Introduction
This project intends to research the spatial distribution of different religions in Paramaribo in 1841. By looking at the locations where people from different religions lived, we can examine if certain religions were concentrated in certain neighborhoods of the city. Our research question is thus: "What is the spatial distribution of religious groups in Paramaribo in 1841?"
This research question can shed more light on the religious landscape of Paramaribo and show us to which extent people of different religions lived among each other or were concentrated in certain neighborhoods in the city. It questions the degree to which religion was a decisive factor in the ways in which the city was organized.
We chose 1841, because in this year, religion is recorded the most in the dataset.To determine the year for which the most data was recorded, we calculated the number of people in the dataset of whom we know their religion in every given year. The graph that is included below, shows us that the most data exists for religion in the year 1841.
Figure 1: Number of recorded people and religion in each year
About the data
It is important to note that the number of people for whom religion was recorded fluctuated significantly from year to year. By choosing the year with the most recorded information, we hope to be able to examine the spatial distribution in the most representative way. However, further examination of the dataset can give us more information on the reasons behind these fluctuations. Were less people recorded in the dataset for these years altogether or was only religion less recorded? What are the reasons for this? If the latter is the case, were all religions equally less recorded or some religions more often than others?
If there is an inconsistency in the recording of religions, meaning not every person wanted to, could (because of eventual disadvantages) or was asked to name their religious affiliation when registering, it is possible to draw wrong conclusions about the spatial distribution of people with certain religions. This could pose significant problems for our analysis or any analysis based on the variable "religion" in the dataset of the Paramaribo ward registers. Therefore, it is important to remember that it remains unclear how representative the data is.
In the previous graph, we also see the total number of people that were recorded each year. As you can see, this number varied significantly over time as well. When you compare the number of people recorded each year with the number of people for whom religion was recorded, you can see that there are some differences in the fluctuation between these graphs. This means that for some years, information about religion is included relatively more in the dataset than in other years. We have not examined wether their were significant fluctuations between different religions as well. As will become clear later, the categories that were for religion in the data set can be problematized as well.
Triplets
In order to retrieve the data regarding religion for the year 1841, we searched for the following triplets:
- Subject = observation (person), predicate = has religion, object = religion [perosnincontext.org/model - hasReligion]
- Subject = source (ArchiveComponent), predicate = date created, object = year [schema.org/ - dateCreated]
To connect these two triplets "PersonObservation" (religion) and "ArchiveComponent" (year) we use the predicate "hadPrimarySource". Briefly summarized, we connect the "person" who has a certain "religion" to the "archival component (source)" with the sub-component "year".
Below, you can find the results when we filter for 1841 and present the data for every recorded religion (there are 44 different religions included in the dataset). Most of the people in Paramaribo in 1841 were Hervormd, followed by Luthers and Hernhutter.
Table 1: Number of people per religion (1841)
Spatial Distribution Data
To examine the spatial distribution of religious groups, our next step consisted of linking the component "wijk and street address" to our existing data. In this way, we can visualize considerable variations between different religions in different neighborhoods. The neighborhoods in Paramaribo are categorized by a letter which go from A to F, and 1e buitenwijk and 2e buitenwijk as additional neighborhoods. Our triplets were:
- subject = PostalAddress, predicate = streetAddress, object = respective address (housenumber + streetname)
- subject = PostalAddress, predicate = wijk, object = respective wijk (A to F +1e buitenwijk + 2e buitenwijk)
additionally
-
subject = PostalAddress, predicate = streetName, object = respective street name
-
subject = PostalAddress, predicate = houseNumber, object = respective house number)
We link these triplets to "religion" over "person" by the predicate "address".
This gives us the following results:
Table 2: Number of people per religion with address (1841)
Religious Categories
Another point that we need to take into account is the labels used to denote different religions. In the table, we see a wide range of these labels. However, what stands out is that separate labels seem to refer to the same religion, such as 'heiden' (2) and 'afgodisch' (4).
Similarly, there are many separate categories that seem to refer to people of Jewish religion, such as those referred to as 'Israeliet' (348), 'Ned. Israeliet' (280), 'Hoogduits Israeliet' (251), and Jewish (2). As this concerns a big number of entries, this would benefit from further data cleaning for clearer visualization.
