In The Semiotics of Performance, Marco de Marinis notes that the field of performance studies has greatly expanded the traditional categories of drama and theatre. “It is obvious,” he writes, “that we are dealing with a field that is far broader and more varied than the category consisting exclusively of traditional stagings of dramatic texts, to which some scholars still restrict the class of theatrical performances.”Footnote ¹ A few scholars of early theatre history have embraced expanded categories of performance. Jody Enders's “medieval theater of cruelty,” for example, rests on a concept of “a theory of virtual performance” that translates “into actual medieval dramatic practice.”Footnote ² Carol Symes's study of the “dramatic activity” suggested by medieval French manuscripts identifies “a vital performative element within the surrounding culture.”Footnote ³ Both writers have shown how new ideas of performance enlarge the category beyond the “traditional stagings” described by de Marinis.

Adopting an approach quite different from the approaches employed by Enders and Symes, I too propose to look for drama somewhere else than on the stage—in this case, in dialogue-rich Anglo-Saxon narrative poems. The performative tradition to which these texts belong has been excluded from drama criticism and theatre history, which have focused too narrowly on texts and productions involving fixed elements of the tradition—actors, paying audiences, costumes, and stages, among others. Taking a semiotic approach based on the work of de Marinis, Jean Alter, Erika Fischer-Lichte, and Keir Elam, I hope to show that narratives rich in dialogue and written for public declamation have a place in theatre history because they meet the essential criteria of “dramatic practice” and “dramatic activity,” terms Enders and Symes use to describe drama that is not self-identified as text for performance. Those criteria include a speaker and an audience; dialogue that requires impersonation; gestures and words that knit the speaker's world to that of the onlookers; the creation of social communication and exchanges of meaning; and a text that establishes a standard of repetition but allows for each realization to manifest unique qualities. Semiotic elements do not in themselves make drama, and, as Paul Bouissac has written, it is not useful to use a semiotic model to recast any and all cultural activity, the better to call it drama. Commenting on Fischer-Lichte's discussion of “signs of signs,” Bouissac asks, “Are not these signs subsumed by a narrative structure which commands and orients the cognitive experience of the spectators? Is it even possible to conceive the signifying material which unfolds on the stage as signs of signs, independently of the broader framework of the institution [of theatre], and of the dynamic structure of the plot?”Footnote ⁴ The answer, obviously, is no. Although theatre and drama depend on the interplay of signs, some form of narrative or plot—epitomized in agon or conflict—must be operative. Other features of the tradition often considered essential to drama, including stages, costumes, settings, and dedicated performance spaces, are either implied in the criteria I have enumerated or are dispensable—as indeed experimental productions have for many years demonstrated.Footnote ⁵

My objective is to get the drama of Anglo-Saxon England off the page and into the world of semiotic communication, the world of ostension. Although the examples used in this essay are drawn from Anglo-Saxon narrative poetry, my argument about dialogue and drama applies equally well to comparable orally performed poems in Middle English and in other medieval vernaculars. For the later periods, which have well-developed performance traditions and dramatic literatures, the extension of drama outside its traditional parameters will at least seem plausible, although dramatic readings of Chaucer's Canterbury Tales remain decidedly out of fashion.Footnote ⁶ It will, however, be considered idiosyncratic, if not positively quaint, to make claims for drama in Anglo-Saxon culture (ca. a.d. 700–1100). The major (and relatively recent) Anglo-Saxon literary histories and encyclopedias do not mention drama, and histories of and introductions to drama do not, it hardly needs to be said, mention the Anglo-Saxons, except to note their liturgical traditions.Footnote ⁷ Thus I survey previous assessments of drama in Old English literature before looking beyond the theoretical limitations of those assessments to the relationship of semiotics to narrative and oral performance. I then analyze two Old English poems, Beowulf and Juliana, to show why the performative world to which these narratives belong itself belongs to our idea of medieval drama.

• • •

In The Literature of the Anglo-Saxons, published in 1949, George K. Anderson devoted just six pages to “A Note on Old English Literature and Drama.”Footnote ⁸ In retrospect, his allowance seems generous, for subsequent literary histories have simply ignored the topic. No plays survive in the Old English vernacular, so Anderson inevitably reached negative conclusions about the drama of the Anglo-Saxons. But he did suggest that some Old English poems “imply drama” or “illustrate a dramatic atmosphere.” More to the point, he described poems as having a “semi-dramatic structure” if they contained dialogue. Among those texts he included were Christ, Genesis B, and two well-known elegies, “The Wanderer” and “The Seafarer.”Footnote ⁹ But the dialogues in those texts were, in Anderson's view, “expository rather than dramatic in purpose,” and none of them satisfied his requirements for “real drama” or “true dramatic dialogue” (207–9).

