By Justus Hillebrand (Independent Scholar/Digital History Consulting)

I remember sitting in a session at the Rural History conference in 2017 in Leuven, Belgium. After presentations on fertilisers and state building from the eighteenth to the early twentieth century, the discussion became increasingly muddled. Scholars from across Europe struggled to find the right terms and concepts to discuss fertilisers in different historical and language contexts. Should they talk about chemical fertilisers, artificial fertilisers, or about fertilisers originating on the farm versus those brought in from somewhere else? They all spoke the same foreign language, English, but could not reach common understandings. Then, Peter Moser, director of the Archives of Rural History in Bern, brought the discussion to a halt. He reminded us that we first needed to understand what different terms meant for the historical actors before we could ask our own questions about what they wrote about fertilisers.

In this blog post, Peter Moser has recently called for a movement in agrarian history towards multilingualism to understand and represent the national particularities in the experiences of people working the land. His point of departure was the fact that researchers across disciplines in Europe and beyond have to publish in English to be understood across national boundaries and be attractive to (European) funding agencies. The lingua franca of English enables the participation of new scholars in the international research world and new connections between them across national boundaries. At the same time, he rightfully bemoans the lack of critical engagement with translations into English especially in agricultural history. The agricultural terminology in historical sources reflects highly particular experiences and environments of farm people in their linguistic and local, regional, or national contexts. The results are terms that often defy translation. The ways translation can erase historical subjectivities leads Moser to his call to add more multilingualism to translation in order to conserve meaning better when comparing agricultural histories of different places and languages.[1]

Comparison, however, is only one part of writing entangled histories. As Jürgen Kocka pointed out in 2003, the key to fulfilling the increasing interest in transcultural and transnational entanglements is combining comparisons and analysis of interconnections.[2] Against this background, I argue that the lingua franca of English and the “language” of data in digital history can be useful challenges in achieving this combination. English and data allow common terms to tell histories that were shared across national boundaries. Reflection on the process of constructing these common language and database terms and drawing conclusions from them is essential to writing entangled histories responsibly and well. I conceptualise this as a three-fold process of understanding, explaining, and (re-)defining:

1. to take apart the historical meaning of terms in different languages;

2. to determine and explain the basic actions or concepts they describe in their historical context;

3. and then to create an English term or a data definition as a kind of “super-category” that describes basic actions and encompasses the meaning best suited to one’s research question – or best suited to negotiate agreement on big questions and concepts in the field.

In the following, I trace this process through two specific examples of my work to raise the larger issue of responsible translation and writing translingual histories. My research on the history of agricultural knowledge connects more directly with Moser’s work. To expand this issue into digital history, I also examine parts of my work with the Ghettos Project of the Holocaust Geographies Collaborative. In this digital humanities project, our research team translates information on ghettos across Eastern Europe contained in the United States Holocaust Memorial Museum’s Encyclopedia of Camps and Ghettos into data rather than text. This example demonstrates how crucial responsible translation and careful development of data categories are for the computational and comprehensive analysis of translingual historical phenomena.

More than just translation into English: the example of agricultural knowledge

The first step in this process is understanding the meaning of terms in the historical sources. In agriculture, many terms have been born out of very specific environmental, linguistic, cultural, and economic contexts and as such, they defy translation. In my own research, where I analyse comparisons and entanglements between late-nineteenth-century New England and Westphalian farmers, the very term “farmer” posed the first challenge – an example Peter Moser also discussed in this blog. A Bauer in German can be translated roughly into a “peasant” in English. However, agrarian authors in Westphalia noticed the derogatory use of Bauer to denote their backwardness and some then counter-stylised the term as an honorary title for those producing the basis of society – food – since time immemorial.[3] Others moved towards the more up-to-date term Landwirt or emphasised their educated status among the “modernisers” of the economic enlightenment with the term Ökonom.[4] In English, “peasant” rang of European feudalism, a stark contrast to the independent “farmer” cultivating American soils, morals, and democracy. Even though suggested as a possible translation for Bauer by an online dictionary, the archaic term “husbandman” did not even appear in New England sources and in any case referred to those engaging in animal husbandry only. So, while the term “farmer” seems basic enough to describe all those engaging in agriculture, it loses in translation the many local and regional shades of practice, meaning, and politics in Westphalia, New England, and likely most other places. By itself, “farmer” as a foundational term would not do.[5]

To communicate my understanding of terms like “farmer,” I had to engage in explanations like the one above. Adding explanations to translations is not a new concept and has been used to analyse agriculture and its institutions across national and linguistic boundaries for a long time.[6] While some may perceive them as digressions from a central argument, I argue that thoughtful, concise explanations are indispensable. They not only help to avoid a loss of meaning, but also force the author to become more aware of the terms and thus of the experiences of the historical actors. Jürgen Kocka called this Verfremdung: A type of estrangement in which “one discovers the case with which one is most familiar as just one possibility among others.”[7] By having to explain terms that need no explanation in one’s native language, we become aware of their full, various, and historical meaning.