Instead of listing each religious denomination individually, we categorize them into four main groups:
1. Protestants: Includes various Protestant denominations such as Hervormd, Luthers, Gereformeerd, Evangelisch, and others like the Waalse Kerk and the Hernhutter. These are all part of the broader Protestant tradition.
2. Jews: This category aggregates various Jewish denominations, including Israeliet, Ned. Israeliet, and different Portuguese and Dutch variants of Jewish communities.
3. Catholics: Encompasses Catholic denominations, such as Rooms-Katholiek and Episcopaals.
4. Others: This is category includes religions or affiliations that do not fall under the main three groups. It includes categories such as "Geen Religie" (No Religion), "Afgodisch" (Idolatrous), and less common religious affiliations that may not fit neatly into the Protestant, Jewish, or Catholic categories.
Table 3: Number of people for four main religious categories
Figure 2: Distribution of Religion per Wijk
Relative Distributions
The following two graphs show the relative distribution of the different religions in each neighborhood. Here we see that the religious constitution of the neighborhoods differed significantly. For example, almost all those that were Gereformeerd lived in the 1e Buitenwijk, whereas people of the religion Luthers made up about 20 percent in each neighborhood. We can conclude that some religions tended to be geographically concentrated, while others were not.
Figure 3: Detailed relative Distribution of Religion per Wijk
Figure 4: Relative Distribution of Religion per Wijk
Visualization on a Map
The historical base map used for the spatial visualization is Plan van Paramaribo by W.L. Loth, a governmental surveyor, originally published in 1898 by J.H. de Bussy (Amsterdam) and included in W.L. Loth’s Beknopte aardrijkskundige beschrijving van Suriname (second revised edition, 1904). The original copy consulted is held by the Koninklijk Instituut voor de Tropen (KIT) and the Library of Leiden University. Although the demographic data analyzed pertains to 1841, the 1898 map was selected as the most detailed and spatially coherent historical source available, providing a reliable representation of Paramaribo’s urban structure for the purposes of this study.
Figure 5: Map of Paramaribo, 1898 (source: wikimedia.org/wikipedia/commons/4/44/Paramaribo1898metwijkindeling)
The process was carried out using QGIS, an open-source geographic information system software. First, the historical map was georeferenced through the QGIS Georeferencer tool.
Following georeferencing, vector layers were created to delineate the wijk boundaries visible on the map. These served as the spatial units for mapping religious populations. Demographic data was visualized by aggregating religious groups according to the number of adherents in each wijk. Only groups with sufficiently large populations were mapped individually. Groups with very small populations, whose spatial patterns could not be meaningfully distinguished at the wijk level, were not included in the individual group visualizations to avoid misrepresenting distribution patterns. For each selected group, a thematic map was produced showing the relative distribution across the districts. Areas without shading indicate districts where no adherents of the specific religion were recorded in the historical sources.
All maps were designed for publication purposes, ensuring appropriate resolution (300 dpi) and a consistent layout style. The original scale indicator and north arrow visible on the historical map were preserved to maintain historical authenticity. Legends were customized to indicate both population ranges and the absence of adherents, supporting clearer interpretation.
Figure 6: Spatial Distribution of Nederlands-Israelieten in Paramaribo (1841)
Figure 7: Spatial Distribution of Hoogduits-Israelieten in Paramaribo (1841)
Figure 8: Spatial Distribution of Israelieten in Paramaribo (1841)
Figure 9: Spatial Distribution of Portugees-Israelieten in Paramaribo (1841)
Figure 10: Spatial Distribution of People without Religion in Paramaribo (1841)
Figure 11: Spatial Distribution of Gereformeerden in Paramaribo (1841)
Figure 12: Spatial Distribution of Rooms-Katholieken in Paramaribo (1841)
Figure 13: Spatial Distribution of Hernhutter Adherents in Paramaribo (1841)
Figure 14: Spatial Distribution of Luthers Adherents in Paramaribo (1841)
Figure 15: Spatial Distribution of Hervormd Adherents in Paramaribo (1841)
Figure 16: Spatial Distribution of Religions in Paramaribo (1841)
Conclusion
This project connected the two variables "religion" and "wijk/address" from the dataset "Paramaribo Ward Registers 1828-1847" with a map as an extra source to answer the research question "What is the spatial distribution of religious groups in Paramaribo in 1841?". It did so by connecting the triplets "Person - hasReligion - Religion" with "Source (ArchiveComponent) - Date created - Year," and "PostalAddress - Wijk - Respective Wijk".
