The Comedy of Errors

by Kent Cartwright

  23 First Folio, sig. H6v (‘Comedies’), comprising CE TLN 1382–1503, 4.4.95 –5.1.37.1

  24 First Folio, sig. H6v, close-up of the foot of column a

  But the major problem on this page occurs at the foot of column a, which carries the SD ‘Runne all out.’ on the last line,14 and the head of column b, which, after a line from the Officer (who should have just exited), offers a second exit direction, this one in both Latin and English: ‘Exeunt omnes, as fast as may be, frighted.’: two SDs in different styles and different columns, but separated by only one five-syllable line of dialogue, calling redundantly for the same exit. That the second employs Latin clarifies little, since Shakespeare wrote SDs in both Latin and English. The possibility that two SDs rather than one might genuinely be intended led the production at Stratford-upon-Avon in 2000 to launch, at the first SD, a Keystone Cops-like chase off the stage and back on to it, and then, after the Officer’s ‘Away, they’l kill vs.’ (TLN 1446), another mass departure to fulfil the second SD (see also Cam2a’s ‘Run all out [and re-enter with others]’). The double chase made for theatrical fun, but the two SDs probably refer to the same exit. Since ‘Runne all out.’ expresses a different sense of theatrical language than does ‘Exeunt omnes’, it may be that the directions are by different hands, especially since the first is the play’s only exit direction not to use ‘Exeunt’ or ‘Exit’ (its closest analogue is ‘All gather to see them.’ at TLN 1815). That first SD succinctly summarizes the second, without worrying about the characters’ emotions (‘frighted’) or their speed (‘as fast as may be’); it is crisp and functional. Perhaps this first SD was added at a later date by someone other than Shakespeare – a theatrical bookkeeper? – with the intention of slightly separating the exits of the women and the Officer, since the latter gets one further line. If the compositor saw before him the two SDs written close to each other on the manuscript page, perhaps on two lines but as if they were collectively one unit, he was faced with the problem of a rather lengthy exit direction and an insufficient amount of space left at the foot of column a into which to squeeze it. If he did not want to put a ‘marginal stage direction’ at the head of a column, as Werstine speculates, then one pragmatic solution was to retain all the language of the directions but to split it up, with part of it placed at the foot of column a, even before the Officer’s ‘Away’, and the rest coming after the Officer’s line. Foakes favours such an explanation.15 Wells and Taylor dismiss it (TxC, 266), although Whitworth offers a conjecture similar to that of Foakes. Still, the explanation does not quite satisfy. Compositor B, just before he set these SDs, seems not to have been bothered about insufficient space, since he left empty lines around the SD at TLN 1440 –1. And if he saw the two SDs as one unit, he could simply have left the last line of column a empty (the standard procedure), begun column b with the Officer’s ‘Away …’ and then beneath that speech-line set the composite exit direction. Perhaps the directions’ different voices (and possibly handwriting) and perhaps their placement in the manuscript made the compositor assume that they were actually two rather than one. Alternatively, maybe the double directions offer an awkward attempt to orchestrate a complex exit, with some in Adriana’s party quickly disappearing, the Officer hesitating but then delivering his own exit line, and any lingering characters, with fears heightened, completing the retreat. Evidence and conjecture cannot take us much further.

  The printer’s copy: authorial ‘foul fapers’? A text suitable for performance?

