A “small” distant reading experiment: Comparing Sara Burgerhart and Willem Leevend

Original post: 30 Jan. 2015

Revisions: In process

Betje (Elizabeth) Wolff and Aagje (Agatha) Deken, who collaborated as life partners and co-authors over many years, are giants in the Dutch literary canon. They were also integrated into the same religious, cultural and political networks that included many Mennonites. Can the tools of digital scholarship (Voyant Tools, in particular) provide any special insights into their loooooong epistolary novels? This post provides readers and me the opportunity to try out answers to the question using two novels: Historie van mejuffrouw Sara Burgerhart (1782), and Historie van den heer Willem Leevend (1784-85).

First, I’ll elaborate on the meaning of “loooooong.” In total, the windows below provide you a glance at 3,828 pages over 9 documents (Sara Burgerhart in 1 volume, plus Willem Leevend in 8) (based on the DBNL sources — DataBase of Nederlands Literature). These pages breakdown as follows: SB 830 pp.; WL1 404 pp.; WL2 372 pp.; WL3 359; WL4 351 pp.; WL5 380 pp.; WL6 382 pp.; WL7 390 pp.; WL8 360 pp.

Sara Burgerhart

Willem Leevend, part 1

Willem Leevend, part 2

Willem Leevend, part 3

Willem Leevend, part 4

Willem Leevend, part 5

Willem Leevend, part 6

Willem Leevend, part 7

Willem Leevend, part 8

Bubblelines for all 9 texts (NEEDS UPDATING)

To make the most of Voyant Tools, it is best to expand each window above using one of the buttons on the top-right of the frames. You can find details of the tools, including instructions, at this link. I am providing windows for only a few tools: 1 word cloud for each of the 9 texts, plus 1 window that provides a comparative analysis of keywords across the 9 texts using the bubblelines tool. You can try new tools either by clicking on terms in the word clouds or along the bubblelines (new analysis options should open up), or by using the export button on the top-right of the tool frames. You might have to reapply the stopword list to the word clouds (the stopwords are not necessary for the bubblelines tool, since you only analyze the text using those concepts that you identify).

If you would like to use a different set of digital humanities software and need a Dutch stopword list (which is now included as part of Voyant), you can download stopwords from my academia.edu page.

Here’s some more background on the two titles. Both novels date from the Patriot era (ca. 1780-1787), a period of intense political upheaval in the Dutch Republic. I don’t know the novels well, but I have read parts of them several years ago while preparing an essay on Mennonites and gender in the early modern era (see the literature list below). For a scholarly point-of-reference related to understanding the works of Wolff and Deken from a digital humanist’s point of view, readers might wish to look at Karina van Dalen-Oskam’s report of her findings from more substantial analysis than I think I will be about to open up here. If you are able to draw any noteworthy conclusions from your explorations using Voyant, please share them, either through this site, or through other research dissemination venues. I’d be happy to hear if these tools help — or are not useful.

If you and I would really like to do a thoughtful analysis of the novels, we need some familiarity with them. I won’t provide any more background other than the briefest technical details above. However, I do believe (as I have outlined in an earlier post about a pamphlet by John Taylor) that tools like those above can help inform a more thoughtful reading, since they can help to quickly gather questions and hypotheses to test with closer reading.

NEEDS UPDATING: The text is the Dutch original, as prepared by and downloaded from dbnl.org (a fabulous online database of Dutch literature). For all its fabulousness, however, the DBNL (DataBase of Nederlands Literature) is neither flawless nor complete. To save time I have downloaded the pdf files provided on the DBNL. Each page of these files has a footer listing the authors and title of the text. These data provide false results in Voyant (particularly in summative tools like word clouds). To control for this, I have added the author and title keywords to the stopword list in Voyant. These words are Betje, Wolff, Aagje, Deken, Historie, Heer, Willem, Leevend, Deel, Mejuffrouw, Sara, Burgerhart, plus a few more minor terms that seem peculiar to this text: doch, i, j, ja, je, jy, r, l, e, ô. Keep in mind that if any of these words are in the text of the novels, they will be excluded from any analysis in which the stopword list is applied — but they will still show up and be a distraction in the corpus reader (the window that displays the entire text in a more conventionally readable form). Other words might be worth excluding, but I will leave this up to the discretion of better trained Dutch speakers.

Here’s a note about the idea for this posting. I gave a presentation on 26 Jan. 2015 at the University of Groningen, and questions and ideas in the discussion after it have inspired me to provide a framework for visitors to this site to try to articulate their own experience-based thoughts on digital text analysis. I would like to express my thanks again to Mathilde van Dijk at the RUG (UGroningen) for her invitation and generous hospitality. I also very much appreciated the contributions from the thoughtful seminar participants.


  • Dalen-Oskam, Karina van. “Epistolary Voices: The Case of Elisabeth Wolff and Agatha Deken.” Literary and Linguistic Computing 29: 3 (2014): 443-51.
  • Driedger, Michael. “Mennonites, Gender and the Rise of Civil Society in the Dutch Enlightenment,” in Mirjam van Veen, Piet Visser, Gary K. Waite, Els Kloek, Marion Kobelt-Groch, and Anna Voolstra, eds., Sisters: Myth and Reality of Anabaptist, Mennonite, and Doopsgezind Women ca. 1525-1900. Leiden: Brill, 2014, pp. 229-249.
  • Sturkenboom, Dorothée. Spectators van hartstocht: sekse en emotionele cultuur in de achttiende eeuw. Hilversum, 1998.
  • Sturkenboom, Dorothée. “Historicizing the Gender of Emotions: Changing Perceptions in Dutch Enlightenment Thought,” Journal of Social History 34 (2000), pp. 55–75.