Storied Statistics: Behind the Numbers of Indonesian Mecca Pilgrims (1851–2016)

Storied Statistics: Behind the Numbers of Indonesian Mecca Pilgrims (1851–2016) Dulce van Vliet, TU Eindhoven Abstract This article examines the historical dataset of Indonesian pilgrims traveling to Mecca between 1851 and 2016 to reflect upon the epistemological and political dimensions of quantification. Drawing on debates in digital humanities, it argues that quantitative historical data, and statistics in particular, are neither self-evident nor neutral but deeply embedded in changing worldviews and governance practices. Using the concept of “storied statistics”, the study discusses how colonial authorities monitored the Hajj pilgrimage to prevent anti-colonial sentiment, later reframing it as an indicator of prosperity. The dataset is situated within a broader research agenda on sustainability and well-being, using the Dutch CBS Monitor as a heuristic model to explore historical developments. By tracing the entanglement of Hajj statistical records with colonial and post-colonial agendas, the article argues that historical quantitative data must be treated qualitatively. Keywords: historical quantitative data, storied statistics, colonial datasets, Indonesia, Mecca pilgrimage, Hajj 1. Introduction At her recent valedictory lecture, historian of technology Ruth Oldenziel reminded the audience that there is no such thing as self-evident or self-explanatory data: data need stories; however, she added, stories could also do with data (Oldenziel 2025). What Oldenziel wished to highlight was the fact that quantification must not be considered a prerogative of objectivity, in the same way that qualification should not promptly equate subjectivity. This issue has been discussed for some time in the context of digital humanities, namely under new concept proposals such Johanna Drucker’s “capta” instead of data, aiming to highlight agency in the production of data rather than the conflation of the phenomenon itself in its recorded observation (Drucker 2011); or Matthew Lavin’s counter proposal of “situated data”, defending that digital humanists should “use the word data mindfully and unapologetically based on the understanding that data are collected, assembled and recorded by people (or their instruments)” (Lavin 2025). In other contexts, data uncertainties and limitations —i.e. the need of further explanation, contextualization, or, one could also say, stories—have been discussed both regarding historical “hard” data (meaning, according to the authors, “readily observable and quantifiable” data, more connected to the natural sciences) and historical “soft” data (attributed to the social sciences). Here, however, the framing of the issue leans into problematizing the long-term historical aspect (multi-centennial timescales), rather than the data per se (Van Bavel et al. 2019). These are, in fact, two different issues that are equally important when thinking about data of the past. Not only was this data “taken”, or “produced”, rather than “given”, but it was taken and produced by historical actors which lived, communicated and created meaning within a different worldview system, or a different episteme, the term Michel Foucault used to describe a language spoken unconsciously in a specific period of history (Krauss 2011, 15–16). There is maybe not a better example of such critical entanglement as in the case of historical statistics. Etymologically derived from the Latin statisticum collegium (“council of state”), the designation also incorporated the meaning of the 18th century German term Statistik, which meant “state affairs” or “state craft” informed by a “science” of politics— meaning scientific inquiries by learned men should be undertaken to help direct the affairs of the state and develop the best kind of policies (Arneil 2020, 744). Statistics shape as well as reinforce specific worldviews: they allow for making sense of the world, not only because they name and organize that world, but because, in doing so, they are actually creating it (Porter 1995, 21). But worldviews change over time, just as the meaning and the stories that one might connect with the numbers recorded in a historical document. This article aims to explore the well-publicized as well as the more hidden stories connected with a specific set of numbers: the historical statistics recorded on the number of Hajj pilgrims departed from the Indonesian archipelago to Mecca between 1851 and 2016, which I have compiled from data published by the Indonesian and the Dutch national statistics offices in several formats (van Vliet 2025). 