DRAFT: DANSE MACABRE AND THE VIRTUAL CHURCHYARD
July 2, 2011
Please note: The published version of this essay will contain illustrations. They do not appear here, as I do not have permission to reproduce them in a public online forum.
 Danse macabre presented itself to its late-medieval audiences as a multimedia phenomenon. In addition to appearing in a variety of static media, including stained glass, cloth, wall painting, manuscript illustration, woodcuts, and possibly tapestry, it may also have existed as a performance tradition. In addition, particular site-specific installations of danse macabre combined within themselves different representational media, including murals, architecture, poetic inscription, and kinetic bodily participation. But in characterizing danse macabre through this efflorescence of media, we must be mindful of Martin K. Foys’ point that the term media is more complicated than it might appear to be: it seems to have become a singular concept when at its foundation it is really always plural, per its inflection (2012 [forthcoming], n.p.). This insight is crucial in understanding danse macabre. In its specific depiction of death as well as its incorporation of different media, danse macabre has to negotiate between singularity and plurality. In this essay, I will argue that danse macabre can resolve its paradoxes of singularity and plurality when we understand it not only to occupy actual space but also to generate virtual space. Virtuality plays a role in danse macabre at the level of the death figure and its representation of danced motion; at the level of the architectural spaces in which danse macabre appears; and finally, at the level of danse macabre poetry, here examined specifically in John Lydgate’s ca. 1426 Dance of Death. At this last level lies the main implication of my argument: when we see danse macabre as virtual, we also recognize medieval poetic form as defined by the interaction of forces in motion on a virtual plane.
 A few well-known examples will help to establish the features of this tradition. The dance of death has possible origins ranging from folkloric graveyard capering to exemplary cautions about the dangers of terpsichorean indulgence. One of its most famous and influential artistic manifestations during the late Middle Ages was the early fifteenth-century mural at the Cemetery of the Holy Innocents in Paris. The paintings presumably featured gesticulating skeletons, some perhaps holding musical instruments, leading a representative of each social estate to death in a choreographed procession. Contributing to the arresting nature of this mural was its placement underneath the cemetery’s charniers, whose open skylights left their human bones visible to the public (Harding, 2002, 101-2). A similar mural was painted on wooden boards and hung around the cloister of the Pardon Churchyard at St. Paul’s in London around 1430 (Barron and Rousseau, 2004, 36). In addition, Guyot Marchant published a set of danse macabre woodcuts in 1485, which may or may not have borne a direct relationship to the Paris murals. Another source of visual information is the Danse macabre des femmes, found in five manuscripts and two printed editions. Because the Danse des femmes could not follow the order of estates that comparable danses including men were able to, the social categories of the women included differ somewhat (Harrison, 1994, 10). But the images depict a range of movement qualities, in the skeletons and their victims, that is somewhat typical of danse macabre as the tradition exists. A more static and less gesticulative choreography for the danse exists in the frescoes at Kermaria, Brittany, probably postdating the Paris mural (Clark, 1950, 30-33). An English mural was also painted (and subsequently whitewashed over by Shakespeare’s father) at the Guild Chapel, Stratford-upon-Avon. Finally, and most evocatively, a 1459 altarpiece panel by Simon Marmion of the Life of St. Bertin depicts a cloister whose walls contain a dance of death mural. This painting of a painting might have been inspired by the dance of death at Amiens, of which Marmion was a native (Clark, 1950, 85). In several danses macabres, poetic inscriptions accompany the visual images.
 At the most basic level, this aesthetic tradition introduces conflicts between singularity and plurality. An early critical debate about danse macabre concerns the nature of the death dancer; namely, are we seeing a single personification of Death, caught at different moments, or several manifestations of individual dead people calling forth the living? Is danse macabre a dance of Death or of the dead? Ernst Manasse challenges earlier work by William Fehse, who saw the skeletons in danse macabre art as “a number of the dead” who compel their living counterparts to join the dance. Manasse suggests that it is Death himself who “assumes the likeness of his victims” from the thirteenth century onward, ultimately creating a “complete identity of the figures of Death and the deceased human being” (1946, 83-84). Questioning the connection between folkloric traditions of the dancing dead and the danse macabre, Manasse suggests that while we cannot rule out the possibility that the dancing corpses represent a plurality of individuals, neither can we rule out the presence of a Death figure as master piper compelling his followers (1946, 96-98). Thus one of the foundational discussions of danse macabre, located in the early German critical tradition examining this phenomenon, concerns the complexity of death’s singularity and plurality. The subtle shadings of the skeletons’ identities – Death, the dead, the dead as emissaries of Death – suggest that the viewer’s perception of the artwork suspended itself between the idea of the one and of the many. In other ways as well, the interaction between singular and plural informs the representation of death’s interaction with the living. Elina Gertsman has recently linked danse macabre with devotio moderna, arguing that the danse’s structure privileges the individual encounter with death and possibly “the self-cultivation of an inner life”; at the same time, however, the absence of a God with whom to engage in the danse macabre creates anxiety about death as a more universal concept (2010, 45). In different ways, then, the danse macabre is constructed of interactions – sometimes conflicts – between structures of plurality and of singularity.
