Chapter 14. Stage Directions Tiffany Stern This chapter tells the history of what have come to be called ‘stage directions’: short, practical performance or reader-oriented instructions, often in pigeon Latin, of unclear authorship, that typically start, end and intersperse a printed play. It is in three parts, beginning in the eighteenth century when the phrase ‘stage direction’ was co-invented by Alexander Pope and Lewis Theobald as a term for non-authorial directives in Shakespeare. It then turns to medieval and early modern plays, asking what a ‘stage direction’ was before it had a title, where it was situated on the page, what language it employed, who may have written it and for whom, and whether indeed it was ‘it’ before two words defined it as one entity. Finally, the essay considers the ‘stage direction’ after the term became established, asking how the rise of the theatre’s ‘stage director’ affected what ‘stage directions’ were on page and stage. Throughout, then, the chapter tells two different and only partially-connected stories: one is the story of the term ‘stage direction’, and the other the story of the thing(s) ‘stage direction’. The invention of ‘stage directions’ The phrase ‘stage direction’ was created by the editor and poet Alexander Pope, and the editor and playwright Lewis Theobald, in the course of a disagreement over a line in Shakespeare’s Henry V. There is, famously, in the 1623 folio text of that play, a passage in which Mistress Quickly says of the dying Falstaff that ‘his nose was a sharpe as a Pen, and a Table of greene fields’ (TLN 838-9). First Folio quotations are taken throughout from the facsimile prepared by Charlton Hinman, Mr. William Shakespeares Comedies, Histories, & Tragedies [The Norton Facsimile] (New York: Norton, 1968), using the through-line-numbers (TLN) of that edition. As the observation does not make immediate sense, Pope, in his 1725 edition of the works of Shakespeare, emended it. He argued that the end of the line was a ‘direction crept into the text from the margin’; that there must have been a ‘Property man’ at that time called ‘Greenfield’; and that the original direction, which he extracted from the speech, will have been ‘A Table of Greenfield’s’. William Shakespeare, The Works ed. Alexander Pope, 6 vols (1725), 1: xviii. Theobald was incensed not just by Pope’s solution – when he published his own edition of Shakespeare’s Works, he introduced the emendation of the line commonly accepted now, ‘and a’ babbled of green fields’ William Shakespeare, The Works ed. Lewis Theobald, 7 vols (1733), 4: 30. – but also by the ignorance of staging it betrayed. The ‘Stage-Direction’, he explained, enlarging on Pope’s ‘direction’ to underline the performance nature of the directive, would never be placed in the text at the moment of need; rather, it would be ‘mark’d … at about a Page in Quantity before the Actors quoted are to enter, or the Properties be carried on’. Lewis Theobald, Shakespeare Restored (1726), 138. Thus ‘stage direction’ was born: it originally indicated an advanced note to a props man that had (Pope) or had not (Theobald) been rammed mistakenly into Shakespeare’s text. Theobald did, however, adopt ‘stage direction’ as his term for non-authorial paratext. In his 1733 edition of Shakespeare Works, he used ‘stage directions’ for dumb shows, writing a scathing gloss to the first words of the dumb show for The Murder of Gonzago in Shakespeare’s Hamlet, ‘Enter a King and Queen very lovingly’. Pointing out that the stars of the ensuing play are not a king and a queen but a duke and a duchess, he expostulated: ‘Thus have the blundering and inadvertent Editors … given us this Stage-Direction’. William Shakespeare, The Works of Shakespeare ed. Lewis Theobald, 7 vols (1733), 7: 295. The blame for the discrepancy, Theobald believed, belonged to John Heminges and Henry Condell, the actors in Shakespeare’s company who, as ‘editors’, had brought the first folio to the press: ‘Royal Coronets being at first order’d by the Poet [Shakespeare] for the Duke and Dutchess, the succeeding Players, … mistook ‘em for a King and Queen’. In this instance, ‘stage direction’ is not a text written by a prompter for a property man, but a text written by actors for readers. In fact dumb shows often had a different heritage from dialogue, and this kind of discrepancy is not unusual. See Tiffany Stern, ‘Inventing Stage Directions; Demoting Dumb Shows’, Stage Directions and Shakespearean Theatre, ed. Sarah Dustagheer and Gillian Woods for Arden Shakespeare (London: Methuen, 2018), 19-43. ‘Stage directions’, then, were from the first understood to be by and for a range of people; what they shared was that they were non-authorial and marred the process of reading Shakespeare’s plays. ‘Stage direction’, and the opprobrium that came with the term, was accepted wholesale by later eighteenth century editors. Of ‘Re-enter fighting, and Macbeth is slain’, the editor George Steevens wrote, in 1773, that ‘This stage-direction … proves, that the players were