The Queen Mab speech, Mercutio’s forty-line flight over the sleeping households of Verona in Act 1, Scene 4, is the most quoted set piece in Romeo and Juliet, and the version readers quote is an editorial assembly. No early printed text contains the speech in exactly the form a modern edition gives it. The speech survives in three early witnesses that agree on its substance to a degree that surprises readers who know the two quartos only by reputation, and differ from each other in arrangement, lineation, connectives and small wording at a density that rewards a magnifying glass. Collating this one speech across the first quarto of 1597, the second quarto of 1599 and the First Folio of 1623 produces a verdict that cuts across the grand theories built on the whole texts. The agreement proves common descent from a written ancestor very close to the second quarto’s copy. The drift, where it occurs, carries the fingerprint of performance and memory rather than the pattern of an author rethinking a finished invention. And the Folio’s speech, derivative almost line by line, warns against counting three witnesses when the evidence really holds two.
The stakes of the collation are set by the speech’s fame. Mercutio invents Queen Mab on the way to the Capulet feast, answering the hero’s dream with a fairy midwife “no bigger than an agate stone” (1.4), and the invention has been anthologized, memorized and parodied ever since. Harold Bloom treated Mercutio as the play’s presiding wit and the speech as the showpiece of that wit, the moment where the tragedy pauses for a virtuoso who can conjure a world out of miniatures. Precisely because the speech is famous, its textual condition is instructive. A passage every reader thinks of as fixed turns out to have reached print in two distinct early shapes, and the distance between those shapes measures the distance between the play as written and the play as performed, remembered and sold.
This article collates the speech unit by unit: the opening address, the chariot inventory, the catalogue of dreamers, the darkening coda and the interruption that ends it. The method is the bibliographer’s, comparing substantive readings, the words themselves, against accidentals, the spelling and punctuation that printing houses imposed. The verdict defended throughout is threefold. The second quarto’s fuller speech stands closest to the author’s papers and rightly supplies the copy-text. The first quarto’s speech is a performed twin, faithful where a performer’s memory is faithful and loose where memory is loose, and its variants are evidence about performance rather than corruption to be discarded. The theories of memorial reconstruction and authorial revision, tested against these forty lines, each explain part of the pattern and neither explains all of it, and the speech is too small and too star-centred a sample to settle the quarrel for the play as a whole.
The witnesses are cited throughout from the collations recorded in the modern critical editions, Weis’s Arden, Evans’s New Cambridge, Gibbons’s Arden and Levenson’s Oxford, checked against the facsimile record of the early copies. Quotations follow the second quarto’s text in modern spelling, cited by act and scene, and are kept brief, since the argument concerns the speech’s transmission rather than its beauties. Where a claim about a variant rests on the editions’ shared record it is stated flatly; where the record thins, at the smallest micro-variants, the article says so and builds nothing on the thin ice. That discipline, claim nothing the copies cannot cash, is the article’s method and its verdict’s warrant, and it is the discipline the grand theories of these texts have not always observed.
The Queen Mab Speech in the First Quarto, the Second Quarto and the Folio
Three printed texts carry the speech through its first three decades. The first quarto, printed by John Danter in 1597, offers the play as “an excellent conceited tragedy,” performed, its title page boasts, “often (with great applause) publicly” by the Lord Hunsdon’s servants. The whole text runs some six hundred lines shorter than its successor, and the Mab speech shares in that economy: the invention is complete, its famous furniture all present, but the joinery is tighter and a number of the second quarto’s lines have no counterpart. The second quarto, printed by Thomas Creede for Cuthbert Burby in 1599, announces itself on its own title page as “newly corrected, augmented, and amended,” and its Mab speech is the fullest early form, the copy from which modern editions descend. The account of that fuller text, its printers and its claims is given in the study of the second quarto of 1599 and its standing as the good quarto, and the present article takes the speech inside that text as its object.
The First Folio of 1623, the third witness, is the least independent. The Folio’s Romeo and Juliet was set from an exemplar of the third quarto of 1609, itself a reprint of the second quarto, corrected in places against theatrical copy. For the Mab speech the consequence is plain: the Folio follows the second quarto’s wording almost everywhere, modernizing spellings, adjusting punctuation and occasionally tidying a line, so that agreement between the Folio and the second quarto counts once, not twice, in any argument about what the author wrote. Where the Folio departs from the second quarto in the speech, the departure is an accidental or a sophistication far more often than a recovery of lost authority. Editors from the eighteenth century onward have therefore treated the speech as a two-witness problem, and that is the shape the collation keeps here.
Between the two real witnesses the relationship is the reverse of what their reputations suggest. The first quarto is routinely called the bad quarto and the second the good, labels inherited from an older bibliography that valued authorial papers over theatrical texts. For this speech, at least, bad is a misleading adjective. The first quarto’s Mab is not a garbled ruin of the second’s. It is recognizably the same invention, in the same scene, performing the same dramatic work, with its most quoted phrases intact. The differences gather in predictable places: in the connective passages between the famous images, in the catalogue of dreamers where the speech becomes a list, in lineation, and in the prose-versus-verse setting that the brief for this study flags as the complication to engage. A text that had been invented afresh by a pirate with a faulty ear would not show this profile. A text carried in performance memory, by contrast, would show exactly this profile, and the later sections test that inference against its rivals.
The dates frame the problem. The play was new in the middle of the 1590s, and the first quarto reached print within a few seasons of the première, while the second followed two years later with its advertised corrections. Two years is a short interval in the life of a text and a long one in the life of a repertory success: long enough for the play to settle into the company’s stock, for Mercutio’s part to become a known attraction, for the speech to acquire the polish of repetition, and for a market to form that a performed text could supply before the authorized book arrived. The sequence of the title pages tells the market’s story in miniature. The 1597 book sells the performance, the applause, the company; the 1599 book sells the text, corrected and augmented, and names the author the earlier book omitted. The Mab speech sits at the hinge of the two sales, the passage a buyer of either book most wanted, and its condition in each is the record of what each sale supplied: the stage’s version in the first, the study’s in the second.
The material survival of the witnesses sharpens the point about evidence. The first quarto survives in a handful of copies, the second in a modest number more, and until photographic facsimiles and digital archives put the pages within every scholar’s reach, collation of the speech meant travel to the libraries that hold the copies and the courtesy of their reading rooms. The Hinman collator’s era was also the facsimile era, and the two together changed the epistemology of passages like this one: a claim about the Mab variants could at last be checked by any reader with the photographs open, and the conversation moved from the authority of the traveller’s report to the discipline of the reproducible record. Modern students meet the speech’s early pages through the quarto archives and the editions’ facsimiles, and the ease of the meeting should not disguise its newness. For most of the speech’s four centuries, its variant state was simply unknown to the readers who loved it.
Why is the second quarto’s Queen Mab speech the fuller text?
The second quarto’s speech is fuller because its copy stood closer to the author’s manuscript, while the first quarto’s copy passed through performance, where the pressure of the stage trims, compresses and reorders. The fuller text preserves lines, half-lines and connectives that the shorter text drops or fuses, and its title page advertises the augmentation as a selling point.
The fuller state of the second quarto’s speech is not a matter of a line or two. The second quarto prints the complete chariot inventory at full stretch, gives the catalogue of dreamers its full sequence of lovers, courtiers, lawyers, parsons and soldiers, and carries the coda on the hag to its full uncomfortable length. The first quarto keeps every one of those units, which is itself the telling fact, but keeps several of them in compressed form, as though the speech had been carried at performance length rather than at paper length. Editors who print the speech, René Weis in the Arden edition and G. Blakemore Evans in the New Cambridge before him, take the second quarto as copy-text for the passage and record the first quarto’s compressions in the apparatus, where they belong: visible, weighed and not followed.
The direction of the difference matters as much as its size. Augmentation of a performed text by an author returning to his papers would produce the fuller second quarto; memorial compression of a written text in performance would produce the shorter first quarto. Both stories predict the same headline fact, a longer speech in 1599, and the headline alone cannot choose between them. What chooses between them is the texture of the variants, examined in the sections that follow: whether the shorter text’s departures look like cuts made by a revising hand with the whole design in view, or like the losses and substitutions of a text remembered under the conditions of the stage. On that texture the verdict of this article falls, and it falls against revision.
Collating the Speech: Where Q1 and Q2 Agree Word for Word
Collation begins with the unglamorous fact that most of the Mab speech is identical in the two quartos. The opening address, “O, then I see Queen Mab hath been with you” (1.4), stands in both texts, and so does the definition that follows: she is “the fairies’ midwife,” and she comes “in shape no bigger than an agate stone / On the forefinger of an alderman” (1.4). The size figure, the civic joke of the alderman’s ring, the fairy function of delivering sleepers of their dreams as a midwife delivers children: all of it is common property of the two texts, word for word at the level of substantive readings. A collation that starts here starts with agreement, and the agreement is the evidence everything else must be measured against, because two texts that had diverged from a lost original by independent corruption would not agree this closely in phrases this odd.
The chariot inventory strengthens the point. Mercutio builds Mab’s coach out of emptied nature: “Her wagon-spokes made of long spinners’ legs” (1.4), the cover from grasshoppers’ wings, the traces from spider’s web, “the collars of the moonshine’s watery beams” (1.4), the whip of cricket’s bone with a lash of film, and “her wagoner, a small grey-coated gnat” (1.4). These are the least forgettable and least guessable phrases in the speech, the images a parody keeps and a schoolroom memorizes. Both quartos keep them. The stability of the exotic nouns is the collation’s control group. Whatever process carried the speech from manuscript to the first quarto preserved the images that carry the invention, and preserved them verbatim, which rules out any theory on which the first quarto’s text was rebuilt from vague recollection of the plot. The reporting, if reporting it was, was reporting by someone who had the words, not the gist.
