Uncategorized

By Jesse W. Torgerson

On behalf of the Comparing Chronicles Team: Churchill Couch, Zaray Dewan, Arla Hoxha, Diana Tran, Tess Usher

The Comparing Chronicles Project took some significant methodological and theoretical steps in 2024-25. The project began in 2023 as an investigation into (1) frameworks of historical time; and, (2) what historians could do with only partially accurate data. By Summer of 2024 the project had turned to nothing less than developing an alternative to the Historical-Critical Method of studying a text. The following narrates a stage on this journey, and a bit of how we got here.

The Comparing Chronicles project was invited by The Flow Project to participate in a panel at the July 2025 International Medieval Congress at Leeds University.

The Flow Project (led by Tobias Hodel at University of Bern and Silke Schwandt at Bielefeld Universität) purues “standardized digital workflows based on existing technology, making it easier for researchers to work with historical sources digitally.” This simple statement contains a significant advance in the Digital History (and Theory) landscape. “Digital workflows” are already a part of historical methods, but without concerted efforts to collaborate with each other scholars find themselves each inventing different but functional versions of the wheel (or, to give a more relevant example, parallel means of extracting machine readable text from handwritten sources).

Panel 544 Digital Data Flows promised difference means of processing medieval documents, and allowed us to present the work done in our Comparing Chronicles project as a unique example of what difference a standardized digital workflow could mean for comparative study of Early Medieval Chronicles from the example of three different versions of the Annales Fuldenses (Annals of Fulda).

The following is an abbreviated version of the remarks made at this panel.

Our Comparing Chronicles Project has been constructed using the web-based relational database tool Nodegoat. We have used the Nodegoat Go license subscription at Wesleyan University Library to create a fully collaborative research environment.

As our team builds this database together we are constantly re-thinking our processes (i.e., our Digital Data Flows) and as a result continually updating our methodology, which takes the form of a new structure to our relational database.

The most significant shift in the past year has been to shift from a TEI-based methodology to one that distinguishes different elements of the text and its structure as distinct but related datasets (or, in Nodegoat’s terms, “Objects”).

Each of the boxes in the above image indicates a distinct dataset. Chronicle is the title of the work (in the next stage of our project it will be a specific manuscript). The Chronicle Entry is the text under each year in that chronicle, in both Latin and (for reference) English, providing the text as it appears in each work (or, eventually: transcribed manuscript).

The next two datasets represent two levels of our distinctive analytical contribution to study of the text. Passage is the division of each annual entry’s text into distinct narrative sections, and the delineation of those narrative sections by the order (“Passage Number”) in which they occur in the annual entry. Thus the “name” of a passage might be AF 2 887 07, where AF 2 means “Annals of Fulda v.2,” 887 is the Chronicle Entry in which the passage occurs, and 07 denotes this as the seventh distinct narrative unit in the entry. We tag the text of the Passage for persons and places (the only use we make of TEI or text-tagging).

Event is where our analysis fully enters the picture. Here we give each Passage a label, in Latin (and translated into English for non-Latin-literate users). The Event name is based on the grammatical phrase which we have isolated as the focus of its narrative. The Event name uses the actual Latin of the text (whenever possible) to designate the central event of a passage. For instance, Celum Apertum is the event name for the 14^th passage under 887 in the third version of the Annals of Fulda (or, in our shorthand: AF 3 887 14):

Et mirum in modum, usque dum honorifice Augensi ecclesia sepelitur, celum apertum multis cernentibus visum est, ut aperte monstraretur, qui spretus terrenae dignitatis ab hominibus exuitur, Deo dignus caelestis patriae vernula mereretur feliciter haberi.

If it is helpful, we have also considered calling Event instead Episode. Finally, we give each Event one (or in extreme cases of ambiguity, two) Event Type labels.

The following extended paragraph explains our use of the Event Type (if you are not concerned about it, feel free to skip down), which is not in fact central to our analysis or project. The important point is that these are helpful to our analyses, but they are not what we are analyzing about the chronicle texts. They are imposed analytical categories, which is why we have made them a distinct part of the database (it is possible to study the text without using these types as an analytic). At the same time we have taken great care and gone through many different versions of these through extensive internal debates before settling on a list of eleven. We have found this list to be sufficient for capturing the different sorts of events which the Annals of Fulda uses to fill out its annual entries. These are: Campaign, Birth/Death, Office/Succession, Meeting, Assembly/Council  Travel/Embassy, Dispute, Celebration, Construction,  Commentary, Phenomenon (note, this is what we assigned to AF 3 887 14, above). It needs to be understood that the Event Type is also NOT the goal of the analysis. These distinctions allow us to identify potential similarities between texts or between entries. In the future they will allow us to make some overall statements about about entries (i.e., an entire year’s entry), or about trends or emphases in different texts as a whole. But ultimately they are simply an analytical tool to understand the text, rather than being core to our argument about the text. All of this follows a central tenet of our database structure: to keep our analytical work (in the Passage and Event and then Event Type datasets or objects) separate from the digital text itself (in the Chronicle and Chronicle Entry datasets).