  Beyond the question of compositorial practices, a second large issue has occupied textual scholars: whether or not the F text of The Comedy of Errors was set from a manuscript in Shakespeare’s own hand, possibly his ‘foul papers’ (i.e. his working draft of the play before it might have been transcribed into a ‘fair copy’ for subsequent use).16 The play exhibits certain anomalies often attributed to an author in the flow of composition – changing SPs and character descriptors, or, in SDs, thinking out loud about characters’ backgrounds or including extraneous bits of narrative. The Antipholus brothers, for example, are first named in a way that suggests that Shakespeare was writing with a copy of Plautus’ Menaechmi fresh in his memory or even open on his desk. The resident (Ephesian) brother is referred to as ‘Antipholis Sereptus’ (TLN 273); that cognomen, close to Latin surreptus, ‘stolen’, recalls Men., in which one of the twins was kidnapped from a market at an early age (see List of Roles, 4n.). Likewise, the wandering (Syracusan) brother first enters the stage as ‘Antipholis Erotes’ (TLN 162); here the Latinate addition is harder to parse, though one of its associations is with the Latin verb errare, to wander (see List of Roles, 3n.). Next time around, that name has become ‘Antipholis Errotis’ (TLN 394), another near miss at a form of errare, before it settles down geographically as, more or less, ‘Antipholus of Siracusia’ (TLN 786; for variants, see TLN 1183, 1811). Speech prefixes can change. Egeon, the Merchant of Syracuse, has ‘Marchant’ for his first SP (TLN 4), but in Act 5 he will be identified first as ‘Mar. Fat.’ (TLN 1671), for ‘Merchant Father’ and finally just ‘Fa[ther].’ (TLN 1762 and subsequently), as if Shakespeare were adjusting terms as, in his imagination, the character became less a merchant and more a father.17 Monikers and SPs can also be shared. In addition to the Syracusan Merchant, F has two other characters identified as ‘a Marchant’, ‘a Merchant’ or ‘the Merchant’ (TLN 162, 981, 1463), implying that overall distinctions were not troubling the author in the moment-by-moment act of composition. In 3.1, a maid called ‘Luce’ (TLN 679) suddenly enters to Syracusan Dromio behind a door and joins him in railing against Ephesian Antipholus; scholars agree that she is the same character who is later described as the wife (or possibly fiancée) of Dromio of Ephesus, but who takes on a different name, ‘Nell’ (see 3.2.110n.). Shakespeare may have changed his idea about the name as he worked through the play, not wanting it in 3.2 to sound too close to that of Adriana’s sister, Luciana, but without feeling that enough was at stake to require him to go back and rewrite dialogue in the earlier scene.

  F also contains in its SDs bits and pieces of information that function less as instructions to the actors than as musings about actions and characters within the narrative: ‘Enter Adriana, wife to Antipholis Sereptus,with / Luciana her Sister.’ (TLN 273 – 4); ‘Enter … Angelo the / Goldsmith, and Balthaser the Merchant.’ (TLN 617–18); ‘Enter … from the Courtizans.’ (TLN 995); ‘Enter … from the Bay.’ (TLN 1073); ‘Exeunt to the Priorie.’ (TLN 1503).18 These directions read as if they envisioned characters within the story rather than actors in the theatre, and locations within the narrative rather than places on the stage; again, they give the impression of the author in the process of composition. Perhaps the most interesting of such directions is ‘Enter … a Schoole- / master, call’d Pinch.’ (TLN 1321–2; see 4.4.40.2n.). The descriptive information about Pinch has no stage utility. Adriana refers to him as ‘Doctor Pinch’, ‘a Coniurer’ (TLN 1330, 4.4.48) and ‘Good Master Doctor’ (TLN 1414, 4.4.123), which might favour identifying him as a pedagogue, although he binds and carries off Ephesian Antipholus and Dromio as if, like a physician, he might effect a cure of their madness. When Shakespeare wrote Pinch’s entrance direction, he may have been defining the character in his mind. With Pinch’s type and manner once established conceptually, the need for vocational exactitude falls away. Likewise, the SD information about Adriana, ‘wife to Antipholis Sereptus’ (TLN 273), serves less a theatrical than a narrative purpose.

  Because the manuscript underlying F seems so close to the story as it is being invented and put to paper, a number of textual scholars have argued strongly that the version must be Shakespeare’s ‘foul papers’, his working draft of Errors, with some light annotations, perhaps made by a bookkeeper with eventual performance in mind – although these scholars maintain that the printer’s copy could not have served for a promptbook, or playbook.19 Such a text would afford a rare opportunity to glimpse the playwright in the act
of creating his original version, before it was, perhaps, marked up by others for theatrical practice. The foul-papers scholars argue that preparation of the author’s draft for the theatre would have involved regularizing Errors’s SPs and names, filling out SDs and clarifying indefinite ones. Because the printer’s copy for Errors is deficient in all these regards, it presumably could not have served for a promptbook. But William B. Long has demonstrated in a series of essays that many surviving playbooks from the era show exactly the kind of variable information that scholars such as Greg believed to be confined only to foul papers.20

