This paper describes the first phase of work in a long-term project to digitize the Stratford Festival of Canada’s collection of Shakespeare promptbooks. The collection is notable for its scale, including over 230 production promptbooks; for its completeness, recording every production of a Shakespeare play (encompassing the entire canon) since the Festival’s founding in 1953; and for its usage patterns, with promptbooks most frequently consulted by Festival artists preparing for new productions (Malone 2013, 66). It also includes the “composite scripts” developed by archivist Jane Edmonds in the late 1990s to help Festival artists compare the textual emendations of past productions (Malone 2013, 68). The composite scripts, which are hand-annotated collations presented in colour-coded, palimpsestic form, prefigure the potential affordances of encoded textual transcriptions for computational analysis and visualization. Malone (2013) sees the potential for a comparative analysis of the Stratford promptbooks to reveal “the dramaturgical heritage” of the Festival (71); and suggests visualizing promptbooks in digital, parallel-text editions as a means of conducting and disseminating this form of research (Malone 2018). In addition to textual data, however, promptbooks also contain rich records of other production phenomena, including technical cues, actors’ blocking and gestures, backstage activity, audience engagement, performance variations, and even unrelated marginalia. This rich data has the potential to significantly expand our understanding of Shakespeare in performance, not only within the Festival context, but also – so long as it is encoded according to reproducible conventions and is interoperable with other data sets – across geographical, historical, and other contexts.
Although other robust promptbook remediation projects exist, including the Folger Shakespeare Library’s extensive microform collection, to date they have focused primarily on facsimile reproduction rather than encoding (Malone 2018, 107; and see our discussions of the Shakespearean Prompt-books of the Seventeenth Century project below). This may in part be a result of the wide variations in stage managerial practices of promptbook preparation, even as attested over the six decades represented in the Festival’s collections (see Malone 2013, 66–8). It may also be a result of the understanding in the field of textual editing of play texts as literary works intended for interpretation in performance, rather than as elements of and ontologically secondary to the performance works to which they contribute (Roberts-Smith et al. 2014, 69–70). As a result of that understanding, editions most often relegate information about performance to commentary in annotations; this is true even in digital series, such as the Queen’s Men Editions, that construe themselves explicitly as performance editions (see Griffin 2018). Similarly, the Text Encoding Initiative (TEI)’s guidelines have a limited ability to account for the kinds of data encountered in promptbooks, as opposed to in print editions of dramatic texts, for which the TEI Guidelines for “Performance Texts” were developed (TEI Consortium 2019).
In this context, the Stratford Festival Online project (a collaboration between the Stratford Festival and the University of Waterloo, founded in 2014 with support from the Ontario Ministry of Research, Innovation and Science) undertook in its initial phase to:
articulate a taxonomy of ontologies of the textual contents of promptbooks – that is, the forms and purposes that make the promptbook texts different from other kinds of literary text – that understands them as instruments of (and hence secondary to) works of performance;
assess the relevance of existing TEI Guidelines for Drama to promptbook encoding.
In the discussion that follows, our central claim is that because the principal function of a promptbook is to help initiate and monitor spatiotemporal events which are more important than the physical book that records them, current TEI standards for digital encoding do not always apply. Consequently, later phases of the project will:
determine the potential for revisions to TEI Guidelines to represent the ontologies of the different kinds of text that appear in promptbooks while enabling interoperability with other major digital editions of Shakespeare;
determine the depth of encoding appropriate to promptbook digitization;
develop and validate an encoding standard for promptbook digitization; and
prototype interfaces for rendering promptbook data.
What is a promptbook?
In this project, we are concerned with the working promptbooks defined by Charles H. Shattuck (1965) as the “book[s] actually used by prompters or stage managers in conducting performances” (5). These are distinct from “souvenir” promptbooks and “performance editions” intended for publication (Shattuck 1965, 5; Osborne 1996, 171); from early modern “playhouse texts” annotated with performance-related information (Werstine 2012, 234); and from other “ontologically complex” witnesses to Shakespeare’s works, which represent text and performance elements simultaneously (Galey 2014, 160), such as, for example, the “stage texts” of interest to the Shakespearean Prompt-books of the Seventeenth Century project. A working promptbook, by contrast, is the central instrument by which a modern production of a performance work remains consistent over a series of individual performance events. The working promptbook is normally compiled by a stage manager during a rehearsal process, and records three principal categories of information required for the stage manager to ensure consistency during the run of a production: First, it records the sequence of actions the stage manager must perform so that other contributors can perform essential actions at the correct time. These are largely technical cues that the stage manager must say during a performance, along the lines of “cue light up right go” or “stand by LX 8–10” (see Figure 2). Second, the promptbook records the actions that others are required to perform, so that the stage manager can monitor them and remind others after a performance if accuracy or timing has slipped. These largely take the form of “blocking” sketches and notes, showing where and how actors are expected to move on the stage (see Figure 3).