Welcome back. The key to our analytical interests—and so our intervention into the text and the history of the text—is the combination of Passage and Event.

Our data creation and analysis thereof is based on the following premises. Chronicles (chronographic, chronological, and annalistic texts) are rarely attributed to a single author, and even when they are they are rarely the product of (a) a single moment and/or (b) a single person’s unique assemblage of information. Chronicles are much more often (1) collective, collaborative enterprises; (2) written over long periods, re-written, both, and more; and, (3) compiled from excerpted, rewritten, partially rewritten, and/or orally-transmitted pieces of information. As a result, what interests us as historians is not a textual-originalist (i.e., the Historical-Critical Method) approach which seeks to reproduce the text as it emanated from the quill of its author-scribe. Rather, what interests us is whatever happens to be the social conditions of the production of the text. This has led to our analytical interest in two things:

• any and all informational overlaps between extant versions (i.e., down to the variant manuscripts) of any chronicle text, since each of those overlaps indicate to us shared knowledge
• shared knowledge in turn indicates the social networks through which that knowledge was shared

Our relational database has shifted its structure over the past two years to allow us to get greater access to these two aspects of the evidence we find the text contains. What we have learned from observing the flow project is to identify where digital tools are and are not useful for this study. Thus far they are useful in two areas: (1) tagging persons and places in each Passage through Named Entity Recognition (nodegoat’s built-in internal process for this is “reconciliation”); and, (2) analysis and visualization of the “informational overlaps” and “social networks” noted above. In the near future we will add (3) Handwritten Text Recognition to our workflow when we turn to using the actual surviving manuscripts for our Latin text rather than the printed critical editions (which we continue to use as we build our database model). Identifying our unique ‘digital flow’ emphasizes how much our work remains tied (by the necessity of our own standards of exactitude) to careful reading practices, double- and triple-checking all of our data as we proceed.

At the time of the IMC Leeds presentation we were able to offer the following.

Having conducted an initial analysis of three different versions of the Annals of Fulda (AF 1, AF 2, AF 3 in our representation), we were able to display the overlap of the entries in each of them (i.e., where the entire text for a year was the same) in the following image.

AF 1 shares all of its entries with both AF2 or AF3, which in turn share a number of their entries with each other but then possess their own distinct entries. This is simply a visualization of the text history as manifest in the MGH critical edition.

What our Comparing Chronicles methodology does to the text is visible in the next image. This is a comparison between the entries for 882 in AF3 (to the left) and AF2 (to the right), which are each represented by the large blue nodes in the bottom corners.

The orange nodes clustered in descending numerical order on the left and right sides each indicate a discreet Event (i.e., narrative episode) in each text. We have activated the persons (bottom, pink) and the Event Types (top, orange) but de-activated the places (blue, floating in the middle) of each version to make it possible to see the encoded relations.

The key element of interest, above, for our purposes are the six Events (purple) where the visualization displays the labels (in English here, rather than in Latin) of the events which we noted as the same event. That is, these five (“Louis III died,” “Comet Appeared,” “Army Returned,” “Northmen burnt Koblenz,” Northmen burnt Trier,” and “Bishop Wala attacked Northmen”) we read as the same core episode even in the text describing or narrating each was different in each text. In numerical terms, AH2 has 18 distinct Events, AH 3 has 20 distinct events, and they share 6.

According to the Historical Critical Method, these are completely divergent textual traditions. According to our Comparative Method, this single year’s entry in each text possesses evidence of a 23-25% overlap in the knowledge network between these two textual communities, as recorded in their respective annals. This is already an exciting an promising result, just from this small test case.

In this upcoming semester we will be applying our most updated methodology to an encoding of a combined Latin-English database of the three versions of the Annals of Fulda, while also applying this same methodology to the Annals of St. Bertin. This work will give us our final methodological prototype, making it possible to visualize what the evidence from a surviving text of a single knowledge network looks like—even when there are three versions of the text spread out amongst the sister monasteries of the monastic house of Fulda—as compared to a distinct knowledge network, that of St. Bertin.