Like readers before and after him, Anderson believed that the strongest links between Anglo-Saxon culture and drama existed in the liturgy of Holy Week. The instructions for the Visitatio sepulchri in the Regularis concordia or “Monastic Agreement,” a customary drawn up in the 970s by Æthelwold, bishop of Winchester, include references to dialogue, costumes, and role-playing.Footnote ¹⁰ Hence this text has been described as “the beginnings of liturgical drama in England.”Footnote ¹¹ Even earlier signs of liturgical drama have been detected in the account of the Harrowing of Hell in the ninth-century Anglo-Saxon prayer collection known as the Book of Cerne, which includes rubrics apparently used to differentiate spoken and narrative sections.Footnote ¹²

Unfortunately, discussions of the liturgical drama of the Anglo-Saxons have only served to isolate the dramatic nature of oral performance from Anderson's “real drama” (209). We see the problem in the recent work of M. Bradford Bedingfield, who has significantly advanced the investigation into the connection between drama and liturgy by using the term “dramatic liturgy.” He has proposed that what he calls “liturgical reenactment” and “ritual expression” are closer to drama than many scholars have allowed. In his view, readers will fail to understand the dramatic nature of such reenactments, including the significance of their description in homilies, so long as they implicitly compare all forms of drama to the “representational drama of the later Middle Ages” (4). This is a point of indisputable importance; indeed, one can scarcely exaggerate the extent to which the representational world of the late medieval cycle plays constitutes a telos toward which all early medieval dramatic activity has been seen as evolving. Yet in reference to Anglo-Saxon drama more generally, Bedingfield's conclusions resemble Anderson's. Given the lack of evidence for secular drama, he writes, “One does not often find the words ‘Anglo-Saxon’ and ‘drama’ spoken together, and for good reason” (3).

I submit that the reason lies not in the performance traditions of the Anglo-Saxons but rather in the limited understanding of drama that informs Anderson's and Bedingfield's arguments—and a great many more discussions, to be sure. While making a case for the importance of dialogue, Anderson observed that “we can have the dramatic without the drama, for the drama, in the usual sense of the word, demands the interplay of two or more characters” (207). Bedingfield imposed a narrower and self-consciously theatrical criterion, claiming that drama “depends upon a performance in which a role is undertaken.” Since there is no evidence about how dialogues were presented (i.e., how those roles were undertaken as theatrical performance), he regarded inquiries into the dramatic nature of Old English poetry as mere “speculations” (3).Footnote ¹³ Bedingfield admitted that Old English narratives, including those incorporated into sermons, “contain dramatically used direct discourse,” but he doubted the usefulness of interpreting Old English dialogues or monologues as “dramatically inspired” (3). It should be clear that such terms as “dramatically inspired” and “dramatically used” fail because they rely on Anderson's “real drama” as the basis for comparison (209). It would seem that dramatic inspiration can come only from texts or enactments whose dramatic nature is unambiguously clear—that is, from plays. Even the term “liturgical drama” itself cannot escape this confusion, since it again positions liturgy as a prototype of “real drama” (Anderson, 209) which culminates in the great late medieval cycle plays.

Other definitions of drama combine performance and text without escaping the telos of representative drama. Lawrence Clopper has observed that “there are phenomena on the edges [of drama] that we might put into the category even if they do not totally conform to our sense of the genre.” One such phenomenon is the use of drama as “a poetic mode,” as a way of “presenting a narrative”: in narrative, the poet speaks; in drama the personae speak; and in a third form, those two voices are mixed. Clopper notes that Bede associated the Song of Songs with drama, identifying the text as an example of the mixed mode.Footnote ¹⁴ Although the Song of Songs lacks what Clopper calls the “visual signs of a dramatic text,” it can be rubricated “to conform to other liturgical manuscripts (responsories, for example)” and so can be said to contain “memorial cues to ancient dramatic texts such as Terence.” However, Clopper concludes that “textually or formally, the liturgy is a drama in the medieval sense; it is not a drama in our [sense] or the ancient one” (9).Footnote ¹⁵ As this comment shows, Clopper drew a clear boundary around drama; however mixed, the modes remain generically distinct. Hence he warned that when considering the Song of Songs and similar texts we should not “think of a script for enactment by persons assuming roles; rather, we should think of it as a formal and visual presentation of responding voices” (9). He adds, “When I use the term ‘drama,’ I mean an enacted script that contains, or, if it is a fragment, at one time contained, an entire narrative; that is, it is a text and a performance” (11; his emphasis). The limits implied by that definition become explicit in the range of texts that Clopper considers. Liturgy excepted, Clopper sets aside as undramatic the discursive modes in which drama is intermixed. He does not, therefore, consider the contribution of orally performed poems to the tradition of drama between the late antique and late medieval worlds.

Depending on its use of signs, however, an “enacted script” that contains “an entire narrative” (or a portion thereof) might well, even within Clopper's strictures, be considered drama. With or without manuscript evidence suggestive of performance, narrative poetry has long been regarded as performed text. Noting that the distinction between reading a play and seeing it performed was often blurred, Rosemary Woolf asserted that in the Middle Ages “all literature was read aloud and dialogue often given vocal impersonation.” She added that “the learned in the twelfth century understood the dramatic form, but did not think of it as exclusive to the stage and therefore inappropriate for reading aloud, either by one reader with a change of voice or by a number of different readers.”Footnote ¹⁶ Likewise, it is plausible to claim that, two centuries earlier, the learned also understood what Woolf calls “the dramatic form” as they experienced it in dialogue-rich narrative poems and that, wherever they heard and saw those performances, they experienced theatre.

