- 1 This article benefited from the kind and stimulating reviews of two anonymous reviewers and the iss (...)
1Digital technologies are now pervasive in our daily lives, and their generalization questions scientific research as much as public authorities. They transform the way in which inhabitants experience their territory, and disrupt the relationships between space and power, i.e. the way in which the territory is produced or governed by institutions. This article aims to shed light on the way in which digital technologies and their everyday use transform the relationship between risk, territory and inhabitants. To do this, we analyze the multiple uses of mainstream digital applications (WhatsApp and Facebook) revolving around the floods caused by the Ocloro River in San José (Costa Rica), as local residents use them to monitor floods and anticipate their possible damage. Although they are not limited to its action, these practices are notably driven by the neighborhood Crisis Management Committee (CMC) of Barrio Luján, a socio-political organisation created by public policy in Costa Rica to facilitate the involvement of populations in risk management actions, in an internationally promoted, bottom-up fashion (Quenault 2015; UNISDR 2015).
- 2 During the course of the fieldwork, there were no other alert groups of this type on WhatsApp (that (...)
2A pioneer country in terms of environmental risk management (earthquakes, volcanic eruptions, floods, etc.), Costa Rica was also connected early to the internet, compared to its regional neighbours (Siles González 2008). Its population is now massively using smartphones and digital applications. According to Brenes Peralta and colleagues (2024), 98% of Costa Rica's adult population uses WhatsApp on a regular basis, making it by far the most used social application in the country, while 76,9% uses Facebook. By following the activity of Barrio Luján neighborhood committee’s Facebook page and, even more so, the conversations on a local WhatsApp group, managed by the inhabitants of the Ocloro watershed, it was thus possible for us to observe in detail and longitudinal the way in which citizens organize themselves in the face of environmental risk, and how the use of these technologies redistributes agency between territorial actors2.
3This article aims to answer the following question: how does the use of mainstream digital technologies reconfigure the participatory turn in the government of risk? We demonstrate how participative risk management, a political mechanism referred to by some authors as “neo-liberal risk governance” (Metzger and Robert 2015), is being reappropriated by populations. This is particularly evident through the use of everyday digital technologies, which empowers residents to develop their skills and results in broader environmental mobilisation, and a reversal of roles between populations and public authorities in the production of risk knowledge. We detail below our theoretical framework, then the materials and methods used to respond to it.
4Our study is at the intersection of three main bodies of literature that we propose to connect: works analyzing citizen engagement in risk management, often anchored in neoliberal governance schemes; on the digitalization of risk management; and the geography of digital uses in urban spaces.
5Literature has shown how contemporary risk management policies shifted to a neoliberal perspective, based on the participation of “civil society”, community engagement and the delegation of responsibilities previously vested in the State to populations or private sector (Tierney 2015; Metzger and Robert 2015). Although the involvement of people in risk reduction is supposed to improve their living conditions as part of an people-centred approach (Wolff 2021), scientific literature shows that, in practice, it can lead to the depoliticisation of social problems and inequalities, and the disengagement of the State (Author, 2019). This can be particularly problematic in the context of the Global South, where the State may fail to support individuals in improving their living conditions, and inhabitants may be forced to undertake risk management themselves (Araos et al. 2021; Rahman and Hickey 2019). Moreover, an analysis of participation experiences within risk management reveals the reproduction of top-down approaches (Cook et al. 2025), in which the main goal remains to raise residents' awareness of the hazard and less to focus on co-constructing solutions with residents. These trends in risk governance are increasingly relying on digital technologies, which are considered a means of facilitating the enrolment of civilian populations (Sakurai and Murayama 2019). Nevertheless, scholarly contributions on digital media and risk management often focus on top-down, institutional initiatives (Torpan et al. 2024). Some research has studied how digital technology makes it possible to model the reactions of populations or to inform them better and more quickly (Matczak et al. 2016; Mees et al. 2016), or to coordinate volunteers during floods (Sharp and Carter 2020). The democratization of digital technologies has in fact been seen as a potential tool for making risk management “data-driven” (Delaney and Kitchin 2023). Beyond social media, other digital devices such as sensors installed in sanitation networks offer, for example, the opportunity to strengthen preventive flood management in the face of more intense precipitation episodes (Keung et al. 2018; Oberascher et al. 2022). The literature also reports on participatory projects that rely on digital technology and resident participation in order to complete data and improve river modeling (Makita 2021; Wolff 2021), occasionally highlighting the ability of populations to spontaneously self-organize through the use of applications during a major crisis (Paul and Sosale 2020). Although a number of works describe the emergence of digital apps, within scientific projects involving populations to improve the collection of data on hydrological risks, they offer little insight on resident-initiated data production and digital uses.
