Slavery in the Electronic Enlightenment Collection

Tessa van Wijk https://orcid.org/0009-0007-2011-8697

Tessa van Wijk is a Research Master student of Literary Studies at Radboud University in the Netherlands. As part of her programme, she spent 4.5 months researching correspondence related to slavery at Electronic Enlightenment.

Part of my research project at Electronic Enlightenment was to conduct a review of content related to the history of slavery and slaveholding of the French and English letters in Electronic Enlightenment.¹ A content review such as this feeds into current debates about racial bias in historical research and data exhaust.²

To conduct this review of content, I started out by doing keyword searches for the terms ‘slave’, ‘slaves’, ‘slavery’, ‘esclave’, ‘esclaves’ and ‘esclavage’ in, firstly, the French correspondences in Electronic Enlightenment. Very quickly, several methodological problems emerged. Firstly, the chosen keywords produced too large a number of results. Secondly, it became clear very quickly that some of the terms, such as ‘esclave(s)’ and ‘esclavage’, were used in a manner that did not relate to my research topic. Indeed, my focus was on correspondences related to the slave trade and chattel slavery, particularly practices by European and North American nations in the eighteenth and early nineteenth centuries. Thus, numerous letters showed that these terms were used in a more metaphorical sense. Voltaire amongst others uses the term ‘esclave’ in this manner:

Another metaphorical use of terms such as ‘esclave’ relates to the political discourses of the time and evokes enslavement to autocratic or absolutist political systems. This is for example the case in Jean Le Rond d’Alembert’s letter to Voltaire, in which he writes the following:

In short, the metaphorical use of terms such as ‘esclavage’ and ‘esclave’ is typical of eighteenth-century French authors. The connotation with the enslavement of African people or the Triangular Slave Trade was a lot less frequently present. Rather, the words ‘esclave’ and ‘esclavage’ are more often defined in opposition to freedom and liberty: “a slave, for the eighteenth century, is someone who was deprived of their freedom, whatever the form or cause of this deprivation.”³

Encountering many letters that used the terms ‘esclave’ and ‘esclavage’ in similar ways, I decided to start using AntConc, a text analysis tool, in order to more quickly be able to identify if a search result was relevant to my project. After feeding the programme a text corpus of all letters in Electronic Enlightenment (many thanks to Mark Rogerson for providing me with this corpus), I could conduct my keyword searches here and use the KWIC (Keyword In Context) tool to identify the textual setting of the keyword for each search result. I proceeded in this manner for both the French and English material in Electronic Enlightenment, searching for words such as ‘slave(s)’, ‘slavery’, ‘esclave(s)’ and ‘esclavage’.

Given the issue of disambiguation and historical usage of terminology, I then decided to narrow my search to include racially insensitive terms such as ‘nègre’, ‘nègres’, ‘negro’ and ‘negroes’. Including these problematic terms is important, because as Jack D. Forbes notes, they, over time, “became synonymous with enslavement.”⁴ To avoid not finding relevant results, it was also important to take into account small spelling differences. Writers might, for example, use an accent acute in ‘nègre’ instead of an accent grave, or spell the plural of ‘negro’ without the added e (so ‘negros’). I therefore also did several wild card searches (e.g. ‘n*gre’) and searches with alternative spellings for the keywords. In the final list of relevant results, choosing to also conduct keyword searches for these racially insensitive terms, meant almost a twofold increase of letters related to, or referring to, slavery in some way.

This increase of relevant search results by including terms such as ‘negro’ in my keyword searches already illustrates one of the problems with this type of research and the corpus. To find material related to slavery, one finds themselves at times having to reproduce and actively search for racially insensitive language.

The next phase of my content review was the systematic extraction of keywords in my datasets. I ordered the relevant search results (310 letters in total) in a spreadsheet. ⁵ 259 of these results were found in English letters and 51 in French letters. In the spreadsheet I included basic information on the respective letter (writer, recipient, date etc.), the relevant sentence(s) including the keyword, and I ordered each letter/result into specific categories: anti-slavery discourse (in a couple cases specified that the letter discusses the reception of (a piece of) anti-slavery discourse), buying of enslaved people, possession of enslaved people, illness and health of enslaved people, freeing of enslaved people, runaway, revolt, selling of enslaved people, slave trade, writing about slavery, pro-slavery discourse, relating to the anti-slavery debate, abolition. There are also two categories that only contain one or two letters, but that could not be put in one of the broader and bigger categories mentioned above. These are the following categories: child of (formerly) enslaved mother, and enslaving. Furthermore, there are two categories that describe letters in which we cannot be certain if they discuss enslaved workers or paid workers of colour; the categories labour by people of colour (containing only one letter), and servant of colour (containing nine letters). Moreover, some letters correspond to more than one category, which leads to ‘merged’ categories such as ‘possession of enslaved people, illness & health.’