Like Bedingfield and Anderson, Clopper rigorously separates medieval and modern ideas of performance. Reinforcing that separation is the lack of material evidence for Anglo-Saxon theatrical performance—that is, vernacular plays, stages or dedicated playing spaces, actors, and self-conscious references to a tradition of performance, all elements that constitute, in Clopper's words, “a drama in our [sense] or the ancient one” (9). One would not want to suppose that the Anglo-Saxons had a tradition of theatre, for of theatres themselves the Anglo-Saxons seem to have known little. The Old English translation of the World History of Orosius, written near the end of the ninth century, contains four references to theatrum, once giving the word an Anglicized spelling with a “þ“(þeater).Footnote ¹⁷ It is doubtful if that word and others used by the Anglo-Saxons, including ludus and palestrarum (both terms were cited by Anderson), point to an understanding of theatre as an architectural term. The Anglo-Saxons might have derived some knowledge of theatre from the works of Isidore of Seville. Donnalee Dox's study of the “idea of Latin theater” distinguishes the theatre as a building and as the process of putting on shows from drama, the latter being a textual phenomenon that comprises narrative structure and genre. Dox shows that those references remained textual rather than social: Augustine had a social understanding of theatre, she suggests, whereas Isidore did not.Footnote ¹⁸ Although the Anglo-Saxons seemingly had no knowledge of theatre as a social event, early medieval authors elsewhere did. The plays of Terence were familiar to the tenth-century nun Hrotswitha of Gandersheim, whose imitations of his works were designed to replace licentious models with chaste ones.Footnote ¹⁹

But the Anglo-Saxons did understand gesture, a key performative component of the late antique tradition. Anderson dismissed the importance of the classical Roman tradition for the Anglo-Saxons. C. R. Dodwell, however, later showed that the Anglo-Saxons understood and employed an important element of the Roman tradition, Quintilian's “language of the hands.”Footnote ²⁰ Dodwell showed that Anglo-Saxons in Canterbury in the eleventh century learned that gestural language through Carolingian manuscripts that reproduced the illustrative vocabulary of Roman playwrights. Those gestures were used to comment on Anglo-Saxon psalter texts. Dodwell compares the gestures to those used by Benedictine monks and recorded in an eleventh-century Anglo-Saxon manuscript (again available at Canterbury). They do not belong to a single tradition, of course, since psalter gestures illustrate emotions, moods, and states of mind, whereas the Benedictine gestures illustrate persons and things.Footnote ²¹ Yet it is clear that the Anglo-Saxons spoke the “language of the hands,” and it is not difficult to imagine that the Anglo-Saxons incorporated those or even other gestures into performances of narrative poetry.

The performer of such poetry in the Anglo-Saxon period, as has long been assumed, was the scop, sometimes but not always using a harp.Footnote ²² His role is described in Anglo-Saxon commentary on the rhetorical tradition of the late classical period. In the Old English Orosius, for example, a text translated from the Latin near the end of the ninth century, poeta is translated as scop. So too is Latin comicus, “a term,” according to Janet Bately, “used of both comic actors and comic poets in classical Latin.”Footnote ²³ In fact, one of those translations of poeta as scop refers to the Roman playwright Terence as “the great scop of the Carthaginians” (Orosius, 107/28). Four Anglo-Saxon glossaries support that important link between poeta and scop, which is in effect a link between the performance of poetry in the hall, a moment frequently described in Anglo-Saxon texts, and acting.Footnote ²⁴

The scop was indeed an actor, a fact implied in several recent discussions of medieval vernacular verse that recognize the performative nature of such recitations. Writing about Welsh poetry, for example, Sioned Davies discusses manuscript evidence that suggests such activity, including the use of rubrication and “dialogue introducers” in narrative poems. Those features “highlight character switching,” according to Davies. “This would surely make more demands on a story-teller/reader who would have to rely on his vocal performance to differentiate between characters. … A vocal performance is implicit in the text: the subject-matter of each tale is highly dramatic, calling for a wide range of voices.”Footnote ²⁵ Similar points have been made about the genre of sermons by Adrian Tudor, who describes “the multiple voices used by medieval preachers.” Tudor suggests that “each voice represents a new or separate role for the speaker, and the linguistic notion of a voice it highlights can easily be imagined to slide gently into the more dramatic notion of voices.” “Obviously,” Tudor concludes, “dialogue cannot be reported in a monotone.”Footnote ²⁶

In reference to Beowulf and the poem known as Widsith, Hugh Magennis has noted that the poems present themselves as the products of oral culture. The speakers adopted the stance of the minstrel and passed on orally “the things they ha[d] heard about.” Thus, according to Magennis, “these poems give the impression of being oral utterances of a minstrel, spontaneous performances, presumably delivered in the hall, which is implicitly evoked as the traditional setting for Germanic verse.”Footnote ²⁷ The impression of spontaneous performance is different from spontaneous performance, of course, and that is the heart of the matter. We may assume that the scop more often than not recited texts that had already been recited many times—and perhaps also texts that had been learned by reading. Ursula Schaefer, who has written extensively about oral performance based on a written text, has proposed that a vicarious voice—the singer's voice—is built into written composition and that that voice “had to be brought to life again by somebody who usually was not this singer.” Maintaining that “all (literary) fiction is ‘staged discourse,’” Schaefer uses the term “vocality,” derived from Paul Zumthor's La Lettre et la voix (1987), to imply “a whole concept of medieval poetic communication” in the idea of “performance.”Footnote ²⁸ For Zumthor, performance was “the temporal coincidence of communication and reception” and a point at which two axes of social communication intersected, one linking “the speaker to the author and [another] through which situation and tradition are united” (Schaefer, 117–18).