6While a number of studies describe top-down initiatives to streamline collective organization in the face of hazards, the ordinary uses of digital technology by city residents have been much less studied. In continuation of the program proposed by Peyroux and Ninot (2019), we wish to understand the urban digital reality in its entirety, through the concept of the “generalized digital”, thus seeking to shed light on the way in which the everyday uses of these technologies transform our perception of the territory, and the way in which citizens coordinate, act together and inhabit urban space. Building on the work of Ash et al. (2018) or Kinsley (2013), we seek to understand how digital technologies do not “undo” the geography of space, but transform its frame, particularly through the use of spatialized digital media in the context of a Global South city, frequently affected by flooding.
7In this perspective, we mobilize the concept of “digital placemaking” (Halegoua 2019; Halegoua and Polson 2021), considering that “digital practices enable people to emplace themselves and others” (Halegoua & Polson, 2021, p.574). We draw inspiration more generally from work in geography which analyzes the use of digital networks as vectors of a collective and possibly political relationship with inhabited space, notably through the construction of commons (Denis and Marquet 2019; Dodge and Kitchin 2013). Following on from Johnson and Halegoua (2014), we are interested in the use of social media for the construction of local communities, with a strong geographical anchor (Loubière et al. 2024).
8A number of recent works have in fact shown how digital social networks are mobilized by residents to make sense of their territory and invest it in an aesthetical and political way. Bronsvoort and Uitermark (2022) describe how the gentrification of a neighborhood in the east of Amsterdam is embodied in the Instagram publications of its new traders and residents, who thus take a direct interest in the change in image they contribute to performance. In a more explicitly political manner, Zur (2024) provides an enlightening analysis of the digital political activism deployed by social groups in a disadvantaged neighborhood in south Tel Aviv, and the way in which digital communication technologies redistribute agency within urban social groups, and opens the way to new forms of competition for the local political agenda. We extend this literature by opening this emerging perspective to a platform that has so far been little studied (WhatsApp), but also to areas from the Global South, much less visible than the large metropolises of industrialized countries, in a less planned and more informal urban context.
9In the same way, in this article we analyze the use of ordinary digital technologies (WhatsApp and Facebook), not technically geolocated, but anchored in the specific territorial sociability of the Ocloro river watershed (Crampton et al. 2013). By doing so, the article aligns with works that study mundane uses of digital technologies (Martin and Dagiral 2016). In the context of this article, studying flood risk management practices in everyday life refers to the ways in which people inhabit their neighbourhoods and the social and interpersonal relationships that develop within them. The aim is to examine how digital technology is incorporated into residents' risk management practices, however banal and invisible they may appear (Brekhus 1998; Certeau 1990). We consider these ordinary uses as political, both in the sense of the collective debate of a subject in a shared arena, and of the indirect, direct or even confrontational questioning of public authorities. In our case study, digital technology, linked to citizen mobilization, increases the residents’ agency, and their capacity to hold public authorities accountable.
- 3 Comité Communal de Emergencia, in Spanish. Furthermore, among the semi-structured interviews carrie (...)
10This article is based on long-term qualitative investigation carried out between 2015 and 2024, first as part of doctoral research on participatory crisis management systems in Costa Rica (2015-2020), then as a follow-up investigation and data collection carried out between 2022 and 2024. We carried out 42 face-to-face semi-structured interviews with institutional representatives (19), and with residents (23) of Barrio Luján, a San José neighborhood with a very active crisis management committee (CMC)3. We also conducted an analysis of institutional archives and realized 15 participant observations of public meetings, field visits with actors, and a 16-hour training on risk management of the Public City Hall to the residents, among others.
- 4 In order to quantitatively and systematically analyze the content of discussions on the Alerta Oclo (...)
- 5 Data loss was caused by changes of smartphone and app updates.
- 6 In order to carry out the semantic analysis, we informed the participants in the WhatsApp group of (...)