The agreement extends to the speech’s argumentative skeleton. Both texts move from the chariot to the itinerary, Mab galloping “night by night” through lovers’ brains, courtiers’ knees, lawyers’ fingers, ladies’ lips and the rest of the sleeping city, and both texts darken at the same joint, where the catalogue reaches the soldier and the speech’s charm curdles into throats cut in sleep. Both print the hag coda, the speech’s most disturbing movement, in substance: Mab as the hag who presses maids and teaches them to bear. And both end the flight the same way, with the hero’s interruption, “Peace, peace, Mercutio, peace” (1.4), and the speaker’s own deflation, the admission that he talks of dreams, “the children of an idle brain” (1.4). Skeleton, furniture and ending are shared. The speech that the literary tradition reads, the flight from whimsy to obscenity to violence that the reading of the Queen Mab speech as Mercutio’s dark dream takes as its subject, exists entire in the 1597 text as well as the 1599, and any account of the variants must begin by honouring that wholeness.
The instrument that turned such comparisons from impression into evidence deserves its sentence here. Charlton Hinman’s mechanical collator, built for his study of the First Folio, let scholars superimpose copies and catch variants the unaided eye slides past, and the culture of exact comparison it created governs modern work on these quartos: every claim in this article about agreement and difference descends from collations of the early copies, checked copy against copy, of the kind the Shakespeare Quartos Archive and the modern critical editions report. Agreement at the level claimed here is not a reader’s feeling that the speeches look alike. It is the recorded outcome of line-by-line comparison, and its weight in argument comes from that discipline.
One refinement keeps the agreement honest. The two quartos agree in substantives far more than in accidentals. Spelling, punctuation, capitalization and the pointing of the verse differ between them on nearly every line, because those features belonged to the printing houses: Danter’s men in 1597 and Creede’s men in 1599 each set the speech in their own house dress. W. W. Greg’s rationale of copy-text, the founding document of modern editing, taught editors to separate exactly these layers, following the copy-text for accidentals while choosing substantives on the evidence of authority. Applied to the Mab speech, Greg’s distinction explains the visual difference between the two early pages, which look strikingly unlike, and the verbal identity beneath the difference, which collation reveals. Readers who open the 1597 quarto and find the speech hard to recognize are meeting its accidentals. The words underneath are the words they know.
The opening line itself, “O, then I see Queen Mab hath been with you” (1.4), is the collation in miniature. It is the speech’s one moment of direct address, Mercutio reading the hero’s face and naming the visitor who has been at his dreams, and it stands in both quartos in the same words, because it is a cue as much as a line: the flight launches from it, and a transmission that lost it would lose the launch. Around that fixed point the whole comparison turns. Everything the speech is famous for behaves like the opening line, fixed; everything it merely uses to travel behaves like the connectives, loose; and the proportion between the fixed and the loose, roughly the proportion between a part’s highlights and its mortar, is the quantity the theories must explain. Stated so, the memorial account explains it in one motion, the revision account explains it not at all, and the printing-house account explains only its smallest change. The opening line’s agreement is thus not a detail but the argument’s fulcrum, the still point from which the drift on either side is measured.
The journey unit, Mab’s progress “night by night” (1.4) through the sleeping city, shows the agreement operating at the level of sequence rather than phrase. Both texts conduct the fairy over the same sleepers in the same order, lovers first, then the professional dreamers, then the soldier, and the order is the meaning: the speech is a descent through Verona’s desires, from love through money to violence, and a transmission that scrambled the order would scramble the sense. Neither witness scrambles it. Whatever else memory did to the speech in the first quarto, it kept the itinerary, and the itinerary’s survival in both texts is the strongest single proof that the two speeches are one speech, carried by two roads from one origin, rather than two treatments of a folk figure who was, as the source studies show, the dramatist’s invention in the first place.
The survival of the hag coda in the first quarto deserves emphasis, because it discriminates among thieves. A pirate assembling a saleable text from the play’s attractions would keep the chariot and the dreamers and drop the hag, the speech’s least lovely and least anthologizable matter, where Mab presses sleeping maids and the flight’s whimsy curdles into sexual threat. The 1597 text keeps the hag. Whoever carried the speech to Danter’s press carried it entire, including the part no excerpt would choose, and that wholeness is the signature of a text transmitted as a part in a play rather than mined as a jewel from a setting. The coda’s survival also anchors the speech’s interpretation across the witnesses: the dark reading of the flight, in which the dream-machinery ends in coercion, is not an editorial imposition on a innocent original but the shape both early texts transmit, and critics who build on the darkness build on ground the collation certifies.
The frame of the speech tells the same story as its core. Mercutio’s flight does not begin from silence: it answers the hero’s report of a dream and his resistance to the night’s revels, and the exchange that launches it, the hero’s “I dreamt a dream tonight” answered by Mercutio’s “And so did I” and the fencing that follows (1.4), stands in both quartos in substance, carrying the same cues in the same order. A reported text often preserves a set piece and loses its launching exchange, because the reporter remembers his own flight better than the cue that fed it. Here the cue survives with the flight, and the survival widens the island of fidelity beyond the speech’s borders into the scene’s exchange, exactly as the reporter hypothesis predicts for material in which both suspected reporters, the hero’s actor and Mercutio’s, were on stage and speaking. The island has shores, and later sections mark them; within them, the agreement of the quartos is the soundest forty lines of common ground the two texts share.
The speech headings deserve their word in the record. Both quartos assign the flight to Mercutio without ambiguity, and neither early text knows anything of the later editorial habit of printing the speech as a detachable poem with a title of its own. In the early pages the speech is an event in a scene, cued by dialogue, interrupted by dialogue, its boundaries set by the drama rather than by the anthology. Collation that forgets the frame mistakes the object: the variants of the Mab speech are variants in a piece of acted dialogue, and their explanation must be an explanation about dialogue carried between study, stage and press, not about a lyric copied between books.
Where the Texts Drift: Substitution, Transposition and the Compressed Catalogue
Against that mass of agreement, the variants cluster in three kinds, and the clustering is the article’s central evidence. The first kind is substitution in the connective tissue. Where the speech joins image to image, with small verbs, particles and transitional phrases, the first quarto repeatedly offers a different, flatter or more familiar wording where the second quarto has the sharper reading. The exotic nouns stand; the mortar between them shifts. This is the profile collators of suspect texts are trained to notice, because it is the profile of a text carried in the ear: memory grips the striking phrase and paraphrases the ordinary one, while a scribe copying a manuscript corrupts at random, striking phrase and ordinary phrase alike, and a compositor’s errors gather by the line and the page, not by the join.
The second kind is compression, and it falls heaviest on the catalogue of dreamers. The catalogue is the speech’s list machinery: Mab drives over lovers who dream of love, courtiers who dream of curtsies, lawyers who dream of fees, parsons who dream of benefices, and a soldier who dreams of throats, breaches and ambuscadoes. Lists are where performed texts economize, because a list can lose a member and still play. The first quarto’s catalogue keeps the sequence and its escalation, the movement from erotic dream to professional dream to military nightmare that gives the speech its architecture, but it carries several of its members in shorter form than the second quarto’s, fusing or trimming the elaboration around the dream’s object. The second quarto’s lawyer, dreaming “on fees” (1.4) with the tickling of the palm developed at full length, stands against a shorter first-quarto counterpart; the pattern repeats at the parson and the courtier. Compression of this kind is directional evidence. A revising author cutting his own list for a tighter second version would leave the shorter text as the later one. Here the shorter text is the earlier in print, and the fuller text advertises itself as augmented, which fits a performed text set against the papers behind it.
The third kind is the divided locus, the single word where the witnesses part company and editors must choose. The speech’s most discussed example is the word for Mab’s team, the tiny draught creatures drawn in the line that modern editions print as a team of “little atomies” (1.4), a word the early texts hand down in divided form and commentators have glossed as atoms, sprites and pygmies. The locus matters beyond its size because it shows what a substantive variant is: one word, no difference of image, and yet a genuine fork in transmission, since one of the early forms must descend from the copy and the other must have arisen in the chain of copying, performance or printing. Editors print the second quarto’s tradition and note the division, and the note, multiplied across forty lines, is what a collation of this speech actually looks like: a stable text carrying a fringe of small, recorded, argued-over differences.
Transposition, the movement of lines or blocks from their order in one witness to a different order in another, completes the standard diagnostic set, and its near-absence in the Mab speech is as informative as its presence elsewhere in the first quarto. The whole-text collation of the first quarto records rearrangement among its memorial features; within the Mab speech, the units stand in the second quarto’s order, opening, inventory, catalogue, coda, interruption, without the wholesale reshuffling that free reconstruction from memory produces in weaker reported passages. The speech’s internal order is load-bearing, since the catalogue’s escalation is its meaning, and the order survives. What varies is wording at the joins, fullness in the list, and the disposition of lines on the page. That is a narrow, specific profile, and the theories weighed later in this article stand or fall by how well they explain it.
What is a transposition, and does this speech contain one?
A transposition is a block of text, a line, a phrase or a whole unit, that stands in one position in one witness and a different position in another. Editors treat transpositions as strong evidence about how a text travelled, because copyists rarely move blocks while memory does. Within the Mab speech, no unit changes place between the quartos.
The qualification matters. At the level of the whole first quarto, transposed material is part of the standard case about that text’s origins, and whole-text studies find speeches and half-scenes standing where the second quarto does not put them. Inside the Mab speech the architecture holds: the chariot precedes the journey, the journey precedes the hag, the hag provokes the interruption. Small units within the catalogue are the only candidates, and even there the recorded differences are of fullness more than of order. A theory of the speech’s transmission must therefore explain a text whose structure memory kept and whose surface memory loosened, which is a more precise demand than the slogan of a garbled quarto suggests, and a more interesting one.