However, our final visualization gives a hint of what is in store. This is a detail from the social network of the Entries and Events in both the Annals of Fulda and the Annales Bertiniani. While these networks are clearly distinct, the Events (in purple) which hang like spiderwebs between the two clouds indicate events which are the same episode. Our methodology already allows us to see that even though these are distinct textual traditions, there are meaningful links in the historical knowledge which they shared and recorded in common.

Annales Fuldenses, Humanist Library of Sélestat MS 11 (Year 855)

By Kathryn L. Jasper

Since its inception the Valle Gianni Field School (or Northwest Bolsena Archaeological Project – NBAP) has found new ways to collaborate with undergraduate students in the archaeological excavation of an imperial Roman villa. In the past three years that collaboration has included Wesleyan students trained in the Travelers’ Lab. The Valle Gianni excavation expands the historical narrative of central Italy during and after the Roman period by focusing on a region that has received little attention from archaeologists and historians. The published report of the initial seasons is available from Fasti Online. The excavation is funded by an ARCS research grant from Illinois State University.

What is the Valle Gianni Excavation?

The archaeological site of Valle Gianni is currently defined by two primary features—the partially exposed remains of a monumental fountain from the Roman imperial period, and some recessed vats associated with wine production. When working on site students experience every part of the excavation process (survey, mapping, excavation, ceramic analysis, etc.). Students thereby actively contribute to recovering the now-lost relationship between the landscape of this region and the centuries of human actors who lived, farmed, and built residences there.

Direct Student Involvement

One of the most exciting and innovative aspects of this project is the direct involvement of undergraduate students in not only the excavation work but in data creation and analysis. The interdisciplinary research-teaching model of the Valle Gianni Field School unites humanities and STEM fields by bringing together the diverse expertise of faculty instructors from four different departments. For students this means an opportunity to not just think about or see the advantages and challenges of interdisciplinary research, but to actually participate in them in real time.

Participating students have the chance to play a role in research outputs, either as co-authors or independently. The Valle Gianni Field School seeks in particular to encourage undergraduates who might not otherwise consider pursuing graduate-level research in developing research expertise and experience. Ultimately, our approach offers a new model for mentorship at Illinois State University, and beyond, through the participation of other schools in the project such as Travelers’ Lab research students from Wesleyan University.

The Role of Wesleyan University Traveler’s Lab Students

The Valle Gianni project requires a database platform which is at once capable of storing diverse types of data while also being capable of visualizing that data in specific ways for analysis. This is a complicated undertaking because of the interdisciplinary nature of the venture, and the resulting range of media and data types which the project generates and accumulates.

To this end, we adopted Nodegoat (see: www.nodegoat.net) to manage our data. Nodegoat gives scholars the freedom to generate and analyze datasets based on their own specific, bespoke data models. Moreover, Nodegoat allows relational modes of analysis, which combine both spatial and chronological data. These features make it ideal as a platform for archaeological data.

Wesleyan students who trained in the Travelers’ Lab have pioneered the integration of Nodegoat with the Valle Gianni project goals. In 2023 two students—Sarah Brown and Sofia Gallegos—worked together to develop the first relational database model for the Valle Gianni data. Then the following year (in summer 2024) Sarah Brown traveled to Valle Gianni and participated in the Summer Field School’s excavation in order to implement that database live for the first time into the excavation workflow, and to incorporate the new data generated that year.

Most recently, in summer 2025 another Travelers’ Lab student from Wesleyan—Alex Williams—spent four weeks on site. Alex developed an advanced understanding of the archaeological process and how data collection functions on the ground.

Advancing Valle Gianni with Relational Databases in Nodegoat

Specifically, in 2025 Alex Williams addressed herself to integrating the two major types of data which the project generates: data describing physical objects, and digital-born data. Digital-born data include geospatial data (vector and raster), digitized photos—such as photos pertaining to photogrammetry used to produce 3-D digital models—and digitized copies of analogue records generated during the survey, excavation, and subsequent analyses (e.g., site notebooks and excavation diaries), satellite images, and semi-automated data recorded in already-structured machine-readable formats (e.g., spreadsheets and databases).

Other found physical objects include stone, tesserae, ceramics, tile, metal objects, slag and metal derivatives, but also soil samples, botanical remains, faunal remains (including osteological remains and shells), and glass. See some representative excavation photographs below.

Alex Williams’ careful attention resulted in successfully reconciling manual recording practices with the semi-automated generation of digital databases and in this tangible way her work significantly improved the Valle Gianni workflow. The contribution of the Travelers’ Lab students to the Valle Gianni data model can be seen in the comparison, schematized below, of the initial data model through 2024 (black outline) and the comprehensively redesigned, rationalized, and simplified model of 2025 (blue outline) as compiled by Alex Williams.