11The analysis of digital uses consisted first of all in a long-term qualitative observation (or “chatnography”; see Käikho (2020) of a WhatsApp group set up by the local committee, in which one of the authors took part (as a local resident and researcher). Second, we carried out a quantitative and semantic analysis of the WhatsApp group “Alerta Ocloro” with the CorTexT software (Breucker et al. 2016), based on four corpora of WhatsApp messages collected between 2016 and 20234. Due to various technical constraints5, we were able to collect 4479 messages over several discontinuous periods of time. The table below summarizes the extent and activity of the different temporal sub-corpora. Corpora 1, 2 and 3 do have a relative continuity, but corpus 4 was created more than two years after the end of corpus 3; its duration (more than a year and a half) allows us to evaluate the more recent uses of the WhatsApp group over an extended period6.
- 7 Dates are given in the format YYYY.MM.DD.
|
Corpus
|
Starts7
|
Ends
|
Duration (days)
|
Posts
|
Average posts per day
|
|
Corpus 1
|
2016.05.30
|
2017.04.29
|
303
|
1927
|
6,4
|
|
Corpus 2
|
2018.06.04
|
2018.09.13
|
101
|
1097
|
10,9
|
|
Corpus 3
|
2019.01.16
|
2019.07.03
|
168
|
838
|
4,9
|
|
Corpus 4
|
2022.01.16
|
2023.08.22
|
583
|
617
|
1,1
|
|
Total
|
2016.05.30
|
2023.08.22
|
1155
|
4479
|
3,9
|
Table 1. Description of WhatsApp sub-corpora
12We also analysed the publications on the committee’s Facebook page, their themes and the conversations they sparked. This offers a broader view of the ecology of digital spaces used by inhabitants to reinforce community organising in the face of floods. Based on archival work, we were also able to carry out a multi-sited analysis of digital uses in other spaces.
13This article shows how digital citizen mobilizations are transforming the relationship of inhabitants to their territory. Studying everyday digital uses, it describes the transition from an alert regime, aimed at immediate risk prevention in an emergency situation, to a broader environmental political mobilization which takes into account multiple aspects of the territory and the factors of urban flooding: the evolution of water flow, but also land use (and its evolution over time), the deployment of infrastructure, and pollutions.
14In this first section, we demonstrate how digital technologies are used in risk management in San José, with a focus on how residents use them for self-organisation in their daily lives, particularly in the event of floods. We base our work on the study of one of the Crisis Management Committees (CMC) created by the Municipality of San José, in the Barrio Luján neighbourhood. The creation of this CMC is part of the implementation of the national risk management policy (Guevara Viquez 2020). While they are set up and monitored mainly by the municipalities, these committees relay on voluntary work from citizens. In the city of San José, CMCs are promoted in neighbourhoods frequently affected by flooding, which is the main threat impacting the capital. Although they do not replace the first responders (police, fire brigade), these committees are supposed to visit the disaster site after an emergency, collect information from affected households and send it to the town hall. Outside emergency situations, they can work with their neighbours to raise awareness and promote prevention. The Municipality of San José initially tried to leverage WhatsApp to foster relationships between CMC representatives and crisis-related public institutions, but the group chats they created often ended up in a vertical, unidirectional communication, aimed at spreading information and instructions to the residents. The WhatsApp group we studied, created by residents themselves, includes CMC participants but is not limited to them, thus bypassing institutional efforts.
15We analyze here how it was built and how its activity unfolded throughout the years. These practices are infrapolitical in the sense of Scott (1990), as they are not explicitly qualified as political in the traditional sense (e.g. a demonstration in a public space), yet they still have political significance. This concept allows us to establish a continuum between practices that are considered mundane in the context of flooding and the everyday relationship with the Ocloro river, and potential political mobilisation (Marche 2012).
- 8 All names in this article are pseudonyms. Individual consent was obtained from the participants in (...)