  Likewise, Paul Werstine has contended that the play’s authorial papers could have been used as a promptbook, that compositorial errors account for many of the editorial problems and that the F text in matters such as names and SPs is quite sufficiently clear and consistent to have been derived from a theatrical playbook. He argues, for example, that the cognomens ‘Erotes’ and ‘Errotis’ for Antipholus of Syracuse would have occasioned no confusion in theatrical practice and that ambiguous SPs, such as Ant. used for both of the Antipholus brothers, are made distinct enough when they need to be, as in the last act where the Antipholuses are onstage together and the text distinguishes them as E. Ant. and S. Ant. Werstine attributes other related inconsistencies to compositorial practice or carelessness. Further, for the duplicate SDs in 4.4 (sig. H6v) – ‘Runne all out.’, ‘Exeunt omnes, as fast as may be,frighted.’ – he maintains that it is difficult to sort out whether Shakespeare wrote one and a bookkeeper another, since both forms have Shakespearean analogues. Werstine concludes that the printer’s copy for Errors could as easily have been a playbook as foul papers and, perhaps more important, that the F text simply resists the rigid distinction between these categories maintained by mid-twentieth-century scholars; the problem, he concludes, may be the assumed dichotomies themselves.21

  Werstine’s perspective has been resisted by Wells and Taylor, who point out that the F text lacks evidence of theatrical use (such as ‘actors’ names, duplicated directions, isolated extra-syntactical calls for essential properties, warning directions’) so that ‘[t]he orthodox view – that the play was set from foul papers – is thus probably correct’ (TxC, 266). The orthodox Chambers–Greg foul-papers argument, however, assumes the difficult task of maintaining a negative; that is, the claim that the manuscript underlying the F text could never have been used as a playbook. The evidence and arguments offered by Long and Werstine about the state of contemporary playbooks run against that conclusion, and Werstine’s analysis of Errors’s textual problems provides a convincing approach. Indeed, Wells and Taylor acknowledge that the F text is to some extent a performance document: they accept the possibility of ‘a few annotations related (indeterminably) to performance’ and see the act divisions ‘as a genuine reflection’ of the 1594 enactment at Gray’s Inn (TxC, 266). Standish Henning, the Variorum editor, concludes that the F text is close to authorial papers yet acknowledges that the difference between foul papers and promptbook is difficult to sustain.22

  Arguments about the nature of the printer’s copy always make their way to those disputed SDs, mentioned previously, on sig. H6v (TLN 1382– 503, 4.4.95 – 5.1.37.1). How do we understand the two duplicative SDs, one at the foot of column a, the other on the second line at the head of column b, repetitious in effect but different in style (see Figs 24 and 25)? The orthodox view, proposed by Greg (Problem, 140), contends that the first of these directions is by Shakespeare, the second a later annotation in the manuscript margin by a bookkeeper, a view that Wells and Taylor accept. According to Werstine, however, both forms of the direction could have been written by Shakespeare. ‘Exeunt omnes, as fast as may be,frighted.’ resembles a SD in F Taming of the Shrew: ‘Exit Biondello, Tranio and Pedant as fast as may be.’ (TLN 2490). That direction is thought to have been Shakespeare’s, and the printer’s copy for Shrew, as with Errors, is considered authorial foul papers. ‘Runne all out.’, notes Werstine, bears some similarity to a SD in a fragment from the first printing of 1H6: ‘ “… they all runne away …” ’ (‘Copy’, 244). It also shares something in manner with a later Errors SD: ‘All gather to see them.’ (TLN 1815; see 5.1.330 SD and n.), which Wells and Taylor consider authorial (TxC, 266). Werstine concludes sensibly that the ‘unusual position’ in the text of this split or doubled SD ‘forbids an editor from using it to draw inferences about the nature of the printer’s copy’ (‘Copy’, 244; see 243 – 4). We are left, then, with the view that the manuscript behind the F text of Errors was probably authorial papers that could have served as a playbook for performance. Perhaps the most important inference that we might draw is that the text of Errors should not be separated from the play’s theatrical life.

