This paper examines Practice-based Research (PbR) as a tool for early period theatre history, arguing for a distinction between PbR and its near relation, Practice as Research (PAR) in terms of the relationship of practice to knowledge-dissemination. In the first part of the paper, I consider the role PbR has played in my own work, and present a preliminary methodology for the application of performance workshops in the study of medieval performance literature. In the second half of the paper, I describe the outcomes of a recent workshop focused on the Northampton Abraham and Isaac and demonstrate the value of PbR for early performance history.
My research on early period performance originates in the vexing question of what happened to all the texts of medieval plays and performances that we know, primarily from the work of REED, occupied the imagination of medieval England. Until recently, the specific focus of my work has been saints’ narratives, especially the legend of St Katherine of Alexandria, one of the most popular of the medieval saints in performance;1 in my investigation of the relationship of ‘lost’ performance histories and extant manuscript texts, my methodology had been chiefly textual. That is, my approach has been to seek to identify performative features embedded within select medieval texts: strategizing about where and in what kinds of non-dramatic texts performative features might appear; describing the appearance and the function of performative features; speculating what these performative features might demonstrate; and proposing how the appearance of the features might challenge traditional medieval theatre historiography. I came to Practice-based Research (PbR) initially as part of the development of a methodology that might allow for the recovery of absent, or ‘lost,’ performance histories through close attention to textual and archival records: in this model, PbR works as a test of my ideas. But, over the last few years, I have been thinking and writing more explicitly about what actual role performance might play in my research, and – still more tricky yet – thinking also about the articulation of that role within the academic discipline I call home (English literary studies).2 My contribution to discussing the performance of The Three Ladies of London is a brief exploration of the following questions: what role might embodied performance play in the research of historically distant texts? what would a methodology based on embodied performance applied to medieval texts look like? and, what should the role of the PbR practitioner / textual critic be? Following a preliminary articulation of the methodology I am proposing, I offer as an example the results of PbR in practice.
‘What methodologies are appropriate to retrieve a theatre history for which there are few traces?’ asks Katie Normington in her discussion of English convent drama.3 Practice-based Research, for me, has provided a framework for thinking about applying embodied performance in the study of embedded performance traces. Baz Kershaw, widely recognized as a leader of the Performance-as-Research (PAR) movement, and director of the now concluded PARIP, the Practice-as-Research-in-Performance multi-institutional project in the UK, describes PAR as ‘creative practices considered as a research methodology in their own right’.4 Though the terms PbR and PAR are used by various scholars nearly synonymously, some theatre practitioners are careful to draw precise distinctions: Robin Nelson, for instance, draws attention to the difference between PbR and Practice as Research or Performance as Research (PAR) by emphasizing that PAR implies that the practice/performance is the research outcome and requires no further theorizing, while PbR implies that the practice/performance is one part of the research that usually leads to further research conclusions.5 Angela Piccini has written that PAR and PbR ‘remain troubling terms that resist close definition. ... PAR and PbR are used interchangeably to suggest a relationship of research between theory and practice, a binary that is increasingly difficult to justify despite the language of “putting one’s theories into practice” commonly articulated at a range of PAR events across the disciplines’.6
When I think about the application of PbR in my research, I am drawn to the way the practice emerged as an antidote to the ephemerality of much practical theatre history, to the variability of audience experiences, as well as to the tension between playing-texts and play-text canonicity, among other things. In this way, there are obvious relevancies to medieval performance studies. As Robin Nelson notes in thinking about the application of PbR in the academic institution, however, ‘One of the major issues ... is whether practice-based arts research can be aligned with established research paradigms or whether, for the academy fully to embrace its outcomes, it requires a shift in the conception of what constitutes research, and even what constitutes knowledge’.7
For me, a textual critic in an English department working on medieval performance practices, the ‘shift in the conception of what constitutes research’ has been profound. Early in my thinking about the value of PbR for my work, I urged a clear distinction between PbR workshops and dramatic reconstructions of early period texts, as well as between PbR and the work of blocking and/or walk-throughs necessary for the development of contemporary performances. In part, this distinction derives from my primary focus on identifying performative elements in non-dramatic manuscripts texts and reclaiming them as ‘artifacts’ or ‘textual relics’ within a more comprehensive set of medieval performance histories. But, even as my own thinking has expanded to include PbR work on existing medieval dramatic texts, I would continue to maintain that PbR differs from the work of dramatic reconstruction or staging preparation in that it frequently advances through workshops with select and/or partial textual focus, and in response to questions not related to eventual productions. Workshops, with an emphasis on improvisation and repetition, for instance, are the central tool of the PbR methodology.8 What remains to be parsed, however, is the relationship of the workshop – of performance research, that is – to the dissemination of knowledge. In contemporary PbR / PAR paradigms, the performance is often understood to be the research contribution; how, then does this differ for early period work? For me, this question captures one of the central anxieties attaching to PbR / PAR: how do we disseminate the knowledge generated by the practice? What is the role of documentation in PbR? How do workshops / performances and the research they embody generate new knowledge or influence scholarly practice without formal means to communicate that research? And, more specifically, are these questions in some way particular to the application of PbR to early period work?
