Widsið and the Digital Age

Sarah Monnier

Abstract

It is a truth universally acknowledged that the Classics, which generally means Greek and Latin language materials, are much more advanced in digital scholarship and resources than Germanic languages. While an enormous amount of material exists online to allow research into Biblical and Classical Greek text, for instance, there is less available for the Germanic languages, including Old English. While the digital world is not devoid of quality Old English digital editions (Electronic Beowulf immediately stands out), there is little connectivity between projects, and with the wider world of scholarship beyond the immediate texts. This project takes Widsið as a prototype for the XML-markup of Old English poetry that allows ease-of-reading for scholars of all levels and language abilities. The goal is to give access to original-language texts to scholars who are not fluent, or perhaps are learning the language, as well as to give wider access to additional information about things mentioned within the poem. Chosen because of its plethora of named entities, Widsið demonstrates the broader possibilities of digital editions, and the future of inclusion within the Perseus Digital Library would give still greater reach to its references. This paper discusses the process of creating this digital text, both as a scholarly edition and as a digital edition, and then explores the greater possibilities available in a digital context.

Introduction

What is Widsið?

Widsið spoke.

When one speaks, what is the implication? If a writer writes, "Jack said", there is a sense that he is merely speaking in a normal way and the writer is conveying that information. However, if the same passage instead says "Jack spoke", it implies that there is significance to the speech. There has been silence from Jack; he was not speaking before (this is not an assumption with "said" as it may be relaying just a part of a continuous speech). There is significance to Jack's speaking; perhaps he made an announcement, or spoke over a device like a phone, or on behalf of somebody. It is not merely saying.

When Widsið opens, it opens with Widsið maðolade: "Widsið spoke". In an oral culture, perhaps this was performed before an opening chord to call the crowd's attention to the performance, or perhaps it was preceded by an anticipatory silence. With the breaking of the silence in the world of the poem, the silence in the primary world is also broken. The poet spoke, and Widsið spoke.

The audience listens, aware of the significance of this statement, aware that the silence has been broken for a purpose, and eager to discover that purpose. The audience has continued to listen, or rather to read, as the poem came down to us in the Exeter Book and has been the study of many brilliant scholars over the last few hundred years.

The Exeter Book itself is an Old English codex, usually dated as having been compiled in the late 10th century. The poems contained therein are older, some perhaps much older, and of these the one usually dated as the oldest of all is Widsið. There is considerable debate among scholars as to how old the poem is. The nature of the poem itself is a catalogue poem, detailing lists of kings, peoples, and heroes, held within a story of a poet who has traveled widely--hence his name: Widsið; literally "wide-journey".

Fragments of the lists may be of varying ages, and later compiled together by a single poet, who is not presumed to be the scribe. It is one of these fragments that Alfred Anscombe dated as far back as 460 AD, and Kemp Malone proposed the compilation of the whole dated no later than the late 7th century. Later authors reject these early dates, but as Neidorf notes that "Frank, Hill, and Niles offer no substantial engagement with the dating arguments of earlier Widsið scholars. The reader is simply left to imagine that there really is no evidence for an early dating of Widsið and that scholarship had operated under a delusion for over 100 and 50 years" (Neidorf 166). Using onomastic evidence and dialectic vocabulary, as well as the absence of the Saxons from the poem, pretty significant oversight if it was written at a later date when the Saxons were a dominant people, Neidorf argues adeptly for sustaining the earlier dating. Weiskott refutes Neidorf's orthographical examples on the grounds that a poet copying lists would copy their original orthography, arguing a later date from a metrical standpoint. Pascual also takes up the argument, siding rather on the early end, also with a metrical perspective. The back and forth between these three terrific scholars is an interesting read and also strongly motivating to scholars who want to dive into this text for themselves.

Thesis Statement

The age and content of Widsið are what make it such an excellent project for digital representation. The long catalogues of names and people, now figures of a shadowy past, can be overwhelming. The fact that the poem, though short at only 143 lines, does not have a strong narrative arc, means that it is not a first choice for students of Old English. Yet the language is clear and systematic, and the repeated syntax excellent for reinforcing linguistic concepts to language learners. When one reads Helm Wulfingum (l. 29), it is apparent that Helm is in the nominative with no ending, and Wulfingum in the plural dative "-um" ending rather than the nominative "-as" to demonstrate Helm "verbed" (in this case weold) the Wulfings. This pattern is then repeated over and over with almost no variation except when Finn Folcwalding [weold] Frēsna cynne (l. 27), or when Hringweald wæs haten Herefarena cyning. These exceptions would then stand out because of their difference from the pattern and introduce new concepts.