Discussion of Results
As an initial, more general answer to the research question, we can conclude that people of Protestant faith (mainly Hervormd, Luthers, and Hernhutter) and Roman-Catholics were equally distributed across the neighborhoods. There are some exceptions to this rule in the case of certain denominations, such as the Gereformeerd who lived mostly in the 1e Buitenwijk. People of Jewish denominations were more commonly located in Wijk A, followed by Wijk B and E. Furthermore, there was a significantly high population of Israelits in Wijk C. At the same time, rather few Jewish settlements existed in 1e Buitenwijk, 2e Buitenwijk, and Wijk D. We conclude, therefore, that despite some tendencies, various religious groups are quite evenly distributed in Paramaribo. If we look at Figure 2, we can even see that within the same household (marked as one street address), people with many different religious affiliations seemed to live under the same roof. This is remarkable as it would support Alex van Stipriaan's paper on religious tolerance in Paramaribo. In his research, he included various citations from Dutch observers confirming that within the same family, members had various religious affiliations. The reason for this was frequent sexual relations between slaves and slave owners, especially in the case of Jewish men having children with non-Jewish women (van Stripriaan 1997, 76f.). Paramaribo was a plantation society, and the first settlers and plantation owners were not the Dutch, but Portuguese Jews who gradually moved from the more rural areas in Suriname to the urban center of Paramaribo after the plantation economy declined in the latter half of the 18th century. From the aforementioned relations between African slaves and Jewish men, a new social and religious group of Jewish Mulattoes formed. However, the white Jewish settlers denied these people the right to officially become a recognized community with their own synagogue as a center. In the eyes of the white settlers, this would have violated the ideas on traditional Jewish identity in Paramaribo (van Stripriaan 1997, 75, 80f.). We can therefore conclude that the large distribution of Jewish people across different Wijken can also be attributed to the inability to establish a centralized community. To show these distributions, we visualized our data on a contemporary map of Paramaribo (1889). While this ensures historical accuracy, we were not able to highlight other important locations and buildings in the map, such as churches or ports. This could have further helped the interpretation of the data.
Tabular Data
We could have answered our research question using tabular data. This would have depended on the structure of the database, meaning according to which variable the database is organized (neighborhoods, individuals, etc.). Given that our data is a ward register and focused on "person", we can assume that we could choose "religion" and "wijk" as variables to create a new database in a first step. In a second step, we would filter for the year 1841. Next, we would have needed to clean the data by calculating the number of people of each religion per wijk to get the information on, for example, how many Catholics and Jews lived in Wijk A. To achieve an overview as in Figure 2, we additionally would have needed to add up all denominations of the protestant church, the Jewish community, etc., and create the four categories "Protestant, Catholic, Jewish, and Other". Furthermore, to calculate percentages for each religion per wijk, we would have needed to calculate the total number of people per wijk and merge it with the database on religion by the variable "wijk". This step of cleaning the data poses the main difference to linked data using triples. While with triples, we could just query for different variables, with tabular data, we would have needed to calculate each variable. Consequently, using Linked Data is more efficient and circumvents the mistakes related to the cleaning of data and merging of data sets.
Outlook
Further research could use our data to connect it to other variables such as "enslavement" or "gender" to further research other reasons for the unique spatial distribution of different religious groups. By connecting our data to "profession", other researchers could look at the interrelations between certain occupations and the religious affiliation of the people who carry out this work. As an example, further research could focus on the location of banks and pawnshops in connection with Jewish settlements.
This data story was created by Meret Bittman, Xiaoli Fan, Daniëlle de Kurver and Alexander Nuijten
