CHAPTER 14

Building Pleasure and the Digital Archive

Dorothy Kim

This essay is an exercise and also a critical meditation on what it means to build and create the Archive of Early Middle English and what the project’s theoretical stakes are in relation to the corpus, the digital platform, the interface. Thus, this is a narrative about the long history of the book and about what it means to translate reading a medieval manuscript from thirteenth-century Britain to reading a mediated version of a manuscript in the twenty-first century. It is the story of medieval to digital remediation, but a remediation that has remarkable feedback loops because it disrupts print as a medium. This essay is also about the archive story of Early Middle English and how making visible the editorial bodies that create the digital manuscript bodies is a form of radical material feminism that reframes the stakes of this digital archive. And so, this essay pivots between the granularity of the codex book and the larger ecosystem of an archive of books.

Drawing from an array of theoretical perspectives—material feminism and especially theories of “intra-action,” postcolonial archive theory, and disability studies—this essay explores the issue of pleasure in critical interface design and multisensory experience in digital reading ecologies, with a focus on the Archive of Early Middle English project. My discussion will turn attention to both the interpretive processes of visualization and the value of developing digital resources that engage senses beyond the visual. I will consider additionally the physical and related lived experiences of building and using archives. What are the alternatives to visual emphases in interface design, and what are the stakes—for nascent and long-standing projects—of creating flexible, stable resources that invite manipulation and change? And what are the stakes for archive theory, digital labor, feminist materialism, and histories of the book? Can what at first glance be seen as traditional and canonical function as a decolonized, feminist, and material ecosystem?

Visual Pleasure, Graphesis, and Histories and Futures of the Book

When the AEME (Archive of Early Middle English)—a digital archive that will eventually include 162 encoded manuscript witnesses to the documentary production of Early Middle English between 1100 and 1350—received a National Endowment for the Humanities (NEH) 2013–2017 Scholarly Editions and Translation grant, the first and completed task Scott Kleinman and I, as co–principal investigators, discussed was the creation of a splash page—both as a placeholder during the project’s coding and back-end development and as the go-to space for anyone who wanted to be directed to the archive.[1] Even though we also created a separate site for encoding guidelines and had an open workspace on GitHub, the splash page became the calling card for the AEME project in development.[2] Even with such a diverse and disparate corpus of archival materials, the visual aesthetics of the splash page was one of the earliest longer discussion items. We, in fact, budgeted and hired a graphic designer, Amy Papaelias (SUNY New Paltz, Art History) to give the “skin” of this interface of our archival portal motion, color balance, functionality, and visual pleasure. As the first task we accomplished with the release of our NEH funds, the foundational importance of organizing our visual calling card and subsequent visual profile speaks to the stakes, importance, and driving frames of visual pleasure in creating and building a digital archive and database. This article is an attempt to unpack the politics and theoretical angles and make transparent our biases in first prioritizing visual pleasure. But it also is a discussion that sketches out what else could be done in the continued development of the digital archive as an area in the long history of the book.

Our archive is not filled with many impressive illuminated and decorated manuscripts. Rather, it was incredibly difficult to even find an image we could use in our logo from the 162 manuscripts in our corpus. But this initial design—decided entirely by considering the visual pleasure of our project team and potential imagined users—became the project touchstone as the scholarly and public profile of the AEME moved across: through not just the splash page but also the temporary encoding guidelines page—though with slightly different but complementary color schemes—to the visual design, program, and swag for the “Making Early Middle English” conference.[3] This latter project for the AEME became the first international conference in the field, funded with a Social Sciences and Humanities Research Council Contact grant. It has also now transformed into the first peer-reviewed scholarly journal in the field, The Journal of Early Middle English, which also utilizes the visual footprint first developed for the AEME project.[4]

The question of pleasure speaks to the importance of desire in archival building and about emotional affect and the dangers of intense affect in hiding archival agendas and the political, national, and social forces often at play in creating them. There is a wish to hide these difficult narratives under the aura of sensory pleasures. And especially with digital projects—whether small, big, or going in multiple directions—pleasure is central in organizing and producing the digital humanities. Visual pleasure, the pleasurable experience of the user/subject, and the ease of the project’s interface are always central in building.[5] There are a number of reasons for visual aesthetics’ importance. They range from the politics of funding and also our own understanding of what will attract, hold, and intrigue digital users/readers/subjects. In a way, whether scholarly or lay, digital projects often function within a rhetoric of seduction. They use sensory pleasure, but almost always visual pleasure, to offer their projects up for consumption, enjoyment, work, and play. Theoretically and practically, the AEME has considered these issues from its inception and particularly from its first real budget expenditure. We have also grappled with the disconnect between the messy, dirty, fragmented quality of our corpus in juxtaposition of our aesthetically delightful and visually attractive splash page and project design aesthetic.

However, there has not been enough discussion about the theories behind what we are building in DH projects and how much theories of the visual and visual pleasure are hidden from view. In Johanna Drucker’s book Graphesis: Visual Forms of Knowledge Production, Drucker tackles the theoretical stakes of digital visuality by explaining that “all images are encoded by their technologies of production and embody the qualities of the media in which they exist. These qualities are part of an image’s informations” whether this be illuminated manuscript, daguerreotype, painting, photograph, or digital image.[6] She highlights how the recent discussions in media archaeology have centered media production and how “reading the matter of media” is how meaning is configured.[7] Digital media environments require multimodal reading, creation, and interpretation. However, digital media then rely more on the histories, theories, and epistemologies of the codex and book than they do on film and video. The issues of layout, marginalia, paratext, columns, table of contents, indexes, and chapter headings are, as Malcolm Parkes discusses in “The Influence of Ordinatio and Compilatio in the Development of the Book,” a development of the medieval scholarly book.[8] These experimental page structures became standard in printed books and eventually in digital texts.

If the codex as developed in the Middle Ages is one of the earlier kinds of informational “interfaces,” then we should consider it as a mediating apparatus: one in which the mise-en-page and material features, its myriad graphic cues explain how to read, use, navigate, and access information in the codex book.[9] Then the digital interface requires us to consider how critical interface design can help us build digital projects that address how this mediating apparatus will change how our readers/users/subjects will interact and create interpretive iterative acts with their reading, access, and navigation of the digital information. What this means is a move away from the codex’s mise-en-page to a film’s visual mise-en-scène to an interactive digital mise-en-système, what Drucker describes as “an environment for action.”[10] A digital mise-en-système is a digital ecology in which the main question posed is how the interface iteratively and at various moment can “enunciate” the subject/user/reader. Interface, then, is a “border zone between cultural systems and human subjects”; it is the codependent space where “speaker and spoken are created.”[11] This interaction in this border zone also pushes back against cinema’s one-directional view of theories of visual pleasure and the gaze into a different model.[12] I believe feminist materialism and the work of Karen Barad on entanglement theory would name this interface site the space of “intra-action.”