16To deal with floods, a citizen early warning system emerged in 2016 in Barrio Luján on WhatsApp, based on the work done by the CMC. This flood prevention alert network has been set up throughout the Ocloro basin and gathers residents who are not exclusively members of the crisis management committee. Intended to anticipate the floods in real time and to organize collective response, this group was created by Alberto8, a young architecture graduate and founder of a local collective for the protection of rivers. Through his work, Alberto knew a number of local figures from the southern neighborhoods of the Costa Rican capital, whom he included in this WhatsApp group, with the initial objective of being able to monitor rising water in real time. The initial objective was for people living upstream of the river to warn those downstream of heavy rainfall and potential overflows to come. The group then grew through the action of residents, in particular the coordinator of the Barrio Luján CMC, by co-opting neighbors living beyond Barrio Luján, directly or indirectly affected by the floods of the Ocloro river, and willing to organize to improve their situation. Through this work of recruitment, the WhatsApp group also got to include students, journalists and interested institutional representatives (in particular representatives of the Human Rights Commissioner, whose role is to ensure that institutions comply with the legal and moral framework in the interests of citizens). During some time periods through the years, municipal representatives were also involved (one elected councilor from San José and one administrator from the municipality of Montes de Oca, for instance).
17Based on the qualitative and quantitative analysis of activity on this remarkably long-lived WhatsApp group, we will show that the group has two main modalities: the most frequent one is routine and consists in welcoming new participants, sharing information regarding weather conditions and the municipality of San José; it involves few messages, often published by the group administrators, two women. The other one is a crisis modality, during heavy rain episodes, and involves the publication of many messages by many different participants.
18One of the authors of this article has been a member of the group almost since its origins, in 2016, and was thus able to carry out a long-term observation of the messages exchanged over these eight years. As we can see in Figure 1, this WhatsApp group is characterized by a significant longevity for a group gathering relative strangers in an instrumental way. Although the average number of messages per day declines over time, it remains relatively high even in the most recent period; it averages 3.9 messages per day over more than three cumulative years of observed activity.
Figure 1. Number of messages per day over the duration of the corpus.
19Despite the discontinuity of our corpus, its longitudinal nature shows us the constancy of the activity, despite a gradual decline, over the eight years of activity. As seen in the figure above, the activity of the WhatsApp group revolves around punctual peaks of activity, during the overflowing of the river, and moments of social and political activity.
- 9 Any members who have not posted are unfortunately inaccessible to us by construction, but qualitati (...)
20The group had 37 members in October 2024, however the analysis of messages since 2016 makes it possible to identify 68 unique group members which posted a least one message on Alerta Ocloro, during the periods studied9. As in many informal digital spaces, contributions to the conversation are distributed in a particularly unequal manner (Hargittai and Walejko 2008). Thus, more than half of the messages were published by Marina and Bethy, two women who, as we will see, play a central role both in the animation of the WhatsApp group and in the citizen mobilization around the Ocloro river.
Figure 2. Total number of messages posted by each member of Alerta Ocloro.
- 10 Bethy was a housewife while raising her two daughters, then set up a home cleaning business, and re (...)
21These two women most often accomplish the general work of coordination in the group: they add new members, welcome them, occasionally remind us of the rules of collective expression, and frequently report incidents, precipitation and other works affecting the flow of the river. Both residents of Barrio Luján for many years, they are around fifty years old, and have a strong social capital within the neighborhood10. They will subsequently play a leading role in citizen mobilization against the work on the watercourse carried out by the municipality of San José and by private actors located on the banks of the Ocloro (an agri-food company and a real estate developer), which they believe to increase the frequency and magnitude of floods.
22Among other notable contributors to this WhatsApp group, we can highlight Daniel, an engineer and resident of the upper Ocloro basin, particularly active during the first years, who plays by the group's rules to the fullest and reports with great regularity the precipitations upstream, in order to warn residents downstream of possible overflows to come. Generally speaking, in addition to the two main leaders of the group, participation revolves around a core of around ten other main contributors, each having posted approximately a hundred messages; then follows the “long tail” of occasional commentators, by far the most numerous. Finally, it is interesting to note that, although he created it, Alberto is not the driving force behind this WhatsApp group, which is largely independent of its initial action: he posted 108 messages only, out of the 4479 archived.
23The concentration of expression around a few central protagonists, in addition to resonating with now classic work in digital sociology, is consistent with the work of Zur (2024), where a a small number of particularly active residents are propelled to the rank of “micro-celebrities activists” (Tufekci 2013). By distributing public speech and facilitating the coordination of collective action, mainstream digital technologies such as WhatsApp and Facebook open the way for the emergence of new figures, who locally replicate the figure of the political influencer. The fact that these WhatsApp “micro-celebrities” are here two women is part of the continuity of work on community organizing and the preponderant place of women in this “often prepolitical action” (Stall and Stoecker 1998).