not even skilful enough to prevent impropriety’. His irritation, again levelled at Heminges and Condell, arises from the confusion between this direction and the one immediately afterwards: ‘Macbeth is here killed on the stage, and a moment after Macduff enters, as from another place, with his head on a spear’. William Shakespeare, The Plays, ed. Samuel Johnson and George Steevens, 10 vols (1773), 4: 530. These illogical directions, Steevens insisted, were ‘unShakespearean’. His fellow editor Edmond Malone agreed, maintaining, in 1790, that ‘many of the stage-directions’ in Shakespeare’s works ‘appear to have been inserted by the players; and they are often very injudicious’. William Shakespeare, The Plays and Poems ed. Edmond Malone, 10 vols in 11 parts (1790), 4: 435. Intriguingly, countries across the continent also started inventing terminology for what we now call ‘stage directions’ in the eighteenth century. Several, borrowing from one another, adopted versions of the ancient Roman word ‘didascaliae’, which dates from around the first century BC: ‘didascalie’ (France); ‘didascalia’ (Italy); ‘didascalia’ (Spain). But the Latin (via Greek) word chosen had ramifications of its own. Originally it had meant the notices about production supplied with Roman plays – when the first performance had taken place; the names of the play’s prompters, composers and actors; who the Roman consul was that year, and so forth: the term did not literally mean ‘stage directions’ as there were none on the Roman texts. So continental ‘didascaliae’ were and are every bit of text that is not the play’s dialogue, including the list of dramatis personae, the title, act and scene divisions as well as ‘stage directions’. European critics referring to ‘didascaliae’ recognise the diversity of paratext generated by a play as a result, and tend not to be concerned with its authorship. But in Anglophone countries, the narrowness of the phrase ‘stage direction’ has isolated a subset of paratext, and unified it as though it has a fixed meaning and author – despite the fact that in its earliest usages, ‘stage direction’ had several meanings, and was thought of as non-authorial. ‘Stage directions’ before the phrase What came to be grouped under the term ‘stage direction’ was a series of passages in early modern plays that resembled one another in terms of page layout and, sometimes, language, but not authorship or readership. Indeed, it was the shared features of ‘stage directions’ that often hid the divergent origins of printed playbooks altogether, which might have, behind them, an authorial text, a text marked up for performance, a text marked up for readers, or a combination of all three. What needs exploration, then, is why and how a series of separate play interventions came to look visually similar and share a common(ish) language. Practices developed for medieval mystery and morality plays explain the placement and language of early modern ‘stage directions’. These plays, handwritten, and not for publication but annual performance, were authored by priests; ‘stage directions’ tended to be added to them later by practitioners, either when mounting a performance, or when recording performance advice for succeeding generations. As ‘stage directions’ were not usually written when the play was, they were likely to be in a different hand from the rest of the play, and placed around the dialogue; they were, further, often ‘boxed’, rubricated (written in red ink), or highlighted with slashes, brackets or dashes so they were not confusable visually with the actual ‘play’ in performance. This layout, separating the poetry to be spoken from the practical advice that was not – though, as Butterworth points out, with ‘considerable inconsistency as to the positions that stage directions occupy on the page Philip Butterworth, Staging Conventions in Medieval English Theatre (Cambridge: CUP, 2014), 4. – aided the running of productions so much that when scribes rewrote plays afresh, they would inscribe dialogue first, and stage directions second, in the margins and ‘separated’ by hand or placement. Stage directions, by their very appearance, were relegated: page layout stated that they were less important than the dialogue they accompanied. Early modern plays inherited layout from medieval manuscript practice, and tended to place the words to be spoken in the central space of a printed page, and the speech prefixes and words for action around the outside, though where on the outside took time to determine. For more, see Linda McJannet, The Voice of Elizabethan Stage Directions (Newark: University of Delaware Press, 1999). The first secular play ever to be printed in English, Henry Medwall’s Fulgens and Lucrece (1512-16), has its directions in the margin and heralded, as is the start of each speech, by pilcrows; Ulpian Fulwell’s Like will to Like (1568) has some stage directions centred and indented and some on the right and boxed or bracketed; Robert Wilson’s Three Ladies of London (1584) has stage directions centred, or