In more general terms, this anxiety has to do with the assessability of PbR and its outcomes. Kershaw has admitted his ongoing resistance to the funding bodies supporting PARIP and their push for recognizable research deliverables: the assessors, he writes, expected that ‘practical creativity must have add-ons to join in the business of knowledge-making’.9 This expectation, he continues, put him in a complicated position:
As PARIP lead investigator, I was queasy about the officially produced paradox that the traces of a creative performance had more value than the event itself. Is an empty plate always the best part of a meal? But I was also intrigued, as the paradox suggested a lacuna (or aporia) at the heart of PaR, and of performance studies as well. Back in the 1950s, the Absurdist playwright Eugène Ionesco pinpointed this more positively when he wrote that ‘only the ephemeral is of lasting value’.10
Assessability is the crux of this discussion, and linked to it, of course, is the question of the relationship of PbR to, or role of PbR in, the academic disciplines within which most practice-based researchers work, or pursue graduate degrees. Roberta Mock recalled key concerns from the 2003 PARIP meeting: ‘What makes an instance of practice “count” as research? Does practice as research involve different methods as a result of its framing as research as distinct from “pure” practice?’11 Each of the contributors to the discussion engaged with the question of documentation and output, not only from the point of view of assessability but also of knowledge creation and communication. Surprisingly, several of them noted their dissatisfaction with the idea that there were no clear expectations for dissemination or documentation of research to coincide with performance practices and/or performances. Drawing on the image of the tree in the forest, John Whelan asks ‘If “research” does not appear in print, has it really happened?’.12 Franc Chamberlain lamented the ‘shortage of formal discussion of performances which were shown as examples of PAR’ at the conference.13
As I have developed my thinking about PbR and medieval performance histories, I have struggled with two aspects of this problem in the definition of the expectations of the methodology. First, I believe that the performance-work is the research, and that the results can be significant. But I also believe that the research generated by the performance needs to be disseminated in ways complementary to (not necessarily more valued than) the performance: that is, in ways commonly distinct from the performance-work. Therefore, I believe there are benefits to defining a PbR methodology more precisely, emphasizing the distinction between PbR and PAR, especially for those of us working with early period texts. I now believe that in Practice-based Research, as the ‘based’ implies, the performance is the central research that leads to the research conclusions; it is the foundation of the research, and thus the key component of the scholarly work. But, ‘based’ also implies that it is not all there is: as compared to PAR, PbR suggests an obligation to think explicitly about ways of disseminating the knowledge the practice generates, though not necessarily solely in traditional print models. Melissa Trimingham recalls her experience at the PARIP conference and argues more strenuously that ‘PAR is doing itself no favours by claiming that “all practice is research”. All practice is relevant to research but does not necessarily contribute to research until it is subject to analysis and commentary using a language that aims to be as clear and unambiguous as possible’.14
I am persuaded from my own recent experiences that this statement is right. Early period performance is not the same, obviously, as contemporary performance practices, the original context for PbR and PAR theory. Contemporary practitioners may create a performance as a piece of research and argue compellingly that the performance is the sole research contribution; further, the ephemerality of that same performance may be seen to be an inherent part of the research. But the practice of performance-research works differently for theatre historians: we conduct our research in the absence of original conditions, audiences, staging technologies, and sensory experiences. And while, arguably, all performances of historically distant texts are instances of performance-research, I am thinking specifically about performance work undertaken directly as part of a research question or line of inquiry. This environment is especially apparent for those, like me, who work in the areas of speculative performance or speculative theatre history. PbR, in this context, is the sole means of developing, exploring, testing ways of thinking, and yet, I would venture, in the end the performance is not sufficient all on its own; the knowledge it generates needs to be disseminated to affect critical practice, especially when the knowledge is related to the broadening of the performance canon. As Trimingham contends, ‘Artistic insight is not necessarily a research outcome’.15 In other words: where is the benefit if the workshops or performance-work I lead result in knowledge about new performance texts but no one knows about it except me and the participants? The performance-work may be the research, but the papers still have to be written both for dissemination and to influence critical practice, as well as for assessability – by my peers, as well as by my institution which keeps track of and assesses my research in terms of ‘deliverables’.