 Jacques Derrida reminds us that this paradox characterizes our relationship to death itself, and in exploring this topic he reveals the instrumentality of formal devices in articulating the paradox. In Apories, he writes:
La mort de chacun, de tous ceux qui peuvent dire « ma mort » est irremplaçable. « Ma vie » aussi. Tout autre est tout autre. D’où une première complication exemplaire de l’exemplarité. Rien n’est plus substitutable et rien ne l’est moins que le syntagme « ma mort ». Il s’agit toujours d’un hapax, d’un hapax legomenon, mais de ce qui ne se dit qu’une fois chaque fois, indéfiniment une seule fois. C’est aussi vrai pour tout ce qui engage la forme grammaticale d’une première personne. D’autre part, les guillemets n’affectent pas seulement cet étrange possessif (l’unicité de l’hapax « ma »), ils signalent l’indétermination du mot « mort » dont au fond on ne connaît peut-être ni le sens ni le referent. (1996, 49)
Ashby Kinch reads this section of Apories as relevant to danse macabre because Derrida approximates the medieval viewer’s problem confronting the corpse in art: “The imaginative corpse of late-medieval art …ask[s] the viewer to identify with a body that is not, and cannot be, his own.” The corpse image “simultaneously evokes our physical discomfort with our individual identity and our participation in the broader human community of the body” (2002, 164). There seems, however, to be another way of reading this passage alongside the paradox of singularity and plurality in the danse macabre. Derrida not only names the paradox here, but also performs a series of engagements with formal elements – at different levels of his writing – as part of the process of communicating the paradox. At the level of his own prose, we see a structure that relies on sequence and concatenation with variation within: Tout [every] autre est tout [completely] autre; hapax, hapax legomenon; une fois chaque fois. These elements suggest that it is not simply the fact of death but rather the form of its reiteration, its status as a link in a chain of similar yet different versions of itself, which produces the paradox Derrida describes. Analogously, the sequential form of the skeleton images in danse macabre introduces the conflict between death as a universal figure and deaths individually and multiply articulated. In addition, at the level of explicit reference to form, Derrida’s passage calls attention to grammatical form and the work of the quotation mark in signaling as indeterminate the multiplicity or singularity of death. Derrida’s account thus foregrounds the constitutive role of form in the paradox of death.
 If form plays a central role in the construction of death’s paradox, we need to consider further what elements lend danse macabre its formal qualities; the most readily available answer to this question is its identity as a dance. On the surface, danse macabre’s engagement with danced practice seems as obscure as the etymology of the word macabre itself. Performance texts, however, as well as contemporary accounts pertaining to churchyard murals, provide some clues about the dance of death’s potential relationship to danced or processional performances. James Clark speculates that the Paris danse macabre shared with other early versions at Lübeck and in Spain a common origin in the drama, a “chorea Machabaeorum” performed during the fifteenth century. While specific references to these performances do not date further back than the middle of the century, Clark (1950, 91-92) and Gertsman (2010, 80, 95) point out that earlier fifteenth-century dramatic texts such as the Pride of Life, Everyman, and parts of the N-Town cycle depict death in ways that draw upon the danse macabre’s conventions. Religious drama and spectacle thus provide us with one way to place these representations of dance within a dynamic performance context. Clark (1950, 93) also mentions an earlier dance of death “masque” performed as part of a wedding in Jedburgh in 1285, citing a description in John of Fordun’s Scotichronicon of a performer “of whom it might almost be doubted whether he was a man or a phantom.” Gertsman (2010, 98) deems this performance a “definitive precursor” to the dance of death. In addition, speculation exists surrounding the possible use of a danced, or simply very animated, representation of death during sermons. One possibility lies in incorporating “dramatic impersonation” into a sermon, with a death figure gliding amongst the congregation; the inspiration for such a possibility lies in a 1429 sermon preached at the Cemetery of the Innocents, which took advantage of the backdrop of the open charnel houses to instill fear of the Judgment (Pearsall, 1987, 61). As Gertsman argues, danse macabre imagery indicates “complex relationships between visual and performative arts in the Middle Ages, and their reception by a spectator” (2010, 98). We therefore need to take account of dance, movement, and performed spectacle as contributors to the form of the dance of death.
 The painted dance of death’s relationship to these embodied performance traditions suggests that theories of the dancing body might contribute productively to the study of danse macabre; such theorizations of the moving body in dance, I will suggest, resolve danse macabre’s paradox of singularity and plurality. The aesthetic philosopher Susanne K. Langer describes dance as an “apparition,” arguing that dance inheres not in the muscles and movements of dancers but rather in the “interacting forces” that these create in the spectator’s perception. Dancers are actually doing something when they dance, but what we see is a “virtual entity” (1968, 78). More recently, José Gil has elaborated upon the idea of the “virtual body” specifically in the dances of Merce Cunningham. Cunningham’s desire to empty movement of conventional emotional or organic content has meant that in his work “the body of the dancer…is composed of a multiplicity of virtual bodies” (2002, 123). For Gil, Langer’s “plane of virtual forces” is a “plane of immanence,” in which each “present gesture” in a dance is “incorporated into a more profound, virtual community” (2002, 125). Two insights emerge from this theorization of the dancing body. First, through the introduction of the virtual, dance accommodates a paradox similar to the one that death sustains. Dance exists both apart from all the particular dancers’ individual gestural manifestations, and at the same time entirely through these individual gestures. Second, as a representation, danse macabre reflects this very quality of dance. The disappearing flesh of the skeletons renders in particularly vivid form the idea of dance as apparition, but more importantly, the painted skeleton parade comments on dance’s ontology. As an attempt to represent movement, the skeletons are not simply a kind of serialized cartoon, but rather a form that produces the virtual in the interactive spaces between its visible representations. On this virtual plane that the mural creates, a dancer exists who is at once ubiquitous as a singular embodiment, and at the same time a collection of individual manifestations of gesture and kinetic experience. As a representation of dance’s virtuality, danse macabre articulates a form that resolves singularity and plurality with each other.
