THIS POST WAS FIRST COMPLETED ON 1 March 2023
The Bigger Purposes of This Post Are Three-Fold:
- to promote digital text analysis as a complement to traditional techniques of close reading, and Voyant Tools as a first-choice for digital analysis programs;
- to advertise my Dutch stop-words list (available in Voyant Tools and in Academia.edu); and
- to give some background about the career and ideas of the unconventional Dutch Mennonite preacher and revolutionary-era activist — François Adriaan (Francis Adrian) van der Kemp.
In this post I have several interactive windows from Voyant Tools. The text-combination I use in each Voyant window includes 3 sermons from the anti-Orange activist and Dutch Patriot Movement leader François Adriaan van der Kemp. Each sermon is from Elftal Kerkelyke Redevoeringen [11 Sermons]. In 1782, when he published the collection, van der Kemp was the preacher of the Doopsgezind (also known in English as “Mennonite”) congregation in Leiden. The sermons were presented at several Doopsgezind congregations in the Dutch Republic. The stop-word list helps make the Voyant analysis manageable and productive.
NOTE: This post is best viewed on a laptop or desktop computer, not a phone or smaller tablet.
Selected Translations of Keywords
For readers who are not familiar with Dutch, here is a selected glossary:
- burgers – citizens
- echtgenoot – spouse
- gehoorzaemheid – obedience
- godsdienst – religion (a complex concept; literally: “service to God”)
- kinderen – children
- liefde – love
- menschen – people (plural of “person”)
- plichten – duties
- tyd – time (era)
- volk – people (in the collective sense similar to “nation”)
- vrouw / vrouwe / vrouwen – woman / women
- vryheid – freedom
- waerheid – truth
Note that the 11 Sermons uses 18th-century spelling conventions. These differ slightly from modern Dutch. For example, in modern Dutch “gehoorzaemheid” would be “gehoorzaamheid”, and “tyd” would be “tijd”.
Why Use Voyant Tools (VT) to “Read” Texts?
VT is an online resource for digital text analysis. I have put “reading” in quotation marks in the title of this blog post, because I am a supporter of traditional, humanistic, scholarly views about close reading; as a result, I am not a one-sided defender of digital methods of reading, nor am I an expert in these methods. That said, I am also open to and curious about newer ways of understanding texts and sources and evidence. Scholars in the digital humanities are on the forefront of thoughtful research and experiments in these areas. Over the last decade, I’ve learned a lot from them in my work as a university teacher and as a researcher. I support the idea that VT and other digital resources can strengthen and help good, high-quality, humanistic scholarship and reading at all levels — from newcomers to university learning to the most advanced and experienced researchers.
I’ll list two reasons why I recommend VT:
- It’s an excellent complement to close reading, especially when we use it to “pre-read” texts — that is, to get an overall picture of patterns in a text or in a large collection of texts. I use pre-reading to generate questions for further investigation and to identify passages to read more closely.
- I also find the visualization tool in VT useful for illustrating conclusions about texts.
I have previously written on this blog site about using Voyant in early modern studies. These articles include more thoughts about why VT is worth using. See the following links:
and
“Reading” John Taylor’s 1641 Pamphlet Using Voyant Tools (2014; updated 2023)
Quick Guides to Using Voyant Tools in Practice
Voyant Tools is easy to use at some levels, but using it well and effectively takes practice! For a good, brief, practical introduction to using VT, see this webpage from the UC Santa Barbara Library.
You will notice from the interactive Yoyant view above that the default analysis window for texts added at https://voyant-tools.org/ includes 5 frames. These frames usually include the tools called Cirrus, Reader, Trends, Summary, and Contexts.
I have tried to modify the 5 frames above for this post. NOTE: It is possible to change the tools that you use in each frame, and it is also possible to include only one tool on a screen. In other words: These tools are not only interactive; it is also possible to use other tools in their place, or in other screen formats.
To learn more about the tools that you can use, and how to use them, go to the official Voyant Tools documentation page. Also see Stéfan Sinclair’s YouTube screencasts at https://www.youtube.com/playlist?list=PLDCADF35691404F54
Other tips and resources to help you learn and use VT are:
- Geoffrey Rockwell and Stéfan Sinclair, Hermeneutica: Computer-Assisted Interpretation in the Humanities (MIT Press, 2016 [2022]).
- https://mitpress.mit.edu/9780262545891/
- See companion resources at http://www.hermeneuti.ca/
In popular perceptions of text analysis, I suspect that the word cloud or cirrus is the most recognizable tool. A big difference between fixed word clouds and the word clouds in Voyant Tools is that VT word clouds are dynamic and interactive. This is true of all digital analysis tools in VT.
