretches of the author’s career, shoulder to shoulder with the dream-comedy, the rhymed history, and the courtly love-comedy. Seeing it there changes the sense of the achievement. The shared sonnet at the feast stops looking like a lucky flourish and starts looking like the experiment of a writer pushing the fashionable lyric into dramatic structure, the same experiment he ran as broad farce in the woods of the dream-play and as political lament in the rhymed history. The tragedy becomes a move in an ongoing argument with form, made by a working dramatist at a specific moment, rather than a timeless utterance delivered from nowhere.

An anchor in the canon chronology

Dating this work does not only fix one drama; it helps hold the chronology of the whole canon in place. The lyrical group of the mid-1590s is dated partly by the works’ relationships to one another, and a confident placement of one member steadies the rest. The twinning with the dream-comedy is the clearest case. That comedy stages a troupe of amateurs rehearsing a Pyramus and Thisbe play, a tale of two lovers kept apart by their families and undone by a fatal misunderstanding at a meeting place, played for broad laughs. The plot is the tragedy’s own material turned to farce. A writer who could mine the same vein for tears and for ridicule in the same stretch of years was working through a single preoccupation in two keys, and dating either work helps date both.

The placement around 1595 therefore does chronological labour beyond itself. It confirms that the author was, in the middle of the decade, conducting a sustained experiment in form across genres: lyric tragedy, festive comedy, rhymed history, and courtly word-play, all sharing a verbal surface and a fascination with patterned eloquence. Move this tragedy out of 1595 and the cluster loosens; keep it there and the whole lyrical phase coheres into a recognisable moment in a career. The date is load-bearing for more than one building. That is why editors treat the question as worth the labour even though it yields only a window, and why the cross-references between the lyrical works recur throughout the scholarship on each of them.

Style as a clock

The dating also vindicates a method that students are often taught to distrust. Stylistic dating sounds soft, a matter of impression, yet it turns out to be the most load-bearing evidence in the whole case. The reason is that style is involuntary and pervasive. A writer can fake a single allusion to suit a date, but cannot fake the rhythm of ten thousand lines. The density of rhyme, the reach for the conceit, the patterned antithesis, these run through the entire work and align it with its mid-decade siblings whether the author intended a signature or not. The dating case teaches a transferable lesson: when documents fail, the grain of the writing is often the surest clock available, and the lovers’ shared sonnet ticks at the rate of 1595.

What the date does not settle

Honesty about the verdict means marking what the window leaves open. A composition around 1595 does not tell us how long the writing took, whether it came in a single push or over months, or whether the author returned to revise. The gap between the short reported first quarto and the fuller second quarto raises the possibility that the text grew or changed between its earliest staging and the better 1599 printing, and verse tests cannot separate an original layer from a later touch. The date fixes a season of first composition, not a finished and frozen object.

Nor does the date resolve which of the lyrical siblings came first. The cluster of mid-decade works coheres in style, but the internal order within 1594 to 1596 is itself debated, and placing this tragedy at 1595 does not automatically rank it before or after the dream-comedy or the rhymed history. The honest position holds the group together while leaving its internal sequence loose. These open questions are not failures of the dating exercise; they are its proper edges. A reconstruction that knows where its evidence stops is worth more than one that pretends to a precision the record cannot supply, and the window centred on 1595 is exactly as confident as the surviving traces allow and no more.

The cliche the date dissolves

The popular framing of the work as a story belonging to no particular time survives partly because no one stops to ask when it was made. Asking the question, and answering it with a real if approximate year, dissolves the cliche. The tragedy was written by a particular man in a particular season working in a particular fashion, racing a particular company’s repertory, in a London gripped by a particular craze for the sonnet. None of that diminishes the work; it sharpens it. The drama is more interesting as the product of 1595 than as the property of eternity, because the year explains the choices the eternity-framing has to treat as miracles.

Why the date is so often gotten wrong

The misinformation around this question is consistent enough to map, and naming the specific errors is more useful than gesturing at vague confusion. Four mistakes recur in thin pages and classroom shorthand, and each has a traceable cause.

The first error is the confident single number with no margin. Pages routinely state 1594 or 1595 as if it were a recorded fact, stripping away the window that every careful editor preserves. The cause is the appetite for a clean answer, and the cost is a false sense of certainty that collapses the moment a reader asks for the evidence. The honest statement is a range centred on 1595, and any source that omits the range is hiding the actual state of knowledge.