We do not usually think of Anglo-Saxon poems as “staged discourse” or as theatre of any kind, but Benjamin Bagby, who has performed Beowulf and Icelandic Eddas and who therefore offers a unique view of performance history, clearly does so. “When performing the Eddic stories or Beowulf,” he has written, “I enter with my voice into a world which is informed as much by the actor's art as by the singer's, and in that world I only rarely make use of the techniques suited to the needs of what we might call lyric song.”Footnote ²⁹ Bagby himself seems to think of the scop as both an actor and a singer. Whether we call them gleoman, scop, leoþwyrhta, or woþbora (the four categories of poet considered by Jeff Opland), when the poets of Anglo-Saxon England recited narrative poetry, they performed it (Opland, 253). Anglo-Saxon narrative poems give the impression of “oral utterances” even though they are not spontaneous but are, rather, “staged discourse”—that is, repeated performances enabled by a “voice” built into written (or orally remembered and recalled) compositions.

A semiotic conception of performance invites us to examine not only the scop and his or her text but also attend to creative activity at the other end of the communication axis, that of the audience. In her important book on public reading, Joyce Coleman borrows Richard Bauman's concept of the “emergent quality of performance,” which he describes as “the interplay between communicative resources, individual competence, and the goals of the participants, within the context of particular situations.”Footnote ³⁰ Coleman adds that, “In a bardic or minstrel performance, everything from the choice of genre and of text, along with many decisions about the emphases within and the length of the text, would be subject to a feedback process.” Thus the “text, and the event, would be different with every performance,” depending on the “performer's skills as composer, actor, singer, and/or editor (of the text as he or she performs it)” (29). That “feedback process” is the heart of semiosis, dramatic information (words, gestures, etc.) circulating between audience and performer in a communicative process that transforms both parties. Alter calls this iconicity. When he asks, “What semiotic competence is really necessary in theatre?,” his answer is that it is “a competence in the use of an iconic code whereby all signs on the stage refer to their mirror image in the imaginary story space outside the stage” (Sociosemiotic Theory, 97; his emphasis). Seen through this “looking-glass distortion,” according to Alter, “the referential story always lacks some precision, and individual spectators must draw on their own experience or imagination to round up its concretization in their minds” (98). His idea evokes the familiar reception model that addresses gaps in narratives that readers must fill by drawing on their own imaginative resources.Footnote ³¹ The text enacted before the audience—whether poem or homily, liturgy or play—is always incomplete: spectators must step in and cocreate it. Theatre, as de Marinis has written, is unlike film or television in that it requires “the real physical co-presence of sender and addressee (the latter is collective, as a rule)” and also the “simultaneity of production and communication.” The public performance of the text, however standard it is supposed to be (however familiar to the audience) is never the same from one recitation to the next; it is repeated as a text, but its dramatic meaning is always subject to conditions of performance, whether they depend on the performer or on the audience, that themselves can never be entirely duplicated. Rather, “every theatrical performance (every single theatrical occurrence) constitutes an unrepeatable, unique event, an ephemeral production that is different each time in spite of all attempts at standardization.” He concludes that “theater, in short, always involves event, as well as code and structure” (Semiotics of Performance, 50–1).

To see that the Anglo-Saxon hall was understood as a scene of performance animated by give and take—although perhaps not as vivid as Coleman's invocation of “a bunch of drunken louts shouting down the performer” (29)—we need look no further than Beowulf. In Heorot, the assembled warriors gather to hear “the noise of the harp, the clear song of the scop” (“hearpan sweg, / swutol sang scopes”).Footnote ³² Later, as mead is served in the celebration following the monster Grendel's death, the scop sings “clear-voiced in Heorot” (“scop hwilum sang, / hador on Heorote,” 496–7). At the celebration that follows the death of Grendel's mother, the scop in the poem tells the story of Finnsburg, a fragment in Beowulf of an independent narrative:

(1063–8)

When the grim account of death and revenge is finished, it is called “the singer's song” (“leoð wæs asungen,/gleomannes gyd,” 1159–60).Footnote ³³

Although “the story of the scop” constitutes a performance, I do not regard it, as described in this passage, as drama, for it contains no dialogue and requires no impersonation by the performer of the poem; the narrative is specifically the scop's story because it is delivered in his voice only, without dialogue requiring that he imitate the voices of others. To see why that point matters, let us consider the relationship between dialogue and narrative from a performative perspective.

In narrative poetry, dialogue holds the key to drama because dialogue is by nature deictic. Elam has defined deixis as direct verbal exchange indexed by pronouns in written or oral form (and hence found in many poems, liturgies, songs, legal exchanges, etc.). Such exchanges create Elam's “dramatic world,” a “here and now” in which both a “you and I” are engaged, a speaker and a receiver. Because it is deictic, dialogue performs an “active” and “dialogic” function rather than one that is “descriptive and choric.”Footnote ³⁴ For Elam, “dialogue is immediate ‘spoken action’ rather than reference to, or representation of, action, so that the central personal, political and moral oppositions which structure the drama are seen and heard to be acted out in the communicational exchange and not described at a narrative remove” (147; his emphasis). Elam stresses the role of dialogue in enabling the audience to see and hear the “central personal, political and moral oppositions” that structure the drama, a point that is underscored by Richard Ohmann's view that “illocutionary acts move the play along.”Footnote ³⁵ Nothing, so far as the text goes, is “acted out” in the scop's account of the Finnsburg episode because the text does not move forward through the intensifying force of dialogue. The scop in the passage above does not act out or impersonate; instead, he stands at a “narrative remove” and merely reports what the poet has already reported. The scop does not use dialogue to intensify the audience's experience or involve them in the give-and-take of meaning that characterizes a semioticized space.