As I noted above, I come to performance studies initially not through dramatic literature, but through critical speculation about the performance histories of texts that survive in non-dramatic form. I have been fortunate enough, however, to use PbR in workshops that explore the performance possibilities of non-dramatic texts as well as in workshops focused on more canonical dramatic literature to explore questions not answerable in any other way but through embodiment, movement, and the attention to spatial relations in performance. Most recently, at the November 2014 meeting of ASTR (The American Society for Theatre Research) in Baltimore, I brought a set of questions drawn from my ongoing project on Trinity College Dublin, MS 432,16 a mid-fifteenth-century manuscript composed of several separate parts assembled probably by one seventeenth-century owner, James Ussher. I have applied PbR workshops to texts in this manuscript before, with encouraging success, but for ASTR 2014 I wanted to bring the manuscript’s distinct version of the Abraham and Isaac play to our group. This play, commonly known as The Northampton Abraham and Isaac (ff 74v-81r),17 is part of a paper manuscript containing various other texts, mainly poems, with a consistent mid-fifteeenth-century hand throughout.18
The Northampton Abraham and Isaac is one of six extant medieval English plays on the theme of Abraham’s willing sacrifice of his son: the other five include the distinct Brome Play (Yale, Beinecke Library MS 365, ff 15r-22r), as well as plays included as part of the cycles of Chester, Coventry, Towneley, and York.19 I am drawn to this particular play for two reasons: one, it is unusually situated in a manuscript of other non-dramatic material which, for me, poses questions about late-medieval compilation and generic hybridity; and two, no other dramatic version of the same story in the English tradition treats Sarah, Abraham’s wife, as fully as this one does. Indeed, Clifford Davidson has claimed that in this version Sarah ‘provides a complicating factor not present in the other British plays on the subject. Abraham would rather face her displeasure than God’s, but still the opportunity is present for the actor playing the role to stress the anxiety that naturally would come of such a conflict’.20
A workshop seemed like a perfect place to explore the figure of Sarah in this play. My first attempts at thinking about this characterization were, not surprisingly perhaps, through language: I considered the speeches given to Sarah (37 lines of the play’s 369), and the way her dialogue worked in tandem with the other speeches and the narrative’s theological centre. But, increasingly, I had questions about the actual presence of Sarah on the stage. In preparation for the ASTR 2014 sessions, Lofton Durham, my co-organizer, and I strategized about ways to explore the consequences of Sarah’s bodily presence on the stage in the context of our workshop. Considering the original manuscript stage directions, we were truly astonished to discover that, despite the quite explicit directions moving angels, Abraham, and Isaac around the stage, Sarah is simply there. She neither enters nor leaves the playing space: she is already present when Abraham returns home after speaking with the Angel; she remains behind when Abraham and Isaac venture out with the servants to the site of the sacrifice; she is waiting where they left her when they return after the angelic intervention. The manuscript specifically directs Abraham and Isaac to travel to her before she speaks to them: ‘Et equitat versus Saram et dicit Sara’ (f 80v). In other words, the dramatic movement can be summed up like this: Abraham is removed from the domestic space; he returns to Sarah; he and Isaac leave the domestic space; they return to her. While reading the play closely can reveal this fact, at least two-dimensionally, the effect of her presence is not evident in the words of the text. The workshop was designed, therefore, to address specifically the question of Sarah’s bodily presence in the space during the sections of the play when she does not participate in the action.
