Towards a Relational Model of the Holocaust Geolocalized Approaches to Perpetration, Complicity, and Knowledge

The Map That Connects the Dots

Imagine you are trying to solve a massive, global mystery, but the clues are scattered across thousands of different boxes in a dusty attic. Some clues are train tickets, some are diary entries, some are official government lists, and others are stories told by survivors decades later. For a long time, historians have been great at reading the individual stories—like a detective focusing on one specific person's diary to understand their fear or their courage. But sometimes, looking at just one story makes it hard to see the whole picture of how the mystery was organized. It's like trying to understand a massive traffic jam by only looking at one car; you know that car is stuck, but you don't know why the whole city grid is gridlocked.

This paper sits at the intersection of history and digital technology, specifically a field called "Historical Network Analysis." Think of this field as a super-powered way to connect the dots. Instead of just reading a list of names, this method builds a giant, interactive web where every person, place, train station, and document is a "node" (a dot), and the relationships between them are "edges" (lines connecting the dots). The paper asks a big question: Can we use these digital webs to map out not just what happened during the Holocaust, but where it happened, who was nearby, and what information was floating around in the air at the time? It's about moving from a storybook view of history to a 3D map view, helping us see patterns of movement, knowledge, and responsibility that are invisible when we only read one document at a time.

The Paper's Big Idea: Building a Digital Web of History

Fabian Schmidt, the author of this paper, proposes a new way to study the Holocaust using these digital webs. He suggests that while personal stories are essential for understanding individual pain, they aren't the best tools for seeing the huge, invisible structures that connected perpetrators, victims, bystanders, and institutions across time and space. To fix this, the paper argues we need to build "geolocalized historical networks."

Imagine taking all the scattered clues—the train schedules, the camp transfer lists, the factory records, and the survivor memories—and feeding them into a computer model. This model doesn't just list them; it connects them. It draws lines between a specific train and the town it passed through, or between a factory and the workers it employed. By doing this, the model can reveal patterns that are too big for a human brain to spot in a single book. For instance, the author shares a personal moment where he realized a deportation train, famous for carrying a young girl named Settela Steinbach, passed less than a kilometer (about 0.6 miles) from his own office in Berlin. While the train's route was already written down in history books, connecting it to a map of his neighborhood suddenly made the event feel real and immediate. It raised new questions: How many other trains passed through that neighborhood? How often did local people see them?

The paper suggests that by building these networks, we can create "knowledge maps." These maps don't tell us exactly what a specific person knew on a specific day (because that's often impossible to prove). Instead, they show us the environment of information. They can show us which towns were right next to forced labor camps, or which train stations had long delays where civilians might have waited and watched. It's like creating a weather map for information: we can't see the wind, but we can see the pressure systems that tell us where the storm was likely to be felt.

What the Paper Actually Does (and Doesn't Do)

It is important to understand what this paper is not claiming. The author is very clear that this method is not a magic wand that will replace the need for reading survivor testimonies or traditional history books. In fact, the paper argues against the idea that we can use these models to definitively say, "This person knew everything" or "This person knew nothing." The model cannot read minds.

Instead, the paper suggests that these networks are a tool to complement existing history. It proposes that by linking fragmented data—like combining a list of prisoners with a map of railway lines and a diary entry from a local resident—we can reconstruct the "relational contexts" of the Holocaust. The paper uses examples from other fields to show how this works. It points to a project called "Sites of Shame," which mapped the deportation routes of Japanese Americans during World War II. That map showed a clear pattern: most people were moved from California to camps in the Midwest, but after the war, they didn't go back to California; they moved to the East Coast. This big-picture pattern was hidden in thousands of individual stories but became obvious when visualized as a network. Similarly, the paper looks at maps of the transatlantic slave trade, which revealed that many enslaved people were moved internally between plantations, a detail often missed in general histories.

The paper explicitly rules out the idea that these models should be used to simplify or "aestheticize" the horror of the Holocaust. The Nazis themselves used maps and diagrams to make their genocide look like a clean, efficient logistical operation. This paper argues that our models must do the opposite: they must reconnect the broken, messy, human pieces of the story to show the true complexity and the human cost.

The Method: From Paper to Pixels

To build these models, the paper outlines a method called Entity-Relationship Modelling. Think of this as a digital Lego set. The "bricks" are the entities: people, places, organizations, and events. The "connectors" are the relationships: "was deported from," "worked at," "lived near," or "saw."

The author notes that while we have a lot of data, it is currently scattered. Some is in government archives, some in museum databases, and some in old books. The paper suggests using technology like Natural Language Processing (NLP)—which is basically a computer reading text to find names and dates—to pull this information together. For example, a computer could scan a diary and find a mention of a train passing by, then link that to a transport schedule to see exactly when and where it happened.

However, the paper is careful to warn that this isn't perfect. The data is incomplete, and some stories might be exaggerated or false. The author suggests that we need to be very careful about what we include in the model. We should focus on events that are anchored to specific times and places, rather than vague rumors. The paper also discusses the ethical side of this work. Since the data involves real people who suffered, the author argues that we should use "pseudonyms" (fake names or ID numbers) in the database to protect their dignity, only revealing real names when absolutely necessary and with care.

The Future: Seeing the Unseen

The paper concludes by suggesting that these tools could help us answer questions that have been hard to ask before. For instance, could we map out how close civilians lived to concentration camps? Could we see how often deportation trains stopped near busy town squares? The author imagines a future where we could ask a computer, "What would it have been like to walk into a pub in a specific town in 1941?" and get an answer based on the network of events happening nearby, rather than just a guess.

While the paper doesn't claim to have solved the mystery of the Holocaust, it suggests that Historical Network Analysis offers a new way to look at the evidence. It turns a pile of disconnected documents into a living, breathing map of relationships. By doing this, it helps us see the "invisible" structures of complicity and knowledge, reminding us that history isn't just a list of dates and names, but a complex web of connections that shaped the world. As the generation of eyewitnesses passes away, the paper argues that these digital webs might be one of the best ways we have left to understand the full scale of what happened, ensuring that the connections between people, places, and events are never lost again.