Another episode in Beowulf, however, showcases “the oral narrator,” who is in this case the young hero himself. Having returned to Geatland, Beowulf recounts his adventures in Denmark to Hygelac, the Geatish king. Even before he describes his own victories over Grendel and his mother, Beowulf refers to the forthcoming marriage of Hrothgar's daughter, Freawaru, to the Heathobard warrior Ingeld. Beowulf predicts that deadly violence will erupt when, at the wedding feast, the Heathobards see armor captured from their families displayed as booty in the Danish court. He then delivers the speech of an “old spear-warrior” of the Heathobards who sees a sword, an “ancestral heirloom,” worn by the bridegroom as he enters the hall. The old man calls the weapon to the attention of a young warrior to whose father the sword once belonged, and asks if the young man recognizes it:

(2041–56)

We must imagine that Beowulf, as he voices the incendiary words of the old warrior, imitates the speaker's voice. The question—“Can you recognize …?”—invites interpretive gestures in its delivery. We hear just one side of the dialogue, but one side is all we need to see how the semioticized space created by dialogue intensifies the audience's experience and involves them in a cocreation of meaning. The scop impersonating Beowulf would, I imagine, move his eyes along the old warrior's line of sight to the weapon that the bridegroom wears and from it to the young Heathobard sitting nearby. The audience, I also imagine, would follow that progression and would also see weapons hanging on the walls of the hall in which this narrative were being performed—or would see something else that might be juxtaposed with the weapons of the scop's speech. No young man would be sitting next to the scop, but such a figure could readily be imagined responding to the old warrior's words, turning to the weapon and, later, taking up a weapon of his own to avenge his father's death. The scop gives the barest outline of such dramatic activity, but it is all that is needed. By “acting out” the poem's words in dialogue, the scop creates a communicational exchange. In other words, he creates drama, and the space within which he is seen and heard is, perforce, a theatre. Whatever he wears serves as his costume; whatever is behind and around him constitutes scenery. These and all other essential theatrical properties are materially present. If they are absent, there can be no public performance, and if they are present, then so too is drama.

This episode forms a brief segment—just 16 of more than 150 lines—of the speech Beowulf delivers to Hygelac's court. Even though most of the passage cannot be considered dramatic by the standard that I have applied to the old warrior's remark, this brief, incisive moment shows how semiotics creates drama within a nondramatic text and generates a mise-en-scène. Joaquín Martínez Pizarro's concept of the “rhetoric of the scene” is useful in understanding the difference between an “acted out” exchange and narrative, between a story that is shown and a story that is told. Pizarro discusses two representative conventions. The narrator stands squarely between his audience and his tale in one, serving as their guide to it. In the other the narrator is effaced; the text represents the action as if it were being seen rather than reported, and description is suppressed in favor of dialogue. That second tradition, in Pizarro's view, produces a visualizing effect, since the reader has the illusion of “witnessing” the events the narrator describes. Gesture and posture become “dramatic elements” that “contribute to the illusion of some kind of visual correlative” of the narrative.Footnote ³⁶ The less that is said about the situation by the narrator, the more must be communicated by dialogue and by gestures and objects. Pizarro argues that “the oral narrator” tries “to become transparent, to vanish from the scene or from his listeners' awareness; by appealing primarily to their dramatic imagination, he invites them to follow an action that does not include him as a judge, critic, or interpreter” (55–6).

When Beowulf conveys the words of the old warrior, the young hero becomes “transparent” and “vanish[es] from the scene [and] from his listeners' awareness,” disappearing into the words of the old man. It is important to remember that my claims for the presence of drama in this scene are framed by the scene's own narrative frame. What we are talking about is, in Pizarro's language, “the illusion of some kind of visual correlative” to the narrative (13). There is, of course, no warrior before Beowulf's audience except Beowulf himself; there are no fictional warriors before the scop's audience except the warriors conjured by his performance. The scop is not operating in a self-consciously theatrical context, since, so far as we know, there was no such context to which he could allude. His account of Beowulf's account cannot, therefore, be compared to Hamlet's use of “The Mousetrap” (act 3, scene 2), but the contrast is instructive. Hamlet explains his play to the stage audience because he must (albeit with malicious irony: “This play is the image of a murder done in Vienna,” he says, and adds, “it touches us not”).Footnote ³⁷ Hamlet reverses the process I am delineating here because he narrativizes the drama he has produced, stepping in to explicate it, materializing before the audience much as the narrator of the Old English poem would seem to dematerialize. The scop can dramatize Beowulf's narrative precisely because he does not have to explain it. He can disappear into Beowulf's words, and Beowulf in turn can disappear into the old Heathobard's words, because the audiences involved, Beowulf's audience and the scop's audience, can see for themselves how events in the narrative resonate with and comment on the world in which the narrative is being told.