If we look at which categories contain the most letters, we notice that these would be the following categories (the number of letters includes those letters that correspond to the respective category as well as one/multiple others): possession of enslaved people (124 letters), anti-slavery discourse (92 letters), buying of enslaved people (44 letters), and slave trade (33). Except for the, still rather big, category anti-slavery discourse, the other three main categories contain letters in which the writer thus does not necessarily explicitly put forward judgements on slavery (i.e. approving of or condemning slavery). Rather, most letters, generally seem to ‘only’ discuss/refer to slavery and the slave trade in a more distanced, emotionally detached, or business-like manner.

Returning to my initial point about bias in digital archives and data exhaust, the small data sample illustrates the limitations of such historical research and indeed a corpus such as Electronic Enlightenment - who leaves archival traces? The top five letter writers with the most letters in the final list of results are Simon Taylor (57), Edmund Pendleton (42), William Cowper (23), William Fitzhugh (21) and Francis Fauquier (15). Except for William Cowper, these men all had interests in the continuation of the slave trade and slavery, and their letters can mostly be found in categories relating to the owning of and trading in enslaved people (e.g. the categories slave trade, possession of enslaved people, buying of enslaved people).

In Edmund Pendleton’s letters, we can clearly see that enslaved people are considered as personal property. Pendleton was an American plantation-owner and slaveowner, as well as an attorney. Several of his letters in Electronic Enlightenment discuss legal affairs, particularly inheritances. The following letter to Jerman Baker clearly illustrates that enslaved people are thought of as property that can be inherited:

The letter here reflects a very business-like linguistic register. There is some disagreement about the inheritance, to which the enslaved belong, and the enslaved are only referred to in this regard. There is no reference to their autonomy or agency as people, only to their value as property.

Simon Taylor who managed Chaloner Arcedeckne’s Jamaican Golden Grove estate, regularly discusses the enslaved workers on the estate in his letters to Arcedeckne. For example, in the following letter, he recommends Acedeckne to purchase more enslaved people for the estate:

Taylor considers enslaved workers a commodity and financial asset. He also writes about making the enslaved “last much longer” and not wishing to push them too hard. This too is proposed in a business-like manner. Making the enslaved ‘last longer’, as if speaking about a perishable good, is not done in their interest, but in the interest of the estate. It ensures there will be workers left to work the land.

The types of letters that showed up during my review of content are therefore mostly letters written by slave- and estate owners. While there is also a rather big number of letters in the anti-slavery discourse category, these are generally still not from the point of view of (formerly) enslaved people or people of colour. Instead, they are from writers such as William Cowper who expressed himself against slavery and wrote poetry and songs about the topic:

Within Electronic Enlightenment, like many other digital archives, the voices of (formerly) enslaved people and people of colour are thus largely absent. During my review of content, I only found one letter written by a (formerly) enslaved person, Ignatius Sancho, ‘Ignatius Sancho to Laurence Sterne, 21 July 1766’. I did also find two letters to another person of colour, Jean Pierre Boyer, president of Haiti, Jeremy Bentham to Jean Pierre Boyer, 10 February 1823 and Jeremy Bentham to Jean Pierre Boyer, 29 December 1822, but these letters thus do not give presence to Boyer’s voice, only that of his correspondent. As such, we are often limited to using accounts by individuals involved in the system of slavery to gain some insight into the slavery past and enslaved people’s experiences.