  25 First Folio, sig. H6v, close-up of the head of column b

  The printer’s copy: the question of performance venue

  That inference invites a related question: whether the F text seems oriented towards public performance, as at the playhouses, or private performance, as at venues such as Gray’s Inn or the court. Almost thirty years separate the composition of Errors in about 1594 and its first appearance in print in 1623; in that interim the play was probably performed in the 1594 – 5 season at the Theatre, played at the Gray’s Inn revels in December 1594, perhaps revived in the 1597– 8 season and again in 1604 – 5 at the playhouses (the first season entailing performance at the Curtain and the last at the Globe), and produced at court before King James in December 1604 (see Introduction). Although the printer’s copy bears signs of authorial composition, it also may have traces of performance (or revision), more likely so if it had been used as a playbook. What kind of venue, public or private, does the printer’s copy reflect?

  One bit of evidence is equivocal but deserves mention: the absence of music. Errors is unusual among the early comedies in having no music within the action, unlike, say, The Two Gentlemen of Verona or The Taming of the Shrew. Plays performed indoors needed pauses for the trimming, mending and replacing of candles, and such pauses would be used, almost as a matter of course, for entr’acte music and other business. That Errors retains no indications of music may reflect the fact that music was treated as something separate and detachable, a notion that squares with performance at a private venue featuring hired music from other sources. In general, however, lost theatrical music is not uncommon, as Stern points out (Documents, 120 –73). Nonetheless, Errors’s anomalous absence of songs, singing or music is worth noting, if inconclusively.

  F’s act–scene divisions also make for tricky evidence. They may have been introduced by Shakespeare in anticipation of the Inns of Court performance (see TxC, 266). Alternatively, they may have been inserted as aids in outlining the complicated action as Shakespeare was drafting the play, without reference to where it would be performed. Or they may have been added somehow at a later time. If the divisions are by one hand, such as Shakespeare’s, then the lack of a scene one indication in Act 2 (scene one designations being present with all the other act headings) makes for an anomaly. Likewise, the act divisions, though sensible, work against the text’s implicitly continuous flow of stage action from 2.2 to 3.1 and from 4.4 to 5.1. Taylor establishes that in plays at the public playhouses from 1592 to c. 1607, continuous flow rather than act intervals was the performance practice (see ‘Act-intervals’). As an odd case, Errors’s act divisions might have been meant for the performance at Gray’s Inn Hall, as Taylor believes (‘Act-intervals’, 45), but that view cannot explain the presence of scene one designations for four of the five acts which lack any obvious performance function. Like the double SDs of 4.4, the act divisions provide little evidence for any particular theory of the text.

  While the staging demands of The Comedy of Errors are modest and flexible enough that they could have been satisfied by almost any venue, the F text does assume certain stage resources. Errors’s most complicated scene for staging is 3.1. In the only close analysis of the configuration of
Gray’s Inn Hall for the presenting of Errors during the 1594 revels, Margaret Knapp and Michal Kobialka conclude that a stage for the revels must have been built between the hall’s dais at one end of the long hall and the temporarily erected scaffolds at the other (and close to the dais itself). Dignitaries would have been seated on the dais, with other spectators seated opposite them, on the scaffolds. The playing space in between would have been ‘about nine feet [c. 2.75 metres] deep and as wide as the hall, that is, thirty-four feet, eight inches [c. 10.5 metres]’ (438), with audiences perhaps not only on two sides (the dais and the scaffold sides) but on all four. Given that the performance of Errors followed immediately after dancing at the end of a long night, it is doubtful that the players would have been afforded time for a disruptive erecting of booth houses on the stage, especially since such booths would block some sight-lines (see Introduction).

  The Gray’s Inn stage also poorly satisfies the needs of the playtext in 3.1 (see Introduction) in other ways. The central physical fact of that scene is the Phoenix’s door: it is locked against Antipholus of Ephesus and his outdoor party, who pound on it for entrance; conversely, the door provides the barrier behind which the indoor party lurks and gibes. The public theatre stages, typically with balcony and central and side doors, are the locales envisioned by the text: Dromio of Syracuse could hide behind the central door (or double doors), while Adriana could emerge on the balcony, as from dining. The stage at Gray’s Inn has no such advantages. Imaginative actors, of course, could have found various ways around any difficulties, but that is exactly the point: for a private venue, the text imposes a staging obstacle that must be overcome. Errors is adaptable enough to be played almost anywhere, but the venue imagined in 3.1 is the public theatre.