Beyond that, the ability of digital technology to provide hyperlinking of texts and information sites, allowing quick access to information on named entities, makes Widsið an appealing project. The potential of a digital edition of a text to make it accessible to casual readers, students, and high-level scholars is a flexibility and breadth of reach that a paper text would struggle or fail to live up to. By incorporating Widsið into a larger digital library, the potential for linking to a variety of primary sources in Greek and Latin, and eventually more in Old English, would also be dramatically increased.

Literary Review

Why Widsið? Of all the prose and poetry in the Old English corpus, what makes this poem such an appealing seminal project? Unlike, say, The Wanderer or The Seafarer, Widsið is filled with references to people and places both mythical and historical. It also is only 143 lines long, making this a manageable example work which can be thoroughly annotated and displayed. For instance, Beowulf also contains plentiful references to named entities that are known from other sources, but its unwieldy 3,181 lines of inconsistent quality make it a significantly more advanced project.

From a literary standpoint, Widsið has little to offer, its narrative moments being more in passing references to other, better-known stories from Norse, Old English, Old High German, or even Roman and Greek texts. But this state as a central meeting point of a great variety of literature from many cultures and languages makes it a central node in an imaginary graph linking references to one another. From Hrothulf of Beowulf one can travel through Widsið's Hrothwulf, to Hróarr of Hrólfr Kraki's Saga. Or through Widsið one can move from Beowulf to the writings of Jordanes, from obvious fiction to dubious history. As a pivot point joining the work of many different fields of literary and historical scholarship, Widsið is an excellent starting point.

There are roughly 137 named entities within the poem, though some of them are mentioned twice. These named entities encompass the categories of individuals, people groups, regions of land, and individual locations. By the way they are grouped, inferences can be made about the associations between various groups of people, and the context of an ambiguous name can help scholars decide on its possible identification. Many locations are known to geography still today, but others, while they can be linked to references in Tacitus or Jordanes, cannot be linked to any modern region or landmark.

Print Editions

While Widsið has received less scholarly attention than some more exciting narratives, it has been well-researched nevertheless, and there are two excellent dedicated print scholarly editions, not to mention numerous editions of the whole Exeter Book and other collections. The earlier standalone edition is by R. W. Chambers and was published in 1912. It is full of excellent notes and extremely detailed commentary. Chambers spends considerable time on the historical details of the text, especially the various peoples and legends. His review of the poem is largely as a historical document that details main heroes and legends that the Anglo-Saxons wanted to preserve. He references works largely by German and Dutch authors and there is much dispute about the divisions of the poem and the various antiquity of these divisions. This view of the poem as a historical document seems reflective of Anglo-Saxon studies and Germanic philology of the period.

The later edition, though not by much, is the Kemp Malone from 1936, revised in a second edition in 1962. Malone focuses much less on the historical aspects of the text, approaching it less as a list of names for dissection and more as a narrative, and a linguistically interesting and purposeful work in its own right. Malone is heavily concerned with the meter of and alliteration of the text in the commentary but his edition still contains a rich and detailed named-entities glossary at the back as well. Neither of these versions is without their controversies, however: the Chambers edition is quite dated now and scholarship has progressed significantly since its publication, while the Malone version contains personal theories and biases that do not align with general scholarly consensus, and he rearranges the poem (in one area fairly significantly) to suit what he believes the correct metrical and alliterative form to be. The rearrangements, while significant, are not out of line with scholarship that had gone before--indeed, Chambers mentions substantial rearrangement and even sections completely removed from the poem by previous scholars trying to approximate the earliest form of the text (or texts). Because this edition aims to present the text as it comes to us, with the scholar making their own decisions regarding the meter or logical ordering of the text, I have not changed any lines or word order.