ranged right, and in roman type (when the dialogue is in black letter); George Peele’s Edward I (1593) has stage directions in italics when the play is in roman type. As these examples show, a shared grammar for stage directions was only slowly developed: past, present and future tense were all experimented with (‘And gaue him a good blow on the buttocke’; ‘he kysseth Diccons breeche’; ‘Here they shall syng)’, William Stevenson, Gammer gurtons nedle (1575), C4r, B2r; John Bale, Kynge Johan (1538) ed. J. Payne Collier (1838), 41. as were hortative, imperative and participle forms (‘Here let Lucar open the boxe and dip her finger in it’; ‘smite him in the neck with a swoord’; ‘pointing to one standing by’) Robert Wilson, Three ladies of London (1584), E1v; Thomas Preston, Cambises (1570), c2v; Ulpian Fulwell, Like will to like (1587), A3r, . Why a simple present tense was ultimately adopted may have been to save words, though it had the invigorating effect of making stage directions happen in real time rather than recall a past performance or predict a future one. As the above also shows, a shared system for the appearance and placement of stage directions also took a while to come into being. Only by the 1590s was the marginal and italic form somewhat settled on. Why that choice was made cannot be traced to a single printer, so may relate to general printing house needs. The words that make up stage directions put pressure on certain letters, particularly capital E for ‘Enter’ and ‘Exit’: extracting these from separate ‘italic’ boxes preserved the full range of roman letters for the play’s dialogue. Since then, directions have tended to be italic, marginal and present-tense, though variants of each remain possible. The page layout finally settled upon, which presents the dialogue as though it is a poem, and the directions as though they are its gloss, has had interpretative implications. Stage directions seem to be a ‘comment’ on the play they flank, rather than part of it, and are often treated with less respect than the dialogue. And, as directions are in a secondary space – a space, moreover, characterised by its emptiness – they seem easy to add to. Plays of the early modern period in second and third editions often contain changes to stage directions though the words of the dialogue are unchanged – Shakespeare’s Richard III in its third quarto (1603), for instance, has additional ‘explanatory’ or necessary (but previously absent) printed stage directions, while the Folger Shakespeare Library’s printed copy of The Two Merry Milkmaids (1620) has early modern manuscript stage directions added onto the text in two different hands. Leslie Thomson, ‘A Quarto “Marked for performance”: Evidence of What?’, Medieval and Renaissance Drama in England, 8 (1996), 176-210. The marginal nature of stage directions made them not simply open to multiple or collaborative authorship: they positively invited it. The patchy Latin of early modern stage directions, too, suggests they can be added to by others – in that ‘stage directions’ across plays sound more like one another than like their plays’ ‘author(s)’. ‘Stage-direction language’ indeed has ‘multiple authorship’ in its origin as it descends, again, from medieval manuscript tradition (not, as might be suspected, from printed classical plays, which did not have stage directions in their earliest versions). While medieval plays, originally in Latin, were slowly translated into the vernacular for audiences, stage directions, which were by and for practitioners only, retained their original language. For more, see T. H. Howard-Hill, ‘The Evolution of the Form of Plays in English During the Renaissance’, Renaissance Quarterly, 43 (1990), 112-145. Over time, English crept in but Latin didn’t entirely creep out, the result being the English/Latin medley that typifies a lot of early modern plays – ‘Exit the Watch. Manet Captain’ reads Marston’s Insatiate Countesse; ‘Exeunt omnes, præter Consta. and Gage’ reads Massinger’s City-Madam. John Marston, The Insatiate Countesse (1613), E1v. Thomas Heywood, If you knovv not me, you know no bodie (1605), D1r. This ‘stage-direction language’ came up with rules of its own. Some verbs conjugated and hence remained ‘Latin’, like ‘exit’/’exeunt’ and ‘manet’/’manent’, and others, like ‘enter’, already at a remove from the Latin ‘intrat’/’intrant’, did not conjugate and behaved in ‘English’ fashion. The word-order of both entrances and exits, however, was firmly Latinate, irrespective of language, with the verb first, and the proper noun second (‘exit Bosola’, for instance, rather than ‘Bosola goes out’). This language, which ensured stage directions resembled one another, but were distinct from all other text, added to the notion that these were unstable passages, open for anyone to write, moderate or change and not tightly linked to an author. It is only the specific functions of (some) ‘stage directions’ of the early modern period that reveal who they were written for, and so hint at who may have written