The second error is treating the 1597 printing as the composition date. This is the most common slip, and the company history exposes it. The first quarto is a reported text printed after a run of performances, so 1597 dates the printing, not the writing, which lies a year or two earlier. Confusing the two telescopes the whole sequence and produces the false claim that the work was new in 1597 when it was already a popular property by then.

The third error is the revival of the earthquake date. Pages drawing on older or fringe scholarship still cite 1591 on the strength of the Nurse’s line, presenting Sidney Thomas’s reading as if it were the consensus rather than a minority position the field discounted long ago. The cause is the line’s surface precision, that tempting number eleven, and the cure is the close reading above: the Nurse is a comic and unreliable speaker, the referent is ambiguous, and the count need not be exact.

The fourth error is the dismissal of the whole question as unknowable and therefore pointless. This nihilist shrug is the inverse of the false-certainty error and just as wrong. The date is not known to the day, but it is known to within a couple of years on a convergence of independent evidence, and that convergence is a genuine and interesting piece of literary detection. Refusing to do the triangulation because it yields a window rather than a point throws away real knowledge. The window is the answer, and a couple of years’ uncertainty on a four-hundred-year-old undocumented manuscript is a remarkably good result.

The four errors share a single root: each substitutes a wish for the evidence. The false single number wishes for tidiness, the printing-date slip wishes for a document to mean more than it does, the earthquake revival wishes for a clue with a number on it, and the nihilist shrug wishes the problem away rather than working it. The corrective in every case is the same discipline the ledger enforces, grading each clue for what it can actually carry and trusting only the place where the strong evidence crosses. Done that way, the question yields a clear and defensible answer, a composition around 1595, held with the right amount of confidence and no more. A reader who internalises the method can date not just this tragedy but any undocumented work, by asking of every clue what kind of evidence it is and refusing to let the most tempting one outrank the most reliable.

A date recovered, not received

The Nurse never meant to date anything. She was remembering an old fright the way old women do, by the year the ground shook, and she handed the conversation a clue she could not have known would outlive the reckoning of her own household. Four centuries later the line still sends readers off to count, and the counting still fails, because she is the wrong witness for the job. The real evidence is quieter and more honest: the sonnet built into a kiss, the company that wore a dead lord’s name for a few months in 1596, a literary man’s list in 1598, and above all the rhythm of the verse, which beats at the rate of the mid-1590s and could beat at no other. Put those together and the answer arrives, not as a fact pulled from an archive but as the place where four lines of evidence cross. The tragedy was made around 1595, by a working writer at the height of a craze, and the date, recovered rather than received, makes the work stranger and better than the timeless framing ever could.

There is a quiet justice in the fact that the play’s most famous false clue and its truest one both live in the verse. The earthquake fails as a date because it is the wrong kind of evidence read the wrong way, a character’s memory mistaken for an author’s signature. The sonnet succeeds because it is the right kind, the involuntary print of a moment pressed into every rhymed couplet, the second Chorus, and the lovers’ shared fourteen lines. The lesson outlasts this one tragedy. When the documents are gone and the manuscripts lost, the writing keeps its own record, and a careful reader counting feet and rhymes rather than years can still hear the decade in which a line was made. The Nurse counts wrong and the verse counts right, and between them they teach how a dateless masterpiece gives up its year to anyone willing to read the evidence it cannot help leaving behind.

Frequently Asked Questions

Q: What is the most likely year Shakespeare wrote Romeo and Juliet?

The likeliest single year is 1595, though it is best stated as a window of roughly 1594 to 1596. No document records the moment of composition, so editors triangulate from the style of the verse, the fashion for sonnets that the work draws on, the cluster of lyrical works it resembles, and the external boundaries set by the 1597 printing and the 1598 mention in Francis Meres. Those lines of evidence intersect in the middle of the decade, with 1595 the point of best fit. Brian Gibbons names 1595 outright in his Arden edition, while Jill Levenson prefers to state a range, and the difference between them is one of caution rather than substance. Any source that gives a flat single year without a margin is overstating what the evidence supports.

Q: Why is there no exact date for the composition of the play?