The soldier’s unit shows the drift at its most characteristic. In the second quarto the soldier dreams “of cutting foreign throats, / Of breaches, ambuscadoes, Spanish blades” (1.4), a triad of military furniture as exotic in its way as the chariot’s, and the unit stands in the first quarto in substance with its wording at the joins loosened in the quarto’s usual manner. The pattern by now will be familiar: the strange nouns hold, the syntax around them gives. Spanish blades and ambuscadoes are not phrases a reporter invents or a compositor guesses; they are phrases a memory keeps because they are strange, and their keeping, inside the speech’s darkest unit, confirms that the fidelity of the first quarto extends to the flight’s violence as faithfully as to its whimsy. The dream-work’s whole arc, charm to threat, is under the same transmission, and under the same verdict.
One further class of difference runs quietly beneath the three argued kinds, and completeness requires its paragraph: the small change of number, tense and connective that alters no image and yet alters the sentence’s gait. Collations of the speech record such micro-variants in their dozens across the two quartos, and editors pass over most of them in silence, since the apparatus would swell beyond any reader’s patience if each were argued. Their collective testimony nevertheless counts. Micro-variants of this order are the statistical background of transmission, the noise floor against which the significant variants, the compressions, the divided team word, the prose setting, stand out as signal. A collation that reported only the striking differences would flatter every theory, because each theory could cherry-pick a fringe; the discipline of the Mab collation is that the fringe is small, the background is ordinary, and the patterns claimed in this article are claimed against the whole recorded field, not against a selection.
The InsightCrunch Mab variants: The Speech Side by Side
The collation argued in prose so far is set out below as the InsightCrunch Mab variants, a side-by-side of the Queen Mab speech in the two quartos, unit by unit, with the Folio’s position and the modern editorial outcome recorded for each unit. The table is the quickest route into the article’s verdict. Read down the first-quarto column and the pattern announces itself: agreement in the famous matter, compression in the list, looseness at the joins, and a setting of the lines that differs from its rival’s. Read down the outcome column and the editorial settlement appears: the second quarto printed, the first quarto recorded, the Folio treated as a descendant rather than a judge. The rows follow the speech’s own order, from the opening address to the interruption that closes it, so that the table also serves as a map of the speech for readers meeting its textual history for the first time.
| Speech unit | Q1, 1597 | Q2, 1599 | Editorial outcome |
|---|---|---|---|
| Opening address to the hero | Agrees in wording | Copy-text form | Q2 printed |
| “The fairies’ midwife” and the agate-stone size | Agrees | Agrees | Unchanged in all editions |
| The chariot: spokes, cover, traces, collars | Exotic nouns agree; connectives vary | Full form | Q2 printed, Q1 variants noted |
| The team word, “little atomies” | A divided early form | Copy-text form | Printed with a commentary note |
| The wagoner, “a small grey-coated gnat” | Agrees | Agrees | Unchanged |
| The journey, “night by night” through sleepers | Agrees in substance | Full form | Q2 printed |
| The dreamers’ catalogue: order of dreamers | Sequence kept; members compressed | Full sequence at full length | Q2 sequence and fullness printed |
| The lawyer on fees, the parson on benefices | Among the most compressed units | Full form | Q2 printed |
| The soldier’s coda of throats and breaches | Agrees in substance | Full form | Q2 printed |
| The hag lines | Agrees in substance | Full form | Q2 printed |
| Setting of the lines on the page | Lineation differs from Q2 | Verse partly set as prose in places | Re-lineated as verse from Q2 |
| Spelling and punctuation throughout | Danter’s house dress | Creede’s house dress | Modernized from Q2 copy-text |
| The interruption and “children of an idle brain” | Agrees | Agrees | Unchanged |
Two rows carry the argument of the whole article. The catalogue row shows the first quarto behaving like a performed text, its list intact in design and trimmed in flesh, while the setting row shows the second quarto behaving like a printed book under production pressure, its verse squeezed into prose at points where the page, not the author, was in charge. Neither row fits a simple morality of good text and bad. The first quarto’s weakness is memorial; the second quarto’s weakness is mechanical; and the speech readers know is the negotiated settlement between a text that was played and a text that was printed, assembled by editors who weighed both. The sections that follow take the three great explanations of the variants in turn, the printing house first, since its effects are the most local and the most often mistaken for deeper causes.
A word of instruction on the table’s language, since its honesty lives in its verbs. “Agrees” means agreement in substantive wording, the words themselves, beneath whatever difference of spelling the two shops imposed. “Agrees in substance” means the unit’s matter, its images and its function, stand in both texts while small wording at the joins divides them. “Compressed” means the unit stands shorter in the first quarto with its design intact. No row uses “corrupt,” “garbled” or “wrong,” because the collation does not find those things in this speech, and the discipline of the verbs is the discipline of the whole study: the first quarto’s Mab is not a failed copy of the second’s but a successful record of a different event, the speech as played, and the table’s columns keep the two events distinct so that the reader can weigh them without confusion.
Two rows repay a second look. The team-word row, the smallest in the table, is the only one where the outcome column must confess a judgement call dressed as a fact: editors print a form of the word for Mab’s draught creatures and gloss it, and the gloss, atoms or sprites or pygmies, quietly chooses what the fairy is, a natural philosopher’s mote or a folkloric imp. The interruption row, by contrast, is the table’s certificate of health: entry and exit agreed, the frame sound in both witnesses, the speech’s dramatic function identical in 1597 and 1599. Between those rows the whole range of the collation’s evidence is spanned, from the fork where authority must be argued to the ground where no argument is possible, and a reader who learns to tell the rows apart has learned the difference between a text’s settled country and its disputed border, which is most of what textual scholarship knows.
Verse Set as Prose: The Mab Speech and the Printing House
The complication the brief for this study insists on is the setting of the speech’s verse. In the second quarto, stretches of the Mab speech that scan and read as verse stand on the page set as prose, run on in justified lines that fill the measure, while the first quarto disposes the same material with its own lineation, differing from the second quarto’s arrangement and from any modern edition’s. A reader who meets the speech first in a facsimile of the 1599 quarto meets a passage that looks like a prose rhapsody with verse trapped inside it. The fact is bibliographical before it is anything else, and it must be explained by the printing house before it is allowed to testify about the author’s manuscript or the actors’ memories.
The standard explanation is economic. Quarto copy was cast off, counted and apportioned to compositors before setting began, and a compositor who found his allotted copy overrunning its allotted pages had remedies: abbreviate, crowd, and set verse as prose, which packs more words into the page than verse lines with their ragged right margins. Peter W. M. Blayney’s work on the printing of King Lear established the model of such analysis for Shakespeare’s texts, reconstructing from the physical page what the printing house did to the copy in front of it, and D. F. McKenzie’s account of printers of the mind supplied the human frame: compositors were piece-workers solving daily problems of fit, not scholars collating an author’s intentions. On that model, prose setting in the second quarto’s Mab speech testifies to pressure in Creede’s shop, a stint where the copy ran long, and it testifies to nothing about the copy itself. Verse in the manuscript can stand as prose on the page whenever the page demands it.
The first quarto’s differing lineation cuts the other way and must not be conscripted too quickly. A performed text taken down or reconstructed would carry the lineation of nobody’s manuscript, since verse lines are units of print, not of speech; an actor remembers phrases and rhythms, not line endings. The first quarto’s arrangement of the speech, which neither matches the second quarto’s verse nor merely copies its prose, is consistent with a copy that arrived at Danter’s shop already disposed by transmission rather than by an author’s hand. But consistency is not proof, and the honest verdict on lineation is that it is weak evidence in both directions: the printing house explains the second quarto’s prose, transmission plausibly explains the first quarto’s disposition, and neither fact can bear the weight of a whole theory of the speech. Editors therefore re-lineate. Modern editions print the speech as the verse its metre declares, following the second quarto’s wording and restoring the line divisions the verse requires, and record the early settings in their textual notes for the specialists who need them.
Why would a compositor set verse lines as prose?
A compositor set verse as prose to save space. Verse lines waste the right-hand side of the page, and prose lines fill the full measure, so a compositor whose cast-off copy threatened to overrun its allotted pages could recover room by running verse on as prose. The choice records the page’s economics, not the copy’s form.
The practice also explains a feature of the second quarto that puzzles first-time readers of the facsimile: the prose setting arrives in stretches rather than at random, because the pressure that caused it was local, a forme or a stint where the count went wrong, and it lifts when the pressure lifts. Bibliographers read such stretches the way engineers read stress marks, as evidence about the object’s manufacture. For the Mab speech the consequence is a discipline of interpretation. Nothing about the author’s draft, the actors’ performance or the speech’s meaning follows from its prose dress in 1599, and any argument that treats the setting as a window on the copy commits the elementary error of reading the printing house as the author. The variants that matter for the speech’s history are the verbal ones, and they are the subject of the theories weighed next.
What re-lineation gives back to the speech is its breath. Set as verse, the catalogue recovers its units, one dreamer to a measure, and the reader hears the acceleration the lines encode, the rush from lover to soldier that performance exploits and the prose page conceals. Set as prose, the same words read as a rhapsody, brilliant and unshaped. The difference is not cosmetic, and it is the strongest practical proof that the 1599 setting is an accident of manufacture: no author writes a crescendo and then hides its stairs, and no editor who restores the stairs is embellishing the text, only unbuilding the press’s economy. The actor’s Mab and the reader’s Mab both depend on the restoration, and the textual note that records what was restored is the toll the edition pays for the rescue, a small print acknowledgement that the speech’s natural shape had to be recovered from its first full printing.