Stacy Alaimo and Susan Hekman have pointed out the volatility of materiality as a location for feminist theory, so much so that “most contemporary feminisms require that one distance oneself as much as possible from the tainted realm of materiality by taking refuge within culture, discourse, and language.”[13] Instead, material feminism proposes that feminist theory must discuss materiality, and particularly the body, as an active agent that includes “lived experience, corporeal practice, and biological substance.”[14] The point of material feminism is “to build on rather than abandon the lessons learned in the linguistic turn,” namely, in this case, “a deconstruction of the material/discursive dichotomy that retains both elements without privileging either.”[15] Thus material feminism rethinks “agency, semiotic force, and the dynamics of bodies and natures.”[16] The most focused energies and the most radical move are to reconsider materiality: the “stuff” of bodies and environments. The “material turn” requires us to take “matter seriously.”[17] Material feminism insists on flattening hierarchies and ontologies; it requires a consideration of how “culture, history, discourse, technology, biology, and the environment” interact without organizing these nodes without giving more power to one or the other.[18] In essence, it is a new way to consider “matter” in relation to “material culture, geopolitical space, food, climate and environment, gender, body, nature, and culture.”[19]

Karen Barad, the theoretical quantum physicist, writes,

The notion of intra’action (in contrast to the usual “interaction,” which presumes the prior existence of independent entities/relata) marks an important shift, reopening and refiguring foundational notions of classical ontology such as causality, agency, space, time, matter, discourse, responsibility, and accountability. A specific intra-action enacts an agential cut (in contrast to the Cartesian cut—an inherent distinction—between subject and object) effecting a separating between “subject” and “object.” That is, the agential cut enacts a “local” resolution within the phenomenon of the inherent ontological indeterminancy.[20]

In essence, Barad’s point in “Nature’s Queer Performativity” is to flatten hierarchies in which everything—human, nonhuman, matter—becomes a constantly shifting component. Within this frame, she “reframes” ideas of causality; and what “intra-acting” ultimately allows is that “‘relata’ do not pre-exist relations, but rather that ‘relata-within-phenomen’ emerge through specific intra-actions.”[21]

What Drucker describes as the “codependent in-betweeness” of the interface where speaker and spoken are born is exactly a description of “intra-action.” And what she has framed in her argument about digital graphesis as part of a long history of reading interfaces is that digital reading has become a mise-en-système with multivocal moments of “intra-action.” She describes this digital shift in the history of reading:

Reading was always a performance of a text or work, always an active remaking through an instantiation. But reading rarely had to grapple with the distinctions between immersion and omniscience—as when we are experiencing the first person view of a video juxtaposed with manipulation of a scalable map, with watching the social network reconfigure itself around a node of discourse even as the node is changing. Digital environments increasingly depend upon a whole series of contingent texts, transient documents, that are created on the fly by search and query, filtered browsing or other results-based displays that last only a few moments on the screen in the stepping-stone sequence of user clicks that move from one ephemeral configuration to the next.[22]

In this digital mise-en-système, beyond the flexible and iterative possibilities of moments of subject and interface “intra-action,” what this digital reading ecology creates is the possibility for extremes of scale.

In this way, though we hear much more about the affordances and utopian heights of big data analytics and visualizations and the theoretical approaches of distance reading as explained by Franco Moretti, or even a discussion from Lev Manovich on the experience of digital visualization at extreme large scale, less is discussed about the potentials of the small, close-reading scale and the possibilities of sifting through the granularity of ever more minute details.[23] For medievalists, that granularity could be in close reading words, to the strokes of a scribal letter. What critical discussions about the digital humanities seem to forget are the possibilities of examining and working with minute granularity—the practice of extreme close reading. Scholars have discussed granularity in e-literature, digital history, and digital media studies who have written about “scalable reading.”[24] In whatever direction digital reading, composing, and writing take, in relation to the new ecology of digital reading, we must think of the interface as “a provocation.”[25]

Thus, the book of the future explodes with different arrays and angles of possibility. It will include reading, writing, annotation, social media; image, sound, tactility; text process, text analytics; small and large data mining and data mapping; the abilities to search, link, visualize, reroute, reconfigure texts and textualities; indexing, displaying; close analysis of pixels and biological properties of vellum and paper and the distance reading of a thousand years of a word—all within a multimodal, multiplatform, intermedial, and remediated digital environment. Drucker writes,

Pages will be temporary configurations based on calls to repositories and data sets. We will “publish” our data trails as guidebooks for the experience of reading, pointing to milestones and portals for in-depth exploration of stories, inventories, and the rich combination of cultural heritage and social life in a global world. The display will take advantage of the n-dimensional space of the screen in ways that combine multiple design visions.[26]

Within the book of the future, what we must understand is that visualization is and always will be an interpretive act. And the interface changes digital reading because of its dynamism by making the act of reading “a set of possibilities we encounter and from which we constitute the tissue of experience.”[27] Digital reading will be located in an ever-changing ecosystem where reader and text will constitute multiple points of “intra-action.” In this way, our bodily senses are particularly heightened in learning to move through this digital reading ecology. Thus, what is most difficult to address is why the emphasis has been on only one sense—sight—over others in digital project-building environments. Why did the AEME decide to invest first in designing the visual signature of its project before anything else?

The Archive of Early Middle English

Bracketed by the Norman Conquest in the eleventh century and the decline of the English populace as a result of the Plague (1348–1450), the Early Middle English period is characterized by its multilingualism and its interaction with cultural developments from Ireland to the Middle East.[28] In addition to four main literary languages (Latin, French, English, Welsh), Britain was also home to speakers and scholars of Greek, Hebrew, Irish, Old Norse, Arabic, and Dutch. This period also witnesses British crusaders’ establishment and loss of colonies in the Middle East, as well as the expulsion of the Jews from England. Literature of the period frequently reflects these cultural encounters among Christians, Jews, Muslims, and heretics. This is a literary world very different from standard views of medieval England; as new scholarship is revealing, this world was multilingual, culturally and racially diverse, intellectually and aesthetically experimental.

Philologists and historical linguists find the Early Middle English period fascinating, for it arguably embraces the most systematic, extreme change in the English language in recorded history. The linguistic shift between 1100 and 1350 is, in many ways, far greater than that which separates Chaucer’s use of language from that of Shakespeare.[29] In addition to internal developments during these centuries, multiple languages heavily influenced English, shaping not just its lexicon but its phonology, morphology, and syntax. The record of dialectical variations increases exponentially, and the unique multilingual and polyglot milieu of Britain makes this period and its materials of great potential interest for scholars working on the integration of cultures.[30] However, many of the period’s manuscripts and texts either have not been edited or exist only in nineteenth-century editions. This makes a systematic, scientific study of data from these texts difficult and in some cases impossible.