Digital Editions

Digitally, Widsið has not been neglected either. Martin Foys's Old English Poetry in Facsimile (OEPF) Project, hosted at the University of Wisconsin-Madison, contains high-quality scans of the text aligned with a facsimile version, as well as an unaligned translation. The OEPF interface makes use of the flexibility of digital text to allow for this visual alignment as well as navigability between poems and sections of the poems, and highlighting to demonstrate what sections of the facsimile edition align with specific portions of the manuscript. For scholars focused on manuscript studies, this is an excellent and well-designed resource.

My Edition

The goal of my edition of Widsið is not to compete with the OEPF version, but to expand on it. I have not created a facsimile edition from the manuscript, nor do I have scans of the text to display. Instead, my goal is to make a version of the poem that can be trusted to reflect general scholarly consensus of the text, aligned with translation and part-of-speech tagging. I have made very few changes to the text as it is derived from the manuscript. A few words that are obviously in error (the equivalent of a modern typo) and universally corrected have been changed, but these are not many. The audience is that of the literary and philological scholar who may or may not have a working knowledge of the language, rather than the digital humanist or the strict historian.

By providing word forms and part-of-speech tagging, as well as a line-by-line alignment of the text, I allow students of the poem to engage with the language closely, but also to take a step back and view the whole poem, aligned between Old English and Modern English, to gain context on the greater content. This approach is amenable to language-learners who wish to engage with the text in the original language without doing their own translation, or constantly swapping back and forth with a glossary, but also allows scholars who wish to study the poem without learning an historical language to still interact with it in the context of that language.

Working with the Historical Text

The Structure of Widsið

The poem opens with nine introductory lines that form a frame narrative for the body of the poem and introduce the scop after whom the poem is named. "Widsið spoke, unlocked his wordhoard." It gives a brief introduction to this well-traveled poet, a man of the tribe of the Myrgings who traveled with Ealhhilde to meet with her husband Ermanric, king of the Ostrogoths. In line 10, Widsið begins his own poem, declaring, "I have heard about many men wielding power over peoples!" "Fela ic monna gefrægn mǣgþum wealdan!"

From here the poem forms a few main sections. Chambers divides the poem into two main sections: the Catalogue of Kings, and the Lay of Ealhhild and Eormanric. He proposes that these separate sections were originally individual works brought together by a third narrative about a scop traveling to Eormanric that conveniently could tie in both poems. Malone divides it slightly differently, with nine lines of prologue, seven of introduction and interpolation by the poet, three fits that contain a thula, and nine lines of epilogue. These divisions are entirely based on his reading of the third-person and first-person segments and how they function within the poem in either a lyrical or formulaic manner. He is less likely to view the internal units as pre-existing large works and tends to view Widsið as more of a patchwork, incorporating various bits and being carefully assembled by a shrewd poet or scribe who added the nine-line prologue and epilogue, creating a pleasing symmetry.

The thulas are very clear, however, and there are obviously three. The first stretches from line 18 through line 35, following the "X weold X" pattern, beginning with Ætla weold Hunum, Eormanric Gotum, (l. 18) and ending with "Offa weold Ongle, Alewih Denum;" (l. 35). This is apparently a good place to break off the list as the next few lines expand on Offa's prowess in battle and continue on to describe other great figures, such as the famous Ingeld, Hroþwulf, and Hroðgar of Beowulf. Even Heorot, that glorious mead-hall of the Danes, gets a mention.

The second thula uses the formulaic first person, saying "mid X ic wæs ond mid X...", going from line 57 to line 87. This second thula, however, has one substantial and two minor interruptions. The first is two lines long and comes after the Burgundians are mentioned, expanding on the generosity of their king. A few more people are named and then the poet briefly expands on the generosity of Ælfwine and Eadwine in Italy. The final brief interruption describes Caesar, ruler of the Greeks and Romans and his many wine-towns.

After this third thula is what could properly be called Chambers's lay of Eormanric and Ealhhild. He starts this section echoing the rhetorical structure of the thula, stating "Ond ic wæs mid Eormanrīce..." (l. 88). This lay stretches until line 111, singing the praises of Ermanric and his wife Ealhhild, when it goes into the final thula. Again, the poet builds artful transition into the thula by mirroring the upcoming pattern in the last sentence of the lay:

"Ðonan ic ealne geondhwearf ēþel Gotena,

sōhte ic a gesīþa þā selestan;

þæt wæs innweorud Earmanrīces"(l. 109-111).