Lofton and I agreed that to make the workshop as effective as possible, we would need to keep our text selection very brief. We decided on two short sections: one, part of Sarah’s response to Abraham’s command that Isaac will accompany him to make the sacrifice (ll 108-27); and two, part of Abraham and Isaac’s exchange after Abraham reveals his sacrificial intention to his son (ll 184-99). This second passage is striking since Sarah does not figure directly in it all; however, she is figuratively present in the dialogue as both Abraham and Isaac speak of her (Isaac asks, ‘But, gentil fader, wot my modre of þis, / þat I shal be dede?’ ll 186-7), and through Isaac’s concern for her (‘Let neuer my moder se my cloþus; / For and she do, withouten othus, / It wol greue her to smert’ (ll 205-7). The group of participants, under our direction, were eager to explore the implications of Sarah’s presence in the play, figuratively, and, physically, in the playing space.21 This is the section we focused on:
Isacc Alas, what haue I displesid þis Lord of blisse,
þat I shal be martyred in þis mysse?
But, gentil fader, wot my modre of þis,
þat I shal be dede?
Habraham She? Mary, son, Crist forbede!
Nay, to telle her it is no nede;
For whan þat euer she knoweþ þis dede,
She wol ete affter but litel brede.
Isacc In feiþe, for my moder I dar wel say,
And she had wist of this aray
I had not riden out from her þis day,
But she had riden also.
Habraham Ye, son, God most be serued ay,
þi modre may not haue hir wille all way.
I loue þe as wele as she doþe, in fay,
And yit þis dede most be do. (ll 184-99)
What emerged through our workshop was significant. Lofton and I arranged our large meeting room at the conference so that it would accommodate a playing space composed of a static domestic space as well as the ‘wilderness’ the male characters venture into and return from. There were no sets in our workshop or costumes, and the characters were drawn from among the participants, who, armed with scripts, spoke and moved through the same set of lines several times, switching places with other participants, and interrupting the ‘action’ to think through what we were witnessing and discovering. If Sarah remains in the playing space while Abraham reveals his fate to her beloved son, we asked, what would Sarah be doing? Is she hidden from view? Possibly, but why would any actor at any time in history give up the dramatic potential of having Sarah’s body visible while the male characters speak her name? The poignancy of viewing Sarah’s body tense as Isaac cries out, ‘does my mother know that I will be killed?’ (‘wot my modre of þis / þat I shal be dede?’) is surely too powerful to yield to other interpretations. Sarah’s bodily presence on the stage during this key scene in the narrative would render the play theologically more complex than other versions, we determined, if only through the emphasis on the mother’s will (‘þi modre may not haue hir wille all way’) as analogous to human will and doubt. In the final lines of the play, Abraham speaks ‘What God comaundeþ say not nay, / For ye shal not lese þerby’ (ll 368-9). In these ways, the figurative and the physical merge.
The power of Sarah’s body – a physical presence made evident through PbR – highlights this play’s remarkable theatricality, giving it depth and energy. The importance of her character develops through speeches and dialogue, certainly, but it depends on the theatrical nature of the drama, on the inherent physicality of bodies in relation to the playing space. These conclusions emerge most effectively through the interpretive possibilities offered through Practice-based Research, through a methodology grounded in performance. Medieval performance tantalizes modern scholars in large part because of its ephemerality; PbR, with its emphasis on performance workshops and complementary critical dissemination, is an important practice both for recovering early period performance histories and for influencing critical practices.
 See Clifford Davidson’s list of performances, drawn from the Records of Early English Drama (REED) and elsewhere: ‘British Saint Play Records: The Saint Plays and Pageants of Medieval Britain’, EDAM Project: Early Drama, Art, and Music, Western Michigan University, http://scholarworks.wmich.edu/early_drama/. Available as pdf.
 I have explored these ideas most fully in the context of the last two meetings of the American Society for Theatre Research (ASTR), specifically in the Medieval Performance Working Group, which I co-organized with Lofton Durham in 2013 (Dallas) and 2014 (Baltimore), and at the recent conference of the Société internationale pour l’étude du théâtre médiéval (SITM; Poznan, Poland, 2013). My particular contribution to the organization of the working group at ASTR has been the introduction of PbR workshops to supplement the paper sessions. I am grateful to the participants in both sets of meetings for their generous feedback on my ideas as they have developed.