 It will no doubt be noticed that virtuality emerges as a central conceptual term in the discussion above, and this term applies not only to the representation of dance itself, but also to the architectural spaces in which danse macabre appears. In what I shall call the “virtual churchyard” as a site for danse macabre, media can both assert their particular qualities and intersect with each other to form kinetic spaces of interaction. Before making this case, it seems necessary to define the terms virtual, churchyard, and media as I invoke them here. I use virtual in the sense that Langer and Gil use it: something that is not actual, but, as Langer says, is not unreal for its lack of actuality and exists through perception (Langer, 1968, 78). I do not make claims here about virtuality as related to computer or imaging technologies. As Rebecca Leuchak has pointed out, the use of computer simulation to reconstruct or conjure the medieval is problematic in its risk of generating a kind of Platonic ideal of a disappeared space (1997, 366). Rather than a virtual reality, I have in mind something closer to what Mary Anne Moser refers to as a “virtual environment” (1996, xviii). As I will hope to show below in my reading of John Lydgate’s danse macabre verses, my sense of the virtual churchyard is something accessible to the medieval spectator encountering the actual churchyard. I focus on churchyards as sites because two important occurrences of dance macabre art and poetry existed in famous, though subsequently destroyed, urban churchyards during the fifteenth century: the cemetery of the Holy Innocents in Paris and the Pardon Churchyard at St. Paul’s in London. Finally, the term media designates the paint, stonework, written inscription, and embodied movement involved in danse macabre installations at the churchyard sites; it also, however, refers to the relationships between one individual medium and another. In discussing the theory of remediation, Bolter and Grusin define medium as that which is constantly containing and being contained by the contexts of other media (1999, 65). Thinking of the churchyard as virtual, I will suggest, elucidates the negotiating, interacting nature of media.
 To some extent, the virtual churchyard opposes the impulse to reconstruct any particular churchyard physically or imaginatively. The Lord Protector Somerset demolished the Pardon Churchyard in 1549, and the Holy Innocents wall painting in Paris was sacrificed to a seventeenth-century initiative to widen the streets. Such familiar narratives of early modern violence toward the medieval make it difficult to conceive of an approach to the destroyed artifact which is utterly untinged by an organizing desire for reconstruction. Before its inclusion in his 2002 Reform and Cultural Revolution, James Simpson’s argument about Lydgate’s Dance of Death appeared in his article “Bulldozing the Middle Ages: The Case of ‘John Lydgate’” (2001). This is a thought-provoking title partly because it does not indicate a positivist or literal desire to recapture what has been lost. Rather, the demolition metaphor subtly intimates how deeply ingrained is our tendency as medievalists to engage our objects of study in a thematics of destruction and reconstitution. Simpson notes that, “With one exception, I do not, of course, suggest that scholars have wished to destroy Lydgate’s corpus, even though our knowledge of much of this work is conditioned by the manner of its survival from destruction” (2001, 216). As he says, the themes of destruction, rescue, and reconstruction inevitably shape the medieval and our encounters with it: “the act of destroying the ‘past’ creates it” (2001, 241). But I propose here to examine the churchyard through a different lens, to understand it not in terms of understanding a lost object through the terms of its destruction or through the vestiges edging those destructive practices. Rather, I consider the churchyard as a virtual object in order to foreground N. Katherine Hayles’ definition of virtuality as “the perception that material structures are interpenetrated with informational patterns” (1996, 4-5). To be sure, Simpson’s work, as well as that of other literary scholars and art historians of danse macabre, insightfully draws out such informational patterns in its readings of dance of death art and poetry. Here I suggest simply that we begin with this thought about informational pattern and interaction, rather than physical or thematic reconstruction. In doing so, we might capitalize fully on virtuality’s interacting forces (in Langer’s terms) to understand what the churchyard was and what it did.
 In the virtual churchyard, for example, human bodily movement – rather than simply bodies themselves – is a medium functioning as part of the installation’s complete multimedia spectacle. Speaking about early dance in particular, Jody Enders has argued that audience roles were not only spectatorial but also participatory and even reciprocating (2006, 143-148). While the viewers of danse macabre art were neither viewing nor necessarily participating in an actual dance, this general sense of kinetic participatory aesthetic might have invisibly informed human interaction with danse macabre art. Recent work has emphasized the role of human presence in the experience of danse macabre art. Gertsman, for instance, focuses on human viewership by exploring throughout her study “the way the danse macabre paintings structure the experience of the viewers, and are, in turn, structured by that experience” (2010, 14). More specifically, she sees this human component in interactive motion. Viewers moved along the Reval painting, an activity that involved experiencing the opposing directionalities of the poetic text and the painted procession; at Chaise-Dieu they also registered a sense of interruption by architecture in the serial contemplation of the images (Gertsman, 2010, 123-124; 131-134). Other, more explicitly choreographed motion was associated with danse macabre art as well. Amy Appleford has demonstrated that the Pardon Churchyard in London was incorporated into civic processions of “political value,” so that in this way as well, the churchyard becomes a space to accommodate and participate in meaningful human mobility (2008, 303).