This last thula is less cohesive than the previous two, with frequent one-line or even two-line interruptions. All in all it stretches thirteen lines, but of those only eight follow the pattern "X sōhte ic ond X". These men are the men he sought immediately before the thula as great warriors, and he ends the thula describing their battles and travels together.

It is at this point that the first-person narrative ends and it is not unreasonable to assume this is where the epilogue begins, wrapping up the frame narrative with which the poem opened. The poet (no longer Widsið, as his speech has ended), states that the life of the scop is that of wandering and seeking out great warriors to immortalize in song to those who will pay for it. The final line declares that "He who works for praise has under the heavens lasting glory" (143). It is, perhaps, up to the reader to decide who has the greater share of the praise in Widsið: the kings and warriors, or the scop himself.

Literature referenced in creating the text

While Kemp Malone's edition of the poem, and his glossary and named entities glossary were of the utmost value in the creation of my own edition, his commentary is what gives life to his edition of Widsið. He calls it a "little masterpiece in structure and style"(Malone), not merely a set of three thulas (the term he chooses for these catalogues) strung together by a stock narrative. The main argument Malone makes regarding the structure of Widsið is that the latter two thulas were joined much earlier and came to the poet as a single entity, which he then combined with the first thula to make a whole poem. This argument hinges on the first person narration of the second and third thulas, but the third person narration of the first thula, which merely states which kings there were, rather than claiming to have traveled among them in any way. This argument is compelling but not without its detractors, not the least of which is Nicholas Howe.

One of the most engaging Old English scholars of the last century is Nicholas Howe, who has written many incredible papers and the fantastic book, The Old English Catalog Poems. This book includes a chapter on Widsið and Deor, which Howe differentiates from other catalogue poems, saying, "Unlike other catalogue poems, they do not limit the speaker to a simple didactic device for the expression of traditional matter"(Howe 166). He agrees with Malone in viewing Widsið as an artfully and intentionally composed poem, giving greater life to the title character through the style and narration. Howe takes issue with Malone's view of the assembly of the poem, arguing that "one must give more credit for originality to the poet and far less priority to his sources", and the first-person pronouns were introduced by the poet himself, necessitated by the invention of the character of Widsið (171). This argument rests on the conception of how poets reuse material (adaptation rather than "scissors and paste"), and of course that the ic pronoun and the physically impossible distances and stretches of time involved necessitates an internal narrator besides the frame-narrative narrator.

The only flaw with this argument rests in how oral storytelling works. Imagine telling a story to children. Howe objects to the verb forms that Malone uses to distinguish the sections of the poem, as pronoun + verb combinations such as "ic gefrægn", "ic sohte", and "ic wæs" require a speaker and therefore raise the question: "who is this speaker? Why has he traveled so much and over such a long time?" While his argument against Malone based on how poets reuse material is not unsound, this particular argument doesn't stand up in light of oral traditional forms. There is, presumably, a poet speaking this poem before the existence of the composed, written form we have received. These forms would make sense coming from a speaker even with the logical leaps of a reciter who claimed to have traveled enormous distances and many generations. One could draw comparisons to the stories we tell children, whether joking about how we wrote on stone tablets and rode dinosaurs to school as children or narrating longer mythical arcs that may use a first person. There is no expectation that the person narrating in first person is telling an unvarnished history of himself. The frame is, necessarily, a sort of "tall tale". And while we see this approach commonly in children's stories, it is less because this sort of fabulous narrative is suitable only for children and more because the innate oral storytelling lives on in the tales that we tell children but could just as easily and enjoyably tell adults. This argument tends to support Malone’s theory, or at least nullify Howe’s objection to it.