The mechanics of casting off deserve one paragraph of precision, since the prose setting is unintelligible without them. Before a quarto was set, the copy was cast off: a workman counted its words or leaves and marked the points at which each page and forme would begin, so that compositors could set their stints out of sequence and the sheets could be machined in the order the press required. The system worked when the count was right. When the copy was verse, with its unpredictable line lengths, or when the count was hurried, a compositor could reach the end of his stint with copy left over and no page to put it in, and the remedies were the ones traced above: crowd the type, abbreviate the words, run the verse on as prose. The second quarto of this play bears the marks of such pressure in more places than the Mab speech, and bibliographers read the speech’s prose stretches as one instance of a book-wide condition, which is the final reason the setting cannot be evidence about this passage in particular. What happened to the Mab lines in Creede’s shop happened, in kind, across the volume, wherever the count went wrong.
Mab Among the Suspect Texts: The Speech Beside the Other Bad Quartos
The vocabulary of this inquiry has a pedigree worth pausing on, because the Mab speech’s behaviour is anomalous within its own family. Alfred W. Pollard’s 1909 study of the folios and quartos coined the category of the bad quarto for early texts he judged stolen and surreptitious, and Peter Alexander’s work of 1929 supplied the mechanism, arguing that certain short quartos were reconstructed from memory by actors for a pirate press. The family assembled under the label includes the first quartos of Hamlet, Henry V, The Merry Wives of Windsor and Pericles alongside this play’s 1597 text, and the family resemblance is real: shortened texts, degraded verse, paraphrase in the minor scenes, stage directions that breathe the air of the playhouse. Set the Mab speech beside its cousins’ famous passages and its quality startles. The 1603 Hamlet’s version of the great soliloquy, “To be, or not to be, I there’s the point” in its garbled opening, shows what memorial transmission does to a star set piece when the chain of reporting is weak: the most famous lines in the language arrive mangled. Mercutio’s set piece, by contrast, arrives with its famous lines intact, and the contrast measures the strength of the chain behind the 1597 Romeo and Juliet, a chain that ran through the part’s own player or through copy close to the papers.
The comparison disciplines the label. Bad quarto names a publishing event, a text issued without the company’s blessing from materials the study did not supply; it does not name a uniform level of corruption, and the texts gathered under it range from the garbled to the nearly respectable. Scholarship since Maguire and Werstine has grown wary of the label’s lumping, preferring to describe each early text’s descent in its own terms, and the Mab collation shows why the wariness is earned. Within a single suspect text, quality varies by scene, by part and by passage, and the variation is the evidence: a label that flattens it into one judgement about one book destroys the data the book contains. The 1597 Romeo and Juliet, judged at the Mab speech, is a competent performed text with a star part carried at high fidelity. Judged at its weakest scenes, it is a sketch. Both judgements are the same book, and the theories that matter are the ones that predict the difference between them.
The Henry V quarto of 1600 offers the closest structural parallel, another shortened performed text in which the great speeches survive better than the connective scenes, and editorial tradition there long ago settled into the posture this article recommends for Mab: print the authoritative text, record the quarto’s readings, and mine the quarto for the stage. The Merry Wives quarto adds the cautionary extreme, a text so thoroughly mediated by memory that its set pieces wobble, proving that the Mab speech’s stability was not guaranteed by its fame. Fame alone does not preserve a text; the Hamlet soliloquy was famous too. What preserves is a short chain of custody, and the Mab collation, read against its family, testifies to a short one: from the papers to the stage, from the stage’s own Mercutio to Danter’s press, with the second quarto’s papers waiting to correct the record two years later.
Memory’s Fingerprint: The Mab Speech and Memorial Reconstruction
The oldest explanation of the first quarto is the boldest: that its text was reconstructed from memory, by actors who had performed the play, and sold to a printer as a performed text for a market the authorized book had not yet supplied. Harry R. Hoppe’s 1948 study of this quarto made the case in its classic form, and George Ian Duthie’s work on the bad quarto of Hamlet had already supplied the diagnostic kit: a memorially reconstructed text betrays itself by paraphrase where memory failed, by anticipation and recollection, phrases arriving early or repeated from elsewhere in the play, by transposed blocks, by metrical breakdown, and by a peculiar unevenness, in which the parts the reporters played stand full and accurate while scenes they watched from the wings shrink to sketches. Hoppe argued that the reporters behind the 1597 Romeo and Juliet included the actors of the hero and of Mercutio, and that prediction is the hinge on which the Mab collation turns, because the speech is Mercutio’s great solo and the theory predicts it should survive in the first quarto in unusually good condition.
The collation bears the prediction out, with one refinement. The Mab speech is not merely present in the first quarto; it is present at a level of verbal fidelity that the first quarto does not sustain for its minor scenes, with its signature images verbatim, its architecture intact and its ending worded as the second quarto words it. Set beside the whole-text profile of the 1597 quarto, where scenes shrink drastically and verse degrades into paraphrase, the speech stands out as a peak of accuracy, and it stands on exactly the peak the reporter theory builds: the set piece of a suspected reporter. Mercutio’s part in the first quarto is throughout among its fullest, and the Queen Mab speech, the part’s showpiece, is its fullest sustained stretch. A theory that predicts a distribution, and finds it, has earned more than a hearing. The drift the collation records fits the same hand: connectives paraphrased while images hold, the catalogue compressed where a list can bear compression, the whole loosened at the joins and firm at the highlights, which is the shape of a speech remembered by the man who nightly spoke it, not the shape of a speech copied, stolen or invented.
The sceptical audit must then be faced at full strength, because Laurie E. Maguire’s Shakespearean Suspect Texts subjected the memorial diagnosis to a statistical discipline it had never met. Maguire tested the canonical suspect texts against the agreed markers of memorial reconstruction and found the markers scattered far more widely than the theory allows: anticipations, recollections and paraphrases occur in texts nobody suspects, memorial criteria convict the innocent, and several parade examples of bad quartos failed her tests or passed them equivocally. Her verdict on the method was severe, and its consequence for the Mab speech is direct. No single marker in the speech, not the catalogue’s compression, not the loosened connectives, not the fidelity of the star part, can convict the first quarto’s copy of memorial origin on its own, because each marker has innocent causes: authors compress their own lists in draft, compositors loosen connectives, and an accurate passage in a reported text may be accurate because the copy behind the whole book was better than its reputation. Maguire’s discipline forbids the older criticism’s habit of pointing at a variant and pronouncing the word memory as though the pointing were the proof.
Adjudicated, the disagreement resolves into a verdict that honours both parties. Maguire wins on method: the Mab variants, taken one by one, prove nothing, and any account of the speech that rests on a favourite variant collapses under her audit. Hoppe wins on the pattern: the distribution of fidelity across the first quarto, peaking in the hero’s and Mercutio’s material and peaking within that material at this speech, is a fact no rival theory predicts as cleanly, and the speech’s specific combination, verbatim highlights, paraphrased joins, compressed list, intact order, is the cumulative profile of a performed part carried in memory rather than of a manuscript copied or a draft superseded. The verdict for this speech is therefore a qualified memorial one. The first quarto’s Mab descends from the play in performance, most plausibly through the memory of its performers, and its variants are performance evidence: precious for the theatre historian, secondary for the editor choosing words. What the verdict does not license is extension. A star’s set piece is the passage a memorial theory would most expect to survive, so the speech is the theory’s easiest test, not its decisive one, and the quarrel over the whole first quarto must be fought on the whole text’s evidence, where the case is harder and the audit stricter.
Paul Werstine’s critique of the narratives editors tell about printed texts adds the final caution, from the opposite flank. Werstine showed how the stories of foul papers and bad quartos hardened, in twentieth-century editing, from hypotheses into scenery, invoked to explain whatever an editor’s copy-text did not, and he pressed for the evidence of the documents themselves against the elegance of the tales. Applied here, his caution trims the memorial verdict to its evidence. The collation shows a speech that is accurate where a performer’s memory is accurate and loose where it is loose; the inference to actors reconstructing the play for Danter’s press is the best available explanation of that showing, and it remains an inference, one the speech supports and cannot seal. The article’s verdict keeps both clauses, and distrusts any retelling that drops the second.
Levenson’s editorial treatment of the 1597 text gives the memorial verdict its institutional form. To edit the first quarto as a document in its own right, with its own apparatus and its own claims on the reader, is to concede in practice what the Mab collation argues in theory: that the shorter text is a record of the play in the theatre, worthy of the same scholarly instruments as the author’s papers, because it answers questions the papers cannot. The papers cannot say how the speech played, how much of it an audience heard, where the performed version economized and where it insisted on fullness. The first quarto, read as its editor teaches, says all of that, and says it nowhere more clearly than in the Mab speech, whose performed shape, highlights at full stretch and catalogue at stage length, is the closest thing the period affords to a promptbook’s ghost. The memorial theory, whatever its excesses elsewhere, bought scholarship that ghost, and the purchase stands.
The economics of the 1597 book complete the memorial picture. A play in repertory was the company’s capital, guarded in manuscript, and the route by which performed plays reached print without authority ran through the people who carried them in memory: the actors themselves, a hired reporter in the audience, or a text assembled from the parts. Danter’s quarto appeared without the usual entry in the Stationers’ Register, a silence consistent with a copy the trade knew to be irregular, and its title page advertised the one asset a reported text genuinely possessed, the performance itself, applause and all. The Mab speech was that asset’s showpiece, the turn a buyer who had seen the play would turn to first, and its high finish in the quarto is therefore doubly explained: the reporter’s own part, and the market’s demand that the famous flight be worth the pennies. A reported text lives or dies by its set pieces, and this one was built, by memory and by commerce together, to live by Mercutio’s.