 The kinetic interactive forces between spectator and spectacle contribute to the churchyard a virtual whole. Rather than structuring itself exclusively as a spectacle to be viewed, the danse macabre also incorporates the signifying capacity of the moving body itself into the entirety of the spectacle. Bodies of course exist as particular and separate, but at the same time, in the manner described by Gil, they contribute to a virtual plane that includes every aspect of the spectacle. As Appleford argues, the London dance of death both alludes to and critiques the association between dance and an ordered and harmonious human society. It negotiates between a civic vision of unity and one of discreteness (2008, 300-301). I would suggest as well that this political dynamic of simultaneous singularity and plurality is possible in part because the space in which it occurs sustains an analogous dynamic for the different kinds of media – kinetic and representational – that constitute it.
 At the same time that the spectators and the installation generate virtual forces of interaction, the particular visual elements of the physical churchyard itself – painted and architectural – also interact with each other. Like the Marmion dance of death, the Holy Innocents mural was painted along the ten arcades of the south wall, the Charnier des Lingères (Harding, 2002, 102). John Stow refers to both the Pardon churchyard and the Holy Innocents cemetery as “cloisters,” again associating them with the cloister arcades depicted in the Marmion painting (Warren, 1971, xxii). Architectural elements thus integrated themselves variously into paintings of the dance, as the fresco from Kermaria and the Marchant woodcuts suggest. In addition to the arcades that these pieces depict, stonework like the Three Living and Three Dead portal, created at the Holy Innocents before the painted panels, also contributed to the imagery and impact of the danse macabre installation. While the relationship between these artistic traditions in uncertain, Paul Binski points out that they share a similar form of anxiety about death, “a moment of instability between two temporal realms”; he sees in both a mise-en-abyme between the living and the dead (1996, 54, 138). Thus we find even in the actual churchyard a persistent integration of, and thematic dialogue between, the painted component and the architectural features of the building containing the murals.
 But if we consider the churchyard as virtual, additional dimensions open up: the interaction between paint and stonework is set in motion, like the moving bodies. As a representation of moving figures, the dance of death tradition suggests the possibility of experiencing all its elements in virtual motion. Certain effects of the juxtaposition between painting and architecture reinforce this possibility. As the Marmion painting suggests, the danse macabre sustains tension between the verticality of the cloister arcade construction and the lateral momentum of the danse macabre mural. Inherent within the stonework is the possibility of monumentality, emphasized by the chapel at the center of the churchyard, which was dedicated to St. Thomas à Becket and his mother St. Anne (Appleford, 2008, 302). Coexisting with this vertical monumentality, however, is the mural, running in a lateral and horizontal direction. As mentioned earlier, it possibly moves in two directions at once in aligning the procession with the verses (Gertsman, 2010, 123-124). The awareness of these visual forces interacting with each other sets the churchyard in motion on the virtual plane. The realm of dance theory can once again help us to explicate this motion through Mark Franko’s term “reverse architecture.” He suggests that dance creates “a space that intends to become.” For Franko, dance’s monumentality lies in its “reverse architecture, taking down what was not there” (2008, 251). Danced movement traces the possibility of a space in which to occur even as it causes this space and trace to vanish. As a representation of dance – and perhaps even a dance at the level of the virtual – the danse macabre mural generates a reverse architecture in the traces of gestures that contribute to and undo the vertical space framing them. The mural and the stone function with and against each other, creating a spectacle that depends on their individual visual programs as well as a virtual space generated by their interaction. Simpson argues that a momentum of “undo[ing]” characterizes the social meaning of Lydgate’s dance of death poem (2002, 55). Setting the poem in the context of a fuller multimediacy shows us how such undoing occurs at the level of the installation’s aesthetics, inhering in the danse macabre’s virtual movement.
 Finally, the virtual churchyard casts into relief for us not the traditional understanding of text/image relations – relations of mutual explication or even opposition – but rather the experience of interactive force generated in the movement between them. The painted scenes of dance or procession were frequently embellished with poetic verses; these are visible, for instance, in the Marmion painting. At Chaise-Dieu, lines appear under the painting, suggesting that that verses would be inserted, but, Kürtz speculates, the painter ran out of time to inscribe the actual words (1975, 81). In this case and in the Marmion altarpiece, where the verses are denoted by lines only resembling writing (but not actually legible), the ingredient of poetic inscription takes on a visual identity separate from legible language itself. As Juliet Fleming has shown, early instances of ornamental writing – or writing incorporated into an architectural site – suggest the indistinguishability of written matter and other materialities (2001, 10). In Marmion’s mediated visual event of a painting of a painting, we see vividly how Fleming’s idea operates. The regular blocks of text running below the mural participate not only in the poetic, but also in the visual, architectural, and kinetic programs of the installation.