In 2013, AEME was awarded an NEH Scholarly Editions and Translation grant in order to create the Archive of Early Middle English (AEME), which will be made freely available to scholars, students, and the public. Initially, we will produce an electronic edition of two Early Middle English manuscripts: Oxford, Bodleian Library Laud Misc. 108 and Oxford, Bodleian Library Junius 1. We also will begin substantive work on an edition of Oxford, Jesus College 29. Our new editions will contain not only electronic transcriptions but also encoded information on names, places, intertextual features, and philological, paleographical, and material features. All information and commentary will be searchable and easily adaptable to use in a variety of digital analytical forms. We also plan as part of the project to include translations.

Our proposed editions of Oxford, Bodleian Library Laud Misc. 108 and Oxford, Bodleian Library Junius 1 will contain the complete manuscript contents in a format that will easily accommodate the addition of new texts after the grant period ends. We have chosen these two manuscripts because their texts are fully available only in nineteenth-century editions and because they are also in high demand by both scholars and students. We believe that they are ideal test cases to fine-tune our editorial methods and publishing platform. Given the conceptual and technical challenges we are taking on for this project, we believe that beginning with manuscripts that have restricted numbers of identifiable individual texts is appropriate for achieving our project goals within the grant’s timeframe. At the grant period’s end, team members intend to continue editing Early Middle English texts to add to the larger archive, as well as to encourage submissions by other scholars.

Our project approaches the challenges of editing Early Middle English texts by treating them in their manuscript contexts as material cultural objects, rather than following earlier scholarship’s tendency to evaluate Early Middle English literature purely in terms of its aesthetic or linguistic value. Rather than invoking the nineteenth-century, Romantic ideal of the authorial/artistic genius, we plan to examine Early Middle English texts first and foremost through the lens of their manuscript witnesses, addressing their larger multilingual, multimedia, and multitemporal contexts. We can examine how multiple texts appearing together in single manuscripts operate in conversation with each other. By focusing on manuscript materiality, we also hope to use the digital platform to think through questions of manuscript mouvance—material variation that includes textual modification, language switches, revision, expansion, replacement, and reorganization—within a three-dimensional platform. Thus, for example, our archive could support the physical analysis of a holograph author’s penchant for gluing, cutting, and physically sewing in his revisions and changes. Likewise, since Early Middle English frequently appears in the same manuscript with non-English literature or literature from preceding and following periods, our focus on manuscript witnesses allows us to create editions that can ultimately encompass texts not traditionally considered Early Middle English but that allow readers to explore ways in which these linguistically and/or chronologically diverse texts interact.

We define the corpus of Early Middle English as all texts occurring in manuscripts containing Early Middle English according to the criteria laid out in Margaret Laing’s Catalogue of Sources for a Linguistic Atlas of Early Medieval English (roughly, those written down between 1066 and 1340, and a few later copies of pre-1300 documentary material).[31] In compiling our list of manuscripts, we use c. 1350 as our end date, for this allows us to include a variety of linguistically Early Middle English material not included in Laing’s catalogue. By these criteria, the total corpus of Early Middle English consists of about 162 manuscripts. Since the archive will include multilingual manuscripts, it will ultimately not be restricted exclusively to Early Middle English language texts, and will thus support growing scholarly interest in the French and Latin literature of England.[32] However, we recognize that this decision may have the inadvertent effect of reinforcing the traditional marginalized status of Early Middle English. By defining our corpus using manuscripts containing Early Middle English, we intend to make Early Middle English the focus of the cultural nexus of medieval England, turning the traditional scholarly approach on its head by shifting the marginal into the center.

The Proposed Editions

In this phase of the project, our proposed editions of Oxford, Bodleian Library, Laud Misc. 108, Oxford, Bodleian Library, Junius 1, and Oxford, Jesus College 29 will contain the complete manuscript contents in a format that will easily accommodate the addition of new texts after the grant period ends. Namely, we will be using TEI-XML encoding to encode the data themselves as a stable encoding language. TEI-XML is the Text Encoding Initiative standard for Extensible Markup Language that is the code used to be machine-readable. The archive itself will have open-source code and will be archived in the Brown Digital Repository.

Laud Misc. 108, a late-thirteenth-century manuscript with entirely English contents, including the earliest version of the South English Legendary, and versions of Havelok the Dane, King Horn, will allow us to do whole-book editing on a manuscript filled entirely with Early Middle English texts. Junius 1, the unusual Ormulum manuscript, will model how to treat a holograph manuscript with three-dimensional textual revision and also postmedieval revisions. Jesus 29, a multilingual anthology, will allow us to explore the editing of a multilingual codex containing Anglo-Norman French, Early Middle English, and Latin, as well as texts from different periods written on different media (paper and vellum).

The AEME has been envisioned as first an archive, a location—albeit a digital location—in which objects have been collected. What we are collecting is the 160+ manuscripts in the Early Middle English corpus. The standard for choosing each manuscript in this corpus is that it includes Early Middle English, whether this be the work of the entire manuscript or a marginal gloss. The whole manuscript is the item unit in our archive—in whatever language, media, or material state a manuscript has or is in. However, the material unit of our archive—the manuscript—has also meant that we are participating in the history of the book in specific ways that need to also address the medieval/digital sensorium.

Medieval/Digital Sensorium and the Long History of the Book

In the narratives of the history of the book, numerous critics have pointed out that the history of print has become the hegemonic center of book history. So much so, in fact, that several medievalists have taken to task Robert Darnton’s definition of the history of the book as “the social and cultural history of communication by print.”[33] Jessica Brantley pushes for a more capacious definition of the book: “the material support for inscribed language, a category that includes rolls, codices and even monumental inscription, both written by hand and printed by many different mechanisms, and also a wide variety of digital media.”[34] And as Alison Walker points out in her article in Digital Humanities Quarterly, “The Boundless Book: A Conversation between the Pre-modern and Posthuman,” where exactly does that leave premodern and posthuman “mediated” textuality? Where does that leave medieval manuscripts and digital texts?[35] Walker further argues that if one decenters print history in the narratives of the history of the book, what we then discern is that “reading technologies from the pre- and postprint eras anticipate the same sort of reader and share similar experiences.”[36]

But strikingly, in our more recent discussion of electronic textuality, multisensory reading practices have turned our gaze back to the medieval world of manuscript textuality. We now hear and touch in order to read in a digital medium, and hopefully this signals a “paradigm shift” happening in digital reading worlds that have migrated away from a “dominant ocularcentrist aesthetic to a haptic aesthetic rooted in embodied affectivity.”[37] Medieval reading practices were not linear, often required vocality to read out loud or sing out loud, ideally required slow and repetitive rereading, were emotive, and involved sound, smell, touch, taste, visual, and even bodily calisthenics. Literally, from how the book is made—from the physical embodiment of vellum or parchment (sheep or cow skin) to the visible remains of hair and flesh side on these writing surfaces—skin is interface. As recent discussions of manuscript materiality and reading have discussed, touch and the body were incredibly central to reading and interacting with a medieval manuscript. One constantly was reminded by the different interface textures between hair and flesh side of a folio as you turned the medieval page.[38] Physical flesh is always present in sound, touch, sight, taste, and smell when one opens a medieval manuscript. Medieval reading invokes emotive, bodily, and multisensory reading practices including touching, feeling, kissing, and licking manuscript parts and pages.[39] Thus, medieval manuscripts inscribe a history of the senses and the reader’s/subject’s/users’ interaction with these fleshly interfaces. Medieval readers have deposited their breath, finger dirt, saliva, and probably bits of their dinners on the vellum page.