There is no evidence in the Beowulf manuscript at this point of rubrication or the “dialogue introducers” or “identifying markers for the speakers” mentioned by Davies and Clopper.Footnote ³⁸ In the manuscript, a new fitt (or chapter), marked by a large O, begins shortly before the passage I have quoted, which stands thirty-nine lines into Beowulf's speech to the court.Footnote ³⁹ But manuscript evidence pointing to direct speech is obscure even in later texts, as in the twelfth-century French materials discussed by Symes, where manuscript clues to performance have often been missed or misunderstood (790–810). John Hines has described the early Middle English verse fabliau “Dame Siriz” as a “performable text, arguably even marked up for performance.” Hines draws attention to the preponderance of dialogue over narrative in that work and in others included in London, British Library, Harley 2253, and suggests that they might have been “literally enacted within the homestead.”Footnote ⁴⁰

The Beowulf manuscript is not marked up for performance, but a poem in another Anglo-Saxon codex, the Exeter Book, might very well be. This is the manuscript containing Juliana, a poem attributed to Cynewulf, one of a few named Anglo-Saxon poets, that survives only here. Katherine O'Brien O'Keeffe notes that, although visual divisions within the poem do not indicate “conceptual divisions” as they do elsewhere in this manuscript, “most heavy punctuation or combinations of points and capitals mark the opening and closing of formal speeches by Juliana and her tormentors,” and that points are “rarely used within a speech.” When used, moreover, such points “set off direct address from the rest of the text.”Footnote ⁴¹ The pointing might have been used to identify formal speeches in Juliana, boundaries of the dialogues observed in performance; O'Brien O'Keeffe also notes that the “text-specificity of punctuation is a broad sign that pointing is subjective and marks a discrete reading of a text” (162), evidence that might also support the idea of a discrete performance of the text.

Cynewulf's Juliana exploits dialogue to far greater effect than most Old English verse. The poem contains an unusually high proportion of dialogue: 290 of 721 lines, or nearly 40 percent of the text, take dialogue form. Moreover, the dialogue is not simply an exchange between a saint and a demon but a debate that closely follows the form of a confession of sins, a verbal pattern that we can assume most educated Anglo-Saxon Christians experienced directly.Footnote ⁴² The dialogue creates a theatrical setting that, I contend, evokes the Anglo-Saxon Christian's theatre of everyday life. Cynewulf achieves this surprising effect by inverting and thus ostending the power relations of the scene. When the captive saint tortures the demon who has been sent to torture her, the demon impersonates a penitent who has been forced to confess.

Since not all readers will be familiar with the saint and her adventures as recounted, a summary is helpful in contextualizing the dialogue. Juliana takes place in Nicomedia (now Izmit, near Istanbul), a city that Diocletian intended to transform into an imperial center to rival Rome. His massive plans for reconstructing the city coincided with an infamous persecution of the Christian population in the early fourth century. In the poem, Eleusius, the city's prefect, declares his intention to marry Juliana. She insists that he must first become a Christian (25–57). Eleusius enlists the aid of Juliana's father, Affricanus (67–88), who fails to persuade Juliana to change her mind. He beats her and sends her to Eleusius (92–157), but he too finds the young girl unmovable. Eleusius orders her stripped and whipped before a crowd, tortures her, and then throws her into prison (140–233). Once in prison, she is visited by a devil who claims to be God's messenger. Juliana subdues and interrogates him (242–530). Her dominance subsides in the course of that interrogation, but Juliana regains the upper hand in their debate, rhetorically at least, when she is taken from prison, brought to trial (539–60), tortured once again, and put to death (605–70). The poem deploys spectacle in ways familiar to us from other early medieval hagiographies: gruesome torture is inflicted on a pious maid but ultimately destroys those who perpetrate it.

Scenes like this have been analyzed often in Anglo-Saxon criticism to establish relationships between a saint's sufferings and the Passion of Christ.Footnote ⁴³ However, the poem's central section (242–530) uses dialogue to shift our focus from the allegorical to the mundane. The devil first appears floating in the air above Juliana—“lyftlacende,” airborne—and speaks “bright from the clouds” (280, 285). But his advantage is swiftly (and, thanks to a textual lacuna, obscurely) lost; for most of the dialogue he is lying at her feet, bound in her grip. This inversion, in which the prisoner imprisons her intended assailant, frames the scene with irregularity. To quote Umberto Eco, the dialogue is “put upon the platform” or ostended by the irregularity of its context. Footnote ⁴⁴ Eco's concept of irregularity takes shape as image, object, and gesture in what Elam calls “signal information” (36–7). “Signal information” grows in inverse relation to the appropriateness of the sign in context; where the sign is most inappropriate, its signal information is greatest. It is not expected that the saint torture instead of suffer torture; we expect her, not the devil, to be bound. The dialogue underscores the paradox, since it pits a saint, who is seemingly rock solid in her righteousness and unvarying in her position, against a demon whose verbal dexterity sometimes seems to draw the saint into his frame of reference and out of her own. The devil's performance establishes a theatre of penance within the poem. His impersonation of a penitent connects the dramatic world (which Elam designates as W_D) to the observing world (which Elam designates as W_O).Footnote ⁴⁵ Demonic protestations of pain and requests for mercy exploit a semiotic relationship with the audience that, however awkward on the theological level, becomes stronger on the dramatic level than the one created between the audience and the saint whom the poem honors.

The dialogue moves through several phases, each of them concerned with disclosing the unknown. Juliana's repeated demands for disclosure meet with the demon's attempts to fend them off or subvert them. The demon behaves like a reluctant penitent. In response to her demand that he explain who he is, for example, the devil reveals some of the crimes he has instigated, including the deaths of Christ, John the Baptist, and the disciples. Then, in response to Juliana's demand, he divulges a fuller confession (352–417), owns up to his role as the “originator of sins,” and describes his actions as “the wounds of sin” (synna wundum, 354). The striking language would have been familiar to Cynewulf's audience—but not from the mouth of a demon. He even describes himself as a “sin-committing one [who] must suffer shame.”