1 Working from this dataset as an example, I will argue that when working with historical quantitative datasets spanning long periods of time, it is fundamental to not only understand, but to make visible the changing historical circumstances throughout the data production process. This is particularly imperative in the case of historical data published by 1 The dataset can be accessed at https://doi.org/10.4121/6ec9e536-afd3-468a-9ca3-8a5b5ad3c199 The several sources are detailed in the documentation available with the dataset. national statistics offices. On the one hand, these are often some of the more readily available materials that researchers can find, since they are archived and preserved by the state apparatus. On the other hand, the same connection to the state, and governance in particular, make this data irredeemably connected to a very specific time and historical context: statistics are gathered to monitor specific policies, to give insight into (by then) current debates, and are firmly embedded in ideological and political projects. These concerns take on a much more crucial dimension when thinking about statistical records of colonial and post-colonial regimes, as is the case of the dataset in question. Historical anthropologist and (post)-colonial scholar Ann Laura Stoler has used the term “storied edges” to talk about the stories that can be found in the in-betweens, in the interstices of colonial archival materials: these are stories that do not fit in the established narrative of the colonial regime, but which are inadvertently preserved in the same archive (Stoler 2009, 143). Expanding on S toler’s concept, I propose to speak of “storied statistics” when analysing the more invisible stories connected to the dataset under investigation. 2. Quantifying Well-being in the Longue Durée As David Armitage and Jo Guldi so lucidly articulated, the mass retro-digitization of archival materials as well as the development of new computational methodologies created an auspicious environment for the historical disciplines to (re)turn to the project of the longue durée, abandoning the long trend of the microhistory perspective which had been popular in the last decades of the 20th century; furthermore, they argued, these were also favourable conditions for a much needed re-engagement of history with ongoing public debate (Armitage and Guldi 2015; Guldi and Armitage 2014). This aspiration is an important point for the research context of the case at hand. The Mecca pilgrims’ dataset was compiled within a research project that looks at how historical metrics can provide insights into well-being and sustainability developments in connection with industrialization and extractivism in the establishment of global supply chains during the early twentieth century until the present. The STONEM project—Sustainability Trade-offs in the Netherlands’ Entangled Modernization, 19002020 2 —investigates these well-being and sustainability developments in particular connected to the flow of resources between the Netherlands and the Global South, namely Indonesia. Central to the project’s approach is the Monitor Well-Being and the Sustainable Development Goals developed by CBS – Netherlands Statistics. 3 The monitor is organized in three main dimensions: Here and Now, which monitors what is happening now and includes indicators such as housing costs, satisfaction with life, (healthy) life expectancy or exposure to particulate matter; the dimension Later, which comprises “savings”, the accumulation (or depletion) of capitals (divided into Economic, Natural, Human and Social capital) that future generations are estimated to have based on current spendings—examples of indicators include education level, trust in institutions, household debt or cumulative CO2 emissions; and the dimension Elsewhere, which aims 2 The project is funded by the Dutch Research Council (NWO) and based at Eindhoven University of Technology. More information can be found at https://stonem.org/ . 3 This monitor was created from a mandate of the Dutch government and conceptualized following the recommendations of the Council for European Statisticians (CES), with the Brundtland definition of sustainable development as its foundational principle (Horlings and Smits 2019). It is based on the following definition: “Well-being concerns the quality of life here and now as well as the extent to which this quality is achieved at the expense of future generations or of people in other countries” (Horlings and Smits 2019). to monitor the impact of the country in the well-being of people living elsewhere in the planet—with indicators such as imports of resources, land footprint or official development assistance. 