The aesthetic beauty and pleasure in building DH projects may hide the interpretive process, but “they are the persistent ghosts in the visual scheme.”[40] But instead of hiding how visualization and visuality of the digital interface organize and interpret informational data, how exactly can we make room to highlight how visualization is an interpretative act? How do we allow room to make transparent the AEME’s choices in its visual design and footprint? If visuality always is perceived as a transparent model of information that helps hide precisely how “constructed” data themselves really are, how can projects escape this building trap? What Drucker advocates is ways to build projects and build visualization models that encourage, highlight, present, and play with ambiguity and uncertainty.[41] By creating a marked space for ambiguity and uncertainty, Drucker argues that this allows digital zones to emphasize and lay bare the centrality of interpretation in the digital project’s constructedness.[42]

What will happen when we move beyond squareness, which has been the guiding shape and visual principal of textual media for over a millennium? What will shift when we move beyond just visual pleasures and consider how the other senses—taste, tactility, and sound—will change the terrain of reading? If the current book is “a momentary slice through a complex stream of many networked conversations, versions, and fields of debate and reference across a wide variety of times and places,” and it is but a “temporary intervention in a living field of language, images, and ideas” in which “each instantiation re-codifies the image of a book as an icon—whether mythic or banal, a treasure or an ordinary object of daily use,” the future book already has taken these intersections and expanded, bent, reformed, and remediated this vision.[43] But for the future book, it cannot just remain fixed on the form of the medieval codex. Instead it will push the boundaries of fluidity and navigating ever-shifting situated contexts connected to “the vast repositories of knowledge, images, interpretation, and interactive platforms.”[44] The book of the future will be a multitudinous event/object: “an interface, a richly networked portal, organized along lines of inquiry in which primary source materials, secondary interpretations, witnesses and evidence, are all available, incorporated, made accessible for use.”[45] It is then rhizomatic: with multiplicities at work, “with no beginning nor end” and always in media res.[46]

As Drucker explains, “we are in the incunabula period of information design.”[47] New frames, new questions, new ways to imagine linked relationships, meandering paths through reading, reading communities are only surfacing. And she is right to posit that “we are learning to read and think and write along rays, arrays, subdivisions, and patterns of thought.”[48] For literary scholars and for readers/subjects/users of DH projects, digital textual data, digital editions, digital writing, and digital rhetoric, digital reading should mean that interpretative acts will be made visible and material and a flexible space of ambiguity will allow for multivocal and rhizomatic writing futures.

The Allure of Visual Aesthetics

So how does this all practically play out in DH project building? Graphesis and the visualization of data constitute the lion-share of DH tool building. If one does a search on the Dirt: Digital Research Tools, under “visual,” you will come up with pages and pages of hits for possible tools to help you visualize data.[49] However, if you put in any other sensorial possibility—sound, touch, tactility, smell, taste—either nothing will come up or you will actually find items like “visualizing sound.”[50] Yet, these are the tools available to most digital projects to do “something” with their coded data. I believe we can prod the ubiquity of graphesis as a default and hegemonic mode of thought which has thus led to a preponderance of such tools/modes in a discipline that often states it is about analysis. However, I do think there has been some small movement to break away and critique this default analytical setting. Recently, there have been a number of projects that have considered how to sonify data; nonetheless, the number of visualization options vastly outnumbers these sonifying options.[51]

Thus, the path always drives us to make pretty maps, as we see in the case of Sexy Codicology’s manuscript maps; and in Angela Bennett’s visualization of Piers Plowman manuscripts; and even the nodelxl map of #medievaltwitter networked range during the International Medieval Congress at Western Michigan University in 2014.[52]

Adam Foster, in a recent INKE (Implementing New Knowledge Environments) post titled “The Political Aesthetics of Digital Humanities Environments,” exhorts the Digital Humanities to “be [more] attuned to the political message of scholarship the new knowledge environments crafted will boast, and consider if they do indeed change the inherent politics of scholarship.”[53] He further speculates that in order to unpack the politics of a digital learning environment, we must ultimately address aesthetics.

This heavy reliance on the visual is particularly prevalent in DH projects; yet, the politics of aesthetics are rarely addressed. As Heather Froehlich commented at a historical corpus linguistic talk in Helsinki for Varieng in 2014,

Corpus linguistics is a very text-oriented approach to language data, with much interest in curation, collection, annotation, and analysis—all things of much concern to digital humanists. If corpus linguistics is primarily concerned with text, digital humanities can be argued to be primarily concerned about images: how to visualize textual information in a way that helps the user understand and interact with large data sets.[54]

Froehlich finishes her talk by asking a provocative question: “If digital humanities currently serves mostly to supplement knowledge, rather than create knowledge, we need to start thinking forward to ask ‘What else can we do with this data we’ve been curating?’”[55] She finishes by pointing out that “digital tools and techniques are question-making machines, not answer-providing packages.” I would like to push this even further with these excellent points and ask, how does creating data become opportunities for question-making? How can we think about knowledge building in ethical, balanced, and critical ways that make DH projects beyond avenues to supplementary knowledge?

Is visualizing data (the DH bells and whistles), the awe-inspiring beauty of visualizations, then the ornament of a digital project? As an avenue of supplementing knowledge, are visualizations a form of digital ornament? In discussions of digital archive preservation, scholars, librarians, and computer programmers have already separated what is primary and what is supplemental knowledge. What is essential to preserve is not the visualization tools but the data in a stable code. The visualization coding and software applications become part of the functionality of the portal or interface, but they are not the priority when thinking about long-term preservation benchmarks. I believe that in order to address visual aesthetics, we must turn to the critical discussions in art history.

Art History, Visuality, and Pleasure

It is from art history, rather than cinema studies, that I wish to frame out discussions of visualization, visual pleasure, and digital environments. In particular, the work of David Brett and his book Rethinking Decoration: Pleasure and Ideology in the Visual Arts help reframe ideas of “decoration and ornament” as “a family of practices devoted mainly to visual pleasure; and treat this pleasure as a family of values, which includes social recognition, perceptual satisfaction, psychological reward and erotic delight (amongst others, all overlapping one another).”[56] He remarks that these are public values because they are in plain view and that further they show individual experience.

