Rhetorical Markers of
Otherness in Medieval Texts
Background
In a 2001 article that appeared in a special issue of The Journal of Medieval and Early Modern Studies, a respected peer-reviewed journal, that focused entirely on the then-controversial issue of the validity of using the concept of “race” when studying difference in the Middle Ages, the historian Robert Bartlett argues that the words gens ("clan") and natio ("breed," "stock," "kind") are “key terms of medieval Latin usage [that]...imply, etymologically, a concept of races as descent groups” (42). Bartlett argues in the article, called "Medieval and Modern Concepts of Race and Ethnicity," that race is indeed applicable to the Middle Ages (an argument that, in 2024, is less controversial) and demonstrates through qualitative reading of several medieval chronicles and letters that gens and natio can be read as linguistic proxies for the modern word “race.” He goes on to argue that populus ("the people") does not indicate the same meaning. Bartlett’s argument is still, twenty-three years later, influential, with Web of Science (whose science emphasis surely means it is undercounting humanities journals) counting 207 citations since then, with many coming in the past few years.
​
Research Question
Does my medieval Latin corpus demonstrate the trend Bartlett argues for? Namely, will collocation results suggest that the words gens and natio imply shared group origin, while populus does not?
​
Method
I amass a very small corpus of seven texts, for a total of almost 90,000 words. Broadly, these texts are medical texts, natural philosophy, and encyclopedias. I lemmatize these texts using the Classical Language Toolkit, a Python library that supports natural language processing for several premodern languages. Then, I upload the lemmatized Latin corpus to AntConc, and run collocations over the three words discussed by Barlett: gens, natio, and populus. After reviewing AntConc’s word frequency list, I also run collocations over the term terra ("land"), which is common in the corpus.
​
Results and Discussion
The collocates for gens are, translated, “people,” “year,” “our,” “name, “his own,” and results after those five are mostly function words that include conjunctions and adverbs. The word terra is the tenth most common collocate. There does seem to be a trend here, a suggestion that in my small corpus, when the word gens is used, the writer is perhaps speaking about his own group, as opposed to another group. Natio is a less common word in the corpus overall (gens occurs 39 times while natio occurs only 13), and its results may show a different trend. While these results are also filled with function words including conjunctions and prepositions, the two content words translate to “other” and "to have". That “other” is the most significant collocate may show that one may write about a natio when referring to someone else, but not oneself. Even Bartlett’s claim about populus seem not to hold up in the collocates. Here, there is certainly a tendency for the word to be used when describing distinctive groups, including words such as gens, Christianorum, and populus (it is not uncommon for a word to be collocated to itself).
​
Words highly associated with terra include many demonyms, indicating that the term is often used when describing a place where a group of people lives. Examples include Voyrat, Karauitarum, Comana, Soldani, Damasci, and Kergis. The results suggest that, at least for the writers in my corpus, there seems to be a concept of associating different groups of people with the land where they live. This trend may in part be explained by the popular climatic theory that said that the climate one lived in affected one’s constitution, personality, and even morality.
These results are very preliminary, but they are exciting because they should remind us that when used carefully, digital tools can help us see our own assumptions and can surprise us with interesting data.
​
The accompanying visualization is a dot plot (above) showing top collocates with the four words: populus, gens, terra, and natio. Contra Bartlett, populus does seem to share some terms with gens and natio, indicating that these words may have been used in similar circumstances. Terra, on the other hand, is associated with a set of words completely unrelated to the other three.