 A contemporary account of the Paris cemetery provides another angle of approach to this interaction of media, an interaction that again differs from the conventional dichotomy of text and image. In an earlier fifteenth-century description of the city of Paris, Guillebert de Metz records that the Holy Innocents contains “paintures notables de la dance macabre et autres, avec escriptures pour esmouvoir les gens à devotion.” Preceding this statement is a short account of paintings at the Collège des Célestins (“paradis et enfer en painture”), and following it is a description of a small spire in the cemetery (“une tournelle en lieu d’un tombel, où il a une image de Nostre-Dame entaillée de pierre, moult bien faicte”) (De Lincy, 1855, 63-64). This account’s rhetorical simulation of a city stroll moves the viewer through these components and moves them around each other. The passing reference to the inscriptions is thus contained within the passage’s larger focus on visual art and architecture. And yet it appears as though the writing is most directly attributed with the power to move people to devotion. The reference thus seems both incidental and significant. In addition, the interjected “et autres” syntactically interferes with any potential purity of correlation between the dance of death paintings and the verses. In these ways, de Metz’s description reflects a perceptual lack of specificity about the role of the inscriptions as distinct from the visual and architectural danse macabre elements. Through its textual component as well, then, danse macabre as cultural tradition or didactic tool exists both within each particular manifestation and in the larger network of vectors generated in all the remediated interactions of text, promenading bodies, archways, and paintings.
 Despite the suggestive indeterminacy of inscription’s role in the installation, however, danse macabre does exist in the form of legible poetry, making it important to consider the specific features of this verse in light of this essay’s larger claims. The last part of my essay will argue that the formal identity of John Lydgate’s ca. 1426 dance of death text reveals itself in virtual space. Lydgate traveled to Paris, viewed the danse macabre at the Holy Innocents, and undertook the project of translating it into English. Some versions of the poem contain a verba translatoris section in which Lydgate admonishes: “Considereth this / ye folkes that ben wyse/ And hit emprenteth / in yowre memorialle / Like the exawmple / whiche that at Parise / I fownde depicte / ones on a walle….” Just as poetic text accompanied the paintings in Paris, so the verses of Lydgate’s poem accompanied the Daunce of Poulys on the cloister walls of the Pardon Churchyard. And in the cases of both Lydgate’s poem and the French text he translated, versions were inscribed on the churchyard murals and copied in books.
 The form of the dance of death poem appears somewhat aesthetically limited when viewed from a modern perspective. The poem’s repetitive structure can feel alienatingly heavy and tedious (some might even say dull) and seems to cause the poem to restrict its horizons to commonplace conventions like estates satire, to which the danse macabre poem seems related. The poem marches through verse dialogues alternating between the figure of death and each of his victims, who generally answer death with resigned statements about their fate. Although its form seems to impart to the poem a pedestrian quality, however, the Dance of Death is hardly an unimportant work in the Lydgate canon. It draws its meaning largely from the context of Lydgate’s interest in material culture, performance, and audience, central topics in Lydgate studies. In this way it is characteristic of Lydgate’s work, which as John Ganim argues is performative and integrated into its surrounding material culture (2008, 166, 178). And yet, despite these attendant themes, the structure of the poem makes it tough going – like many of Lydgate’s other works – for modern readers.
 And like many other poems in preserved in manuscript, the Dance of Death presents additional challenges to the reader in the form of complicated sets of variants. The most readily available edition of the poem, Florence Warren’s for the Early English Text Society, is problematic in its way of handling the different manuscript groups. It manipulates the order in which the stanzas appear in one major manuscript group (B) in order to create the semblance of a parallel text with another manuscript group (A) (Warren, 1971, xxix). This editorial decision results in a highly mediated experience of the text. Subsequent to Warren’s edition, M. C. Seymour placed the extant manuscripts into four sub-groups, mentioning additional details about the variants (1983-85, 22-24). Appleford has helpfully generated a chart comparing the characters in the French version to those in the Ellesmere (A) and Lansdowne (B) manuscripts of Lydgate’s version, but as she points out, the manuscripts in their entirety “show significant variation” and require updated study and comparison (2008, 296, 312 n. 29). Furthermore, the question concerning which version of the poem accompanied the murals arises but is difficult to answer. Appleford suggests that because the B group contains manuscripts that designate the poem as the Daunce of Poulys, it is likely that this version accompanied the paintings (2008, 295). At the same time, as she and Warren note, Trinity College Cambridge MS R. 3. 21, not part of this group, contains a statement about the words being painted in the cloister at the request of John Carpenter (Warren, 1971, xxvi). Pearsall speculates that the presence of two manuscript groups indicates revisions made in response to Carpenter’s 1430 request to inscribe the walls with the poem; however, he acknowledges that others besides Lydgate might have contributed to the poem and that “[i]t would be difficult, in such circumstances, to talk about stages in a process of revision”(Pearsall, 1987, 62-63; see also Hammond, 1921, 250-251). Neither the characters who explicitly mention dance (like the Lady of Great Estate) nor the initiating words of the translator appear across all groups; it is difficult to say whether the inclusion of such features would make a version more or less likely to be the one appearing with the wall painting. For all these reasons of manuscript variance, the poem is challenging to work with, and much scholarship concerning the Dance of Death (including the present essay) deals with an incomplete picture of it.