Creating the translation

The first step in creating a digital edition is to become thoroughly acquainted with the text in its original language. The high-resolution scans available on the Exeter website allowed me to consult the manuscript as necessary, but mainly I worked from three existing and highly acclaimed print editions. These were the Krapp and Dobbie Anglo-Saxon Poetic Records (ASPR) version, the Campbell edition, and the Malone edition. A lightly XML-tagged version of the Krapp and Dobbie text is available from the Bodleian website. The Campbell and Malone texts were not available in a digital form, so these were created by comparing the ASPR to the physical versions of these and correcting it to match. Then, assuming the ASPR to be a reasonably standard and reliable version, I compared the rich and detailed commentaries provided by Malone and Campbell to make informed choices in my own edition. By the time I had an edition I was happy with, I was comfortable working within the structure and language of the poem.

One of the first goals after getting familiar with the poem in Old English with Malone's glossary and Bosworth-Toller in hand (or in browser, in the case of Bosworth-Toller), was to create a simple translation of the poem, a process that is always trickier than it seems. If the words are assembled without regard to form and part-of-speech, this word salad can be turned into a passable poem that is quite frequently wrong. Because Old English relies heavily on inflection to convey meaning, the word order is deceptive, even nonsensical to a modern English-speaker. My aid in this matter was the Joseph Wright Old English Grammar, an old but still very reliable work. While I do possess a hard copy, the Bosworth-Toller website links each word to the entry in which it appears in Wright's Grammar, displayed from the Internet Archive. After this, some refining was necessary, as certain terms are disputed.

Making choices for the glossary

The translation complete, I proceeded to tag each individual token (in this case words, though that is not always the smallest unit of text in a work) with the lemma, part-of-speech, gloss, and any extra information necessary. Part-of-speech can be quite simple or immensely complicated, depending on the goal of the project. It is not difficult to simply tag things as "noun", "verb", "adjective", etc., but this has limited usefulness in an inflected language like Old English. It is preferable to list whether the noun is, say, nominative and plural or if a verb is strong or weak. Beyond this, one can label verb stems from proto-Germanic, or add the gender of the nouns, adjectives, and other gendered parts of speech. In my edition my main goal was to provide a medium between the two extremes. Malone's tagging is extremely dense and thorough, essential for close linguistic study, but much of that tagging is for people interested in historical linguistics who already have a strong knowledge of the Germanic and proto-Indo-European language families.

After about 15 lines of word-tagging, I grew overwhelmed by the length of time this required, especially as I was skipping named entities so that I could return to them and add information about the person, place, or people-group to which the noun referred. Switching tack, I moved to focusing on them exclusively, a much more manageable project, and extremely interesting. The variety and interest of the named entities were the main reason I selected Widsið as my prototype of a digital Old English text. It was essential that some amount of information be included in the reading environment itself, enough to give a reader enough background about the named people, places, and groups that the poem had context and meaning, but not so much as to overwhelm the reader who wanted to focus on the poem itself. This involved summarizing considerable information from Chambers's edition which spends considerable time on the people and legends that form the backdrop of the lists of Widsið and Malone's glossary that, while less rich than Chambers, is still thorough and lengthy.

Up until this point I had been working with printed copies of both editions, but to finish the word tagging after the named entity tagging was complete, I used that most basic of digital text tools: OCR, or Optical Character Recognition. Working from a scanned version of Malone's Widsið, and aided by the Bosworth-Toller and Wright's Grammar, both also in digital form, I was able to use search functions to make much easier the hunt for each word, with numerous spelling quirks and the characters such as ð, þ, or æ that are not native to modern English. The ease with which I was able to finish the tagging once I was in a fully digital environment reinforced my appreciation for the flexibility and potential in a digital text edition.

Working with the digital text

What is a digital text?

What differentiates a digital edition from a print edition? This question at first has a simple, obvious answer: one exists in paper and ink, the other on a computer screen. While not factually untrue, this does not begin to do justice to the history and possibilities of either edition. A print edition has features that a digital edition should not pretend to, and a digital edition has possibilities that the constraints of print media could not attain.

A print edition is limited by the medium of the codex form. The vast ideas of mankind cannot be contained on a page-by-page system, yet for hundreds of years, this was the main and best way of containing and conveying information to as wide an audience as possible. Therefore, choices had to be made. An excellent example of this exists in the Widsið body of scholarship, in the differences between the Chambers and the Malone scholarly editions. Both of these editions display a deep knowledge of the language, the history, and the text. In dealing with the difficulties of displaying the poem in its original language while also annotating it for linguistic features, historical facts, and references to other scholarship and literary references, however, the space constraints demanded they make choices, and they made very different choices.