them. For instance, there is a subset of so-called ‘stage directions’ that aren’t about staging at all. These are directions for the theatre’s scribe; they might therefore logically be called ‘scribe directions’. For more on scribe directions, see Tiffany Stern, Documents of Performance (Cambridge: CUP, 2009), 154, 181-4. Scribe directions include the note in Thomas Kyd’s The Spanish Tragedy against a letter beginning ‘For want of incke receive this bloudie writ’. It reads, ‘Red incke’. Thomas Kyd, Spanish Tragedy (1592), E1v. This direction is for the writer of the stage letter, who is being instructed to use red ink in order to create a property letter that looks ‘bloody’ to the audience. Usually such directions are placed around ‘stage scrolls’ (papers to be read on stage); indeed, the titles that often herald such scrolls – ‘the letter’, ‘the riddle’, ‘the proclamation’ – are probably directions telling the scribe to create those particular documents. See Tiffany Stern, ‘Scrolls’ in Documents of Performance in Early Modern England (Cambridge: CUP, 2009), 174-200. Other paratexts are apparently for ‘stage keepers’ and/or ‘prompters’. Examples include the direction in L. S.’s Noble Stranger: ‘Enter Plod with a Boxe, in which are little pieces of paper rold up: A Table set forth’. L. S., Noble Stranger (1640), G3r. This may instruct a stage keeper to create and fill such a box; it certainly tells a stage functionary, probably a prompter, to make sure that a box is supplied to the actor and a table is brought onstage. It bears a similarity to the directions which, as Theobald had pointed out, are for props or people to be made ready for a future stage moment. The printed Quarto of Shakespeare and Fletcher’s Two Noble Kinsmen has such ‘advanced’ directions, asking for ‘2. Hearses’ to be ‘ready with Palamon; and Arcite: the 3. Queenes. Theseus: and his Lordes ready’. John Fletcher and William Shakespeare, The Two Noble Kinsmen (1634), C3v. These directions tell a backstage person, again, probably the prompter, which people and things are to be gathered and made ‘ready’ at the correct doors for a forthcoming entrance. None of the directions described so far are for an actor. Probably, indeed, actors were never the direct recipients of early modern stage directions, as they learned plays from individual ‘parts’, not the full play text. Though actors’ parts did contain stage directions, and though these might have been extracted from the full play – rendering them, on the full play, further ‘scribe directions’ – parts may equally have contained different, actor-specific directions. With only one English professional actor’s part surviving from the period, for ‘Orlando’ in Greene’s Orlando Furioso, there is not much information to go on: nevertheless, the part’s directions are brief and Latinate against the full play’s explanatory English. Thus the ‘Orlando’ part has the one-word Latin direction ‘Inchaunt’, The ‘Part’ of Orlando, Dulwich MSS 1, reproduced http://www.henslowe-alleyn.org.uk/images/MSS-1/Article-138/08r.html (accessed 27th July 2017). where the Orlando printed playbook (admittedly of the play in a different version) has ‘Hee drinkes, and she charmes him with her wand, and [he] lies downe to sleepe’. Robert Greene, Orlando Furioso (1594), G1r. It is worth recalling that playbook directions for entrances and exits are of most immediate use to the prompter, who needs to direct stage traffic, as are action directions like ‘whispers’ or ‘dies’ – as the prompter needs to know when not to prompt because staged silence is required. Printed playbooks, at one remove from actors’ texts, are similar to, and are sometimes copied from, the text that was in the prompter’s hands. Warren Smith, ‘New Light on Stage Directions in Shakespeare’, Studies in Philology, 47 (1950), 173-181 (178). It is likely that many of their directions are for, and sometimes by, the prompter. Then there are directions that would never work in performance. They include the ‘massed entrances’ found in several early modern plays including some by Jonson, Shakespeare and Middleton – in which every person who is to speak is apparently made to ‘enter’ at the start of the scene – which have been attributed to the habits of Ralph Crane, the scribe, as he tried to give plays on the page a classical aspect. They also include the directions that eighteenth century editors noticed with such condemnation, like the instruction in Macbeth for ‘A shew of eight Kings, and Banquo last, with a glasse in his hand’ (Macbeth, TLN 1657-8). As, in the speech to come, the first king is said to resemble Banquo, while the last is said to carry a looking glass, Theobald protested that ‘The Editors’ (Heminges and Condell, again) ‘could not help blundering even in this Stage-Direction’, which, indeed, does contradict the dialogue. Shakespeare, Works ed. Theobald, 5: 443. Such directions seem to be (wrong) attempts to help the reader visualise the staging; as they are non-performable, the word ‘stage’ sits oddly on them. They are ‘reader