24What do people talk about in this WhatsApp conversation? We can outline the semantic landscape of the conversation within the group, and distinguish different registers of expression, which we detail below.
Figure 3. Semantic analysis of discussions on Alerta Ocloro (translated to English)
- 11 The following four clusters are less important, comprising between 4 and 32 keywords, which actuall (...)
25We can identify three main modes of expression for the members of Alerta Ocloro, materialized by three main clusters in purple, green and blue11. These partially overlap the group's two modalities (emergency crisis management and routine). In purple, a first cluster brings together 64 keywords, with a connotation of collective mobilization on the subject of flooding. First of all, it includes the toponyms of the places where the residents affected by the floods live (Los Yoses, a sub-district of San José; districts of Curridabat, San Pedro; municipality of Montes de Oca), which shows how this digital sociability is anchored in a shared territory. A set of technical terms also appears, referring to urban hydraulic infrastructure and communications routes (bridge, street, avenue, pipe, basin, grid, gabions), as subject of complaints, and critical analyzes (reporting of road works, or their absence; the accumulation of waste on a grid etc.). Eloquently, we find in this same cluster the lexical field of public policies (municipality, support, commission, prevention, plan), and sometimes that of controversy or legal battle (complaint, issue), in which the above-mentioned municipalities are questioned for their (in)action. Finally, considering the two most frequent terms of the cluster (water and house), this first cluster clearly marks the expression of infra-political talk, bringing together the terms by which the members of the WhatsApp group monitor the extent of the damage caused by urban flooding, and seek to provide collective responses.
26The second cluster, in green, is numerically almost as important as the first one (n=62). It shows a vocabulary of sociability, anchored in the urban territory, and thus brings together terms designating both individual and groups, including the two most frequent ones, neighbors and neighborhood, but also Mr., Mrs., families, people, community, citizens, companions. In this cluster, the group mainly speaks about itself: terms appear which stand for the information process and the communication itself (information, message, chat, meeting, report), which suggests an intense activity of coordination and relay of information. Finally, we find in this cluster a good number of civilities (please, thanks, greetings, welcome), as well as, bordering the first cluster, terms which could also belong to it, and concern the general vocabulary of flood risk (risk, danger, sewage, problem, floods). This cluster thus reveals a second mode of expression, which refers to the organization of the WhatsApp group chat itself, and to the collective relationships that emerge in the face of the recurring problem of flooding. These first two clusters are the densest and most numerous, and that they are located close to each other, implying the sharing of numerous mutual links. Conversely, the next cluster is located at an opposite pole of the graph: its terms are more specific and more rarely used in the same message with those of the first two clusters.
27The third main cluster is structured around the lexical field of meteorology; it mainly refers to alerts issued by the National Meteorological Institute of Costa Rica, whose bulletins are often copied and pasted within the group (most often by the two leaders or by a journalist who is in the group). There we find technical vocabulary, typical of the Institute and not of the inhabitants themselves: rainy conditions, precipitation instead of rain, for example, but also areas of vulnerability, atmospheric instability. Certain terms reflect intense climatic events (intertropical convergence, heavy downpour), presenting significant risks (landslide, saturation), in the face of which the population is urged to exercise caution (step from, safe place) and to keep informed (Institute, page, warning). In these messages, the WhatsApp group acts as a more traditional transmission belt for official alerts issued by State scientists, which its members insert directly into the ordinary sociability of the group.
28The semantic analysis thus shows the particularly versatile aspect of the Alerta Ocloro group. Three main registers of expression coexist, which respond to three purposes pursued by its members: monitoring the river and identifying short and medium-term action prospects to remedy the problem of flooding; organize the sharing of information and the recruitment of new members, particularly at the most local level, that of the Ocloro watershed or municipality; finally, transmit weather warnings issued by public authorities in the event of major events. This versatility, orchestrated by a small number of central contributors, partly explains the longevity of the group and testifies to its infrapolitical nature, in the sense of Scott (1990): far from being a simple channel for the authorities, it organizes the cooperation of the river's residents, and allows them engage in “placemaking”, e.g. to make sense of their lived territory (Halegoua and Polson 2021).