 This complicated manuscript status might, however, represent not only a difficulty but also an indication of the importance of approaching the poetic danse macabre tradition by way of the same themes of virtuality and remediation that we applied to the visual tradition. Doing so, I suggest, reveals the possibility of a medieval conception of poetic form not always evident to us as modern readers, a form born of interactions among versions rather than the tracing of particular versions. Two potential objections to this approach immediately arise. First, one might say that I am using what is here a nonspecific device of “interaction” to fan away lazily all the particular manuscript problems like so many gnats. I would respond that I have used the readings of the danse macabre installations and of the virtual churchyard to create a conceptual foundation for suggesting that the manuscript tradition’s complexity has implications for the poem’s form. Second, one might say that I am simply replicating Paul Zumthor’s familiar model – also built upon a dichotomy of singularity and plurality – of manuscript mouvance. For Zumthor, the identity of the literary work is “dynamic” and “exists outside and hierarchically above its textual manifestations” (Zumthor, 1992, 45-48). To this second objection, I would reply that I am looking specifically at form and structure, rather than textual identity in a larger sense. I have argued elsewhere that dance can provide forms for visualizing and conceptualizing the movements that take place in mouvance (Chaganti, 2008, 77-78). In the case of Lydgate’s Dance of Death, I propose that a new form emerges when the poem is considered in terms of virtuality and the shadows of danced movement within it. Despite the assertive structure of social hierarchy that the poem seems to espouse, it also offers an alternative to this in characters who appear and disappear between versions. This phenomenon initiates other possible trajectories to define the poem’s form. The forces of interaction in the virtual spaces between versions constitute the form of the poem as powerfully as its explicit device of serialization does. It might be useful to understand this poem’s participation in the larger danse macabre tradition not by choosing one manuscript or manuscript group as intended for the churchyard, but rather by considering the vectors in play between stanzas and versions, creating a form for the poem that is as three-dimensional as the churchyard.
 I shall close by reading a passage in the poem to illustrate some of these points. Manuscripts in the A group, including Ellesmere and the Trinity College manuscript, contain a dialogue between death and a tregetour:
Maister Jon Rikelle / some tyme tregetowre
Of nobille harry / kynge of Ingelonde
And of Fraunce / the myghti Conquerowre
For alle the sleightes / and turnyng of thyn honde
Thow most come nere / this daunce to vndurstonde
Now3t mai a-vaile / alle thi conclusiouns
For dethe shortli / nowther on see ne londe
Is not deceyued / be noon illusiouns. (Ellesmere l. 513-20)
This stanza provides a medieval language for the concept of the virtual. The tregetour testifies to a secular medieval capacity to perceive and experience that which was not actual. And this tregetour in particular, Lydgate points out, creates the force of his virtual images in the “turnyng” of his hand (l. 516); as with the clapping of the Franklin’s tregetour, insubstantial but perceptible images are created and dissolved in the spaces of movement. This section’s language further elaborates on the idea that its substance resides in the interactions between points rather than with any one medium or materiality. In the following stanza, the tregetour uses the word “Legerdemeyn” (l. 526). This French response to the English “turnyng of thyn honde” replicates between the two stanzas the dynamic space created in the translation project as a whole, a constitutive interaction between linguistic media.
 Present in some versions and absent from others, the tregetour illustrates finally how form emerges in the virtual space of interaction. In the A version, the tregetour follows the Juror and Minstrel, whereas in the B version these characters are followed by the “Famulus,” a “seruant or officer” (Lansdowne l. 449). The two characters together generate a complicated informational pattern between them. The famulus is a type, with no differentiated identity; however, the concept of minutely differentiated social and economic hierarchy inheres within him. The language of these stanzas develops this idea with references to “pore and riche,” (l. 431); “offices / & profites” (l. 460); and “service” (l. 464). The tregetour, meanwhile, embodies great social specificity in having been named and attached to the king (making him unusual among the characters). And yet, where the famulus speaks the particularizing language of rank, the tregetour’s discourse lacks these terms, seeming to suggest instead a more generalized or unified social vision – the terms of distinction for him are the socially neutral “nowther on see ne londe” (l. 519). Juxtaposing these two characters throws into relief the juxtaposition of two different formal rationales. The tregetour subtly represents an alternative to the poem’s formal principle of hierarchy, a principle recognized in the famulus. But the language of the poem also reveals that we are dealing with something more choreographically complex than the lining up of one character who is conventionally hierarchical with one who is less so. Rather, the tregetour and famulus generate a kind of virtual crisscross between them in their shifting attachments to social differentiation and generality. These patterns give the space between them motion and force. Seeing the poem’s form as an integrated virtual whole allows it to accommodate these different possibilities and thereby to multiply its directionalities and dimensions.
 Scholars of danse macabre have sometimes attempted to imagine what kind of thought it inspired. Some (Manasse, 1946, 98; Bowers, 1948, 118) have seen the danse as expressing compulsion, or an impulse to repetition. Pearsall counters that dances of death signal not obsessive behavior but rather the need for admonition, a move away from the more explicitly psychological reading (1987, 67). In some ways, this disagreement only confirms what we already suspect – that the content of any particular medieval viewer’s thought is impossible to know. What the danse macabre might indicate more reliably, however, are medieval forms for seeing and knowing. Virtuality reveals the forms through which medieval audiences thought about cultural phenomena like the danse macabre. Aware of the interacting forces among bodies, texts, and other media, medieval audiences could perceive virtual spaces that contained social meaning and the fullness of poetic form.
 For the most recent and comprehensive account of danse macabre art throughout Europe, see Gertsman, 2010. Oosterwijk and Knöll, 2011, was not yet available at this draft’s submission, but see Oosterwijk, 2004, 2008. Kürtz (1975) enumerates different manifestations of this tradition, including a nineteenth-century account of a tapestry in the Tower of London (145). Gertsman (2010) offers an analysis of the didactic goals associated with the dance of death tradition, discussing a “poetics of mortality” that shapes the understanding of death (15, 180) and detailing the relationship between danse macabreand preaching (45-49, 164). Appleford’s (2008) civic and political reading of the London dance of death sees it “as an aesthetic image of political community that emphasizes diversity and temporality, an imagination much more relevant to the city’s political structure than the common metaphorization of the social body as an incorporated, organic whole” (291).