The marker-by-marker application of Duthie’s kit to the speech, promised by the method and owed to the sceptics, runs as follows. Paraphrase: present, at the connectives, in the milder form of flatter wording for sharper. Anticipation and recollection: effectively absent within the speech, no phrase imported from the play’s later scenes, no image arriving before its cue. Transposition: absent at the level of units, the architecture unmoved. Metrical breakdown: present at the prose-set stretches of the rival text rather than in this one, and charged there to the press. Unevenness by part: present at the level of the whole quarto and visible in miniature, the speech outstripping its own scene’s frame in finish. The scorecard is mixed in exactly the way a genuine performed text should be mixed, strongest where the theory is strongest, silent where the theory makes no claim, and innocent of the florid corruptions that disgrace the family’s worst members. Hoppe’s case, audited by Maguire’s rules, survives as a pattern judgement; it would not survive as a checklist, and the difference between those two survivals is the present state of the question.
One more consideration, drawn from the theatre rather than the library, supports the pattern. A set piece of this kind, a solo flight applauded nightly, leads a double life in a repertory company: it is spoken in performance and it is quoted, imitated and shown off beyond performance, in the tiring house and the tavern, wherever actors display their wares. Such a passage is rehearsed by the culture around the play as well as by the player within it, and its words acquire a fixity that quieter dialogue never earns. The Mab speech’s fidelity in the first quarto may therefore overstate even its own reporter’s memory, since the speech was memorable to everyone who heard it, and its survival is the survival of the most over-learned forty lines in the play. The caution trims the memorial argument’s triumph without touching its substance: the speech was always going to be the first quarto’s best page, and the theory that predicted a best page here merely had the easiest of its predictions confirmed.
Revision’s Second Thoughts: Testing the Authorial Explanation
The rival explanation deserves its strongest form before its sentence. On the revision hypothesis, the first quarto preserves an earlier authorial state of the play, Shakespeare’s first version of the speech, and the second quarto prints his revised and expanded second thoughts. The model has a distinguished modern precedent: the two-text King Lear, where Stanley Wells and Gary Taylor persuaded a generation that the quarto and Folio texts embody distinct authorial versions, and editors of the Oxford Shakespeare printed both. If Lear can exist in two authorial shapes, the argument runs, a Romeo and Juliet in two shapes is no scandal, and the Mab variants would then be the record of a young dramatist improving his most famous invention, tightening here, expanding there, between a first draft and its revision. The hypothesis flatters the first quarto, promotes its variants from errors to readings, and would require editors to weigh its Mab speech as an authorial alternative rather than a performed derivative.
Tested against the collated speech, the hypothesis fails on the direction and texture of the variants. Revision, as the Lear precedent defines it, is patterned: cuts that remove a design’s excess, additions that sharpen its emphases, changes that can be narrated as intentions. The Mab variants between the quartos refuse that narration. The first quarto’s compressions fall on the catalogue’s elaboration without altering its design, its substitutions flatten the connectives without redirecting any image, and its text shows none of the local rethinking, the swapped image, the reconsidered ending, the re-aimed joke, that authorial second thoughts leave behind. A dramatist revising the Mab speech had forty famous lines to improve and, on the revision hypothesis, improved them by making their joins plainer and their list shorter in the version printed first and then reversed every one of those decisions in the version printed second, restoring fullness everywhere and changing almost nothing that matters. That is not a pattern of revision. It is a pattern of loss and recovery, and loss and recovery is the biography of a text that left the papers, lived on the stage and was imperfectly carried back to print, while the papers themselves went to Creede’s press in their fuller state.
Two details of the collation press the point home. The speech’s most inventive matter, the chariot built from spiders’ legs and moonshine, is identical in both texts; revision works hardest exactly there, on the invention itself, and there is no revision there. And the speech’s frame, the opening address and the closing deflation about dreams and the idle brain, is likewise stable; a reviser re-aims frames, and the frame never moves. What moves is the middle distance of the speech, its connective wording and its list’s flesh, the matter a memory drops and a reviser has no motive to touch. The verdict follows the evidence: for the Queen Mab speech, the variants between the quartos are transmission variants, not authorial ones, and the first quarto’s speech has the authority of a performance, not of a draft. The hypothesis is not absurd in itself, and the Lear precedent forbids dismissing two-text authorship in principle; it is simply not what these forty lines contain.
The draft hypothesis draws strength from a romance that should be discounted before the evidence is weighed: the romance of the young Shakespeare, the apprentice dramatist whose first sketches the quartos supposedly preserve, improving his craft in public until the masterpieces arrive. The romance flatters a developmental story the canon elsewhere refuses, since the plays of the mid-1590s arrive in their authoritative texts already masterful, and it asks the first quarto to be a window into a workshop that no other document shows. Windows into workshops are precious when they exist, and the desire for one is understandable; but desire is not evidence, and the Mab collation, the one place where the window should be clearest because the invention is most famous, shows no workshop at all. It shows a finished invention, identically finished in both texts, and a performed copy of it losing flesh on the road to print. The romance explains the hypothesis’s popularity better than the hypothesis explains the variants.
The Lear contrast, stated in full, shows the gap the revision hypothesis must cross. Between the quarto and Folio Lears stand differences of design: whole episodes present in one text and absent from the other, a trial scene cut, the play’s closing lines reassigned from one speaker to another, hundreds of lines’ worth of divergence that an interpreter can narrate as a mind returning to its work. Between the two Mab speeches stand differences of degree: the same episodes, the same order, the same speaker, the same jokes, differing in the flesh on the catalogue’s bones and the mortar between the images. To call both phenomena revision is to empty the word. The Lear texts argue; the Mab texts merely vary, and variation without argument is transmission’s signature, the wear a speech takes travelling roads the author never walked. Editors who print the two Lears as versions honour a real distinction; an editor who printed the two Mabs as versions would be manufacturing one.
Accidentals and Substantives: The Printing House Inside the Speech
Every variant discussed so far has been a substantive, a difference in words. The pages of the two quartos also differ in a thousand accidentals, spellings, punctuation marks, capitalizations, elisions, and those differences, properly understood, discipline the whole inquiry, because they mark the boundary between what a printing house could do to the speech and what it could not. Danter’s compositors in 1597 and Creede’s in 1599 worked to different house habits, and the same sentence of Mercutio’s emerges from the two shops dressed in different orthography, pointed to different rhythms, its verse disposed by different economies. An editor who modernizes, as all the editions named in this article modernize, strips both dresses away and prints the substantives in modern spelling, an act of mercy to readers that quietly destroys the evidence of manufacture. The collator’s first lesson, learned from Greg’s rationale and practised by Fredson Bowers and his school, is to read through the dress to the words, and to charge the dress to the shop.
The distinction carries a verdict about agency. Compositors could misread a word, set a synonym from the pull of a familiar phrase, drop a connective, transpose letters, and sophisticate a hard reading into an easy one; the Mab speech’s fringe of small variants is compatible with such shop-floor noise at a low level, and no argument in this article requires every first-quarto variant to be an actor’s rather than a printer’s. What compositors could not do is redesign a speech: preserve its architecture entire, keep its rarest images verbatim, compress its catalogue in performance-shaped ways and loosen its joins in memory-shaped ways. That pattern is larger than the shop. The printing house explains the texture of each page, the dress, the prose setting, the hundred small accidents; it does not explain the relation between the two texts, and the attempt to make it do so, occasionally heard when the memorial theory falls from fashion, fails on the same distribution that convicts memory, since shop noise falls evenly and this speech’s variants fall in the shape of a part performed and recalled.
A further bibliographical subtlety belongs in any complete collation of the speech: copies of a single early edition are not all identical. Stop-press correction, the practice of halting the press to amend a type error and then binding corrected and uncorrected sheets together indifferently, means that two surviving copies of the second quarto can differ at a given line, one carrying the error and one the fix. Hinman’s Folio work made press variants famous, and the discipline applies to the quartos as well: a collation of the Mab speech is strictly a collation of copies, and a variant attested in a single copy may be a press state rather than a reading of the edition. Editors record such states where they matter, and the Mab speech’s apparatus, in Weis’s Arden edition and in the collations behind it, distinguishes the edition’s text from its copies’ accidents. For the argument of this article the point is a caution with a cutting edge: the evidence base for forty lines of verse is a handful of surviving copies, each an individual manufactured object, and certainty about any single small variant should be held at the strength the copies allow.
The accidentals of the speech repay one direct look, because they carry a kind of evidence substantives cannot. The two shops spell the fairy’s world differently, point Mercutio’s rush differently, and divide his sentences at different breaths, and those differences map the compositors’ stints and habits across the passage as surely as a fingerprint maps a hand. Where the second quarto’s pointing hurries the catalogue into breathless runs and the modern editor must repoint to rescue the sense, the page is confessing the speed of its manufacture. Bibliographers prize such confessions precisely because they are unselfconscious: no compositor set out to testify about casting off, yet the testimony stands in every crowded forme. The Mab speech, so often read as the play’s freest flight of fancy, is in its early pages among the most manufactured objects in the canon, and the manufacture is legible to any reader taught where to look, in the dress the words wear rather than in the words themselves.
What separates a substantive variant from an accidental one?
A substantive variant changes the words and so can change the meaning: a different noun, verb or phrase where the witnesses divide. An accidental variant changes only the dress, the spelling, punctuation or capitalization of the same words. Editors judge substantives for authority and take accidentals from the copy-text, because printers imposed the dress.