Plays in the 1590s were commercial property of acting companies rather than dated literary works, so authors rarely signed and dated manuscripts, and those manuscripts are in any case lost for the entire Shakespeare canon. The surviving traces of a work are accidents of preservation: a stationer chose to print the 1597 quarto, and a copy happened to survive. Composition itself left no paper trail. What survives are external mentions, which fix the latest the work can be, and the internal evidence of style and allusion, which narrows the earliest plausible point. Between those bounds editors reconstruct a window. The absence of an exact date is normal for the period and reflects how the early modern theatre treated scripts, not any special mystery about this particular tragedy.

Q: What is the earthquake clue in Romeo and Juliet?

In the third scene of the first act, the Nurse fixes Juliet’s age by recalling that it has been eleven years since an earthquake. Some readers have argued that the tremor is the London earthquake of April 1580, which would place the writing of the scene around 1591 by simple subtraction. The reading goes back to a 1949 argument by Sidney Thomas. Most editors reject it. The Nurse is a comic and unreliable speaker whose rambling has to be cut off; the earthquake she means is ambiguous, especially given the work’s Italian setting; and the count of eleven need not be exact. The clue is the most cited and least trustworthy of the dating signals, valuable mainly because it keeps resurfacing rather than because it proves anything.

Q: Does the earthquake line mean the play was written in 1591?

Almost certainly not. The 1591 date depends on three improbable conditions holding at once: that the Nurse is a precise chronometer rather than a comic rambler, that her earthquake is unambiguously the 1580 London shock despite a Verona setting and famous Continental quakes such as the 1570 Ferrara earthquake, and that her count of eleven years is exact rather than approximate. Each assumption is weak, and stacking all three is unjustifiable. Against the 1591 reading stands a coherent body of stylistic and external evidence pointing to the mid-1590s. The field discounted the early date decades ago. It survives chiefly in thin pages that mistake the surface precision of the number eleven for genuine dating evidence, when in fact the line characterises the Nurse rather than dating the manuscript.

Q: How does the sonnet craze help date the play?

The work is saturated in sonnet practice: the Prologue is a sonnet, a second chorus opens the second act, and the lovers build a complete sonnet together at their first meeting. That commitment to the form ties the work to a datable fashion. The publication of Philip Sidney’s sequence in 1591 set off a London craze for sonnet sequences that ran hard through the early and middle 1590s, with the major printed sequences crowding 1592 to 1595, before the form tipped toward parody later in the decade. A dramatist treating the sonnet as a living, prestigious mode, rather than an exhausted one, was working at the peak of that fashion, which sits in the middle of the decade. The verse, in other words, carries the fingerprint of its moment, and that fingerprint points to roughly 1595.

Q: What does Francis Meres tell us about the date?

Francis Meres published Palladis Tamia in 1598, a commonplace book comparing classical and modern writers. In a passage praising Shakespeare he listed the comedies and tragedies then current and named Romeo and Juliet among the tragic works. The entry is a fixed boundary: by 1598 the work existed, was attributed to Shakespeare by a literary observer, and was reputable enough to cite alongside the canon. Meres sets the terminus ante quem, the latest the work can possibly be. He does not narrow the earliest point, and his mention sits a year behind the 1597 printing, so his real function is corroboration. As an independent witness with no stake in any printing, he confirms that the work was established by the end of the decade, which fits a mid-decade composition comfortably.

Q: Why does the 1597 quarto not settle the date?

The first quarto carries the date 1597 and advertises performances by the Lord of Hunsdon’s servants, but neither fact dates the writing. The text is short and garbled, widely regarded as a reported or memorially reconstructed version rather than an authorial copy, the kind of edition long called a bad quarto. A reported text reaches print only after a run of performances substantial enough to be memorised, so the 1597 printing is a floor on staging, not on composition. The work could have been written a year or more before the performances the title page advertises. The quarto is strong evidence that the drama existed and was popular before 1597, and silent on exactly when it was first set down. Confusing the printing date with the writing date is the single most common dating error.

Q: What were the Lord Hunsdon’s Men, and how do they help?

The acting company Shakespeare belonged to was the Lord Chamberlain’s Men, but it carried the name Hunsdon’s Men during a brief interval. Henry Carey, the first Lord Hunsdon, was Lord Chamberlain until his death in July 1596, and the company held its prestigious title through him. When he died, his son George Carey inherited the barony but not immediately the chamberlainship, so the players were simply Hunsdon’s Men until George secured the office in spring 1597, when the older title returned. The 1597 quarto records performances under the Hunsdon’s Men name, which dates those stagings to roughly mid-1596 through early 1597. The detail corroborates that the work was on the boards and popular before printing, and it is valuable because the patron’s title had nothing to do with the drama and could not have been shaped to fit a dating theory.