Note that some digital analysis tools are better for analyzing a single text or for treating a series of texts as though it were one text. Other tools are better for analyzing a series of texts or several parts of a single text.
Some Initial Analysis of 3 of the 11 Sermons
In this section I will recommend two VT tools for pre-reading texts (with pre-loaded interactive windows), and I will highlight an initial observation about the three sermons included in these analyses.
My first recommendation is the Topics tool. Below is the Topics tools with the three sermons from van der Kemp’s 11 Sermons:
Try clicking on each row in the Documents and Topics columns to see how the function. This tool shows how key terms-clusters vary across texts. The terms in each colour-code row are related to each other in some statistically noteworthy way.
Another vary useful tool for pre-reading texts is the TermsRadio:
This tool shows how key terms vary in significance from text to text. Each column in this tool represents one of the three sermons, and the hierarchy of terms indicates the frequency in each text. Try moving your cursor over the various terms to see how the tool functions.
An initial observation after exploring the three sermons using these and other tools is that the three sermons are very different in content (not a surprise, given that sermon II is about the duties of wives, and another is about the duties of preachers / teachers to tell the truth — but, nonetheless, they do have some textual connections.
Below is an example that shows that the phrase “suffering obedience” (“lydelyke gehoorzaemheid”) is a word-combination that comes up in each of the three sermons. In the graph below, there are three columns. The first represents the sermon on the duty of wives; the second represents the sermon on the duty of preachers to tell the truth; and the third column represents the prayer-day sermon on America as a model for the Dutch Republic. “Obedience” is much more frequent in the first of these (remember: a sermon about the duties of wives); and “freedom” is much more frequent in the last of these (a sermons celebrating the American Revolution). So far, few surprises that traditional humanistic reading of the printed versions of these sermons could not reveal just as (or even more) effectively.
The insight that I can highlight with Collocates in VT is that the terms lydelyke and gehoorzaemheid are related to each other in a statistically noticeable way, and that the combination appears with the same relative frequency in each of the three texts. The Trends tool (which visualizes variation across texts or portions of a single text) also helps highlight this observation.
Two Voyant Output Windows for 3 Sermons from _Elftal Kerkelycke Redevoeringen_ (1782)
The question I am curious to investigate now is how F.A. van der Kemp defined “suffering obedience” in each of his three sermons included in this analysis. The meaning of this combination of terms can shift significantly from the very gendered discussion in the first of three sermons about wives to the third sermon about Dutch national freedoms (possibly again gendered, but, if so, probably in a masculinized way?). A bigger question is if (and how) van der Kemp uses these terms in the other 8 sermons (and in other writings; and if and how the terms are used by other Dutch writers of the era). More traditional reading of the sermons and other sources, and / or more digitization of the sermons is necessary to answer these kinds of questions.
For me as a scholar of Mennonite studies, I also wonder how “suffering obedience” might be related (if at all) to “suffering toleration / yielding” (lijdelijke verdraagzaamheid”) — the latter is a term combination that shows up in 19th-century Dutch literature. It is worth noting that verdraagzaamheid (toleration / yielding / submission) was a subject of intense debate among Doopsgezinden / Mennonites in the Dutch Republic around 1700.
Technical Notes
Please note that the sermon collection analyzed in these posts are only 3 of 11 sermons! Why not more? Short answer: Technical difficulties! Longer answer: Transcribing early modern texts into clean, corrected, digital formats is NOT easy! F.A. van der Kemp’s 11 Sermons are available in digital format. At first glance, they look to be transcribed. WRONG! It took me hours and hours to prepare the 2nd, 10th, and 11th sermons — the ones shared in this post.
Here’s an example of uncorrected, machine-transcribed text from page 182 of the 11 Sermons (from the sermon on Jewish political history — for more details, see the bibliographical notes below):
Onder een onweder van donder en blixem
ten, welke zy verplicht waren te onderhouden• alle gewigtige zaeken tot den ftaet betrekke! hk alle netelige Gefchilftukken, moeften to God gebragt worden, die dan door den Ur * e > Thummim, door den mond eens Prophe2 of op eenige andere wyze, antwoorde
Nog duidelyker fpeuren vinden wy van deze Godsregeenng in de beloften en Ldreign
v^eleke” ^7 ~g
, j «wcriey gelujc, verwinning
md?ï£ f n’/reede en overvJoed> – *4 *7 de geboden der wet onderhielden j: allerfcy
?anv*
You do not need to know a single word of Dutch to guess correctly that significant portions of the passage above are a nonsensical mess! For examples in English, see the blog post that Danny Samson and I put together in 2015.