We regard the accessibility of our data as a critical component of our project. The new developments and components of our workflow have added new facets to the curriculum. These ensure that our data is available, accessible, and as easy to use and understandable (“readable”) as possible. Nodegoat actively helps make this possible. The platform will allow us to transfer data to a public-facing website with intuitive user functions.

The development of our work and processes continues. We continue to refine our databases and their relationality, which in turn opens up new opportunities for new student researchers as well as new types of student research: database design is now a part of the program curriculum.

We invite any students interested in helping with this aspect of the project (data management and web design) to apply and inquire about participating in the project, whether during the summer field school or during the academic year. And, of course: all aspiring archaeologists (to right: ISU student Hannah Torkelson) are always welcome to come and help us move dirt around the Italian countryside!

By Diana Tran

In May 2024, we started with Daniel Feldman’s and Arla Hoxha’s event types as a basis to categorize events into ten different types, as shown below:

•  Mission (External)
• Petition (Internal)
• Campaign
• Phenomenon
• Birth/Death
• Office (previously titled Succession)
• Dispute
• Celebration
• Commentary
• Construction

In Summer 2025, Churchill Couch and Tess Usher worked with Principal Investigator Jesse Torgerson to update the event types. From the original ten event types, they have added an eleventh and updated certain event types to ensure higher accuracy. The new event type categorizations are as follows:

1. Travel/Embassy, replacing Mission’s purpose
2. Petition being split into Meeting and Assembly/Council

According to Professor Jesse Torgerson, here are the definitions of  the new event types:

1. Travel / Embassy: Ultimately this is about tracking the movement and exchanges between those who run political collectives – kings, popes, emperors, etc.. “Travel” here is only being noted not because there is movement implied in the action, but because the action IS movement. E.g., the King going to Worms is in and of itself an event worth noting. Berengar (e.g.) going to Worms is not an event in and of itself (rare exception: a nobleman such as Berengar is ACTING like a King, i.e., putting together a rebellion), but rather will be part of another action such as a Campaign
2. Meeting: Here the event action is characterized by subjects of the text getting together for the purpose of exchanging something, or making some kind of communication. Examples include people bringing a petition to the King, or the king asking a nobleman to come to court for something not particularly negative. One issue to pursue / keep thinking about. Currently, when one side in a campaign comes to meet with the other side and negotiate something (e.g., terms of surrender) we have tagged this as a Meeting.

Meeting is meant to encapsulate all meetings within the chronicles. These meetings don’t necessarily need to be antagonistic, but need to have two or more Persons involved. Professor Torgerson does acknowledge a possible conflict in interest in the example he gave for Meeting could be counted as Embassy because of the meeting between two warring factions (political collectives) as a result of a campaign

Lastly, 

1. Assembly/Council: The event action is defined typically by the word ‘assembly’ or ‘council’ within the event itself. It is typically an official meeting organized by an official, such as a king, emperor, or a papal leader. 

Petition was deleted in favor of Assembly/Council and Meeting for the sake of more specificity in our categorization. The move to create Assembly/Council is a callback to Daniel Feldman’s tag general assembly. This time, we’ve broadened the scope of this tag to include council meetings as well (based on context) and the specific keyword ‘assembly’ is a common indicator to use this tag. 

A change we’ve made is that one passage can be tagged with two event types by splitting the passage into two events and tagging both events with the relevant event type. 

Thus, the new event types are:

1. Birth/Death
2. Office/Succession
3. Celebration
4. Construction
5. Travel/Embassy
6. Meeting
7. Assembly/Council
8. Campaign
9. Phenomenon
10. Commentary
11. Dispute

We are hopeful that this will be our final iteration of the event type categorizations, however, if major issues appear as we continue to utilize this methodology in Fall 2025, we will revisit it.

By Diana Tran

A History:

In Spring 2024, we had written up a methodology to show newcomers to the Traveler’s Lab how to tag in Nodegoat. Under this methodology, the tagging would occur in the body of the chronicle entry. The body of the chronicle entry would be fully tagged like so:

The tagger would go through the entire entry and tag the People and Places first before beginning to tag the events, one by one in the chronicle entry. With this methodology, problems would come up when an event was too long to tag and/or was prone to mistakes, especially as the tagger would be looking through thousands of words to tag. Context was lost during long rounds of tagging, so an ‘emperor’ would not be tagged as the correct one, which needed to be rectified by a second round of tagging. We regarded these two problems as major time sinks. After meeting with the NodeGoat developers, they suggested we cease tagging in the body of chronicle entries and instead utilize the cross-reference function more often. 