Although stage managerial practices have varied significantly over the decades documented in the Festival’s collection, and do not demonstrate consistency until the 1980s (Malone 2013, 66–7), stage managers have continuously used the same convention for recording the timing of the actions required in a performance. Events are keyed by various mechanisms – including asterisks, numbering systems, and arrows – to words spoken by actors on the stage. Hence, the third major category of information recorded in a promptbook – along with what the stage manager should do and what the stage manager should monitor – is when in the sequence of onstage utterances the stage manager’s and others’ actions should take place (see Figure 4). To put this another way: a promptbook records a series of utterances to which intended performance events are mapped in a relative temporal sequence.
Taxonomy of ontologies in a promptbook – Working draft
Following the usual practice in text encoding projects, we began our exploration of the scope and demands of encoding the Stratford Festival promptbooks by generating a taxonomy of the ontologies of data that are available for encoding. They are as follows:
The timeline of a performance. While this is mainly verbal (the utterances spoken by actors), it may also include indications of temporal periods before a performance, silences during a performance, intermissions, curtain calls, and temporal periods after a performance.
Performance events. These are the actions undertaken by the stage manager (cues spoken) or monitored by the stage manager (actors’ blocking movements through space, specific gestures, and the unstructured gesture sequences often referred to as “business”).
Divisions of content in the promptbook. These include the entire performance, the units of conceptual stage managerial time represented by an opening, conceptual temporal subunits represented visually within an opening, and conceptual content units explicitly identified (such as scenes).
Marginalia not relevant to performance.
Ontologies of content relationships
The ways in which events can be linked to the timeline. These may be momentary: events that occur at as precise a moment as the space between two uttered syllables, as in the cue for LX 8 to “go”, which is spoken between the syllables “ven” and “ted” of the word “prevented” in an utterance of the character of King Lear (see Figure 4). They may also be linked to a period of time spanning a number of lines. They may be individual events, simultaneous events, or overlapping events. And they may occur in all or only some performances or types of performance, such as, for example, only in performances that are student matinees. These content relationships are rarely recorded verbally, but more often symbolically, including by means of glyphs, numerals, drawings, and spatial proximity on a page. Such symbols are ordinarily generated by individual stage managers or are learned as apprentices. There are no global standards for the use of glyphs, symbols, or keys nor for their interpretation, but these markings are usually uniform within a single promptbook. While stage management guides and handbooks do exist with examples of useful symbols (see Ionazzi 1992), practices vary widely: indeed, as noted by Malone (2013), over their careers individual stage managers often generate their own personal shorthand to indicate stage movements, which may or may not resemble another system. Importantly, these notes must be clearly enough understood by the stage management team for any individual production that another stage manager could interpret them in case of emergency. Possible future applications of this project include comparative analyses of stage managerial notation practices among stage managers and over time.
These are the variables that describe content in sufficient detail that it is usefully machine-readable. Examples may include “where” (useful for searching blocking notes), “who” (useful for searching for particular actors’ gestural habits), and so on.
Editorial interventions. These may be interpretive interventions made at the level of encoding, i.e. determinations of the nature of the content that result in encoding it in one way rather than another; or they may be explanatory commentary in editorial annotations.
Provenance of the digital edition and its sources. This is the category of information normally recorded in the metadata in the TEI header.
Relationships to related content, physical or digital, within or outside the archive. This category includes not only potentially interoperable texts, but also catalogue records for material objects such as those held in the extensive costume and property collections housed in the Stratford Festival Archive, or records held in other archives related to, for example, the work of actors who have performed both at Stratford as well as in other theatres.
Existing TEI conventions: Strengths and limitations
The Shakespearean Prompt-books of the Seventeenth Century project (Evans 1997), housed in the Bibliographical Society of the University of Western Virginia, is a good example of an application of existing TEI to pre-modern performance texts. It offers collated bibliographical descriptions of the material artefacts it is transcribing and links its descriptions to facsimile images of the originals (see Figures 5 and 6). This approach is consistent with, and well-supported by, traditional editorial practice and existing TEI protocols for dramatic texts, which allow us to encode the bibliographical characteristics of dramatic texts and also to acknowledge performative elements suggested semiotically by the texts. The most obvious first example is the stage direction. The TEI Guidelines define the “stage” element in this way:
<stage> (stage direction) contains any kind of stage direction within a dramatic text or fragment.