29We draw here a qualitative analysis of how the group chat enabled collective empowerment and fostered political action, through the description of the concrete unfolding of flood alerts in the WhatsApp group, and of accountability practices.
- 12 District of San José where the author of the message lives.
30One of the main uses of the group is for participants to keep other members informed of the advent of heavy rain, particularly upstream of the river (in the east of San José), in order to warn residents downstream (neighborhoods of Los Yoses and Barrio Luján, in San José) of the possibility of flash floods. A multitude of messages are indeed devoted to describing the rain, its intensity, its characteristics and its location, based on the places of residence of the various participants. “Here on Avenue 14, for now the rain continues to fall but it does not overflow the gutters, and the sidewalks are not yet covered”, publishes Rodrigo on 09/10/2018 at the beginning of the afternoon, a time when showers generally begin during the Costa Rican rainy season. These messages, often short and concise, nevertheless contain precise details on the location of the rains (like here, with the avenue number), but also on its intensity (is it weak, heavy? Is it stormy, with thunder and strong winds?), or even its consequences in the urban space (are the streets flooded? Is the sidewalk submerged? Are some houses flooded?). Monitoring of rising water levels is often carried out in real time, particularly in the event of intensified rain: “in Francisco Peralta12, light rain” indicates Marina on 09/12/2018 at 2:18 p.m.; six minutes later, the same person updates their report with a new message: “heavier rain”.
31The occurrence of particularly intense episodes is an opportunity for rich exchanges which help document the rising water levels in real time, as observed in this sequence dated June 13, 2018, between 7 and 8 p.m.
Figure 4. Discussions on the Alerta Ocloro WhatsApp group, June 13, 2018
32This sequence is typical of the activity peaks which punctuate the life of this group chat during heavy rainfall. We can observe how the group reports, over a one-hour sequence, the rising water levels in the urban environment and its consequences on the inhabitants. For this, the inhabitants of San José and the neighboring communities to the east of the city mobilize on the one hand a form of experiential knowledge, as a form of testimony: photos of their street, their house, how covered is some stone in the river, i.e. visual reports of the intensity of precipitation; but also in-house measuring devices such as the rain gauge installed at Marina’s, one of the two leaders of the group, which allows her to quantify in real time the level of precipitation, as an indication (among others) of the severity of the rainfall.
33This discussion also shows how embedded the group chat is in the territory. Participants are regularly asked to specify where they are talking from: reports on the intensity of the rain and flood are constantly indexed to the name of a neighborhood, a city, or even a street number or a landmark of the urban space (the clinic, the bar, etc.). These same landmarks serve both to monitor the extent of the rainfall and the flooding, and to guide the police during their intervention, with precise indications on the state of the roads and on passable paths.
34A whole geographical network activates during this episode of precipitation, a network which even goes beyond the limits of the WhatsApp group, since one of the members asks his son to confirm information. Likewise, the way in which Marina and Bethy indicate to the police where to intervene suggests other parallel exchanges, outside the group chat, between the two women and other neighbors in the neighborhood affected by the floods, without however being present on Alerta Ocloro. We thus note the once again central role of these two “micro-celebrity activists” (Tufekci 2013), who act as the interface between residents, citizens mobilized via the WhatsApp group, and the public authorities.
35In this respect, it is remarkable to see how the municipal police of San José insert themselves into the discussion: silent most of the time, they intervene while the precipitation is intensifying and several houses are affected by the floods. The agents then follow the instructions produced by the local solidarity network to head towards the most affected homes. Unlike the top-down approach that the public authorities had sought to promote, the police are here mobilized by residents and guided by their information, in an informal and fluid dialogue enabled by the WhatsApp group. As observed by Zur (2024), social media also allows residents to make their issue visible, and even to become essential for public authorities, gaining power on the modalities and framing of public interventions. This is what we show in the next section.
36Beyond mere emergency management, Alerta Ocloro plays an interesting role as it supports an empowerment dynamic for its members, who use it to collectively seek longer-term solutions to the problem of urban flooding. These solutions then very often involve a relationship of negotiation or even conflict with public authorities (municipalities of San José and Montes de Oca in particular).