A Passage: 

Rather than tagging in the chronicle entry, the events would be split into Passages:

Then the individual passage would be tagged, Person/Place, and then Event. The Passage would be numbered  by sequence, and the naming pattern would be the chronicle entry followed by what passage number it was for said chronicle entry. Shown below is what a complete, tagged passage would look like:

and the accompanying chronicle entry:

New System:

Over the summer, Churchill Couch, Tess Usher, and Principal Investigator Jesse Torgerson developed a new methodology of tagging. 

The tagger is instructed to read through the chronicle entry and manually split up the chronicle into passages (preferably on a separate document website) like so:

Once this is done, the Latin taggers can add the Latin version and re-input the entire text file back into the chronicle entry, like so:

The passage would be created with both the English and Latin versions in one passage:

Persons and Places are tagged within the passage. However, there is an ongoing discussion as to whether to tag the Latin or the English version, or both. We won’t be tagging the event within the passage now. The “Date Start” is actually not a chronological thing. The ‘year’ is the chronicle entry year, the ‘month’ is the version of the annual [i.e. AF1 = 01, AF2 = 02], and the ‘day’ is the numbering of the passage [Passage 8 = 08]. Something to note in the titling of the passages should be [0X] for any number that is less than 10 due to a coding/decimal issue that messes up the order. 

The event is going to only be written in Latin and the passage will be linked to the event via the ‘subobject’ within Event as can be seen here:

The event will take the event type. Multiple events can take the same passage, as can be seen in here:

The Latin terminology has different syntax and sentence structure than in English, so what would be one event type in the English translation could be two or three within the Latin syntax.

 

By Jesse W. Torgerson

On behalf of the Comparing Chronicles Team: Churchill Couch, Zaray Dewan, Arla Hoxha, Diana Tran, Tess Usher

 

The Comparing Chronicles Project took some significant methodological and theoretical steps in 2024-25. The project began in 2023 as an investigation into (1) frameworks of historical time, and (2) what historians could do with only partially accurate data. By the summer of 2024 the project had turned to nothing less than developing an alternative to the Historical-Critical Method of studying a text. The following narrates a stage on this journey, and a bit of how we got here.

The Comparing Chronicles project was invited by The Flow Project to participate in a panel at the July 2025 International Medieval Congress at Leeds University.

The Flow Project (led by Tobias Hodel at University of Bern and Silke Schwandt at Bielefeld Universität) purues “standardized digital workflows based on existing technology, making it easier for researchers to work with historical sources digitally.” This simple statement contains a significant advance in the Digital History (and Theory) landscape. “Digital workflows” are already a part of historical methods, but without concerted efforts to collaborate with each other scholars find themselves each inventing different but functional versions of the wheel (or, to give a more relevant example, parallel means of extracting machine readable text from handwritten sources).

Panel 544 Digital Data Flows promised difference means of processing medieval documents, and allowed us to present the work done in our Comparing Chronicles project as a unique example of what difference a standardized digital workflow could mean for comparative study of Early Medieval Chronicles from the example of three different versions of the Annales Fuldenses (Annals of Fulda).

The following is an abbreviated version of the remarks made at this panel:

Our Comparing Chronicles Project has been constructed using the web-based relational database tool Nodegoat. We have used the Nodegoat Go license subscription at Wesleyan University Library to create a fully collaborative research environment.

As our team builds this database together we are constantly re-thinking our processes (i.e., our Digital Data Flows) and as a result continually updating our methodology, which takes the form of a new structure to our relational database.

The most significant shift in the past year has been to shift from a TEI-based methodology to one that distinguishes different elements of the text and its structure as distinct but related datasets (or, in Nodegoat’s terms, “Objects”).

Each of the boxes in the above image indicates a distinct dataset. Chronicle is the title of the work (in the next stage of our project it will be a specific manuscript). The Chronicle Entry is the text under each year in that chronicle, in both Latin and (for reference) English, providing the text as it appears in each work (or, eventually: transcribed manuscript).

The next two datasets represent two levels of our distinctive analytical contribution to study of the text. Passage is the division of each annual entry’s text into distinct narrative sections, and the delineation of those narrative sections by the order (“Passage Number”) in which they occur in the annual entry. Thus the “name” of a passage might be AF 2 887 07, where AF 2 means “Annals of Fulda v.2,” 887 is the Chronicle Entry in which the passage occurs, and 07 denotes this as the seventh distinct narrative unit in the entry. We tag the text of the Passage for persons and places (the only use we make of TEI or text-tagging).