Several writers have noted the parodic nature of the devil's confession; John P. Hermann has compared the poem to the rhetoric of the Anglo-Saxon penitentials.Footnote ⁴⁶ The demon's familiarity with that idiom is a ruse, of course, but also an instance of semiotic irregularity, since the demon deliberately ironizes a commonplace of the observing world (Elam's W_O) within the dramatic world (Elam's W_D). Although the poem is about the trials (formal and otherwise) of Juliana, it is the devil's trial that is ostended in the dialogue. In this scene his body and his voice are the poet's—and the scop's—strongest “visual correlative” (Pizarro, 13) and most potent signs. It is not the suffering saint who is the center of attention; rather, the poet directs the scop's and the audience's energies to the devil, whose impersonation of the sinful penitent has put the spiritual struggles of daily Christianity at the center of the drama.

The devil even parodies confessional prayer:

(443–54)

As hypocrisy goes, this is difficult to match. The devil has already boasted that he caused Christ's death (304–6); here he seeks to profit by the grace it brought into the world. He entreats Juliana as a sinner might entreat a confessor, using a mode that the audience would surely have recognized from other texts, including the pious Old English poem known as “A Prayer”Footnote ⁴⁷ and the heroic narrative Judith. Footnote ⁴⁸ The devil's request “that you be merciful to me, a needy one” (þæt þu miltsige me þearfendum [449]) would also have reminded the poet's audience of confession. Anglo-Saxon penitentials contemporaneous with the manuscript of Juliana contained dialogues written with the intention of performance. Although not literally “designed for stage production” (to refer to Elam's requirement for drama), these dialogues were unquestionably intended for performance in private confession.Footnote ⁴⁹ However, the power relations in the penitentials were opposite to those shown in Juliana since, in the penitentials, the priest offered mercy and the penitents did not request it.Footnote ⁵⁰ The priest asked the penitent if, as one who sought forgiveness, he would forgive others: “Will you forgive each of those who have ever offended against you?” (Wylt þu forgyfan ælcon þæra þe wið þe æfre agylton?). If the penitent agreed, the priest proceeded to ask for mercy for the sinner and for himself: “If he says, ‘I will,’ say to him then, ‘May God be merciful to you and help me that I may be permitted (to do so)’” (Gif he cwyð “ic wylle,” cweð him þænne to: “God ælmihtig gemiltsige þin, and me geunne þæt ic mote”).Footnote ⁵¹ In the penitential rite it was the priest who directed the penitential. At this point in their dialogue the devil has seized the initiative from Juliana and begun to undermine her authority by professing the truths for which she has been imprisoned. His objective is to turn her from true belief, and he thinks he can best accomplish that aim by impersonating a true believer. It is remarkable to see how closely Cynewulf allows the demon to parrot confessional piety, as no doubt the expedient penitent often did. The rhetorical and dramatic success a good interpreter might have brought to the demon's role is worth contemplating.

Juliana forces yet another confession from the demon, his most extensive (460–530). He has already confessed wrongs of historical significance, including the deaths of Christ and the apostles (290–306). But that confession, which grandiosely claims responsibility for the Fall of Adam and Eve (499–505), is different, for it also includes many sins that would have been recognized as the daily disasters of Anglo-Saxon life. Such references strengthen his connection to the daily lives of those in the scop's audience. They include bloody deaths (475–7), deaths at sea (478), deaths by hanging “on the high gallows” (on hean galgan [483]), and the deaths of those caught in drunken revenge in winsele, in the hall of joy (483–8). All of these acts and circumstances were familiar to the Anglo-Saxons, and not just from their knowledge of Beowulf. The demon's catalog cleverly mixes sufferings inflicted on Christians with sufferings they inflict on others. We may assume that Juliana, like Beowulf, was performed in a hall, in the presence of an audience of Christians gathered for entertainment. That audience would find that dialogue in the religious poem, like dialogue in the heroic one, correlated to their experience.

It is clear that the demon talks like a penitent, impersonating a penitent's prayers and confessing to evils such as those some Anglo-Saxons confessed. The awkward similarity between penitents (i.e., the audience) and the demon is reinforced by his ostended posture. Bound at his confessor's feet, his voice apparently trembling with emotion, the demon mimics the posture and behavior required of Anglo-Saxon penitents. They were not bound, of course, but as the penitential quoted above shows, their behavior in confession was scripted. They were told to weep and prostrate themselves before their confessors, according to the instructions offered by the introduction that accompanies two eleventh-century vernacular penitentials. Sincere contrition required the sinner to “prostrate himself before [the priest] with utmost fear of God and humility, and in a weeping voice beg that he prescribe for him penance for all those faults that he has committed against God's will.”Footnote ⁵² The demon does not simply borrow the rhetoric of the penitentials, then, but, through the skill of the scop re-creating his words, parodies the scene of confession itself. The demon's performance (operating through the performance of the scop) offered Anglo-Saxon Christians a glimpse of the theatricality of their spiritual lives: using language and words prescribed by the penitentials, he demonstrates both insincere confession and feigned sorrow for sins.