His work in theorizing decoration and ornament looks at Pierre Bourdieu’s schemata of perception with a little nod to Kant, but relies heavily on John Dewey’s “naturalistic account of experience as a relationship between an organism and its situation—an account which does away with subject/object dichotomies in favour of an interactive model of perception and meaning.”[57] Thus, Brett’s arguments about reframing visual pleasure in relation to decoration and ornament are precisely centered on an individual’s “experience” with the natural, visual, decorative world. In this way, his theoretical points fit well into the immersive, interactive, yet highly visual worlds of DH projects because visual pleasure is about centering the individual experience and point of view.

I now wish to turn to the AEME to consider how this medieval DH project remediates medieval manuscript textuality into digital textuality and what that world looks like for the history of the book, for multimodal reading, for the postcolonial archive stories of the building of this medieval manuscript archive, and finally, what the theoretical implications are of building this world. What does the AEME’s choices in digital database design say about decolonial, feminist material, and multivocal archive building? What is the AEME’s archive story, and how is it figured as an embodied archive?

The State of Early Middle English Studies

When linguistic and literary scholars have described the Early Middle English period (roughly ca. 1100–1350), their collective evaluations have labeled it “one of the dullest and least accessible intervals in standard literary history, an incoherent, intractable, impenetrable dark age scarcely redeemed by a handful of highlights.”[58] J. A. Bennett and G. V. Smithers, embarking on an edition of extracts of Early Middle English literature in 1966, found little to challenge “the traditional view that the reigns of William [the Conqueror] and his sons mark an hiatus in our literature and the widespread literary use of the vernacular that is such a distinctive feature of Anglo-Saxon culture.”[59] Even scholars who recognize shifting aesthetic standards nonetheless dismiss Early Middle English literature on the basis of principles laid down in the nineteenth century, when much of this material was first (and often last) edited. Thus, Early Middle English is imagined as a literary wasteland in which “the débris of an old literature is mixed in with the imperfectly processed materials of a new.”[60] Even when scholars try to depart from these paradigms, there is a tendency, as Christopher Cannon observes, to view Early Middle English texts in terms of a “profound isolation from immediate vernacular models and examples, from any local precedent for the business of writing English.”[61] For Hahn, the period has a reputation for “aridity and remoteness,” and for Cannon, the consequence is “literary history’s general sense that there is nothing there, since the lack of continuous tradition has so generally (and subtly) been equated with a lack of literature.”[62] But the Early Middle English period was in fact a time of intense linguistic change, literary experimentation, and textual production that juggled regional specificities, genres in process, and multilingual interactions with verve.

From an explanation of Early Middle English, one can see how disruptive and difficult to pin down the period and its manuscripts/texts are. Even if one just takes a quick sample look at the manuscripts, you can see their vast range and often illegibility. If one examines samples only from the first three manuscripts the AEME will edit—Oxford, Bodleian Library MS Misc. Laud 108; Oxford Bodleian Library, MS Junius 1 also known as the Ormulum; and Oxford, Jesus College MS 29—you see the lack of uniformity. If you add the page containing a fragment of the early Middle English lyric Worldes Blisce, only preserved on a scrap of vellum slotted in sideways at the end of Cambridge, Corpus Christi Library MS 8, the interface mise-en-système of this manuscript’s archives are multiple.[63]

The contours and shape of the archive and the corpus are difficult to narrate. Early Middle English is zone betwixt and between, a literary eruption, an epistemological disruption of linear narratives of literary history, manuscript production, and stories of continuity. In the 162-manuscript corpus, a little over fifteen items are entirely in Early Middle English. The rest are in multilingual compilations. In addition, the most popular Middle English text from the period was the product of female anchoritic patronage rather than a monastic milieu or a royal court. The period has no masculine epic like Beowulf or the Nowell Codex for the Anglo-Saxon period, no visual splendor, no court poet like Chaucer or Gower. Instead, we have plucked the image of Laȝamon’s decorated initial, based on visual cues of Jerome in Jerome Bibles, as writer/as coder but primarily because there were so few visual fields available in our corpus. So the question is, how does a team, mostly of women, rethink a digital archive of disruptive objects?

The importance of the visual point of view as I have discussed with Brett and art history is also the underlying framework that is building the spine of our archive. Our encoding documents reveal that we are primarily focused on our XML encoding schema. This choice that we have made to focus on XML is about sustainability in code, practicalities of work flow, but also I would argue a theoretical choice. Recently several projects have begun to use Resource Description Framework (RDF) as the base building block of their projects. But the best way to explain the different forms of data modeling and how they will have an effect on the constructedness of the data themselves requires a description of data model choices.

Data Models and the Semantic Web

The best quick and easy explanation with graphs on data storage models for the semantic web is available from Linked Data Tools.[64] In this modeling of data, what one has to understand is how information (and in the case of literary and historical databases this is usually a textual set of data) in data modules is organized. There are currently three different kinds of data schemes that can be easily explained by a visual diagram.

First is the relational database that usually is built with programs like MySQL and MS SQL.[65] MySQL is the one of the most popular open-source relational database management systems (RBDSM). It underwrites sites including WordPress, Facebook, and Twitter. It is a model of data organization that thinks through relations and links. Thus, it thinks about data units and their organization through a relational model, a network.

The second data model is a hierarchical one using TEI-XML (Text Encoding Initiative–Extensible Markup Language). This data model is usually the one used most for DH edition projects because the hierarchical model allows for a tiered data organization structure that accounts for the organization of books—the book, the author, the chapters, the sentence or line, and so on. This organization, of course, is about the layout and format, the mise-en-page of the codex that was developed in the Middle Ages. In TEI-XML, you can organize a data informational structure that has a very schematized guideline. It is easy to identify textual units including chapters, sections, and lines.

The third data model is RDF. This model is an arbitrary object relations model; in other words, there is no schematized structure or relation networked connection. Instead, it works more like a blank sandbox in which you place the various digital objects in arbitrary relations. The usefulness of this model is about digital objects—like manuscript pages—in which you can do mock-ups that are as close to the original as possible. In other words, as the example from Stanford’s Shared Canvas demonstrates, it is as near to surrogacy to the original data space as possible.[66] It also allows for different kinds of organizations. For instance, TEI-XML works best when items have lines, but what do you do with handwritten manuscripts in which handwritten marginalia, drawings, charts, doodles, and other nontextual material are presented all over a manuscript page? Because of its canvas/sandbox frame, RDF models allow the possibility of making units of code in relation to marked-out zones or areas.

I am walking through the schematics of this because it brings up the question of what data models for a digital edition and eventually an archive will mean in terms of how readers of the editions and archive understand the interpretive architecture already built into the digital item that the reader or participant will be working, playing, and reading. And a discussion of the database choices also explains visual pleasure, as explained in David Brett’s Rethinking Decoration, in which visual pleasure is precisely about an individual’s experience with his or her environment.