Event is where our analysis fully enters the picture. Here we give each Passage a label, in Latin (and translated into English for non-Latin-literate users). The Event name is based on the grammatical phrase which we have isolated as the focus of its narrative. The Event name uses the actual Latin of the text (whenever possible) to designate the central event of a passage. For instance, Celum Apertum is the event name for the 14^th passage under 887 in the third version of the Annals of Fulda (or, in our shorthand: AF 3 887 14):

Et mirum in modum, usque dum honorifice Augensi ecclesia sepelitur, celum apertum multis cernentibus visum est, ut aperte monstraretur, qui spretus terrenae dignitatis ab hominibus exuitur, Deo dignus caelestis patriae vernula mereretur feliciter haberi.

If it is helpful, we have also considered calling Event instead Episode. Finally, we give each Event one (or in extreme cases of ambiguity, two) Event Type labels.

The following extended paragraph explains our use of the Event Type (if you are not concerned about it, feel free to skip down), which is not in fact central to our analysis or project. The important point is that these are helpful to our analyses, but they are not what we are analyzing about the chronicle texts. They are imposed analytical categories, which is why we have made them a distinct part of the database (it is possible to study the text without using these types as an analytic). At the same time we have taken great care and gone through many different versions of these through extensive internal debates before settling on a list of eleven. We have found this list to be sufficient for capturing the different sorts of events which the Annals of Fulda uses to fill out its annual entries. These are: Campaign, Birth/Death, Office/Succession, Meeting, Assembly/Council  Travel/Embassy, Dispute, Celebration, Construction,  Commentary, Phenomenon (note, this is what we assigned to AF 3 887 14, above). It needs to be understood that the Event Type is also NOT the goal of the analysis. These distinctions allow us to identify potential similarities between texts or between entries. In the future they will allow us to make some overall statements about about entries (i.e., an entire year’s entry), or about trends or emphases in different texts as a whole. But ultimately they are simply an analytical tool to understand the text, rather than being core to our argument about the text. All of this follows a central tenet of our database structure: to keep our analytical work (in the Passage and Event and then Event Type datasets or objects) separate from the digital text itself (in the Chronicle and Chronicle Entry datasets).

Welcome back. The key to our analytical interests—and so our intervention into the text and the history of the text—is the combination of Passage and Event.

Our data creation and analysis thereof is based on the following premises. Chronicles (chronographic, chronological, and annalistic texts) are rarely attributed to a single author, and even when they are they are rarely the product of (a) a single moment and/or (b) a single person’s unique assemblage of information. Chronicles are much more often (1) collective, collaborative enterprises; (2) written over long periods, re-written, both, and more; and, (3) compiled from excerpted, rewritten, partially rewritten, and/or orally-transmitted pieces of information. As a result, what interests us as historians is not a textual-originalist (i.e., the Historical-Critical Method) approach which seeks to reproduce the text as it emanated from the quill of its author-scribe. Rather, what interests us is whatever happens to be the social conditions of the production of the text. This has led to our analytical interest in two things:

1. Any and all informational overlaps between extant versions (i.e., down to the variant manuscripts) of any chronicle text, since each of those overlaps indicate to us shared knowledge
2. Shared knowledge in turn indicates the social networks through which that knowledge was shared

Our relational database has shifted its structure over the past two years to allow us to get greater access to these two aspects of the evidence we find the text contains. What we have learned from observing the flow project is to identify where digital tools are and are not useful for this study. Thus far they are useful in two areas: (1) tagging persons and places in each Passage through Named Entity Recognition (nodegoat’s built-in internal process for this is “reconciliation”); and, (2) analysis and visualization of the “informational overlaps” and “social networks” noted above. In the near future we will add (3) Handwritten Text Recognition to our workflow when we turn to using the actual surviving manuscripts for our Latin text rather than the printed critical editions (which we continue to use as we build our database model). Identifying our unique ‘digital flow’ emphasizes how much our work remains tied (by the necessity of our own standards of exactitude) to careful reading practices, double- and triple-checking all of our data as we proceed.

At the time of the IMC Leeds presentation we were able to offer the following.

Having conducted an initial analysis of three different versions of the Annals of Fulda (AF 1, AF 2, AF 3 in our representation), we were able to display the overlap of the entries in each of them (i.e., where the entire text for a year was the same) in the following image.

AF 1 shares all of its entries with both AF2 or AF3, which in turn share a number of their entries with each other but then possess their own distinct entries. This is simply a visualization of the text history as manifest in the MGH critical edition.

What our Comparing Chronicles methodology does to the text is visible in the next image. This is a comparison between the entries for 882 in AF3 (to the left) and AF2 (to the right), which are each represented by the large blue nodes in the bottom corners.