 Both have been put forward as theories explaining the origin of the danse macabre as a convention. Eisler has pointed out that danse macabre skeletons regularly carry the equipment of grave diggers, and also that they sometimes appear to be not skeletons, but figures wearing skeleton costumes, suggesting their depiction of a folk ritual involving grave diggers pantomiming and possibly dancing (Eisler, 1948, 187; Meyer-Baer, 1970, 310). On the idea that the danse macabre was used as a tool of figuration, an image preachers could invoke to instill fear of bodily excess as represented by dancing, see Freeman, 2005, 18, 29. Binski remarks that “A thirteenth-century council at Rouen forbade dancing in churchyards” (1996, 56).
 On the specific detail of skeletons carrying musical instruments, see Meyer-Baer, 1970, 300.
 Gertsman, 2010, 4-5, sees the woodcuts as preserving the mural, but Appleford, 2008, 287, cautions that this relationship is conjectural.
 See Davidson, 1988, and Puddephat, 1960, esp. 30 on the whitewashing and 33 on the figures and colors in the frescoes:
The sixty participants in the dance were arranged in pairs, each consisting of a sprightly cadaver with a reluctant victim in its grasp. There were two pairs of characters in each compartment. The figures were lighter in tone than the rich scarlet background against which they were represented in attitudes which suggested that they were moving to the left along a floor paved with vermilion and black tiles, which formed a chequered pattern of squares in the upper tier but of lozenge shapes in the lower.
Davidson, however, critiques some of Puddephat’s records (1988, 19-20).
 As Pächt argues, “It is the very essence of continuous narrative to render changes visible by means of comparing the same person in different moments or states” (1962, 8 n. 1).
 Many scholars have commented on this etymology; see, for example, Huet, 1917-1918, 148-58; Eisler, 1948, 187-225; DuBruck, 1958, 536-43; Sperber, 1958, 391-402; Warren, 1971, xvi-xviii; Pearsall, 1987, 62; and Taylor, 1991, 4.
 Stone from the St. Paul’s cloister was used to build the original Somerset House (Clark, 1950, 11). On the Holy Innocents’ destruction, see Gertsman, 2010, 6.
 Simspon sees Lydgate’s dance of death poem as staging a constantly mobile renegotiation of authority among the various sites it circumscribes: “God, Death, the Angel, the poem’s patrons, its translator, and its auctor” (2001, 238).
 I use the term “architecture” loosely here to encompass everything from the columns and archways visible in the Marchant woodcuts (which may or may not be part of the scene of the dance, given their juxtaposition with the flowered background), to the Reval and Lübeck dances of death containing buildings and cityscapes in their backgrounds (Gertsman, 2007, 43-52). Unlike the iconic modern depiction of the dance of death at the end of Ingmar Bergman’s The Seventh Seal (1957), which takes place on an empty hilltop with a background of clouds, many medieval dances of death use the imagery of buildings and architecture to organize their visual information.
 Warren, 1971, 2 , l. 17-20; subsequent line numbers will appear in the text.
Appleford, A. 2008. The Dance of Death in London: John Carpenter, John Lydgate, and the Daunce of Poulys. Journal of Medieval and Early Modern Studies 38.2: 285-314.
Barron, C. M. and M.-H. Rousseau. 2004. Cathedral, City, and State, 1300-1540. In St. Paul’s: The Cathedral Church of London, 604-2004, 33-44, ed. D Keene, A. Burns, A. Saint. New Haven: Yale University Press.
Binski, P. 1996. Medieval Death: Ritual and Representation. Ithaca: Cornell University Press.
Bolter, J. D. and R. A. Grusin. 1999. Remediation: Understanding New Media. Cambridge, MA: MIT Press.
Bowers, R. H. 1948. Iconography in Lydgate’s ‘Dance of Death.’ Southern Folklore Quarterly 12.2: 111-128.
Chaganti, S. 2008. Choreographing Mouvance: The Case of the English Carol. Philological Quarterly 87.1/2: 77-103.
Clark, J. M. 1950. The Dance of Death in the Middle Ages and the Renaissance. Glasgow: Jackson.
Davidson, C. 1988. The Guild Chapel Wall Paintings at Stratford-upon-Avon. New York: AMS Press.
De Lincy, L. 1855. Description de la ville de Paris au XVe siècle par Guillebert de Metz. Paris: Auguste Aubry.
Derrida, J. 1996. Apories. Paris: Galilée.
DuBruck, E. 1958. Another Look at ‘Macabre.’ Romania 79: 536-43.
Eisler, R. 1948. Danse macabre. Traditio 6: 187-225.
Enders, Jody. 2006. Death by Dance. Mediaevalia 27.1: 135-53.
Fleming, Juliet. 2001. Graffiti and the Writing Arts of Early Modern England. Philadelphia: University of Pennsylvania Press.
Foys, M. K. 2012 (forthcoming). In Media Res: Media Matters in Anglo-Saxon England. In A Handbook to Anglo-Saxon Studies, n.p., ed. J. A. Stodnick and R. R. Trilling. Malden, MA: Blackwell.