Chambers's text is characterized by footnotes on each page, often so extensive that the poem itself is limited to only four or even a single line on that page followed by a large block of notes. The benefits of this choice are clear: fluidity of reading is sacrificed for the sake of referencing convenience. Malone makes the opposite choice, printing the versions of his poem in a single chunk followed by notes on the text referring back to line numbers. This decision allows the poem to be seen at once and its brevity to be taken in as well as its overall structure, something that Malone makes much of. In study, however, it necessitates a constant flipping back of pages to compare the notes to the poem. Neither method is superior, they simply emphasize different goals of the editor and allow for different reading experiences.

A digital edition does not have page constraints. The poem can easily be displayed entirely on one page with no flipping or turn. Hyperlinks mean that one can switch from text to notes without hunting for line numbers. This is a feature that many ebook readers have used to great effect, while still maintaining similar sensations to that of print media. An ebook, however, chooses to be held to the constraints of a book with only minor convenience additions to provide a familiar experience to a reader. A truly digital edition does not need to bother with that sense of familiarity. As digital libraries have grown and evolved, they have been able to shake off the lingering habits of print media and embrace more fully the potential of the digital medium.

References and other digital libraries

There are thousands of fantastic digital libraries and reading environments in various states of development. For projects focused on the Medieval period, the Medieval Academy of America is an excellent resource with a variety of types of digital projects catalogued there. These projects can take an enormous variety of forms. The Early English Laws project, housed at the University of St. Andrews, for instance, worked with a relatively small number of documents in a limited scope, meaning that the project had a definite end date and has been completed. Electronic Beowulf, housed at the University of Kentucky, is also in a form of completion, where maintenance and changing web standards have necessitated updates, but the main goal of the project has been attained. Other projects, such as Menota (Medieval Nordic Text Archive) at the University of Copenhagen, are ongoing in perpetuity, with no limit on types of texts so long as they are Nordic in nature, either in Nordic languages or in Latin. Scholars can and do contribute to this project regularly. Many projects are not text-specific or are more manuscript focused. The Electronic Beowulf project is one of these that emphasizes aligning text with manuscript pages, but the concept was expanded by the Old English Poetry in Facsimile mentioned earlier, whose goal is to provide all Old English poetry in manuscript and facsimile edition. Finally there is the enormous and ongoing Perseus Scaife library, housing an enormous body of Greek and Latin texts, some with translation into Farsi. This library has existed for decades now and has grown to allow a variety of reading environments and connections between texts. Because of the ambition of this project, it has the potential to continue ad infinitum.

XML and tagging

How does a digital edition come into being, then? The answer to this question obviously varies greatly from text to text, but the process for this Widsið began with finding a few authoritative versions of the poem, as well as comparison with the manuscript. Scans of the manuscript are available at both the Old English Poetry in Facsimile project, as well as the University of Exeter Digital Humanities lab. It was not my goal to produce a facsimile edition from the manuscript, so I only used the scans for reference as I built my own edition and did not work directly from them for extensive periods. I also gathered Chambers's, Malone's, and the Anglo-Saxon Poetic Record's. From here I did a light markup of each in XML (eXtensible Markup Language) for ease of alignment and comparison.

XML is the foundation of most digital editions, as it allows the writer to add machine-readable tags and then write scripts in a programming language (mine is Python) to do large-scale work on the text that would be onerous to do manually. A sample below shows the markup on my own edition of Widsið.

Display choices and flexibility

While this edition does not, due to time constraints, display the wide capabilities of digital editions as much as I should like, the ability to experience the poem, with its modern alignment, as a single monolith, as well as the ability to read it with notes on each individual lines, does to an extent demonstrate the ability. Additionally, there exists a single glossary page and a single named-entities glossary page to recreate the experience of that section of a text at the back of the book. They can, however, be accessed by hyperlink now, and the search bar enables rapid locating of the desired word or entity. While a side-by-side display of text in two languages is not particularly novel, and has existed in physical book form for centuries, the line-by-line view is a fine-grained view of the poem that would be ridiculous in a book but is simple to render digitally. My particular rendition was inspired by the Open Epictetus view of the Enchiridion.