The orange nodes clustered in descending numerical order on the left and right sides each indicate a discreet Event (i.e., narrative episode) in each text. We have activated the persons (bottom, pink) and the Event Types (top, orange) but de-activated the places (blue, floating in the middle) of each version to make it possible to see the encoded relations.

The key element of interest, above, for our purposes are the six Events (purple) where the visualization displays the labels (in English here, rather than in Latin) of the events which we noted as the same event. That is, these five (“Louis III died,” “Comet Appeared,” “Army Returned,” “Northmen burnt Koblenz,” Northmen burnt Trier,” and “Bishop Wala attacked Northmen”) we read as the same core episode even in the text describing or narrating each was different in each text. In numerical terms, AH2 has 18 distinct Events, AH 3 has 20 distinct events, and they share 6.

According to the Historical Critical Method, these are completely divergent textual traditions. According to our Comparative Method, this single year’s entry in each text possesses evidence of a 23-25% overlap in the knowledge network between these two textual communities, as recorded in their respective annals. This is already an exciting and promising result, just from this small test case.

In the upcoming semester we will be applying our most updated methodology to an encoding of a combined Latin-English database of the three versions of the Annals of Fulda, while also applying this same methodology to the Annals of St. Bertin. This work will give us our final methodological prototype, making it possible to visualize what the evidence from a surviving text of a single knowledge network looks like—even when there are three versions of the text spread out amongst the sister monasteries of the monastic house of Fulda—as compared to a distinct knowledge network, that of St. Bertin.

However, our final visualization gives a hint of what is in store. This is a detail from the social network of the Entries and Events in both the Annals of Fulda and the Annales Bertiniani. While these networks are clearly distinct, the Events (in purple) which hang like spiderwebs between the two clouds indicate events which are the same episode. Our methodology already allows us to see that even though these are distinct textual traditions, there are meaningful links in the historical knowledge which they shared and recorded in common.

By Arla Hoxha

During summer 2024, Lab Manager Arla Hoxha ’25 continued developing Traveler’s Lab Comparing Chronicles Project through a QAC (Quantitative Analysis Center) Summer Fellowship. Throughout the summer, she experimented with different statistical methods and software and explored how new methods might be utilized to better compare different manuscripts from the Annals of Fulda. The summer research process culminated in a poster presentation session summarizing the progress of Chronicles up to that point, as well as visual representations of our work.

The presentation started off by giving a definition of the event unit for a general audience unfamiliar with our work. In recent years, there has been an effort in the field to shift the focus from the chronological development of historical texts and towards the development of the narrative. Events do not always show up in a narrative as they occur chronologically; they are compiled by an author who chooses what events to include and how to arrange them. Studying chronicles using ‘years’ as units reveals little about the chronicle and even less about its author as it is too broad and unable to capture the nuance of meaning in the language and the way text is ordered. The intervention of the Chronicles Project so far is the alternative unit of observation; the event defined as a string, spanning from a few words to paragraphs, with a central theme (event types), consistent named tag entities (characters and setting), terminating with a change of temporal identifiers or agents in the narrative. We expand more on these definitions in our previous methodologies.

Using a narrative-focused approach in the study of chronicles and the unit of the ‘event’ we explore chronicle entries from the Annals of Fulda—the text which continues to be our main focus—and determine how events differ on the manuscript level. The focus of the summer research was reducing events to their differentiating components and using different text analysis tools to compare these components across events to determine event similarity. We experimented with Python text mining libraries such as spaCy to filter the entries. spaCy is an open-source library used for Natural Language Processing (NLP) in Python. The program developed takes a csv file with the event titles, years and manuscripts. It then splits the title into its components, then using spaCy’s tokenization feature it tags each component into parts of speech and then tries to match entries based on named entities, verbs, etc. This method of filtering can also easily be accomplished through Nodegoat. The idea behind using event titles for the comparison is that they are supposed to capture the event and utilize specific verbs from the event that are representative.

An interesting function spaCy allows is comparing words and giving their similarity through a percentage (cosine similarity model), which can be utilized to compare two similar verbs used in an event title—this could be used to match events that are potentially the same although they have been differently labeled, forgoing the issue of human error. ‘Northmen attacked’ and ‘Northmen plundered’ do not have the same label, nor are their passages textually the same. But we can filter events by year and look for different manuscripts, then differentiate between events based on their type categorizations; checking for cosine similarity above a certain value can help us determine whether ‘plundered’ and ‘attacked’ have a similar meaning and therefore ‘Northmen plundered’ and ‘Northmen attacked’ would be understood as referring to the same event. However, this functionality comes with its own problems; it is limited by the library’s vocabulary and its efficiency is undermined by lack of accuracy. Moreover, even though the verbs used in the titles are important this method overemphasizes the way events are titled over other elements. The accuracy can be increased by training a module using data that is specific to our project. For now, we can go through the process of determining if two verbs are similar enough using a human reader, which is a slower but more accurate process. Although this is beyond the scope of our project as of now, it might be interesting in the future to train a module that is specific to the Chronicles Project, which could prove useful in automatizing part of the process of detecting the same event.