 See Jim Davis, Katie Normington, and Gilli Bush-Bailey with Jacky Bratton, ‘Researching Theatre History and Historiography’, Baz Kershaw and Helen Nicholson (eds), Research Methods in Theatre and Performance (Edinburgh, 2010), 86-110.
 Baz Kershaw, ‘Performance as Research: Live Events and Documents’, Tracy C. Davis (ed.), The Cambridge Companion to Performance Studies (Cambridge, 2008), 23.
 Robin Nelson, Practice as Research in the Arts: Principles, Protocols, Pedagogies, Resistances (Houndmills and New York, 2013), ‘Modes of PaR Knowledge’, 117. See also the introduction to Hazel Smith and Roger T. Dean (eds), Practice-led Research, Research-led Practice in the Creative Arts (Edinburgh, 2009).
 Angela Piccini, ‘An Historiographic Perspective on Practice as Research’, Studies in Theatre and Performance 23 (2003), 192.
 Nelson, ‘Modes of PaR Knowledge’, 114, rightly observes that Humanists are much more comfortable when working with clear and rational argument, and – more to the point – the institutions which evaluate and assess our research and its outcomes are more able to recognize those outcomes in the form of clear and rational argument; PbR, he perceives, raises problems in this regard: ‘The particular question ... is whether anything might be called a rigorous research method which not only does not present itself in terms of rational argument but which might not even be put into words’ (115).
 See for instance descriptions of the workshop practices employed by Lynette Hunter and Peter Lichtenfels, especially in their work on The Humorous Magistrate in ‘(Un)Editing with (Non-)Fictional Bodies: Pope’s Daggers’, Jacqueline Jenkins and Julie Sanders (eds), Editing, Performance, Texts: New Practices in Medieval and Early Modern English Drama (Houndmills, 2014), 171-97.
 Kershaw, ‘Performance as Research’, 25.
 Roberta Mock, in discussion with Franc Chamberlain, Simon Ellis, John C. Whelan, and Nicholas Till, in ‘Reflections on practice as research following the PARIP conference, 2003’, Studies in Theatre and Performance 24.2 (2004), 129.
 Ibid, 134.
 Ibid, 130.
 MelissaTrimingham, ‘A Methodology for Practice as Research’, Studies in Theatre and Performance 22.1 (2002), 54.
 Ibid, 54.
 Hereafter TCD 432; olim Trinity College MS D.4.18.
 Cf. Norman Davis (ed.), Non-Cycle Plays and Fragments, Early English Text Society s.s. 1 (Oxford, 1970) 32-42; notes xlvii-lviii.
 Cf. Norman Davis (ed.), Non-Cycle Plays and the Winchester Dialogues: Facsimiles of Plays and Fragments in Various Manuscripts and the Dialogues in Winchester College MS 33 (Leeds, 1979), 33. The play under discussion is located in the third part of the manuscript, ff 59-86 though with pages out of order in the current binding.
 There seems to be no extant evidence that the ‘Northampton’ play was ever actually performed in Northampton, even though it would seem to be a good workable play. It is so-called because of some of the non-dramatic material in the same manuscript section. Alexandra Johnston, who has started the work on the REED Northampton volume, generously shared some of her preliminary research with me in 2012; this early research has confirmed that no records indicating a performance of this play – or much else for that matter in the area – appear to be extant.
 Clifford Davidson, ‘The Sacrifice of Isaac in Medieval English Drama’, Papers on Language and Literature 35.1 (1999), 48.
 I am very grateful to my co-organizer, Lofty Durham, and the participants in the Medieval Performance Working Group at ASTR 2014 for their enthusiasm for PbR and their willingness to participate in the workshop we held, as well as for all the feedback I’ve received in the last two years on my written explorations of the theory behind the methodology. The contributors to last year’s sessions included Carolyn Coulson, Susannah Crowder, Erith Jaffe-Berg, Heather H. Jennings, Brann Munro, Jenna Soleo-Shanks, and Claire Sponsler.