Data Layers and Archival Points of View

The AEME Guidelines specify that manuscripts in the archive will be composed of at least four layers of representation to support user interaction and workflow: image, facsimile transcription, diplomatic (uncorrected) transcription, and critical transcription.[68] If one or more of these are unavailable, they can be supplied with place fillers. This basic structure allows for further layers of representation to be added, such as a translation. Image metadata and transcription texts will be searchable. We describe each of these features in greater detail below.

1. 1. Image: Access to manuscript digital facsimiles containing Early Middle English texts. AEME-held images will initially be photographed as 24-bit 600 dpi max TIFF files. These will be converted to lower-resolution JPG200 files for service on the AEME platform. Images integrated through LOD may vary in format and quality, but the platform will be able load the most common formats.
2. 2. Facsimile Transcription: An encoded mock-up of page elements that can then stand in for missing images (such as when part of a manuscript has not been digitized) and which can further serve to categorize page elements into searchable objects for comparative analysis.
3. 3. Diplomatic Transcription: A more or less literal transcription of the text for readers interested in the scribal representation of the text. Coded in TEI-XML following AEME markup guidelines.
4. 4. Critical Transcription: A transcription of the text including various types of editorial intervention, including modernized punctuation and capitalization, editorial corrections and notes, and contextual information (glossary references, geolocation tags, etc.). It is anticipated that the critical layer will be suitable for student readers. Coded in TEI-XML following AEME markup guidelines.
5. 5. Translation: The AEME platform will accept translations of texts, which will be displayable in the same way as diplomatic and critical transcriptions.
6. 6. General Search and discovery of digital images and texts.

With the exception of the facsimile layer, we are primarily coding in TEI-XML. However, there are examples of projects that have begun using RDF as the primary editing space. For example, the Shared Canvas project out of Stanford University is creating an RDF editing platform. It has done a mock-up—if you look at slide 53 in this slide-share you can see their mock-up of Worldes Blisce from the Parker Library CCCC MS 8.[69] What they are demonstrating is that they can overlay the coding for the edition on top of the manuscript like a palimpsest and then encode an audio file so that the lines can be sung when one clicks the edited line. They use RDF as their main building framework. AEME has decided to work with TEI-XML for everything but the facsimile layer, which is in fact a layer that fits this particular data model.

We have made a choice to privilege experience over ideas of “objective” data; we have prioritized visual pleasure. RDF splits information into grid units (or zones) on a screen, and each square unit of data in the grid is moved and read in this way. Thus, I would argue that RDF is actually the digital heir of Dürer’s grid in which objects are broken up into individual grid units in order to produce visual perspective.[70] This visual perspective gives the artist the God’s eye view of the world. And as the famous image of Dürer’s grid reveals, what gets broken up are not just landscape images and objects but also people, especially women. As art history has discussed, perspective can often be violent to these objects, and it is often women objectified behind the grid.[71]

AEME has discussed RDF, but we have chosen to go with TEI-XML because it allows us to see the narrative of individual editors. We have prioritized the experience and interaction with the material from the point of view of each reader. We have privileged individual archive stories, rather than imagining the possibility of algorithmic objectivity in building the spine of the archive. We have decentered the archive, flattened subject/object relationships, allowed for a multiplicity of views; we have built in room for the individual editor/editor who currently and in the future will work and play in our archive location.[72] We have attempted the actual digital building praxis of a decolonized and feminist archives manifesto. In my mind, it is the difference between RDF as the genealogical child of Dürer’s perspective grid versus the narrative choices of individual interaction. We have chosen the path of the Wife of Bath; we have taken “experience” over algorithmic authority.

You can see this working even on a micro level with our decisions. For example, in an early discussion about editorial frameworks, Scott Kleinman (co-director) sent a question out to the group to discuss:

Verse-initial letters (often highlighted by shadow gaps, rubrication, and the like) can be difficult to identify as capital or lower case. I have drafted the following suggestion for handling this phenomenon in the Guidelines:

Verse-initial letters should always be capitalised inside <reg> tags (i.e. the critical representation). The representation in the diplomatic layer can be problematic because it is often difficult to distinguish capitals from non-capitals in this position. AEME leaves it up to the editor’s discretion to decide ambiguous cases. Future versions of these Guidelines will list some best practices to aid in decision-making. A useful tool would be a list of suspect letters.

Please comment on this guideline. Does it seem adequate to you? Can we begin compiling a list of letters that should be called to the transcriber’s attention?[73]

As this micro-discussion point shows, we have at every turn privileged individual interaction with the manuscript and data as the final say in our decisions. The last email round was about capital letters in a manuscript and how we have indeterminancy problems and what we may want to do about that. We came up with the suspect letter list and then the narrative discretion of each editor. The Archive of Early Middle English then is focused on how archival manuscript bodies interact with editorial bodies. We are in essence capturing the experience of editors with the physical and the digital manuscripts and how this interaction happens. We are recording their aesthetic and visual pleasure. My last point about Shared Canvas’s publishing possibilities also brings up the question of how to get away from privileging the visual in DH projects. Where is the space where we can interrogate these issues? I believe this will really come from Disability Studies in Digital Humanities building. And this shift to Disability Studies is where the AEME will develop into more robustly capturing the process and experience of its editorial/user/player bodies.

Disability Studies

In current disability studies, the term used to discuss designing web environments with disability issues in mind is “universal design.” Adeline Koh wrote about her experience at the Accessible Future Workshop in Austin 2014 and critiqued numerous issues with the idea of universal design.[74] But it is the Twitter conversation on the hashtag that brings up the most resonant issues in relation to making visible the agendas of our digital projects and the issues of the senses in digital project building. She explained on Twitter her discomfort with imagining “universal” accessibility as a default to be the priority benchmark. She explained through the lens of postcolonial theory and criticism whether “universal” should be desired by all when, in fact, postcolonial writers often resist writing in “accessible” colonial languages as a form of resistance. She explained that “the drive towards universal and ultimate accessibility for everyone and everything, seen in this light has parallels with a colonial impulse to observe, survey, control, force open.”[75]

Universal design becomes too close to the ideas of one-size-fits-all in Enlightenment political liberalism. It also references current critical race discussions about how postracial ideas of the “universal” erase difference. The term “design” itself already problematizes this because it values visual aesthetics, and even the term “architecture” has connotations and agendas centered on the 1 percent, and on masculine visuality and the gaze.[76] In my interactions with the Twitter discussion, I pointed out that the goal of disability studies should be to disrupt the very idea of “accommodation” in order to reconceptualize how building computer programs or actual buildings should be disrupted and rethought from the disability studies angle. We should stop thinking only as ableist+accommodations, but rather from disability studies’ point of view to rethink the critical possibilities. What if the mainstream angle was the angle of disability studies? In digital humanities work for instance, classic textual data modeling is a simple wordle (http://www.wordle.net) word cloud with colors and shapes. Why can’t the norm of textual data analysis be a sound cloud or a textural cloud that allows you to print a 3-D model to touch?