directions’. Harder to sort out is who wrote these varied paratexts. The ‘reader directions’ have been said to be by ‘editors’ preparing texts for the page, but could in fact be by anyone from playwright to compositor in the printing house; other, scribal and more ‘stagey’ directions may be by playwrights, prompters, or other stage functionaries, or a stage-focused compositor. Playwrights will probably have written some of the directions, though surviving manuscripts suggest they did so haphazardly; other professionals seem often to have added to what was there. W. B. Long, ‘“A bed / for Woodstock”: a Warning for the Unwary’, Medieval and Renaissance Drama in England, 2 (1985), 91-118. Only a particular kind of direction is certainly authorial: the ‘implied’ stage direction, embedded in the language (like ‘weepe / Not, gentle boy’), which does not take stage direction form at all. Francis Beaumont, Phylaster (1620), C4r. The other type of stage direction sometimes thought to bear the hallmark of a playwright’s authorship is more questionable. ‘Fiction’ directions like ‘Witches vanish’ (Macbeth, TLN 179) – ‘vanish’ here meaning, in staging terms, ‘exit’ – are often said to originate in a playwright, because they come from someone deeply involved in the story’s narrative. But such directions may equally, of course, be particularly intense ‘reader directions’, or, alternatively, may have been theatrically explicable. It is always easier to say who directions are for than who they are by. With ‘scribe directions’, ‘stage keeper directions’, ‘prompter directions’ and ‘reader directions’, as well as ‘implied directions’, however, what is clear is that there is no single type of text that is a ‘stage direction’ – indeed, as shown, not all surviving such directions are even about staging. Perhaps that is why, in the period, there was no collective terminology for these paratexts. Only ‘entrances and exits’, are spoken of as a staging unit – famously, in Shakespeare’s As You Like It, men and women on the stage of the world ‘have their Exits and their Entrances’ (TLN 1120) – and only entrances and, sometimes, exits were on occasion further extracted onto special documents, ‘backstage plots’. For more on backstage plots see Tiffany Stern, ‘Backstage-Plots’ in Documents of Performance, 201-231; for more on entrances see Mariko Ichikawa, Shakespearean Entrances (Basingstoke: Palgrave, 2002), passim. The lack of a broader terminology for ‘stage direction’ underlies the notion that instructions and recollections, imaginings and fact, texts for scribes, texts for prompters, texts for readers are not one thing, even if made to share layout and language. What is ironic is that the phrase used to categorise them has hidden the range of what they actually are. ‘Stage directions’ after the phrase A change in the theatres of the late nineteenth century altered stage directions forever. The prompter, whose job had been to facilitate performance of the play as received, was relegated; a new role was created, that of the actor manager, later called the ‘director’. It was the job of the director to give a unique, creative, personal interpretation to the plays he or she was overseeing – which often contradicted, or left little space for, the playwright’s own artistic stage vision. The battle for creative ownership of the play, between playwrights and directors, was born. It took place in stage directions. Nineteenth-century playwrights, usually denied access to rehearsals, began to use stage directions to dictate what they wanted from performance in an attempt to lead, rival or stymy ‘stage directors’. Directors then and ever since have largely ignored these intrusions. Modernist theatre practitioner and theorist Edward Gordon Craig, who expressed his ideas in ‘theatrical’ dialogue form, declared as early as 1905 that authorial stage directions were ‘an offence to the men of the theatre’: Stage Director: If to gag or cut the poet’s lines is an offence, so is it an offence to temper with the art of the stage-director. Playgoer: Then is all the stage direction of the world’s plays worthless? Stage Director: Not to the reader, but to the stage-director, and to the actor – yes. Gordon Craig, The Art of the Theatre (Edinburgh and London: T. N. Foulis, 1905), 29-30. Almost one hundred years later, in 2003, director Jean Schiffman explained how she had been taught to ‘cross out stage directions on the first reading’; acting and teaching director Amy Glazer was instructed likewise, being told ‘it’s a sign of a bad actor to even look at stage directions’. Jean Schiffman, ‘Taking Directions’, Backstage (5 March, 2003), https://www.backstage.com/news/taking-directions/ accessed 31st July 2017. Against this, a few possessive playwrights fought a counter battle. Samuel Beckett, author of two plays that consist only of stage directions, Act Without Words I and Act Without Words 2, wrote detailed and prescriptive directions for all his performances, his Endgame instructing the actor of Hamm phrase-by-phrase what to