Greg’s rationale, the 1942 argument that still structures the editions on the shelf, draws the line where the printing house’s authority ends. Printers were licensed, by the practice of their trade, to normalize the dress of whatever copy reached them; they were not licensed to rewrite, and where their setting rewrites, by error or sophistication, the editor’s duty is to detect and repair. The Mab speech shows both halves of the rule at work. Its accidentals in both quartos are the shops’ property and no editor defends them. Its substantives divide along the lines the earlier sections traced, and there the editor must judge: the second quarto’s readings carry the authority of its copy, the first quarto’s carry the interest of performance, and the Folio’s, where they differ from both, carry the suspicion due to a descendant caught tidying its ancestors. The framework is a century’s editorial common sense, and the speech, small as it is, exhibits every category the framework names.
How Editors Assemble the Queen Mab Speech Readers Know
The speech in a modern paperback is a construction, and its blueprint is consistent across the major editions. The copy-text is the second quarto. Brian Gibbons, in the second Arden edition, built the play on the 1599 text and consulted the first quarto throughout; G. Blakemore Evans, for the New Cambridge, did the same with an apparatus that lets a reader reconstruct the early evidence; René Weis, in the Arden Third Series edition, prints the second quarto’s speech with the fullest modern collation of the passage, weighing first-quarto and Folio variants in the commentary and admitting a first-quarto reading only where the second quarto’s text limps. Jill Levenson’s Oxford edition likewise rests on the second quarto, while Levenson’s wider editorial work on the 1597 text, treating the first quarto as a document worth editing in its own right rather than as a quarry of variants, represents the strongest modern form of the case for the shorter text’s dignity. Stanley Wells and Gary Taylor, whose Oxford Complete Works set the methodological weather for a generation, supplied the frame in which such choices are argued: copy-text chosen on the evidence of descent, emendation defended reading by reading, and the early texts’ claims heard even when they lose.
The disagreement among these editors is not over the speech’s words, where their texts converge almost completely, but over the first quarto’s standing, and it is worth adjudicating because readers inherit its consequences. On one side stands the traditional settlement: the first quarto is a derivative, reported text, useful as a record of performance and as an occasional corrective, but without authority where the second quarto is sound. On the other stands the revisionist courtesy, extended by Levenson’s editorial practice and by the modern theatre’s appetite for the shorter, faster 1597 play: the first quarto as a version, with its own integrity, to be read and even played as itself. For the Mab speech the adjudication is this. As a source of readings, the traditional settlement holds, and holds easily: the collation shows a performed derivative, and no editor should print its compressions against the second quarto’s fullness. As a document, the revisionists hold: the first quarto’s Mab is the closest surviving approach to the speech as an Elizabethan audience heard it from the stage, compressed at performance tempo, and scholarship that treats it only as a failed copy of 1599 wastes its best evidence. Both holdings can be true because they answer different questions, and the confusion of the two questions, authority and interest, is the commonest error in popular accounts of the bad quarto.
The working scholar’s Mab, therefore, lives in several places at once: the reading text in Weis or Evans, the facsimiles of the two quartos for the page’s own testimony, the apparatus for the fringe of variants, and a notebook, paper or digital, where the collation’s rows accumulate. Keeping such a collation, with the speech’s units in order and the witnesses’ readings aligned, is precisely the labour the InsightCrunch Mab variants table abbreviates, and research notes of this kind repay a durable home: a personal research archive such as VaultBook keeps the collation beside the editions it was drawn from, so that the next passage collated joins a growing record instead of a scattering of slips. The point is methodological rather than commercial. Textual scholarship is cumulative at the level of the individual reader, and the reader who has collated Mab owns a standard against which every subsequent claim about the quartos can be tested.
What no modern editor does is print a conflated Mab, a speech assembled by picking the favourite line from each witness. The conflation temptation is real, because the first quarto occasionally preserves a reading that charms, and the Folio a tidiness that soothes, but the discipline of copy-text exists to refuse it: a text assembled by taste is nobody’s historical document, neither the author’s nor the stage’s nor the printing house’s. The speech readers know is the second quarto’s speech, modernized in dress, re-lineated where the press squeezed it, emended at a handful of argued points, and shadowed in the apparatus by its performed twin. That is the editorial assembly the series thesis names: the cliché of the fixed famous speech dissolves, under collation, into a documented set of choices, each defensible, none inevitable, all invisible until the witnesses are set side by side.
The digital turn has changed the speech’s scholarly habitat once more. The Shakespeare Quartos Archive and the Internet Shakespeare Editions place the early pages themselves before any reader with a screen, the 1597 Mab and the 1599 Mab a click apart, and the effect on argument has been wholesome: claims about the passage are now made in the presence of the evidence, and a student can test this article’s collation against the photographed pages in an afternoon. The editions named here were built in the older habitat, on travel and facsimile, and their convergence with what the screens now show every reader is a further check on the verdict. When the witnesses became universally visible, the settlement they support did not change, because the settlement had been reached by looking at exactly what the screens display.
The Folger Shakespeare, edited by Barbara A. Mowat and Paul Werstine, the edition through which most students now meet the play, prints the same settlement in its most widely read form: a modernized second-quarto speech, its early variants summarized in notes pitched to the general reader, the first quarto honoured as a performance document in the textual essay. The convergence of the Folger with Arden, Cambridge and Oxford on the speech’s words, reached by editors of different schools across four decades, is itself a datum for the collation’s verdict. Editorial convergence happens when the evidence admits one reasonable outcome, and it breaks down, as in the two-text Lear, when the evidence sustains two. For the Mab speech the evidence sustains one text and one shadow, and every serious edition prints accordingly. The remaining editorial work on the passage is the work of the notes, the keeping of the shadow visible, and it is there, rather than in any fantasy of a better conflated speech, that the first quarto’s claims are now prosecuted and, in their proper court, allowed.
Why do editors follow the second quarto for this speech?
Editors follow the second quarto because its copy stood nearest the author’s manuscript, and the collation of the speech confirms the standing: the fuller text carries the complete invention at full length, its variants’ rivals explain themselves as performance loss or press noise, and no witness offers a competing authority of the same kind.
The Folio’s claim fails first, since a descendant cannot outvote its ancestor. The first quarto’s claim fails second, on the texture examined in this article: a text may be a precious performance record and still be a derivative one, and derivation is what the loosened joins and the compressed catalogue record. What remains is the 1599 text, not because it is flawless, its prose-set stretches and compositorial accidents forbid that word, but because its flaws are the flaws of manufacture, corrigible by editing, while the first quarto’s differences are the differences of a different mode of existence, the stage’s, which editing cannot convert into the author’s page without destroying the evidence it values.
The Speech’s Afterlife in Print: Q3, the Folio Page and the Editors’ Centuries
Between the second quarto and the modern edition stands a chain of reprints that the collation must account for, because each link explains some feature of the text readers inherit. The third quarto of 1609 reprinted the second quarto page for page in substance, passing the speech onward with its readings intact and its accidentals renewed, and it was an exemplar of that third quarto, corrected against theatrical copy, that the Folio’s compositors set in 1623. The Folio page therefore shows the speech at two removes from the author’s papers and one remove into the theatre’s keeping: its agreements with the second quarto are inheritance, its occasional disagreements are the corrections of a playhouse document or the errors of a new shop, and its authority, for this speech, is that of a witness to performance practice c. 1623 rather than to composition c. 1595. Collators who cite the Folio’s Mab are citing the speech’s stage afterlife, and the distinction disciplines every argument that counts heads among the witnesses.
The eighteenth-century editors, Nicholas Rowe, Alexander Pope and Lewis Theobald in succession, received the speech through the Folio tradition and treated it as all their age treated Shakespeare: they regularized its metre where the early lineation offended them, smoothed its punctuation into periodic order, and began the long process by which the speech’s accidentals became invisible and its substantives canonical. Their work fixed the Mab of the anthologies, the speech as a detachable jewel, and fixed it so successfully that the textual plurality this article describes vanished from common knowledge for two centuries. The recovery of the quartos as objects of study, the nineteenth century’s facsimiles and the twentieth century’s bibliography, restored the plurality: the first quarto’s Mab re-entered scholarship as evidence, the second quarto’s page was photographed and collated, and the speech acquired the double existence it now has, a single famous text in the reading editions and a documented family of texts in the scholarship. The whole-text bearings of that recovery, the way the two quartos of the play diverge across five acts, are traced in the comparison of what the Q1 and Q2 differences reveal across the whole play; the present study’s contribution is the single speech at magnification, and the magnification shows the family likeness and the family quarrel in miniature.
The parallel with the play’s other famous solo flight is instructive, and this run of studies draws it deliberately. Juliet’s potion speech and her “Gallop apace” soliloquy have their own variant histories, examined in the comparison of the Gallop apace variants across the early texts, and the pattern there, a stable core of famous lines carried by a variable connective tissue, rhymes with the Mab pattern closely enough to suggest a general law of these texts: the lines a culture memorizes are the lines transmission preserves, in the study, on the stage and in the printing house alike, while the lines that merely carry the sense are the lines that drift. The Mab speech is the law’s clearest instance because its famous matter is so dense and so strange. Nobody paraphrases moonshine’s watery beams. Everybody, given a night and a stage, paraphrases a transition.
The anthology afterlife completed what the eighteenth-century editors began. The Mab speech entered the quotation books and the elocution manuals as a self-contained recitation, severed from the dream argument that provokes it and from the interruption that ends it, and in that severed form it conquered the culture: generations met Queen Mab as a poem about a fairy, full stop, and the play around her became the poem’s footnote. Textual scholarship’s restoration of the variants belongs to the same movement as the modern stage’s restoration of the frame, productions that play the speech as an act of aggression inside an argument rather than as a decorative aria. Both restorations return the speech to its conditions, and the conditions include its texts: a speech twice printed in its author’s lifetime, in shapes that differ, was never the monolith the anthologies sold. The monolith was a product, manufactured by excerpting, and the collation is the audit of its manufacture.