Q: What is the difference between the dating of composition and first performance?

Composition is when the text was written; first performance is when it was first acted. They are not the same, and the gap between them can be a year or more. A drama could be composed in 1595 and first staged the same season or later, and the surviving records capture neither moment directly. For this tragedy the earliest documented performances are the applauded public runs advertised on the 1597 quarto, dated by the Hunsdon’s Men name to 1596 and early 1597, but those runs were clearly not the premiere, since a reported text presupposes a dense performance history. The actual first performance is lost. The dating exercise recovers the season of writing, roughly 1595, and the season of the documented run, 1596, but never the opening night itself.

Q: Which scholarly editions should I trust on the dating question?

The standard scholarly editions are the most reliable guides. Brian Gibbons edited the second Arden series in 1980 and commits to 1595 on stylistic grounds. Jill Levenson edited the Oxford single text in 2000 and prefers a cautious mid-decade window over a single year. G. Blakemore Evans edited the New Cambridge text and reaches a similar mid-decade placement. Rene Weis edited the third Arden series in 2012 and arrives near the same ground while weighing the textual situation of the early quartos. All four discount the 1591 earthquake date. Reading two of them against each other, Gibbons for the confident year and Levenson for the careful range, gives the fullest picture, because the small disagreement between them exposes exactly how much the evidence supports.

Q: Why do some sources say 1594 and others say 1597?

The 1594 figure usually reflects an editor pushing toward the earlier edge of the stylistic window, emphasising the work’s kinship with the lyrical pieces of that year. The 1597 figure almost always reflects the error of mistaking the first printing date for the composition date. Since the 1597 quarto is a reported text printed after a run of performances, its date marks the printing rather than the writing, which lies earlier. The spread from 1594 to 1597 across different sources is therefore partly a genuine difference about where in the window to place the writing and partly a confusion of printing with composition. The reconciling answer is that the work was written around 1595, staged in 1596, and printed in 1597, so all three years describe different events in one short sequence.

Q: How reliable is stylistic dating?

More reliable than its soft reputation suggests. A single allusion can be faked or misread, but the rhythm and texture of ten thousand lines cannot be. The density of rhyme, the reach for conceited imagery, and the patterned antithesis run through the entire work and are involuntary, a sensibility rather than a decision. That pervasiveness is what makes style a good clock. It aligns the tragedy with its mid-1590s siblings, the dream-comedy and the rhymed history, by a measure that cannot be selectively quoted to support a thesis, because it either runs through the whole text or it does not. The main limit is that style dates a phase rather than a calendar year, so it narrows the field to the mid-decade without naming a precise twelve-month span. Within that limit it is the most load-bearing evidence in the case.

Q: What other plays did Shakespeare write around the same time?

The tragedy belongs to the lyrical period of the mid-1590s, sharing its verbal signature with A Midsummer Night’s Dream, Richard the Second, and Love’s Labour’s Lost, all clustered in 1594 to 1596. The kinship is striking with the dream-comedy in particular, whose Pyramus and Thisbe burlesque works the same raw material of doomed young love as broad farce, as if the author handled the same plot in two registers at once. The rhymed history offers the same heavy rhyme and formal patterning in a political key, and the courtly love-comedy shares the delight in verbal wit and set-piece eloquence. Seeing the tragedy among these neighbours is itself part of the dating argument, since their dates and its dates support one another, and it reframes the work as a move in an ongoing experiment with form rather than an isolated masterpiece.

Q: Is the second quarto of 1599 relevant to dating the play?

For dating composition, only indirectly. The second quarto advertised itself as newly corrected, augmented, and amended, and it is roughly half again as long as the 1597 text, carrying the fuller version from which modern editions chiefly descend. It postdates both the first printing and the Meres mention, so it adds nothing to the boundaries. Its relevance is what it implies about the earlier text: the existence of a short reported quarto by 1597 confirms that performances were frequent and popular enough to be memorised and pirated, which fits a work already a year or two into a successful run. The 1599 text is central to understanding what the play says and how it was transmitted, but for the question of when it was written it sits well inside the window already established.