4 We want to use this monitor as a heuristic model to investigate historical sustainability trade-offs. The advantages of using this monitor as a blueprint for historical research is that it discourages cherry-picking single issues and pushes for charting and exploring historical interactions between trade and living conditions, working conditions and personal and social changes, but also environmental pressures and economic development. By consolidating economic, environmental and social concerns into a data model, the monitor also connects with environmental historian S verker Sörlin’s Anthropocene Weltanschauung argument, in which a data model and a particular worldview embody and reinforce one another (Sörlin 2025). Furthermore, the monitor provides a helpful frame of reference, a common language to be used between policy makers, governmental bodies, the press, civil society, and academic researchers, namely historians. The project builds on previous research that reconstructed historical monitors for the Netherlands between 1850 and 2010 (Lintsen et al. 2018). This research showed that the monitor had the potential not only for conveying long-term dynamics, but also for making visible changes in the social agenda over time, in terms of what is important for historical societies. For instance, while extreme poverty was a concern of late nineteenth century Dutch society, climate pressures and natural environments are the most 4 The monitor is published annually on the website of Statistics Netherlands. See, for example, the 2025 Monitor at https://www.cbs.nl/en-gb/visualisations/monitor-of-well-being-and-the-sustainabledevelopment-goals. important topics today. And these are not independent developments: while industrialization helped lift populations from poverty, it did so at the costs of the environment (Smits and Lintsen 2018). While the monitor was originally created for the Netherlands, its focus on the Brundtland definition of sustainable development, as well as its support of the proposals of the Council for European Statisticians (CES) and its alignment with the United Nations SDG’s, make for convincing arguments that could support, theoretically, the adoption of this model to monitor well-being and sustainability in other parts of the world. My idea was to try to reconstruct an historical monitor for Indonesia, at sub-national level; my motivations included wanting to try to “flip the monitor”, so to speak, in order to give center stage to the periphery (from the perspective of the Netherlands) and to place the Netherlands in the Elsewhere dimension of the monitor. My first step, then, was to explore and locate historical data that I could use as relevant indicators for the thematic sections of the monitor. 3. Historical Statistics and Big Data of the Past Datafication has been defined as the process of putting a phenomenon in quantified form so that it can be analysed and tabulated, as well as a kind of dematerialization that converts natural phenomena into symbolic material that can be indexed and searched. It is usually associated with the most recent stage of digitalization, where big data and algorithms are used for “quantifying life” through digital information, often for economic purposes (Wickberg et al. 2024, 2). The process of datafication and quantification is central to the well-being and sustainable development monitor. But while digital technology was instrumental to create the datafied world of the present, data streams of different yet considerable sizes have existed for a very long time. In fact, as digital humanists Frederic Kaplan and Isabella di Leonardo have argued, one might even be able to talk about Big Data of the Past (2017). Kaplan and di Lenardo claim that data acceleration regimes can be found in ancient cities of Mesopotamia, where administrative documents were produced on argyle tablets, compiling information on economic, diplomatic and commercial exchanges. Thousands of years later, nineteenth century industrialization saw the rise of information tracking systems such as the cadaster or the census. According to the authors, the fundamental characteristic of any historical recording technology that makes one able to talk about data acceleration, is their capacity to deal with an open-ended stream of information and reorganize it to fit a given information paradigm, creating new relationships between them: we call them regulated representations. (…) man-made material documents governed by a set of production and usage rules, that stand for something else, typically a complex event or phenomenon (2017, 4). With the well-being and sustainability model functioning as our current information paradigm, it was upon this regulated representation—organized in dimensions, themes and indicators—that I aimed to, in Kaplan and Di Lenardo’s terms , to “redocument the past”, which implies recollection (choosing and rejecting data, which will produce distortions by amplifying certain issues and diminishing others) and remapping (bending the historical data to fit the current regulated representation) (2017, 8). But these, eventual, current transformations must nevertheless contend with distortions not made by the present paradigms, but by the historical paradigms that shaped their production. And with this I particularly mean