What the Mab Collation Proves, and What It Cannot Prove
The findings of the collation can be stated as five propositions, each held at the strength its evidence allows. First, the two quartos’ Mab speeches descend from a common written ancestor, and that ancestor stood very close to the second quarto’s copy: the verbatim agreement in invented, unguessable phrases admits no other explanation. Second, the first quarto’s speech is a performed derivative of that ancestor’s text, carrying the distribution of fidelity, peak accuracy in the star part, loosened connectives, compressed list, intact architecture, that performance transmission produces and that neither scribal copying nor compositorial error produces. Third, the variants between the quartos in this speech are not authorial revision: they lack direction, design and motive, and the revision hypothesis, respectable elsewhere in the canon, finds no purchase in these forty lines. Fourth, the second quarto’s oddities in the speech, above all its prose-set stretches, are printing-house events, evidence about Creede’s shop and not about any manuscript, and the Folio’s text is a descendant whose independent testimony concerns the stage of 1623, not the study of the mid-1590s. Fifth, the speech readers know is an editorial assembly on a second-quarto base, and its familiarity is the familiarity of a settlement, stable because editors converged on it, not because any early page contains it entire.
A sixth proposition belongs to the speech rather than to the texts, and it is the one the literary reader carries away. The transmission history certifies the interpretation. Because both witnesses, by roads as different as a performer’s memory and an author’s papers, transmit the same arc, whimsy darkening through the catalogue to the soldier and the hag, that arc cannot be an accident of either road: it is the invention itself, the thing Shakespeare made, attested twice over by the only two independent channels the period provides. Readings of the speech that build on its darkness, its coercion, its dream-theory, stand on doubly witnessed ground, and readings that treat the flight as detachable ornament stand against both witnesses at once. Collation, which began in this study as bibliography, ends as a warrant for criticism: the variants show what in the speech is negotiable, the connectives and the catalogue’s flesh, and what is not, the design that every road from the 1590s preserves.
The calibration the speech provides extends beyond this play, and students of the other suspect texts use it whether they cite it or not. A scholar approaching the garbled soliloquy of the first Hamlet quarto, or the sketch scenes of the Henry V quarto, needs a standard for what good memorial transmission looks like before pronouncing on bad, and the Mab speech is that standard in its purest available form: a star part, a famous flight, a short chain of custody, and a fuller authoritative text against which every loss can be measured to the word. The losses, measured, prove small and patterned; the gains of the comparison prove large, since the whole vocabulary of the debate, paraphrase at the joins, compression in the lists, fidelity at the highlights, was learned from passages behaving as this one behaves. Forty lines in a comedy of love and death thus anchor the bibliography of an era’s drama, and the anchor holds because the collation beneath it was done at the level of the line, the copy and the press, where anchors must be set.
The limits bind as tightly as the findings. The speech is forty lines in a play of some three thousand, and it is the worst possible sample in one respect: a star set piece, the passage every theory expects to behave exceptionally. Memorial reconstruction would predict its survival; revision would predict its polish; the printing house would predict its fame would protect it. A sample that every theory predicts cannot discriminate among theories for the whole text, and this article’s verdicts are therefore issued for the speech, with the whole-text quarrel expressly reserved. The first quarto contains scenes where its text collapses far below the Mab standard, and scenes where it preserves matter the second quarto lacks; the second quarto contains its own famous cruxes and duplicated lines elsewhere in the play; and the theories built on those whole texts must answer to those whole texts. What one speech can do, and this one does, is calibrate the instruments: it shows what memorial drift looks like at its gentlest, what the absence of revision looks like, what press interference looks like, and it gives every larger claim a control case to be checked against.
The series thesis completes itself here. The cliché reads one fixed speech, the set piece as anthology piece, identical in every mouth and every edition, a jewel cut once and set forever. The collation restores the variants and the cliché dissolves: in its place stands a documented history, an author’s invention carried to the stage, held in a performer’s memory at the highlights and loosened at the joins, printed short in 1597 and full in 1599, squeezed into prose by a press under pressure, reprinted, corrected, regularized, edited, and handed to modern readers as an assembly whose seams show only under magnification. The speech loses nothing by the restoration. Queen Mab survives her collation exactly as she survives Romeo’s interruption, by being more interesting than the dream she mocks: not a fixed text but a transmitted one, and all the more worth the forty lines of attention for having lived, demonstrably, in more than one shape.
For editors and teachers the collation carries a practical charge. An edition that prints the speech without signalling its variant state teaches a falsehood by omission, the falsehood of the fixed text; an edition whose apparatus records the first quarto’s twin teaches the truth at the cost of a footnote’s attention. The best modern editions pay the cost, and the classroom use of the Mab variants table in this article is meant in the same spirit: forty lines are few enough for a student to hold two versions in view at once, and the exercise of comparing them teaches more about how Shakespeare’s texts exist than any general lecture on bibliography. The speech is the ideal school of collation precisely because its famous core refuses to move. Against a ground of agreement, every small difference becomes visible, arguable and instructive, and the student learns the discipline’s central skill, which is not the hunting of errors but the weighing of witnesses.
There is a last implication, for the play’s readers rather than its editors. Knowing that the Mab speech lived two early lives changes the reading of neither, and changes the status of both. The 1599 speech remains the words to quote, cite and memorize; the 1597 speech becomes something rarer than a rival, a recording in print of the flight as the first audiences heard it, at performance tempo, from the first Mercutio’s mouth, with its list trimmed to the length a stage allows and its highlights polished by nightly applause. The cliché of the fixed set piece asked for one speech and received two, and the doubleness is not a scandal in the text but the ordinary condition of a play that lived, as living plays do, in more places than its author’s study.
Frequently Asked Questions
Q: What is collation, and how is it done on the Queen Mab speech?
Collation is the systematic comparison of two or more early texts, word by word and line by line, to record exactly where they agree and where they differ. On the Queen Mab speech it means setting the first quarto of 1597, the second quarto of 1599 and the First Folio of 1623 in parallel, unit by unit, and logging every substantive difference in wording alongside the accidentals of spelling and punctuation. Modern collation descends from Charlton Hinman’s mechanical collator, which superimposed copies to expose variants invisible to the naked eye, and its results appear in the apparatus of editions such as René Weis’s Arden text, where each recorded variant lets a reader audit the editor’s choices against the surviving evidence.
Q: What does the second quarto’s title page promise, and does the Mab speech bear the promise out?
The second quarto’s title page promises a text “newly corrected, augmented, and amended,” and the Mab speech bears the promise out in substance. Against the 1597 printing, the 1599 speech is augmented in the plainest sense, fuller in its catalogue, fuller in its connectives, carrying lines and half-lines the earlier text compresses away, and it is corrected in the sense that its readings repeatedly prove the sharper and more characteristic where the two divide. The promise was a bookseller’s advertisement, and advertisements of the period promised much, but collation shows this one kept: the augmented speech is the fuller and better text, and modern editors’ reliance on the second quarto for the passage rests on that verification, not on the title page’s word.
Q: Which words of the Mab speech are identical in the first and second quartos?
The identical words are the speech’s famous matter: the fairy midwife definition, the agate-stone size figure, and the chariot inventory of spinners’ legs, grasshoppers’ wings, spider’s web traces and moonshine collars, together with the grey-coated gnat who drives the team. Both quartos carry these substantives verbatim, and both carry the hag coda and the closing exchange about dreams in substance. The identity is the collation’s control group, because such phrases could not be reinvented by a reporter or a pirate: their survival in the shorter text proves descent from a written ancestor close to the second quarto’s copy, and it confines the first quarto’s variance to the connective wording, the catalogue’s fullness and the page’s dress, where its readings loosen and compress.
Q: What is the atomies crux in the Queen Mab speech?
The atomies crux is the divided early reading in the line naming Mab’s draught team, the tiny creatures that draw her chariot through the night. Modern editions print the team as “little atomies,” a word glossed variously as atoms, minute sprites or pygmy beings, but the early witnesses hand the word down in divided form, and commentators have argued the locus for centuries because no gloss quite settles what the creatures are. The crux is a true substantive variant: a single word, no difference in the image of the tiny team, and yet a genuine fork in transmission, since one form must descend from the copy and the other must have arisen along the chain of performance, copying or printing that carried the speech to its pages.
Q: Does the first quarto omit any lines of the Queen Mab speech that the second quarto prints?
The first quarto omits no whole unit of the speech, and that negative fact is among the collation’s most important findings. The address, the midwife definition, the chariot inventory, the journey through the sleepers, the catalogue of dreamers, the soldier’s darkening, the hag coda and the interruption all stand in the 1597 text. What the first quarto does is compress within units: lines the second quarto gives at full length appear fused, trimmed or carried in shorter form, most visibly in the catalogue, where the elaboration around a dreamer’s fee or benefice shrinks. Editors describe the relation as compression rather than omission, and the distinction matters for theory, because a memorially carried text loses flesh in exactly this way, while a censored or radically cut text loses limbs, and the Mab speech has lost none.
Q: Where does the first quarto’s lineation of the Mab speech diverge most sharply from the second quarto’s?
The sharpest divergence is not a place but a principle: the two shops disposed the same verse by different economies, so that line divisions an editor now prints as verse stand differently divided, fused or run on in the early pages, with the divergence thickest where the second quarto squeezes its verse toward prose and the first quarto follows its own arrangement. Neither early page can be read as a map of the author’s lineation, since compositors, not authors, owned the disposition of lines under the pressure of cast-off copy. Modern editions therefore re-lineate the speech according to its metre, print the verse the words declare, and leave the early dispositions to the textual notes, where they testify about manufacture rather than about the poem.
Q: Why is part of the second quarto’s Mab speech set as prose?