Q: Could the play have been written before 1594?

It is possible but unlikely, and almost no editor argues for it apart from the discredited 1591 earthquake reading. A date before 1594 would sit at the very early edge of the sonnet craze and slightly ahead of the lyrical cluster the work resembles, opening a longer and otherwise unexplained gap between composition and the documented 1596 performances. The stylistic evidence pulls toward the middle of the decade rather than its opening, and the external boundaries do nothing to require an early date. The honest lower margin of the window is around 1594, reflecting the earliest the lyrical style and sonnet saturation comfortably allow. Anything earlier rests on the earthquake arithmetic, which fails for the reasons set out above, so the practical floor is 1594 with 1595 as the better estimate.

Q: Why is dating an undated manuscript even possible?

Because writing leaves fingerprints beyond the explicit date a manuscript would carry. External mentions fix outer boundaries: a work cannot be later than the first record of it, here the 1597 printing and the 1598 Meres list. Internal allusions can point to datable events or fashions, like the sonnet craze the verse draws on. And style, the involuntary texture of the verse, aligns a work with others whose dates are independently fixed. None of these alone yields a precise year, but together they converge. Dating an undated work is therefore an exercise in triangulation, the same logic a detective uses with circumstantial evidence: no single clue proves the case, but several independent clues pointing to the same span make a conclusion that a couple of years’ margin cannot shake. The convergence on roughly 1595 is exactly such a result.

Q: Does it matter to a reader of the play when it was written?

It matters because the date dissolves a cliche and sharpens the work. The framing of the tragedy as belonging to no particular time treats its boldest choices, the sonnet built into a kiss, the spoiler Prologue, the collision of farce and tragedy, as timeless miracles. Placing the writing around 1595 reveals them instead as the experiments of a working dramatist at the height of a sonnet craze, racing a company’s repertory and handling the same material as comedy and history in the same stretch of years. The drama becomes a move in an argument with form, made at a specific moment, which is more interesting than the eternity-framing allows. A reader who knows the date reads the choices as choices rather than as the inexplicable utterances of a writer outside history.

Q: How do the 1592 to 1594 plague closures bear on the date?

Plague shut the London playhouses for long stretches between 1592 and 1594, scattering and reorganising the acting companies, and out of that disruption the Lord Chamberlain’s Men formed in 1594. Since the first quarto records the tragedy as performed by that company under its Hunsdon’s Men name, the staging belongs to the troupe’s life, after the 1594 reorganisation. If the drama was written for that company, composition most plausibly follows its 1594 formation, because the closed theatres and the flux among troupes offered no stable home for a new script in the preceding two years. This institutional history pushes the earliest plausible point up to around 1594, matching the lower edge of the stylistic window. It also weighs against the 1591 earthquake date, which would put composition before the very company whose name the first printing carries.

Q: Did Shakespeare write the sonnets at the same time as the play?

The two projects overlap in period without being the same act of writing. Francis Meres noted in 1598 that Shakespeare’s sugared sonnets were circulating in manuscript among private friends, which shows the poet was deep in the form by the later 1590s, and the dense, confident sonneteering inside the tragedy belongs to the same creative phase. The play, dated to roughly 1595, sits inside the years when the author was steeped in the sonnet, and that shared immersion is part of why the drama treats the form so fluently, opening with one, bracketing its first movement with a second, and sealing the lovers’ meeting with a third. The sonnet collection itself was not printed until 1609 and its individual poems resist precise dating, so the safe statement is that the play and the poet’s sonnet practice share a moment in the middle of the 1590s rather than a single date.

Q: Could a single document ever settle the composition date?

In principle yes, in practice almost certainly not. A dated authorial manuscript, a precise performance receipt, or a contemporary letter naming the work as newly written would fix the date directly, and such documents survive for some plays of the period through Philip Henslowe’s account book. None survives for this tragedy, because the company that owned it left no comparable ledger and the manuscripts of the whole canon are lost. The realistic hope is not a smoking-gun document but a refinement of the existing evidence: a sharper account of the verse statistics, a firmer date for a neighbouring work, or a fresh reading of the company history that narrows the window. Barring an unlikely archival discovery, the answer will remain a reconstruction centred on 1595 rather than a fact, and that is the normal condition of dating early modern drama rather than a special defeat in this case.