@type indicates the kind of stage direction. Suggested values include: 1] setting; 2] entrance; 3] exit; 4] business; 5] novelistic; 6] delivery; 7] modifier; 8] location; 9] mixed (TEI Consortium 2019, 7.2.4)
The list of suggested values offered in this definition demonstrates that TEI does not distinguish between things that theatre-creators can do and audiences can witness (such as “entrance”, “exit,” and “business”); things that audiences can understand by inference from what they witness (such as “setting”); or meta-things audiences can infer by inter-textual comparison (“novelistic”). The latter category also crucially collapses the difference between text and performance, since audiences do not normally actually perceive stage directions in performance, and hence cannot determine whether a stage direction may have been novelistic in its style. The TEI Guidelines note that “a satisfactory typology of stage directions is difficult to define [… I]t is not uncommon to mix types within a single direction” (TEI Consortium 2019, 7.2.4), and this is demonstrably true in the case of literary stage directions such as those offered in the guidelines’ list of examples. It is, however, demonstrably untrue in the case of promptbooks, which consist entirely and only of descriptions of things that theatre-creators can do and audiences can witness. In other words, promptbooks consist on the one hand only of stage directions, and on the other hand only of one particular type of stage direction. We argue, consequently, that the concept “stage direction” – at least as it is construed in the TEI Guidelines – is fundamentally textual and perhaps also literary. We might even say that it is principally a bibliographical tag, analogous to the typographical conventions that make it easy for readers to recognize stage directions on a page; its main function is to help readers locate semiotic information not classified as speech.
A second illustrative case is the “move” element, which the TEI Guidelines describe in this way:
The move element is intended to help overcome the fact that the stage directions of a printed text may often not provide full information about either the intended or the actual movement of actors on stage. It may be used to keep track of entrances and exits in detail, so as to know which characters are on stage at which time. […] For stage-historical purposes, a @perf attribute is also provided; this allows the recording of different move elements as taken in different performances of the same text. (TEI Consortium 2019, 7.2.4)
The goals of the “move” element are explicitly to enable editorial annotation, either of an editor’s interpretation of the stage movements implicit in text and stage directions, or of a description of the actual stage movements undertaken in a particular past production, which can be identified using the “perf” attribute. There is of course no requirement that all “perfs” referenced be the same “perf”; in other words: the theatre-historical gesture is towards collation of multiple past performances rather than reconstruction of a single past performance. This is, again, although consistent with editorial practices, nonetheless ontologically different from the blocking instructions for “moves” in a promptbook, which do not describe what an editor thinks might happen or knows has happened in various productions in the past, but rather describe what must happen in order for this production to manifest in the future. We think of the “move” element as an example of interpretive semiotic annotation.
Conclusion: Why TEI?
Existing TEI conventions are, of course, not intended to supply instructions that a reader can follow in order to generate a non-textual work of performance in time and space. So, why engage TEI at all? In this project, our answer is that we want to be – and we want our texts to be – in conversation with the field of text encoding more broadly. In the most practical terms, we are hoping that TEI will be a vehicle for making the digital texts we create intelligible to and even interoperable with texts created by other digital projects. Our most immediate collaboration to date has been with Janelle Jenstad and Martin Holmes at Internet Shakespeare Editions (ISE), whose texts will be the first that we are able to display in the parallel-text visualization proposed by Malone (2018). We are not, however, naive about the intellectual labour involved in expressing what we see as a concept incompatible with existing TEI in TEI; in fact, we wonder whether what will emerge from our project might be a markup language developing out of TEI but actually unique to promptbooks, similar to the work done by John A. Walsh on establishing a markup language particular to graphic novels, called Comic Book Markup Language (Walsh 2012). Nonetheless, we see this work as the central advantage of working in TEI. As Julia Flanders (2016) puts it, TEI is “not a perfect language for collaboration, but a […] mechanism for negotiating about language: in fact, a mechanism for negotiating dissent” (74). Our understanding of the ontologies of the contents of the modern Shakespeare promptbooks in the Stratford Festival collection is essentially a dissenting one in the world of textual editing. Although promptbooks from one perspective may represent the words that comprise Shakespeare’s plays, annotated variously as a typical early printed text one would find in a library archive might be, the purpose and interpretive framework of a promptbook is entirely different from those precedents. Since a promptbook is a series of cues, recorded by a stage manager, which describe events taking place in time and space on the stage, and since the stage manager is primarily responsible for ensuring that those events do take place as described, the stage manager is both the primary author and the primary audience for a promptbook. The source and provenance of the words of Shakespeare’s plays are almost irrelevant in this context. For a stage manager, the words and markings on a page function less urgently as units of meaning in themselves, and more urgently as the temporal markers by which actions taking shape in the space of the theatre are organized. Word-marked cues do not comprise an absolute temporal system, of course, because, even though a company like the Stratford Festival will deliver extremely consistent running times for individual performances in a production run, words will not be spoken at exactly the same pace in every performance. However, the words in a promptbook do describe the parameters of repeatable relative sequences of events. Consequently, in a promptbook, text is not primarily text but rather a relative sequential navigation system; it is a spatiotemporal structural language, as opposed to a semiotic language (representing concepts) or a bibliographical language (remediating material signs). This ontological difference between language as text encoders normally understand it and the language of promptbooks warrants, from our perspective, a re-evaluation and extension of the principles of TEI beyond bibliographical and semiotic concerns to spatiotemporal concerns. We want to begin with TEI, because we want to participate with the TEI community in a negotiation of our dissenting understanding of the ontology of texts. TEI gives us a metalanguage through which to negotiate.