Code and HTML generation

The main home of this digital edition of Widsið is in a Github repository. The repository at github.com/Lorehouse/Widsið contains folders which hold Python scripts, notes, HTML (Hypertext Markup Language) files, CSS (Cascading Style Sheets) files, XML documents, and anything else required or used in the course of project creation.

The main markup documents are in the body of the repository, not in a directory, and can be easily reviewed on their own, but within the scripts folder are the various Python scripts that generate the HTML that is displayed on the website. While it takes many smaller scripts to generate each section of each page, the scripts necessary to generate, say, the named entity glossary, have been put into a master script that calls each function as necessary, starting from the base XML markup, and transforming it into the web page that is displayed. The importance of running the whole process every time is so that changes only need to happen at the root Old English or Modern English XML, and then pushed in their various forms through the pipeline to generate the HTML that the website shows.

This root-to-production pipeline also means that once another file, such as a different poem, is pointed to as the root file, it can also be pushed through the pipeline and displayed on the website with slightly different web addresses. In this way the project is easily extensible after initial mark-up is done. To demonstrate the possibility of this, I have created a page for Beowulf, and any number beyond this can be added as well.

With the addition of more texts in future, the possibilities also expand. Currently all the features of the site assume a single poem, and everything points internally. The glossary is exclusively lemmas from Widsið, and there is no linking between Widsið and any other texts yet. With additional texts, the overall glossary would expand, but the reader could select between the main glossary and the poem-specific glossary. But rather than extending this single site to become a source for exclusively Old English literature, it would be preferable to add it to another terrific library where existing infrastructure exists to link within the texts. Widsið would benefit from linking to various Latin and Greek texts such as Tacitus or Herodotus, giving greater primary-source texts to provide context for peoples and places.

The Future

In thinking about the possibilities for this digital edition, it was critical to look into the existing digital libraries, which are plentiful, to understand both what was possible and what was important. The key decision was from the beginning identifying the audience that might benefit most from this edition. One of the issues facing digital editions is that "it is possible that these changes have been driven less by the concrete use of new media than by assumptions about their potential" (Rasmussen). The ability to consult with Dr. Gregory Crane of Tufts University and creator of the Perseus Digital Library, one of most well-established and respected digital libraries in existence, was invaluable in discerning and prioritizing. Dr. Crane, in a meeting on 18 February, 2025, discussed the problems inherent in critical scholarly editions, namely that the commentary is so dense as to render the text almost a tertiary after-thought, when in fact what students, scholars, and enthusiasts most need is as accessible a text as possible. This means access to it in the original language, the modern language, and a computer-actionable format, such as XML. Including translation aids and using the most powerful digital tools available in the creation of texts allows us to offer a vast quantity of high-quality texts efficiently. In considering Rasmussen's three categories of reader roles, reader, user and co-worker, as applied to digital scholarly editions, I had to consider what sort of reader, user, or co-worker would join me on this endeavor.

The reader to whom I geared this project was the student or the hobbyist. This person would not expect to contribute in any way and would be here simply for the most practical, accessible version of the text -- a free, open-source, and reliable version that aided them in their language journey. This reader would get the greatest benefit from the simple side-by-side alignment and possibly the named-entities glossary, which would allow them to read the modern English text fluidly.

The user is somebody with a more philological purpose. They want to see how this text functions within itself and within the greater body of Old English poetry. For this user, the line-by-line view, with its stricter translation of individual words, its wooden translation, and the access to lemma and part-of-speech tags will be more useful. This along with the named-entity information would allow them to understand the historical societies and contexts included in the poem, and be a jumping-off point for their own further research.

The co-worker is the reader for whom I have done the least preparation, but for whom I have the greatest future hopes. This person would be an enthusiastic student of Old English and want to annotate and contribute their own texts and translations to the project. Because there is no graphic interface for the co-worker, everything would require a low-level knowledge of XML and possibly Python. The scripts I have written, however, are as broad as possible to allow for a variety of possibilities. Once the poem was properly tagged, it is hardly more than the push of a button to generate the HTML page that displays on the site. Every time the text was updated, it could easily be freshly generated and pushed to the site again, allowing for constant updates.