While the material relating to the slavery past in archives is often limited to such accounts, by taking a critical approach to them, they can still give us insight into the history of slavery, even if they are also problematic. When studying these types of materials, it is important to take what Cailtin Rosenthal calls a ‘slow approach’. It is paramount that we acknowledge the biases embedded in them. According to Rosenthal, “[s]cholars who have relied on slaveholders’ data have tended to answer slaveholders’ questions.”⁶ This data contains slaveholders’ values which are then easily (accidentally and blindly) replicated. As such, when wanting to move past ‘answering slaveholders’ questions’ and to gain insight into enslaved people’s lived experience, we need to take a ‘slow’ approach that recognises the biases inherent to the archive.⁷. While a more quantitative approach, such as my review of content, can be useful in exploring archives and researching the history of slavery and slaveholding, this should be a first step. Afterwards, we must return to the results with a more qualitative, critical and slow approach. For example, by close reading the instances in which the keywords appeared. Furthermore, it is important that we balance accounts by individuals involved in the system of slavery with materials in which the voices of (formerly) enslaved people and people of colour are present. After I conducted my review of content, Electronic Enlightenment added over 190 letters from The Letters of Charles Ignatius Sancho (Broadview, 2015), thereby taking an important step in making these voices in their collection more present.

1. It is important to note that France and Great Britain have different histories and historical contexts of slavery. France first ended slavery in 1794, several years after the revolution. It was however reinstated by Napoleon in 1802. Slavery was finally permanently abolished in 1848. Great Britain put an end to the slave trade in 1807 and adopted the Slavery Abolition Act in 1833. In 1838, the system of apprenticeship, which required formerly enslaved people to work up to 45 hours per week for their former owners and without compensation, was ended and the formerly enslaved people officially gained freedom. There were, however, also links between France and Britain during this time. The French anti-slavery movement, for example, was largely inspired and supported by the British one.
   • Lawrence C. Jennings, “The Interaction of French and British antislavery, 1789–1848”, Proceedings of the Meeting of the French Colonial Historical Society, vol. 15, 1992, p. 81–91.
   • Gad Heuman, “The Apprenticeship System in the Caribbean”, New West Indian Guide / Nieuwe West-Indische Gids, vol. 97, no 3–4, 2023, p. 230.
2. See for example Caitlin Rosenthal, “Lessons From the ‘Data Exhaust’ of Plantation Slavery”, Harvard Data Science Review, vol. 3, no 2, 2021, p. 2–9.
3. Michael O’Dea, “Qu’entend-on par esclavage au XVIII Voltaire, Rousseau, Diderot,” in Littérature et esclavage. xviiie-xixe siècles (ed. Sarga Moussa), Paris, Desjonquères, 2010, p. 40. My translation.
4. Jack D. Forbes, “The Use of the Terms ‘Negro’ and ‘Black’ to Include Persons of Native American Ancestry in ‘Anglo’ North America,” Explorations in Ethnic Studies, vol. 7, no 7, 1984, p. 13.
5. It is difficult to pinpoint how many search results (meaning also including non-relevant ones) showed up in total. This firstly is due to the many different keyword searches, for different terms, conducted. Secondly, AntConc also multiplicated results of letters if these letters contained the keyword searched for multiple times. For example, a letter containing the word ‘slave’ three times, would show up three times in the results lists. Lastly, multiple letters showed up in more than one of the keyword searches. This would be the case if they contained multiple of the keywords searched for (e.g. if a letter includes both the words ‘esclave’ and ‘nègre’ or contains both ‘slave’ and ‘slavery’). To facilitate an easy overview of the amount of French and English letters in Electronic Enlightenment related to slavery and the slave trade, letters that showed up multiple times (because they included more than one keyword or one of the keywords multiple times) were not multiplicated in my spreadsheet, but rather only included and categorised once, with all of the different hits in their respective sentences provided in the last column of the spreadsheet.
6. Cailtin Rosenthal, “Lessons From the ‘Data Exhaust’ of Plantation Slavery”, Harvard Data Science Review, vol. 3, no 2, 2021, p. 2.
7. Cailtin Rosenthal, “Lessons From the ‘Data Exhaust’ of Plantation Slavery”, Harvard Data Science Review, no 2, 2021, p. 2–9.

---

• BLOG

• ---

• NEWS
  • ◊ current
  • ◊ archive
• ARTICLES
  • ◊ current
  • ◊ archive

help : login

• ELECTRONIC ENLIGHTENMENT
• ABOUT
• BLOG
• BROWSE
• SEARCH