Nevertheless, as my research pointed towards the commonalities and entanglements between New England and Westphalian farmers, I wanted a common term that could adequately describe those engaged in agriculture in both regions and languages. Definition, or in this case, redefinition of fundamental terms was my solution.[8] This redefinition had to be particular to my research question, and in fact became part of my argument itself.

I analysed and named how different groups of actors produced, communicated, and negotiated agricultural innovation. In this respect, not all those who farmed were the same. Based on their self-descriptions, I developed analytical categories that communicated the differences in the ways those working the land produced knowledge for different goals. “Agricultural improvement” was a historical term also used in English-language research that described those farmers looking positively towards new farming knowledge and practices from elsewhere. However, German sources and historiographies had no such uncomplicated shorthand that could have served my analysis, even though they described the same class of largely well-to-do farmers. Those I termed “improvers” German contemporaries described as “intelligent” or “rational,” both of which dripped with condescension toward dumb and backward Bauern. Rather than silently accept simple translations, my reflections on the historical meanings of these terms in both German and American contexts allowed me to select, exclude, and add meanings useful to my analysis. I redefined the terms “farmer” and “improver” as opposite ends of a spectrum of actors engaged in agriculture, whether only in management or working the land with their own hands. Improvers had the means to invest time, learning, and money into trials of new methods for a net profit and social standing as “modern” among their peers. They had the ability to adapt knowledge from places far removed from their place, whether a field a continent away or an experimental plot attached to a chemical laboratory. Farmers did not have the means to learn, adapt, or even trust knowledge far removed in distance or context. Farm families looked at their neighbors and tried the new ideas, methods, and materials they saw and talked about on their own farm to make ends meet. Most historical actors and their experience were somewhere in between these analytical opposites. With this definition particular to my argument, I explicitly excluded the connotations of “farmer” or “Bauer” outlined above while summarising what the sources taught me about the experience of historical actors. Rather than attempting a bad translation and losing historical meaning, I defined analytical terms based on historical sources and my argument to sharpen the point of my analysis.[9]

(Re-)defining historical terms to fit one’s analysis and historical experience is possible in any language. English just happens to be the lingua franca that provides reach and credibility to researchers across linguistic boundaries today. Not by coincidence, this three-part process is also fundamental to the digital humanities where translation of text into data provides the basis for creating structured data about the past. Like English, data speaks across languages and nation states, if defined well.

The “language” of data: the example of Holocaust ghettos

Humanists in general, and historians in particular, are often wary of data. Very aware of the complex and often problematic production of data in the past, historians are rightfully careful in their use of historical data. In agricultural history, much of the knowledge production used to industrialise farming from the mid-nineteenth century onward privileged scientists’ data production over farmers’ know-how and narrative. This marginalised farmers who were critical of industrialisation.[10] In Holocaust history, producing data about concentration camps, ghettos, and the victims within them can reproduce, rather than critically examine, the racist Nazi gaze of seeing its victims only as a statistic or dots on a map rather than people.[11] If historians reveal data production in the past as problematic, why should they repeat these mistakes by producing data of their own? As other digital historians, I argue not to throw out the baby with the bath water. Awareness of the problematic past of data production is an asset to improve it, not avoid it.