These three roles–reader, user, co-worker–are not discrete individuals but "they should rather be regarded as modalities" (Rasmussen). The greatest hope for this Widsið edition is that its initial accessibility to the reader would inspire greater curiosity, allowing the casual visitor to become a user who could reference the work repeatedly and gain greater understanding. The final hope for at least a few of these users would be to instill a desire to contribute and continue this work. There can be no doubt to the academically-minded that the greatest thrill of engaging with a text or an edition is when one can become part of its process. By making the transition from reader to user to co-worker simple and graspable, I hope to support budding scholars in their own growth and enthusiasm.

This leads us to the future of this project. What hopes for the future can be safely cherished? This edition is already prepared for incorporation in the Perseus Digital Library, and all the files will be available for anyone who wants to use this edition to build upon. Perseus has a long history of contribution by scholars, and my hope is that we can build within that existing structure rather than requiring me to create something new, incrementally, from scratch. Because Perseus already contains a wealth of Greek and Latin texts, many of which share references within Widsið, I hope that the linking of these references might be made simpler. Also, by working within such a strong existing framework, I don't create yet another site for scholars to visit; this is a single location where one can interact with a multitude of texts.

In the longer term, I hope that not only I, but other scholars who want to join me, can mark up more and more of the Old English Poetic Corpus and easily integrate them into Perseus. There is simply not enough lifetime for me to do this on my own, but slowly and surely I believe there is great potential for all of these texts, in a reliable scholarly form, to be collected in one place, and surrounded by the digital tools to allow richer close scrutiny and distant study.

Summary

Academic

The goal of this project has always been to allow literary, linguistic, historical and casual scholars to more easily access the text of Widsið in its original language. This project is a prototype of the possibilities in store. Language learning, one of the biggest barriers to studying historical texts in their original form, is made easier by providing in-line translations and part-of-speech markup. A student who has no Old English can appreciate the language but still understand the text.

Digital

The digital environment opens up possibilities that print media couldn't accommodate. The ability to draw connections between other documents and primary sources as well as academic or popular texts (Wikipedia, the most accessible source of information, is used as an example here) makes the text a central node from which other avenues of research become possible. Specialization is given greater depth by giving very specific information on the text, but generalizing is also given greater breadth by linking ideas available in other texts and on other sites.

Future

While Widsið itself is still a small text in a wide world, the extensibility of the project opens up the possibility to include not only other Old English texts, but texts in other languages. The power of digital form means that linking to other academic resources and research is easy. The text itself can be as central or peripheral as the particular research topic demands. In addition, the rise of generative “Artificial Intelligence” programs, such as ChatGPT, Claude, DeepSeek, Gemini, just to name a few, has opened up a whole new field of possibilities, not the least of which is the ability to translate and lemmatize texts with startling speed and accuracy. On the website for this thesis is included a table of all the words in Widsið, along with Claude 3.7 Sonnet’s analysis of the part-of-speech, word stem, and gloss. Very little correction would be needed to make this response exactly what was needed. This would enable speedy inclusion of a variety of Old English texts in great detail.

Thesis

Not only are all these possibilities made manifest by the concept of a digital edition, people at all levels of academia can find primary texts more accessible and enjoyable. A Tolkien scholar who wants to read Cynewulf's Crist so that they can witness the beauty of "Eala éarendel engla beorhtast" in the context in which Tolkien first encountered it, or a fabulous monster fan who wants to meet Grendel in Hrothgar's hall, or a humble but faithful person of faith who finds comfort in the story of Cædmon and wants to read the creation of the world as set out by the ancient poet can do so in an environment like this without any prior study. The Old Norse student who finds their favorite characters existing in an Old English world as well can easily navigate to those spaces and re-encounter these heroes and villains in a new light and a new cultural understanding.

Conclusion

A digital text is the seed of a digital library, a library unrestricted by the length of the shelves, the expense of color printing, the size of the page, or the amount of hands a person has on their body to hold all the books. While it in no way replaces the value of a physical (and not back-lit) volume, it allows for expanded study and access to the greater knowledge of other scholars. While sitting and reading contemplatively from a paper book will probably never go out of fashion, those who long for greater depth, more knowledge, and broader horizons can find that it is all within reach starting with a digital edition and blossoming outward to the wider world.

Works Cited