Part of the second quarto’s speech stands as prose because the printing house needed the room. Verse lines waste the right margin of the page, prose lines fill the measure, and a compositor whose cast-off copy threatened to overrun its allotted pages could recover space by running verse on as prose. The setting arrives in stretches, matching the local pressure of a forme or a stint, and it lifts when the pressure lifts. Bibliographical analysis of the kind Peter W. M. Blayney practised on King Lear reads such stretches as stress marks of manufacture. Nothing follows from the prose dress about the author’s manuscript, which the metre shows to have been verse, and editors restore the lineation the verse requires while recording the early setting for the evidence it gives about Creede’s shop.
Q: What does an editor do with Mab lines that the early texts set as prose?
An editor re-lineates them. The words are printed as the verse their metre declares, with line divisions restored where the early page ran them together, because the prose setting is charged to the printing house and not to the copy. The decision is recorded, not hidden: the textual notes of editions such as the Arden and Cambridge texts register the early disposition of the passage, so that a reader who needs the 1599 page’s own arrangement can recover it. What the editor does not do is preserve the prose as though it were an authorial choice, or rewrite the words to fit the restored lines. Re-lineation changes the speech’s look on the modern page and nothing else, which is precisely Greg’s distinction at work, accidentals surrendered to the copy-text’s heir, substantives untouched.
Q: How does the First Folio’s Mab speech descend from the quartos?
The Folio’s speech descends at two removes. The Folio text of the play was set from an exemplar of the third quarto of 1609, itself a reprint of the second quarto, corrected in places against theatrical copy. The Mab speech follows that descent: the second quarto’s wording inherited through the reprint, the spelling and punctuation renewed by a new shop, and a scatter of small changes that are best explained as playhouse correction or fresh error. The consequence for collation is strict. Agreement between the Folio and the second quarto is a single testimony, not two, and a Folio departure carries weight only as evidence of the play in the theatre around 1623. Editors accordingly cite the Folio’s speech as a descendant with a stage history, valuable for that history and disqualified as an independent judge.
Q: What did the Hinman collator change about comparing texts like the Mab speech’s?
Charlton Hinman’s mechanical collator changed comparison from an act of reading into an act of verification. The machine superimposed two copies of an edition and blinked their images against each other, so that any difference, a turned letter, a corrected word, a whole variant line, announced itself to the eye instead of hiding in the similarity of the pages. Hinman built it for the First Folio, and its discipline spread to the quartos: claims about agreement and difference in a passage like the Mab speech now rest on recorded, checkable collations of copies rather than on a scholar’s impression of resemblance. The change matters because the arguments built on the speech, about memory, revision and the press, are only as strong as the variant record beneath them, and the collator made that record exact.
Q: Why does the memorial-reconstruction theory predict that Mercutio’s great speech would survive well in the first quarto?
The theory predicts it because of who was suspected of doing the remembering. Harry R. Hoppe’s classic study of the 1597 quarto argued that the text was reconstructed by performers, among them the actors of the hero and of Mercutio, and a reporter’s own part is the part he knows nightly, in his bones, at performance pitch. A star’s set piece should therefore stand in a reported text at a fidelity the surrounding scenes cannot match, and the Mab speech does: its signature images verbatim, its architecture intact, its ending worded as the fuller text words it. The prediction was made from the whole quarto’s unevenness before this speech was collated at magnification, and the collation’s confirmation of it, in the part and at the speech where the theory points, is the strongest single argument the memorial account possesses.
Q: What evidence in the Mab collation tells against a purely memorial origin for the first quarto?
The fidelity itself, at one remove. A text reconstructed purely from memory, under theatre conditions, should show memory’s characteristic failures inside the star part as well as outside it: anticipations, where a phrase arrives before its place; recollections, where the play’s later language leaks backward; and paraphrase of the famous matter itself when the ear tires. The Mab speech shows loosened connectives and a compressed catalogue, but its invented core, the chariot of spiders’ legs and moonshine, stands verbatim, and its order never breaks. That profile suggests a transmission with writing somewhere near it, copy consulted, a part written out, a book within reach, rather than unaided recollection across a whole play. Laurie Maguire’s audit of the memorial markers makes the same caution systematic: the markers that remain are weaker than the theory’s parade cases require.
Q: Could any Mab variant between the quartos be Shakespeare’s own revision rather than transmission error?
In principle the possibility cannot be excluded for any single variant, and in this speech it persuades for none. Revision, on the model the two-text King Lear made respectable, leaves a narratable pattern: a design tightened, an emphasis re-aimed, an image reconsidered. The Mab variants between 1597 and 1599 refuse narration of that kind. The shorter text’s differences flatten connectives and trim the catalogue’s flesh without touching the invention, the frame or the ending, and the fuller text restores fullness everywhere rather than improving anywhere in particular. A reviser with forty famous lines in hand does not spend his revision on mortar. The honest verdict assigns the speech’s variants to transmission, performance loss in one witness and press noise in both, while granting that the category of authorial revision remains open for other passages of the play on their own evidence.
Q: How does René Weis’s edition present the Mab speech’s variants?
René Weis’s Arden edition prints the speech from the second quarto as copy-text and presents the variants where an editor should: in the apparatus and commentary that shadow the reading text. The first quarto’s compressions and divided readings are recorded at the lines where they occur, the Folio’s departures are noted with their derivative standing made plain, and the prose-set stretches of the 1599 page are registered so that the re-lineated verse can be audited against its copy. The presentation embodies the modern settlement this article defends: the reading page gives one speech, the second quarto’s, modernized and intelligible, while the scholarly page preserves the plurality, letting any reader test the editor’s choices against the witnesses instead of taking the assembly on faith.
Q: What is the InsightCrunch Mab variants, and what pattern does it expose?
The InsightCrunch Mab variants is the side-by-side table in this article that sets the Queen Mab speech’s units, in their spoken order, against the first quarto, the second quarto and the modern editorial outcome. The pattern it exposes is the collation’s verdict at a glance. Down the first-quarto column runs agreement in the famous matter and compression in the list; down the second-quarto column runs fullness of text with patches of press-driven prose setting; down the outcome column runs the modern settlement, the 1599 text printed, the 1597 text recorded, the Folio treated as a descendant. A reader who absorbs the table owns the article’s argument in miniature: one invention, two early shapes, and an editorial tradition that weighed both before handing down the speech now quoted as fixed.
Q: Do the two quartos agree on the ending of the Mab speech, where the hero interrupts?
They agree completely, and the agreement is structural as well as verbal. Both texts end the flight with the hero’s interruption, the plea for peace addressed to Mercutio, and both follow it with the speaker’s own retraction, the admission that he talks of dreams, the children of an idle brain, begot of nothing but vain fantasy. The ending matters to the collation because frames are what transmission usually breaks: a reported text drifts at its edges, where memory’s grip loosens entering and leaving a set piece. Here the edges hold in both witnesses, entry and exit alike, which corroborates the article’s profile of the first quarto’s speech, a performed text carried by someone who had the whole shape, not a snatch of famous lines torn from a scene half remembered.
Q: What is a press variant, and could one hide inside the Mab speech’s quartos?
A press variant is a difference between copies of the same edition, created when the press was stopped mid-run to correct a type error and the corrected and uncorrected sheets were bound together indifferently. Two surviving copies of one quarto can therefore read differently at the same line, one carrying the error, the other the fix. A press variant could certainly hide inside the Mab speech’s evidence: a small divided reading attested in a single copy of the second quarto might be a stop-press state rather than a reading of the edition. Editors and collators guard against the trap by comparing multiple copies where they survive, and by recording states in the apparatus, so that an argument is never built, unknowing, on one copy’s accident.
Q: Why do editors modernize the spelling of the Mab speech, and what is lost in the process?
Editors modernize because the editions serve readers of the play, and the early dress, the long s, the variable spellings, the pointing by breath rather than by grammar, bars the door to all but specialists. Greg’s rationale licenses the practice: accidentals were the printing house’s property, normalized by each shop to its own habit, so no authorial intention is betrayed when a modern shop normalizes them again. What is lost is evidence, the visible manufacture of the early pages, the house styles that let bibliographers assign stints and date practices, and the period texture of the words on their first pages. The loss is real but recoverable, since facsimiles and quarto archives preserve the originals, and the modern settlement, a modernized reading text audited by an apparatus and backed by the photographed witnesses, keeps both goods at once.
Q: How should a student cite the Queen Mab speech when the quartos differ?
A student should cite the speech from a named modern edition, by act, scene and line, and name that edition wherever the argument touches the words themselves: the Queen Mab speech at Act 1, Scene 4 in Weis’s Arden text, or in Evans’s, or in Gibbons’s Arden, is a determinate object, while the speech cited from no edition is a blur of witnesses. Where an argument depends on a variant, the citation must go behind the edition to the witness, first quarto or second, and say so, because the editions converge on the speech’s substance and diverge in their apparatus. The discipline costs one parenthesis and buys exactness: every claim about the speech becomes checkable, which is the whole ethic of the collation this article performs.
Q: What can the Mab speech’s transmission not tell us about the whole play’s texts?
It cannot settle the whole-text quarrel, and pretending otherwise is the error this article has resisted throughout. The speech is forty lines of some three thousand, and it is a biased sample, a star’s set piece that every theory expects to behave exceptionally: memory protects it, fame protects it, anthologies protect it. The first quarto contains scenes that collapse far below its standard and matter it alone preserves; the second quarto contains cruxes and duplications the speech never dreams of; and the theories of memorial reconstruction, revision and foul papers must answer to that whole evidence or not at all. What the speech gives the larger argument is calibration, a control case where the instruments can be checked, not a verdict to be smuggled from forty lines into five acts.