Returning to the main topic of comparing events, two events are the same as long as they speak of the same event; that is the same event type (see Diana Tran’s Event Type methodology here), with the same named entities. Whether they are textually the same is less important. We found events that spoke of the same occurrence by filtering (through the method described above as well as Nodegoat filters) for events with the same title happening in the same year but different manuscripts. An instance of this is ‘Pope Hadrian dies’ in year entry 885 for both manuscripts 2 and 3 in Annals of Fulda:

The passage differs between manuscripts, but the central idea captured in the title remains: ‘Pope Hadrian died.’ Despite the first event including more details surrounding the death of Pope Hadrian and the difference in length of the two event entries, both refer to the same event. Beyond the categorization under the same title, the events in both manuscripts are cataloged under event type ‘Birth/Death’ and both have ‘Pope Hadrian’ listed as a principal actor. The way the event is named is helpful in communicating the ideas of the passage and helping us identify them correctly. Here we see how ‘Even Types’ can be a powerful tool in determining event similarity as well as understanding distribution of events across time. We believe that the categorization of events by event types will increase the accuracy of determining ‘same’ events.

Of course, over relying on Event Types presents the caveat of human error that is built into this categorization. We observe that chronicle year entries with largely the same events have different distributions of event type tags:

The data in use is the cross-referenced chronicle transcript from the different manuscripts of the Annals of Fulda stored through the Nodegoat environment. The parsing of the data and visuals were completed using Google Sheets for data management and plotting libraries in R.

What is the importance of determining events that are the same across manuscripts and chronicles? By pointing out the similarities between events we are able to discern their differences as well and start asking questions about authorship and the context in which different manuscripts emerged. The same methodology we have been using on events from different manuscripts of Fulda is to be applied to events that pertain to different but overlapping chronicles in the future.

Through the work done this summer, determining event similarities we found overlap in a few events from different manuscripts. We attempted to tag text from different manuscripts describing the same event under the same event tag, so we tried to list the passage of events from different manuscripts under the same event object entry. Perhaps the most important result from our work this summer was noticing the issues with our old model of event categorization. Our old model required the passage to be written out on the event description; in the case of two events we would have to do so for both passages and so on depending on the number of manuscripts that describe the same event. Our future goal is to be able to cross reference the English text with the original Latin, this would require us to list another passage under the event entry. We realized that this methodology would become increasingly more insupportable. Moreover, the work comparing chronicles in this way made apparent certain redundancies in our data. Most notably, named entities appeared twice; as tags in the chronicle object and cross-referenced in the object descriptors of events.

To reduce the redundancy in our data and to make the process more efficient moving forward, we implemented a mass restructuring. The new objects are as follows:

Chronicle Entry is now the object that contains passages. Passage is the object where named entities are tagged—previously this was done in Chronicle Entry. Now events are linked to the chronicle entry through the Passage object. The passage is cross referenced in the object description of the event, but the event is not tagged in Passage as to reduce redundancy. Named entities on the other hand appear as tags in Passage, but they are not cross referenced in Event, because Passage already contains them, and Passage is cross listed in Event.

Here is an example of the Passage object: ‘in entry’ is the link to the Chronicle entry, which in turn is linked to the object ‘Chronicle’ which stores the references to the source texts, i.e.: the English translations of Annals of Fulda. The text is the passage, as well as the text of the event in question. Passage number shows where the passage/event occurs in relation to the other events in that same year, emphasizing the progression of narrative over chronological progression, one of our main goals that we have previously had issues representing in Nodegoat.

The event no longer has description tags for ‘Places’ and ‘Person’ because those are tagged in Passage and can be cross-referenced through it. This also makes the work of referencing multiple passages under the same Event much easier; instead of writing down multiple events in the description of the Event object, they can be cross-referenced. This will also aid us in the eventual transition to Latin. The only elements directly listed under Event are the event title, Chronicle entry, and in the sub object, the Event Type.

Other smaller but not insignificant improvements to the model were cleaning up the data of the Person and Places objects where each entry was filled out with relevant information that was previously missing, such as birth and death days for Person and coordinates for Places. In the process is also the development of types of locations to categorize Places.