one spectacularly prolific data acceleration regime: colonialism. Colonialism and statistics have an entangled history: they could further the “principle of improvement” of nineteenth century social reform ideals; as well as to be used to identify “deviations from the norm” as “deviations from the mean” (Arneil 2020, 751; Stoler 2009, 30). Furthermore, numbers could convey a semblance of objectivity and fairness, speak the universalist language of Empire, all while providing authority to governing bodies (Porter 1995). The first efforts to produce statistics by colonial administrators in the Netherlands Indies date back to the early nineteenth century: these efforts aimed to understand what could be subject to taxation and focused only on the island of Java (Van der Eng 1996). With the creation of specialized departments in the colonial administrative structure from 1855 onwards—departments of Finance, Public Works, Education, Religion and Industry, Justice, and, Agriculture—the production of statistics became their own responsibility. This gathered information was fundamental for the creation of a “paper colony” that could be ruled directly from the metropole, in The Hague (Jeurgens 2015). The data collected by these administrative departments was published in Colonial Reports and also added to the Dutch Annual Statistical Yearbook published by the CBS. By 1925, the mandate to produce colonial statistics grew to such an extent that a centralized office was created in Weltevreden, a suburb of Batavia, now Jakarta: the Centraal Kantoor voor de Statistiek (CKS). By 1941, 700 people worked at the CKS (Van der Eng 1996). During the Japanese occupation and later the war for independence, data collection was intermittent. After independence, the CKS merged with the Kantor Penjelidikan Oemoem (KPO) and formed the Kantor Pusat Statistik (KPS), under the department of economic affairs. By the 1960s, with the mandate to produce population This is, of course, particularly critical in terms of colonial data. Digital historians are well aware of the challenges to decolonize the archive (Zaagsma 2023). And specific proposals such as Data Envelopes for Cultural Heritage Data provide important frameworks to reflect upon the positionality of the stakeholders involved in creating meaning from archival materials, from historical producers to contemporary annotators (Eskevich and Luthra 2024). However, this perspective must be applied not only to qualitative but also to quantitative historical data, such as statistics produced for purposes of (colonial) administration and governance. What was measured, how it was measured, the categories, classifications, and level of detail were all determined by the concerns and priorities of historical political regimes. These datasets reflect the worldview of their producers—and how that worldview evolved over time. The Mecca Pilgrims dataset is a particularly illustrative example of the multiple narratives behind the numbers: these figures are not only linked to economic and social developments but also reveal the shifting perspectives and public framing of their colonial producers, moving from a tool of surveillance and control to an indicator of well-being. In fact, these historical records may tell us more about the people who created them than about the phenomenon they sought to represent. Contemporary public debate thrives on numbers and quantification. But according to Armitage and Guldi, historians might be partly to blame for this state of affairs: retreating to microhistory, they “stopped writing for the institutions of the world government” and “economists took their place” (2015, 237). Returning to the public arena, the historian must be able to, in the words of Ruth Oldenziel, put forth “the poetry of data and the precision of stories” (2025). Because, contrary to prevailing assumptions, the boundary between quantitative and qualitative inquiry—between numbers and stories—is not as rigid as it is often framed. As data scholar Jacqueline Wernimont reminds us, (…) those who work regularly with quantifications understand the paradox that the ultimate test of quantitative methods and results… is not quantitative, but qualitative: do the numbers give us something we can work with?(2021, 430) References Armitage, David, and Jo Guldi. 2015. “The Return of the Longue Dure e: An AngloAmerican Perspective.” Annales. Histoire, Sciences Sociales - English Edition 70 (2): 219–47. https://doi.org/10.1017/S2398568200001126. Arneil, Barbara. 2020. “Origins: Colonies and Statistics.” Canadian Journal of Political Science 53 (4): 735–54. https://doi.org/10. 1017/S000842392000116X. Centraal Kantoor voor de Statistiek. 1932. Indisch Verslag. Landsdrukkerij. https://historisch.cbs.nl/detail.php?nav_id=7-1&index=&id=391097833. Davis, Hannah. 2020. “A Dataset Is a Worldview.” TDS Archive, March 7. https://medium.com/data-science/a-dataset-is-a-worldview-5328216dd44d. Drucker, Johanna. 