The authors have no competing interests to declare.
Conceptualization: JRS, MK, TM
Methodology: JRS, MK, TM, MH
Software: MK, JT
Formal Analysis: JRS, MK, TM
Investigation: MK, TM
Resources: LG, JJ
Data Curation: TM, MK
Writing – Original Draft Preparation: JRS, MK, TM
Writing – Review & Editing: JRS, MK, TM
Supervision: JRS, MH, JJ
Project Administration: JRS, JJ, LG
Funding Acquisition: JRS
Evans, G. Blakemore. (ed.) 1997. Shakespearean Prompt-Books of the Seventeenth Century. Bibliographical Society of the University of Virginia. Accessed March 11, 2019. http://bsuva.org/bsuva/promptbook/.
Flanders, Julia. 2016. “Challenges of Collaborative Standards for Digital Humanities.” In Collaborative Research in the Digital Humanities, edited by Marilyn Deegan, and Willard McCarty, 67–79. Abingdon: Routledge.
Galey, Alan. 2014. The Shakespearean Archive: Experiments in New Media from the Renaissance to Postmodernity. Cambridge: Cambridge University Press. DOI: http://doi.org/10.1017/CBO9781139629201
Griffin, Andrew. 2018. “Text, Performance, and Multidisciplinarity: On a Digital Edition of King Leir.” In Shakespeare’s Language in Digital Media: Old Words, New Tools, edited by Janelle Jenstad, Mark Kaethler, and Jennifer Roberts-Smith, 84–104. Abingdon: Routledge.
Ionazzi, Daniel. 1992. The Stage Management Handbook. New York: Betterway Books.
Malone, Toby. 2013. “‘Distract Parcels in Combined Sums’: The Stratford Festival Archives’ Stage-Managerial Collections.” Canadian Theatre Review 156: 66–71. DOI: http://doi.org/10.3138/ctr.156.012
Malone, Toby. 2018. “A Digital Parallel-text Approach to Performance Historiography.” In Shakespeare’s Language in Digital Media: Old Words, New Tools, edited by Janelle Jenstad, Mark Kaethler, and Jennifer Roberts-Smith, 105–24. Abingdon: Routledge. DOI: http://doi.org/10.4324/9781315608747-9
Osborne, Laurie E. 1996. “Rethinking the Performance Editions: Theatrical and Textual Productions of Shakespeare.” In Shakespeare, Theory and Performance, edited by James C. Bulman, 168–86. Abingdon: Routledge. DOI: http://doi.org/10.4324/9780203360071_chapter_10
Roberts-Smith, Jennifer, Shawn DeSouza-Coelho, Teresa Dobson, Sandra Gabriele, Omar Rodriguez-Arenas, Stan Ruecker, Stéfan Sinclair, Paul Stoesser, and Alexander Kovacs. 2014. “SET Free: Breaking the Rules in a Processual, User-Generated, Digital Performance Edition of Richard the Third.” In The Shakespearean International Yearbook 14, Part 1, Special Section, Digital Shakespeares: Innovations, Interventions, Mediations. Edited by Brett D. Hirsch, and Hugh Craig, 69–99. Farnham, Surrey: Ashgate. DOI: http://doi.org/10.4324/9781315264226-4
Shakespeare, William. 1988. King Lear. Directed by Robin Phillips, and stage managed by The. John Gray. Promptbook. Stratford Festival Archive.
Shattuck, Charles H. 1965. The Shakespeare Promptbooks: A Descriptive Catalogue. Urbana: University of Illinois Press.
TEI Consortium. (eds.) 2019. TEI: P5 Guidelines for Electronic Text Encoding and Interchange. Oxford: Text Encoding Initiative Consortium. Accessed March 11. http://www.tei-c.org/Guidelines/P5/.
Walsh, John A. 2012. “Comic Book Markup Language: An Introduction and Rationale.” Digital Humanities Quarterly 6(1). Accessed March 11, 2019. http://www.digitalhumanities.org/dhq/vol/6/1/000117/000117.html.
Werstine, Paul. 2012. Early Modern Playhouse Manuscripts and the Editing of Shakespeare. Cambridge: Cambridge University Press. DOI: http://doi.org/10.1017/CBO9781139103978