The key is understanding and situating the knowledge production practices of digital historians. The “scholarly primitives” of digital history are the same as in non-digital history: Historians select sources and information in them, synthesise this information, and arrange it into an interpretation of the sources. [12] The central difference between digital and non-digital historians is the shape these interpretations take to enable analysis and narrative. Imagine going into the archive or reading an encyclopedia not with a blank notepad, but with a stack of specifically structured spreadsheets. Rather than pondering idiosyncratic notes to write narrative, database historians analyse their spreadsheet notes computationally to create tables, graphs, and maps which then help them write narrative. Understanding, explaining, and (re-)defining carefully develops the definitions of the spreadsheet columns. The heart of every database are the definitions of its table columns, also called data fields, and the conventions of how to interpret source information into them consistently. They determine the very structure of information that enables analysis and narrative. Defining database fields is the same process as defining terms in English and likewise must be made explicit to remain credible. Matthew Lavin has recently described this as “situating the data.” Extending Donna Haraway’s concept of “situated knowledge” to “situated data” allows digital historians to emphasise that all knowledge is created by humans (and their instruments) in a specific context with specific goals, whether in the humanities, the social sciences, the natural sciences, or information science.[13] Transparency, historical expertise, and self-reflection are assets to the development of historical databases. They help confront the pitfalls of data creation and access the potential of building historical databases: scaling up analysis beyond the volume of information which the human mind can take in and process – and beyond national boundaries.[14]

With the Holocaust Geographies Collaborative, I co-developed a database capturing all Holocaust ghettos across Eastern Europe as recorded in volume 2 of the United States Holocaust Memorial Museum’s Encyclopedia of Camps and Ghettos. One of our central concerns turned out to be the term “ghetto” itself. The encyclopedia included all places which sources ranging from historical research to Nazi reports and survivor testimony have called a “ghetto.” Struggling with a definition, encyclopedia editor Martin Dean settled on the “ghetto” as “a place where the Germans concentrated the Jews.”[15] However, the authors of the 1,142 Encyclopedia entries used descriptors of ghettos differently. We discovered that ghetto types were not low-hanging fruit, saving us from the arduous work of understanding, explaining, and defining, but the very starting point of the process. Some described instances where German administrators prohibited Jews from leaving their town as an “open ghetto,” while others did not typify these instances as ghettos at all. Others used “open ghetto” to describe neighborhoods that Jews could leave in general, during certain times, or not at all – if they stayed within the bounds of the settlement. By contrast, “closed ghettos” could be places that Jews were allowed to leave only during specific times, for specific tasks, or with specific permits. These places were often marked by an enclosure varying from walls, fences, or barbed wire to mere signs. Some “closed ghettos” did not allow Jews to leave at all, which led some authors to describe them as “sealed ghettos.” Within this complex web of terms, also shaped by German source terms like geschlossenes Ghetto (verbatim “closed ghetto”), Judenviertel (verbatim “Jewish quarter”), or Judenwohnbezirk (verbatim “Jewish residential district”), we understood that for a database capturing Holocaust ghettos comprehensively and spatially, these terms would not do.

When we discussed the many situations in which the type names were used, we had to explain the historical meanings and events “on the ground” to untangle terminologies. Through exploration of these basic actions, we moved away from trying to typify “ghetto” to recording the process of ghettoisation. We realized that “the imposition and enforcement of spatial forms of restriction that were intended to control the location and mobility of Jewish bodies [had] three interlocking spatial aspects: the declaration of one or more holding places, restriction of movement, and means of enclosure.”[16] In the database, this meant we defined fields capturing aspects of these three parts. For holding places, we recorded their “structure(s) or space(s) where Jews were confined or required to stay for any period of time by non-Jewish authorities or actors.”[17] Categories for this field included large areas like whole settlement or neighborhood down to house(s) or market square. We defined “type(s) of enclosure erected or instituted by non-Jewish authorities or actors to enforce control over Jewish bodies in one or more holding places”[18] with values like fence, geographic feature, or guard(s). These definitions of database fields and their corresponding categorical values were the result of the long process of collaboratively understanding, explaining, and defining parts of ghettoization in the light of our spatial research questions.

Thus, to create the basis of a comparative and entangled history of Holocaust ghettos, we consciously managed the challenging interpretations of turning source information into data. The central part of database development was reflecting again and again on our own assumptions in light of historical evidence and context as well as having to explain ourselves to each other.


Multilingualism is one solution to the loss of meanings in translation and the “homogenisation of perspectives.”[19] It can also be the foundation of understanding, explaining, and (re-)defining the meanings of the very terms we use to tell comparative and entangled histories. As historians, we have the responsibility to write narratives shaped by the particularities of people’s places, experiences, and languages while at the same time revealing commonalities and shared histories. English and the “language” of data are not barriers but opportunities. Constructing common terms, whether in English or database fields, requires we develop an awareness of language, even of those we have spoken for a lifetime, and of the nuanced meanings words convey. Collaboration – that is the need to negotiate meanings with others – makes this process more visible and active. The results are categories or concepts of analysis specific to and productive for our shared questions. We should not shirk this task but embrace it as foundational to historical research.