Another technical note: The most frequent words in all Latin languages are probably not the most significant ones. Here’s an example showing two analysis frames, each with a cirrus / word cloud of the three sermons I use in this post:
A Picture of Two Word Clouds from the 3 Sermons — One without Stop-Words, and the Other with Stop-Words Applied
The left-hand image shows the most frequent terms across the three texts; the right-hand image shows the most frequent terms — excluding “insignificant” words. The common term in text analysis for these insignificant terms is “stop-words”. You can modify your stop-words in Voyant Tools for each analysis!!!
VT did not include a Dutch stop-words list in its early variations. This frustrated me around 2015, when I started to experiment with digital text analysis of early modern Dutch texts. F.A. van der Kemp’s 11 Sermons were a test-case for me. I decided to build my own Dutch stop-words list that included modern and early modern Dutch spelling variations. It took me many, many more hours to produce a Dutch stop-words list; I did build upon other online digital humanities resources from Dutch scholars — for which I am very grateful — but the early modern spelling variations that are included in my stop-words list and the VT Dutch list are of my compilation. In short, the stop-words list pre-loaded in Voyant Tools is the one I created (thanks to the late Stéfan Sinclair for including these stop-words in VT!). I have shared the list on Academia.edu. I used F.A. van der Kemp’s 1782 sermon series as a model for the list. This might explain why I have included “bizonder” (van der Kemp’s spelling) instead of “bijzonder” (modern Dutch spelling).
Suggestions for improving the list are welcome!!!
NOTE: You can choose to apply / use or not apply stop-words in a VT analysis. Some research questions (especially ones related to the analysis of writing styles) might require that no stop-words are applied.
The Sources
Here is the bibliographical information for F.A. van der Kemp’s 1782 sermon collection:
Fr. Adr. van der Kemp, Elftal Kerkelyke Redevoeringen (Leiden: Leendert Herdingh, Leendert, 1782).
- Location: Amsterdam, Universiteitsbibliotheek (O 60-33)
- Persistent URL: http://resolver.kb.nl/resolve?urn=dpo:8570:mpeg21
Included in the book are the following sermons (I have edited the 3 highlighted sermons and they are the ones included in the Voyant windows):
- Redevoering over ‘t Huuwlyk (Sermon … about marriage). Starting at page 1
- … de plichten der Vrouwen (… about the duties of women / wives). … 21
- … de plichten der Mannen (… about the duties of men / husbands). … 47
- … de plichten der Ouders (… about the duties of parents). … 69
- … de plichten der Kinders (… about the duties of children). … 92
- … de voordeelen eener vroege Gods-vrucht (… about the advantages of an early fear of God). … 115
- … de plichten der Heeren en Dienstknechten (… about the duties of lords and servants). … 133
- … der Overheeden en Onderdaenen (… about the duties of rulers and subjects). … 155
- … het veranderen der Republik. Regeeringsvorm in eene Koninglyke by de Jooden (… about the change from a republican form of government to a royal one among the Jews). … 176
- … de plicht om de waerheid te spreeken (… about the duty to speak the truth). … 200
- … gehouden op den Bededag van den 27. Febr. 1782 (… held on the prayer day of the 27th Feb. 1782). … 223
Here are links to the texts of each sermon:
Historical Background
For a biography of F.A. van der Kemp, visit HERE.
For more about the 11 sermons, visit HERE.
Some readers might consider van der Kemp to have been a bad Mennonite because he supported armed resistance against the Orange family regime in the Dutch Republic. My view is that such interpretations are counterproductive in good historical analysis because they involve using partisan theological, ethical, political positions from today or from the past to condemn rather than to try to understand people such as van der Kemp.
For an outline of a curious partisan attack from the 1780s against van der Kemp, see my post about “the ghost of Menno Simons”.
For more about the politics of Mennonite and related adult baptizing groups in early modern Europe, read my essay on “Anabaptists and the Early Modern State” in Stayer and Roth (eds.), The Companion to Anabaptism and Spiritualism, 1521-1700 (2006) and James Urry, Mennonites, Politics, and Peoplehood (2006).
For an introduction to early modern cosmopolitan Dutch Mennonites and Doopsgezinden, read Piet Visser’s essay in Stayer and Roth (eds.), The Companion to Anabaptism and Spiritualism, 1521-1700 (2006), and my essay in Mennonite Quarterly Review on “The Year 1625, the Dutch Republic, and Book History” (2023)
For an overview of Doopsgezind-Mennonite philosophical contributions to Enlightenment cultures, read https://www.academia.edu/93611083/Aufkl%C3%A4rung_Enlightenment_particularly_in_Dutch_Mennonite_history_
— FINIS —