These are questions I have asked Rick Godden and Jonathan Hsy as I edited the revision of their article “Universal Design and Its Discontents” for Disrupting the Digital Humanities. In particular, Rick Godden explains:

As an entry point to my reflections on Universal Design, I want to first think about some of the ways that Digital Humanities (DH), Disability Studies (DS), and Universal Design (UD) productively converge using recent discussions about the physical act of hand-written notes as an opening example. This is not unusual in a bid to consider the necessity of UD; however, I also want to use this example in order to begin to disorient some of our understandings of UD. Although UD arose out of a real social and political response to the disabling aspects of everyday life for People with Disabilities, I want to suggest that the “Universal” in UD can carry with it some unintended and unexpected assumptions about normalcy and our physical orientation to the world. . . . But what they are also doing, whether intended or not, is participating in “compulsory able-bodiedness,” where “normal,” “best,” and “able-bodied” ultimately occupy the same subject position.[77]

I am interested in how to disorient digital humanities and DH projects. How does a multiplicity of views through the experience of a multiplicity of different bodies help rethink the future of the AEME? As I said in my comments, architecture is too burdened with art in the hands of the elite; instead, I prefer to consider the process of intelligent, critical building. And in the discussion on Twitter, I reply to several of the conference participants and ask if we can’t begin to go beyond “accommodations.” Instead, we should center the point of view of disability studies as the vision of the digital humanities project. From disability studies scholarship, I believe we can begin to think through issues of tactility, sound, and other sensory perceptions that will help us rethink our digital tools and our digital agendas. For instance, why use word clouds only, or even sound visualized clouds? Why aren’t there more options to allow us to analyze poetic data orally by the loudness of various repeated words? Medieval manuscripts are media devices that record multimedia and multimodal experiences. They are visually laid out, but the reader and user is asked to bring the data to life, to make them sing, dance, move. In fact, the world of medieval manuscripts is an early vision of a functioning mise-en-système.

It is through the angle of disabilities studies that we can decenter the focus on visuality that has been central to the digital humanities, to move away from “ocularcentrism.” Instead, the critical possibilities may include a move away from the classic textual data modeling as a simple wordle word cloud w/colors and shapes. Instead, we may be able to begin imagining the norm of data analysis through a sound cloud or a texture cloud that allows you to print a 3-D model to touch. This possibility of different data analytics has only just begun to surface. For example, a recent blog post discusses a class that has taken a tactile path in data analytics in collaboration with art practice, literary studies, and DH.[78] This beautiful and jagged blazon sculpture highlights where other data analysis vis-à-vis the other senses may go, but I believe further critical discussions especially in relation to critical disability studies will help form critical discussions about these models. Thus, an individual’s polyvocal, multibodied pleasure becomes the entry point to navigate an ecosystem.

Skyscraper versus Snake

So if the Archive of Early Middle English is a repository of individual editorial experiences interacting with the digital and physical materiality of objects, then what is our final goal? Our goals are decidedly rhizomatic—here, I am specifically referring to Deleuze and Guattari’s theory of “multiple, non-hierarchical entry and exit points of data representation and interpretation” that Adeline Koh and Roopika Risam have pointed out as the possibility of digital archives—for a number of reasons. Wide use and access are the AEME’s goal. But so is wide building from what I would describe as the project’s eventual stable skeleton. I have argued about this difference in building a digital project with other digital humanists in the past. AEME does not want to architecturally design a skyscraper that requires years of digging out the basement and substructure of a digital project only to find out that years after building, the technology and materials have changed and the skyscraper will not be able to stand without major changes. In this way, much of early DH project focused on monumental architectural designs that took years and never quite finished what they wanted before technology, money, and general sustainability made their DH skyscrapers half-finished ruins on the digital landscape.

AEME’s main goal is to create a stable yet flexible manuscript skeleton mostly built by TEI-XML to allow for future users and builders to graft, “enflesh” on layers to our stable frame whatever world they wish to create, interconnect, recreate, form, deform. So instead of a skyscraper, I imagine the Archive of Early Middle English and its archival building goals as more akin to the flexible spine of a snake. We wish to build a digital portal that fulfills the possibilities of interface “in-betweeness” where speaker and spoken are created. We want to create a mise-en-système that critically remediates the mise-en-système/mise-en-page of medieval manuscripts, their somatic reading practices, their worlds. We want to create an ecology that can be a provocative catalyst for cascades of intra-action between participant, text, community, image, sound, and so on. Whether they be students creating narrative bitstrips with our translated Havelok the Dane or historical linguistics adding layers of linguistic markup, or musicologists visualizing the notation models in the early Middle English corpus, the builders of AEME would like all of this to happen and the archive to change and be added to, used, and played with in these different ways. In this way, the AEME will never be a “finished” product, but a system always in flux. It will be, as the “TwitterEthics Manifesto” discussed, always in process. In the end, the archive’s goal is access and use to the widest range of people so they can create a multiplicity of experiences in the database.

Archive Stories

I would like to end by relating AEME’s archive story. If AEME is an archive of objects and to create this archive is to essentially create a visual, multimedia narrative of Early Middle English, what is that story? In Antoinette Burton’s edited collection Archive Stories: Facts, Fiction, and The Writing of History, she writes that the underlying issue at stake in the volume is that the claims to “objectivity associated with the traditional archive pose a challenge which must be met in part by telling stories about its provenance, its histories, its effect on its users, and above all, its power to shape all the narratives which are to be ‘found’ there.”[79] In this way, they claim they are constructing “self-conscious ethnographies of one of the chief investigative foundations of History as a discipline.”[80] They emphasize the critical importance of these “archive stories,” these ethnographies that explain everything about how an archive was created, used, and experienced, in order to highlight that “all archives are ‘figured.’”[81] What Burton means by this is not just self-conscious creation but that archives “all have dynamic relationships, not just to the past and the present, but to the fate of regimes, the physical environment, the serendipity of bureaucrats, and the care and neglect of archivists.”[82]

As for the archive’s appeal, so much of it is centered on sensory experiences and the romance of history they invoke, whether they be the actual dust one breathes as Caroline Steedman writes, or the habitus, the experience of the archive itself—the silence, the tension, the smell, the feel of the archival matter/material as Arlette Farge writes in her classic The Appeal of the Archive.[83] Farge writes herself about the experience of the historian/archivist as waiting for that moment when “the sheer pleasure of being astonished by the beauty of the texts and the overabundance of life brimming in so many ordinary lives” grabs hold of the archivist. And the archive itself is governed by emotion: “To feel the allure of the archive is to seek to extract additional meaning from the fragmented phrases found there. Emotion is another tool with which to split the rock of the past, of silence.”[84] In this way, the appeal and often the drive of the archive are a single-minded sensorial drive toward pleasure through vision, touch, smell, and sound. In this way, the physical and digital archives are mimetic in their drive for sensory pleasure. However, they diverge in their possibilities for access and viewpoint. One is built to exclude and be about power; the other has the potential to be multiple and disrupt hierarchical power structures.