This necessitates collaboration with other experts. But just as we should be reluctant to let a language translator unfamiliar with the historical context identify the peculiarities of meaning in historical sources, we should not outsource database design to IT departments unconcerned with the histories we study. Giving historical information to IT professionals as a black box yields data we can neither fully understand nor develop to its full potential. We either need to understand databases and history, or work with experts who do. Whether understanding, explaining, and (re-)defining between languages or into data, historians need to collaborate in the process of negotiating the basic terms we use to write histories.

[1] Peter Moser, “Empowerment of new actors and homogenisation of perspectives? The ambivalent effects of European research funding on academic work in the field of rural and agrarian history. A call for a debate.” Worlds of Related Coercions in Work Blog, October 29, 2021,

[2] Jürgen Kocka, “Comparison and Beyond,” History and Theory 42, no. 1 (2003): 39–44,

[3] See e.g. Johannes Dornseiffer “Vom Bauernstande: Plaudereien für die Winterabende,“ Sauerländisches Volksblatt 91, 12 Nov, 1879, 2.

[4] For the use of Ökonom in the economic enlightenment, see e.g. Verena Lehmbrock, “Lob Des Handwerks: Wissenstheorie Heute Und Bei Albrecht Daniel Thaer (1752-1828),” Zeitschrift Für Agrargeschichte Und Agrarsoziologie 62 (2014): 30–41.

[5] For example, Sigmund von Frauendorfer analyzed this difficulty in translation in 1929 already and settled on defining European peasants and American farmers as “gradations or varieties of the same economic category,” (642) similar enough to compare them. Sigmund von Frauendorfer, “American Farmers and European Peasantry,” Journal of Farm Economics 11, no. 4 (1929): 633–42,

[6] For an early example, compare Frauendorfer again, Sigmund von Frauendorfer, “Development Methods and Results of Agricultural Economic Research in the United States,” Journal of Farm Economics 10, no. 3 (1928): 286–311,

[7] Kocka, 41.

[8] As it has been for many, see e.g. Frauendorfer, 1929.

[9] Justus Hillebrand, “To Know the Land with Hands and Minds: Negotiating Agricultural Knowledge in Late-Nineteenth-Century New England and Westphalia” (2021). Electronic Theses and Dissertations. 3409.

[10] See e.g. Deborah Kay Fitzgerald, Every Farm a Factory: The Industrial Ideal in American Agriculture (New Haven: Yale University Press, 2003).

[11] Compare e.g. the discussion of this problem in Anne Kelly Knowles et al., eds., Geographies of the Holocaust, (Bloomington: Indiana University Press, 2014).

[12] John Unsworth’s term of “scholarly primitives” is summarized nicely by the Arguing with Digital History working group, see Arguing with Digital History working group, “Digital History and Argument,” white paper, Roy Rosenzweig Center for History and New Media (November 13, 2017):

[13] Matthew Lavin, “Why Digital Humanists Should Emphasize Situated Data over Capta,” Digital Humanities Quarterly 15, no. 2 (June 15, 2021); Donna Haraway, “Situated Knowledges: The Science Question in Feminism and the Privilege of Partial Perspective,” Feminist Studies 14, no. 3 (1988): 575–99,

[14] See also Robert Crease, Elyse Graham, and Jamie Folsom. “Database Thinking and Deep Description: Designing a Digital Archive of the National Synchrotron Light Source.” Digital Scholarship in the Humanities 34, no. Supplement_1 (2019): i46-i57.

[15] Martin Dean, ed., Encyclopedia of Camps and Ghettos, 1933-1945: Ghettos in German-Occupied Eastern Europe, vol. 2 (Bloomington: Indiana Univ. Press, 2012), XLIII.

[16] Anne Kelly Knowles and Justus Hillebrand with Paul B. Jaskot and Anika Walke, “Integrative, Interdisciplinary Database Design for the Spatial Humanities: the Case of the Holocaust Ghettos Project,” International Journal of Humanities and Arts Computing, Volume 14, Issue 1-2, 77.

[17] Ibid, Supplementary Material, “Holocaust Ghettos Physical Characteristics Spatial Table,” 8.

[18] Ibid, 10.

[19] Moser.