Franko, M. 2008. Mimique (originally published 1995; repr. here with Franko’s new introduction Mimique Revisited). In Migrations of Gesture, 241-258, ed. C. Noland and S. A. Ness. Minneapolis: University of Minnesota Press.
Freeman, M. 2005. The Dance of the Living: Beyond the Macabre in Fifteenth-Century France. In Sur quel pied danser?, 11-30, ed. E. Nye. Amsterdam: Rodopi.
Ganim, J. M. 2008. Lydgate, Location, and the Poetics of Exemption. In Lydgate Matters: Poetry and Material Culture in the Fifteenth Century, 165-83, ed. L. H. Cooper and A. Denny-Brown. New York: Palgrave Macmillan.
Gertsman, E. 2010. The Dance of Death in the Middle Ages: Image, Text, Performance. Turnhout: Brepols.
—–. 2007. Death and the Miniaturized City: Nostalgia, Authority, Idyll. Essays in Medieval Studies 24: 43-52.
Gil, José. 2002. The Dancer’s Body. In A Shock to Thought: Expression after Deleuze and Guattari, 117-127, ed. B. Massumi. London: Routledge.
Hammond, E. P. 1921. The Texts of Lydgate’s Danse Macabre. Modern Language Notes 36: 250-251.
Harding, V. 2002. The Dead and the Living in Paris and London, 1500-1670. Cambridge: Cambridge University Press.
Harrison, A. and S. Hindeman. The Danse Macabre of Women: MS Fr. 995 of the Bibliothèque Nationale. Kent, OH: The Kent State University Press.
Hayles, N. K. 1996. Embodied Virtuality: Or How to Put Bodies Back into the Picture. In Immersed in Technology: Art and Virtual Environments, 1-28, ed. M. A. Moser. Cambridge, MA: MIT Press.
Huet, G. 1917-1918. La Danse macabré. Le Moyen-âge 29: 148-167.
Kinch, A. 2002. The Danse Macabre and the Medieval Community of Death. Mediaevalia 23.1: 159-202.
Kürtz, L. 1975. The Dance of Death and the Macabre Spirit in European Literature. Geneva: Slatkine.
Langer, S. K. 1968. The Dynamic Image: Some Philosophical Reflections on Dance. In Aesthetics and the Arts, 76-82, ed. L. A. Jacobus. New York: McGraw Hill.
Leuchak, R. 1997. Imagining and Imaging the Medieval: The Cloisters, Virtual Reality, and Paradigm Shifts. Historical Reflections 23.3: 349-69.
Manasse, E. M. 1946. The Dance Motive of the Latin Dance of Death. Medievalia et Humanistica 4: 83-103.
Meyer-Baer, K. 1970. The Music of the Spheres and the Dance of Death: Studies in Musical Iconology. Princeton: Princeton University Press.
Moser, M. A. 1996. Introduction. In Immersed in Technology: Art and Virtual Environments, xvii-xxv, ed. M. A. Moser. Cambridge, MA: MIT Press.
Oosterwijk, S. 2008. ‘For no man mai fro dethes stroke fle.’ Death and Danse Macabre Iconography in Memorial Art. Church Monuments 23: 62-87, 166-68.
—–. 2004. Of Corpses, Constables, and Kings: The Danse Macabre in Late-Medieval and Renaissance Culture. Journal of the British Archaeological Association 157: 61-90.
—– and S. A. Knoll, eds. 2011. Mixed Metaphors: The Danse Macabre in Medieval and Early Modern Europe. Newcastle upon Tyne: Cambridge Scholars.
Pächt, O. 1962. The Rise of Pictorial Narrative in Twelfth-Century England. Oxford: Oxford University Press.
Pearsall, D. 1987. Signs of Life in Lydgate’s Danse Macabre. In Zeit, Tod und Ewigkeit in der Renaissance Literatur, 58-61, ed. J. Hogg. Salzburg: Institut für Anglistik und Amerikanstik.
Puddephat, W. 1960. The Mural Paintings of the Dance of Death in the Guild Chapel of Stratford-upon-Avon. Birmingham Archaeological Society Transactions 76: 29-35.
Seymour, M. C. 1983-85. Some Lydgate Manuscripts: Lives of SS. Edmund and Fremund and Danse Macabre. Edinburgh Bibliographical Society Transactions 5.4: 10-24.
Simpson, J. 2001. Bulldozing the Middle Ages: The Case of ‘John Lydgate.’ New Medieval Literatures 4: 213-242.
—–. 2002. The Oxford English Literary History Volume 2, 1350-1547: Reform and Cultural Revolution. New York: Oxford University Press.
Sperber, H. 1958. The Etymology of Macabre. In Studia philologica et litteraria in honorem L. Spitzer, 391-402, ed. A. G. Hatcher and K. L. Selig. Bern: Francke.
Taylor, J. H. M. 1991. Que signifiait danse au quinzième siècle? Danse la Danse macabré. Fifteenth Century Studies 18: 259-77.
Warren, F., ed. 1971 (1931). The Dance of Death, Edited from MSS Ellesmere 26/A.13 and B.M. Lansdowne 699, Collated with the Other Extant MSS. EETS o.s. 181. New York: Kraus Reprint.
Zumthor, P. 1992. Toward a Medieval Poetics. Trans. Philip Bennett. Minneapolis: University of Minnesota Press.