However, this romance is helped by the archive’s inaccessibility. Archives, in the words of Michel Foucault, were “documents of exclusion” and “monuments to particular configurations of power.”[85] And the archive itself is both potentially a “mundane workplace and a panopticon of intense surveillance.”[86] It is from postcolonial studies where we have seen the archive used politically and socially and often becoming a “technology of imperial power, conquest, and hegemony.”[87] Archival making, revision, and erasures are not neutral, objective acts, but rather usually critical, rhetorical, and shaped by nationalist, political, and social agendas. Burton asks how the personal encounters and experience of archive uses matter in constructing archival histories. How in fact do gender, race, ability, religion, and sexuality and the negotiations between archival objects and archival users and the power dynamics therein play out?

These questions, in many ways, may be a bit surprising to some because we are discussing primarily the building of the Archive of Early Middle English. How can an archival build of 162 medieval manuscript objects with items of early Middle English text be considered politicized, socially charged, or even shaped by postcolonial and national discourses? Yet, the archive we are discussing shapes the story of a historical colonial power, and it is this archival build that is being funded by the National Endowment for the Humanities, an arm of the U.S. government. The archive’s time period, 1100–1350, is historically a moment after the Norman Conquest, a temporal eruption of colonialism and conquest. In addition, the project directors and editorial team are made almost entirely of women, several with intersectional identities related to race, sexual orientation, and religion. In many ways, our archival stories inform our individual archival experiences that we are telegraphing into this digital archive.

There are reasons to make archival labor—usually hidden from view—clear and apparent. Consider one of the main sources of concomitant scholarly digital big data utopia and critical angst: Google Books and its mass digitization project. It has only been recently that any considerations about mass digitization have addressed the issue of the gendered and often racialized labor that has powered this project. In Shawn Wen’s recent article, “The Ladies Vanish,” she discusses the invisible, separated, and underpaid labor happening in Silicon Valley and specifically on the Google campus.[88] She writes about the army of invisible, segregated, usually female and either Latina or Black labor that arrives in the middle of the night (4 a.m.) and leaves in the afternoon (2 p.m.). They are separate and not equal to the rest of the computer engineers and daily staff at Google. They do not mingle; they are never seen but hide in a different building or on separate floors. These are the women who do the painstakingly detailed work of digitizing the world’s knowledge also known as Google Books and the Google Books Library Project. They are Google’s army of “mechanical turkers,” so named after the eighteenth-century automaton robot, a chess player automaton unveiled in 1770 in Austria that had a human inside of the machine working its parts. As Wen writes,

Of course books don’t digitize themselves. Human hands have to individually scan the books, to open the covers and flip the pages. But when Google promotes its project—a database of “millions of books from libraries and publishers worldwide”—they put the technology, the search function and the expansive virtual library in the forefront. The laborers are erased from the narrative, even as we experience their work firsthand when we look at Google Books.[89]

The vision of a worldwide, accessible, digitally available library of scanned books rarely gets the same attention as the realities of gendered, racialized, and the underpaid labor that produce these products “magically” for the world. In arguably the biggest Big Data project for the history of the book, the material bodies of these “turkers” are rarely examined in relation to the digital paper bodies they scan and digitally release to the world. They are part of the invisible digital laborers that Lisa Nakamura discusses in “Economies of Digital Production in East Asia: iPhone Girls and the Transnational Circuits of Cool” who power our digital lives.[90] Our digital archive futures are being built on the backs of the invisible labor of women of color around the world or in this case in Silicon Valley itself. Yet, these unknown, unseen, and uncredited women are the ones perfecting this big data future of the world’s library. As Shawn Wen explains, “Relying on data from mechanical turkers, computers have dramatically improved in recent years at facial recognition, translation, and transcription. These were tasks previously thought to be impossible for computers to complete accurately. Which means that mechanical turkers (mostly women) teach computers to do what engineers (mostly men) cannot on their own program computers to do.”[91] These are the women perfecting the algorithmic perfection of optical character recognition (OCR), the process driving the continued improvement in the search accuracies of Google Books.

The Archive of Early Middle English is a DH project that in evaluating the critical stakes of its own work will and must address the archive stories of the labor behind it. Our signatures, our digital paper trails of work, our streams of online discussion on ASANA, GitHub, and listservs are parts of the project’s archive that are being preserved in the files of our archival creation. How ironic would it be for a digital archive—whose main influential and popular text was fueled by the drive of female readers (i.e., Ancrene Wisse) and where so many other manuscripts have had scholarly questions about the possibilities of female scribes and “authors” (i.e., Katherine Group and The Owl and the Nightingale) that have left no visible mark of gendered ownership, authority, or labor—vanish its own digital editors, graphic artists, and builders? The mostly women behind the Archive of Early Middle English are the physical bodies driving the machine, but we plan to speak, to write, to sign our work, and to leave our records to explain how we have “figured” in this digital archive in progress. Our bodies matter to the imprint we leave on our digital archive.

I would like to begin this process by telling my own archive story. My archive story must include the fact that as an Asian woman continuously traveling to Britain to visit these archives, I have acquired a number of library cards from the Bodleian, Cambridge, Lambeth Palace, British Library, and so on. Yet, my passport’s steady entrance into Britain and undoubtedly my “suspect” racialized body, not to mention my markedly Californian accent, have regularly made me a suspicious body at Customs in Heathrow. “Why,” they ask “are you coming to Britain?” I say “for business.” “What kind of business?” they ask. I reply, “To do research in the manuscript archives.” The interrogation goes on usually until I begin pulling out the sheer number of specialized library cards in my wallet. This archive story speaks acutely to the power dynamics, the odd-accented postcoloniality, the negotiations at play in working and now building the Archive of Early Middle English.

The Archive Story of Early Middle English

I would like to end by addressing that final question: what is the story of Early Middle English? Because of the indeterminate, varied, disparate, and in-flux status of the manuscripts in the corpus, because the Early Middle English period is an epoch of intense change and also what I would call a moment of mass experimentation, the archive story of Early Middle English is one of intra-action. I believe it is the story of indeterminancy, of slippery desire, of frustrated pleasures. It is the story of experimental multilingualism; experimental genre forms; the emergence of the lyric, the romance, history, debate, sermon, prose guidebook, and first drama in Middle English. This, then, is what the story of Early Middle English is: it is local, contingent, unformed, still forming, difficult to categorize, difficult to create clear-cut distinctions between causality, agency, space, time, matter, discourse, responsibility, and accountability. But by creating an archive, by editors who are recording their experience with the archive, we are precisely enacting “an agential cut” that allows us to define the archive as forever in progress, forever local and dialectal, always indeterminate.

Notes

Bibliography