do: My... dog? (Pause.) Oh I am willing to believe they suffer as much as such creatures can suffer. But does that mean their sufferings equal mine? No doubt. (Pause.) No, all is a— (he yawns) —bsolute, (proudly) the bigger a man is the fuller he is. (Pause. Gloomily.) And the emptier. (He sniffs.) Samuel Beckett, Endgame (New York: Grove Press, 1958), 2. Performers given permission to put on Samuel Beckett’s plays are contractually obliged to follow his stage directions by his estate. But even Beckett was not able to prevent the American Repertory Theatre’s 1983 director-led production of Endgame; instead, he had a bitter note inserted, by legal demand, into the programme: ‘[This] production which dismisses my directions is a complete parody of the play as conceived by me’. Legal insertion in American Repertory Theater’s programme for Endgame (1984). As, with the rise of the director, stage directions became ever less significant in production terms, so they became correspondingly more powerful on the printed page. New copyright laws – an international law of 1887 in Europe, an 1891 law in US – partly brought this about. Before the laws, playwrights had relied on performance for payment, and often avoided publication, as theatre companies could legally perform any play once it was in print. But the new laws protected plays in print from unsanctioned performance. Now playwrights started to conceive of two lives for their dramas, one on the stage, one on the page. The result was the stage direction written with the page specifically in mind: the ‘literary’ stage direction – or, rather, the ‘literary direction’ as there is often little stagy about them. Bernard Shaw’s plays feature almost exclusively ‘literary directions’. His Man and Superman (1905), for instance, has directions that are sometimes four pages long; they are generally unactable: Hector Malone is an Eastern American; but he is not at all ashamed of his nationality. This makes English people of fashion think well of him, as of a young fellow who is manly enough to confess to an obvious disadvantage without any attempt to conceal or extenuate it …’ Bernard Shaw, Man and Superman (New York: Brentano’s, 1905), 61. This background information, written with Dickensian aplomb, gives the character of Malone a rich pre-history, allowing the play on the page to rival, or be seen as a version of, a novel. A consequence of playwrights’ new focus on stage directions is that the term once again changed meaning. ‘Stage directions’ came to be thought of as texts by playwrights for actors (that they are often for readers has not made it into conventional definitions). A 1929 Dictionary describes stage direction as ‘a direction printed or written with a play, as to the manner in which it is to be acted’, and similar such definitions have been supplied ever since. Funk and Wagnall’s New Standard Dictionary of the English Language (1929), 4: 2361. Even the OED, which traces the term ‘stage direction’ only back to the 1790s, calls it ‘a direction inserted in a written or printed play where it is thought necessary to indicate the appropriate action, etc.’ ‘Stage, n.’, OED Online. Oxford University Press, June 2016. Web. 30 July 2017. In fact, stage directions remain as mixed in authorship and intention as before, depending on the version of the text published. ‘Acting editions’ of plays, like the those published for Samuel French, largely concern particular performances, and are likely to be taken from production prompt books including theatrical ‘stage directions’; ‘literary’ versions are likely to preserve an authorial text with authorial stage-directions, though, if also post-production, may well include performance notes too. What hasn’t changed is the appearance and placement of stage directions: they keep, by and large, the format developed for them in the medieval and early modern period, and are often marginal, italic and, in terms of ‘enter’ and ‘exit’, in Latinate word-order. As a result, stage directions are still made to broadcast their nature as secondary texts, open to revision and change; they tend to be treated as such. This can be seen in modern editorial practice: editors of historic plays, reverent to dialogue, often add to or modify stage directions, while also excluding them from line numbering. They render these tiny texts authorially suspect, and hard to quote, as a result. Stage directions have always been in every respect – placement on the page, look, phraseology, authorship (and hence moment of creation and intended reader), necessity and treatment – awkwardly different from the dialogue they surround. It is the term used to define them that has enforced, and latterly brought about, notions of their authoriality and purpose. Only when we realise that a stage direction is not always for the stage, is not necessarily a direction, and does not have the consistency over time that its appearance suggests, will we come nearer to decoding this – or, rather, these – fascinating and variable souvenirs of stage and page.