The dialogue between Juliana and the devil would have taxed the resources of the most capable scop as it moved from speaker to speaker and as the devil moved from one attitude to another. His shiftiness obscures the central ideological and psychological difference that the poem is expected to maintain: the separation between victim and victimizer. The devil's position as the bound one together with his canny use of confessional speech make it difficult to focus on the saint as a victim of persecution (which she is) and on him as tormentor (and persecutors' agent). He repeatedly contrasts her authority as a judge to his lack of freedom and power, doing his best to undermine her position with his words as he adapts his voice—his different voices—to his needs. The devil is an actor, and the scop performing his role needed to make the devil's theatricality clear. The devil's parody of the rhetoric of confession implicates the sincerity of all Christian penitents, whose emotional involvement in confession served as an index to their moral candor. There is a clash between what we have been told about the demon (by the poet and by the demon himself), and what we see and hear, which is a prisoner squirming at the feet of his interrogator and struggling to find some way out of his dilemma. He momentarily becomes the protagonist. To respond to the scene properly, the audience must resist the demon's words and the spectacle of his suffering by recalling instead the earlier frame of reference created by the saint's righteousness. The saint's words are temporarily superseded by the power of the devil. What actor, given the choice, would take the saint's role over the demon's?

Earlier and later in the poem, Cynewulf declaims as a historical narrator, making the poem's ideological lines clear. But when he relies on dialogue, the poet creates a scene wherein predictable truths become subject to the dynamics and contingencies of performance, inviting, at least momentarily, other interpretations. The possibilities of such interpretations would have greatly increased in a performance of Juliana read with different voices—not read with the magnificent bardic Anglo-Saxon declamation one imagined in the olden days, but rather with as much (or more) interpretive latitude as we would grant to Beowulf as he addresses his king and court. In performance, form follows function. If we think only about the clarity of the poem's ideological positions—and that, I would say, has been the chief concern of readers ranging from Rosemary Woolf to John Hermann—we miss its point, which is that those positions are not, after all, so clear. It is to Cynewulf's credit that his villain manifests so much power. The devil will finally be publicly humiliated in the realm of spectacle, where his fate is a foregone conclusion. But behind closed doors, he is a formidable foe. The less he is described, the more he must speak, and the more he speaks, the murkier the poem's moral lines become.

In his discussion of the overlap between “drama” and the categories of lyric, narrative, and exposition, Anderson argues that Anglo-Saxon poems exhibiting “the potentialities of dialogue” tended to be lyrical, and that poems in which “fancy and emotional expression can have freest play” sometimes manifested the “semi-dramatic structure” suggested by dialogue.Footnote ⁵³ Anderson might have been thinking of the “dramatic voices” attributed to “The Seafarer” and “The Wanderer” by John C. Pope and others who assert that these are poems in which the parts of a divided self stand in dialogue.Footnote ⁵⁴ Although it is not a lyrical poem, Juliana makes great demands on the scop's powers of “emotional expression.” It is a poem, like Beowulf, rich in drama because it is rich in the semiotic properties of impersonation. Both poems, we must assume, were performed in a setting that reinforced the similarities between the dramatic world and the world of the observers.

The heart of my proposal for expanding the scope of theatre history to include the public performance of Old English verse rests on the communicative nature of that event, specifically as indexed by the texts' use of dialogue. How close such texts are to contemporary ideas of performance we can see in the words of Joachim Fiebach:

Fiebach's claims seem manifestly relevant to the performance of Anglo-Saxon narratives that are rich in dialogue—or would seem relevant, were it not for the fact that readers are predisposed to reject the idea of drama among the Anglo-Saxons for the simple and wholly unsatisfactory reason that they wrote, so far as we know, no plays.

One of the most encouraging developments in theatre research is the isolation of the trappings of nineteenth-century theatre to their own period; it was, after all, because medieval plays lacked such trappings that they were not considered plays for so long. According to Hardin Craig, whose views now seem impossible to credit, “the religious drama had no dramatic technique or dramatic purpose and no artistic self-consciousness. Its life-blood was religion, and its success depended on awakening and releasing a pent-up body of religious knowledge and religious feeling.”Footnote ⁵⁶ Medieval drama has long been recognized as being as rich in “dramatic technique” and “artistic self-consciousness” as modern drama. Rather than focus on such features, however, whether in self-identified dramatic texts or in the cultural world around them, I have followed scholars and critics who study performance. In the light of their work, we can hardly avoid the conclusion that the Anglo-Saxon scop's recitation of dialogue-rich verse qualifies as theatre and that the texts that guided his performance belong to the corpus of medieval drama.

Anglo-Saxonists need the category of drama, I believe, not because Anglo-Saxon literature is somehow incomplete without it but because our ideas of the Anglo-Saxons' experiential worlds are, at best, extremely limited. An understanding of public poetry as dramatic text that informed theatrical performance is, in the first instance, a contribution to our knowledge of the performative experiences of the early Middle Ages. Drama is thought to have disappeared between the late antique period and the late Middle Ages. I hope to have shown that we have failed to see the continuity of the dramatic tradition because we have allowed representational theatre to stand for all dramatic activity. The continuity emerges when we apply the theoretical principles and the practical criticism of performance to early medieval narratives. It remains to be seen how other dialogue-rich Old English texts, including penitentials, demonstrate the semiotic environments of the Anglo-Saxons and illuminate their forms of drama—which, although we have overlooked them, have been in full view all the time.