2011. “Humanities Approaches to Graphical Disp lay.” Digital Humanities Quarterly 5 (1): 1–21. Eskevich, Maria, and Mrinalini Luthra. 2024. “Data-Envelopes for Cultural Heritage: Going beyond Datasheets.” Proceedings of the Workshop on Legal and Ethical Issues in Human Language Technologies @ LREC-COLING 2024, 52–65. Guldi, Jo, and David Armitage. 2014. The History Manifesto. Cambridge University Press. Haccou, J. F. 1947. Nederland en Indonesië. H. E. Stenfert Kroese’s Uitgevers Maatschappij N.V. https://resolver.kb.nl/resolve?urn=MMKB28:011581000:00001. Istahar, Hasbiyani Kamilah, Kirana Putri Aulia, Windy Dermawan, and Akim Akim. 2024. “Indonesian Hajj Diplomacy To Increase Hajj Quota During The Joko Widodo Presidency: 2015-2023.” Dynamics of Politics and Democracy 2 (2): 39–60. https://doi.org/10.35912/dpd.v2i2.1884. Jeurgens, Charles. 2015. “Networks of Information. The Dutch East Indies.” In Exploring the Dutch Empire. Agents, Networks and Institutions, 1600-2000, edited by Ca tia Antunes and Jos Gommans. Bloomsbury Academic, an imprint of Bloomsbury Publishing Inc. Kaplan, Fre de ric, and Isabella Di Lenardo. 2017. “Big Data of the Past.” Frontiers in Digital Humanities 4 (May): 12. https://doi.org/10. 3389/fdigh.2017.00012. Krauss, Rosalind E. 2011. Under Blue Cup. MIT Press. Lavin, Matthew. 2025. “Why Digital Humanists Should Emphasize Situated Data over Capta.” In The Companion to Digital Humanities in Practice, 1st ed., by Constance Crompton, Laura Estill, Richard J. Lane, and Ray Siemens. Routledge. https://doi.org/10.4324/9781003327677-10. Lintsen, Harry, Frank Veraart, Jan-Pieter Smits, and John Grin. 2018. Well-Being, Sustainability and Social Development: The Netherlands 1850–2050. Sp ringer International Publishing. https://doi.org/10.1007/978-3-319-76696-6. Oldenziel, Ruth. 2025. “The Poetry of Data, the Precision of Stories.” Valedictory Lecture. Eindhoven University of Technology, November 6. https://assets.w3.tue.nl/w/fileadmin/content/universiteit/Academische_plechti gheden/Redes/2025/Invitation%20valedictory%20lecture%20prof%20Ruth%2 0Oldenziel.pdf. Porter, Theodore M. 1995. Trust in Numbers: The Pursuit of Objectivity in Science and Public Life. Princeton University Press. Ray, Sandeep. 2023. “What Did the Haji Jawa See in Mecca? A Film from 1928 as Primary Source.” Historical Journal of Film, Radio and Television 43 (4): 1023–49. https://doi.org/10.1080/01439685.2023.2193044. Smits, Jan-Pieter, and Harry L intsen. 2018. “The Great Transformation an d the Questions.” In Well-Being, Sustainability and Social Development: The Netherlands 1850–2050, edited by Harry Lin tsen, Frank Veraart, Jan-Pieter Smits, and John Grin. Springer International Publishing. https://doi.org/10.1007/978-3-31976696-6_2. So rlin, Sverker. 2025. “Planetary Environing: The Return of Boundaries as a Category in Global Environmental Governance.” The Anthropocene Review, January 23, 20530196241308865. https://doi.org/10.1177/20530196241308865. Stoler, Ann Laura. 2009. Along the Archival Grain: Epistemic Anxieties and Colonial Common Sense. Princeton University Press. Van Bavel, Bas J. P., Daniel R. Curtis, Matthew J. Hannaford, Michail Moatsos, Joris Roosen, and Tim Soens. 2019. “Climate and Society in Long-term Perspective: Opportunities and Pitfalls in the Use of Historical Datasets.” WIREs Climate Change 10 (6): e611. https://doi.org/10.1002/wcc.611. Van der Eng, Pierre. 1996. “Historical Economic Statistics in Indonesia: Conti nuity and Change.” Newsletter of the Asian Historical Statistics Project 3 (October): 12–15. Vliet, Dulce van. 2025. “Historical Dataset: Indonesia Mecca Pilgrims Statistics, 18512016.” Version 1. With TU Eindhoven , Department of Industrial Engineering and Innovation Sciences. 4TU.ResearchData, December 4. Dataset/xlsx; source list and raw dat a/xlsx. https://doi.org/10.4121/6EC9E536-AFD3-468A-9CA38A5B5AD3C199. Vredenbregt, J. 1962. “The Haddj. Some of Its Features and Functions in Indonesia.” Bijdragen Tot de Taal-, LandEn Volkenkunde 118 (1): 91–154. https://doi.org/10.1163/22134379-90002165. Wernimont, Jacqueline. 2021. “Quantification.” In Uncertain Archives, edited by Nanna Bonde Thylstrup, Daniela Agostinho, Annie Ring, Catherine D’Ignazio, and Kristin Veel. The MIT Press. https://doi.org/10.7551/mitpress/12236.003.0047. Wickberg, Adam, Susanna Lidstro m, Amanda Lagerkvist, et al. 2024. “The Mediated Planet: Datafication and the Environmental SDGs.” Environmental Science & Policy 153 (March): 1036–73. http s://doi.org/10.1016/j.envsci.2024.103673. Zaagsma, Gerben. 2023. “Digital History and the Politics of Digitization.” Digital Scholarship in the Humanities 38 (2): 830–51. Zeilinger, F. A. 1933. Kapitaal en kapitaalvorming in de inheemsche maatschappij van Nederlandsch-Indië. Veenman. https://resolver.kb.nl/resolve?urn=MMKB05:000033974:00007.