37Through the chat, participants often share photographs and videos of the river: during floods, as seen above, but also outside of moments of crisis, in order to document the material state of the river, in particular the infrastructures which surround it: grates, pipes, bridges, sewers and various canals. One of the issues identified by the collective lies in the congestion of the many grates which dot the course of the Ocloro, and whose congestion with various waste encourages the overflow of the river by generating micro-dams which prevent the normal flow of water. It is a photo sent by Gerardo, one of the group members, on June 1, 2016, which triggers the sequence that we transcribe below.
Figure 5. Photo and discussion on the WhatsApp group Alerta Ocloro, 1st June 2016
38Here, the mainstream nature of the technology is particularly decisive. It is because they are constantly equipped with their smartphone, which combines camera and messaging applications, that Alerta Ocloro participants are able to capture and transmit images of the river, as they move through the city. These handcrafted photographs and videos contribute to enriching the understanding of the flooding phenomenon for local residents, provided that they are reindexed within the spatial framework of the city, and the political framework of the municipality. This is what the active members of the group pursue, again at the interface between local residents and public authorities: faced with the publication of photos in the group, Bethy and Marina strive to link the images to specific locations, so as to be able to challenge the municipal teams responsible for infrastructure maintenance. As the message aimed at depriving municipal agents of their usual “excuse” suggests, the conflict with the municipality varies over time, but is always on the verge of re-emerging. This form of participatory engineering of water networks thus aims to compensate for the action - or inaction - of public authorities deemed to be failing with regard to the problem of flooding.
Figure 6. Activity and percentage of messages with photo or video over time
39The use of photography and video is regularly encouraged by the two local leaders: “we need your observations and/or photos of retention tanks, green roofs, wastewater treatment systems [...]. We will propose to include all of this in the development plan which will be adjusted by the municipality of San José, to mitigate the impact of large constructions.” (message posted by Bethy, 06/10/2016). As shown in Figure 6, although the number of messages per day decreases over time, the proportion of messages with images increases, as a result of the acculturation of the participants in this documentation process. As we will detail later, the accumulation of these images of the river also plays a fully political role in the sense that they are explicitly collected with the aim of proving, in the short to medium term, in the legal conflicts between the residents and public authorities.
40Through multi-sited analysis, this last section shows how digital networks are used to support not only grassroots crisis management, but longer-term political mobilisation through the dissemination of information to local populations, or even geographical data from lay research. The WhatsApp group Alerta Ocloro operates within an extended social network supported by digital technologies, including Facebook. These digital technologies allow the sharing of testimonies and experiences, as well as the archiving of data, which residents do to reinforce community empowerment and to challenge public institutions. This section demonstrates how social media uses contribute to broader environmental mobilisation and transforms the role of inhabitants in producing knowledge about the Ocloro River.
- 13 https://www.facebook.com/profile.php?id=100066194006298
41Common digital tools can facilitate the dissemination of geographic data, such as aerial photographs (Gautreau 2016). In our case, this dissemination aims at facilitating 'public' debate on the watercourse transformation, in particular through the CMC Facebook page, created on September 2014, which currently has 741 followers13. Unlike WhatsApp, the Facebook page does not play a role in crisis management. It is used to share information about the impact of flood events and to provide updates on the CMC's activities, particularly the progress made in discussions with the municipality. This page also provides an opportunity to share knowledge about the river and the neighbourhood. Figures 7 and 8 show, for example, the publication of cadastral maps and historical aerial photographs of Barrio Luján. These images are intended to show the alteration of the Ocloro bed, sixty years ago. Figure 7 shows the cadastral photo, first sent by the CMC coordinator to a research centre at the University of Costa Rica. Figure 7 shows how the CMC Facebook page shares the publication of this research centre, which also publishes a map showing the modification of the meander of the watercourse by redrawing the route of the old watercourse in the current landscape of the neighbourhood. As indicated in the commentary that introduces the publication, this modification is interpreted by the CMC as one of the causes of the increase in flooding that affects the residents. Figure 8 publishes an aerial photograph from 1955 showing the course of the watercourse in its lower left corner.
Figure 7. Cadastral maps and map shared on the CMC Facebook page, on 20-11-2015. The original image has been modified. Using Illustrator, we cut and translated the Spanish text.
Figure 8. Publication of an aerial photograph from 1955 on the Facebook page of the CMC, on 2-12-2015. The comment states: “Here is a 1955 map of Barrio Luján where you can see [...] an avenue of trees that marked the original crossing of the Ocloro River [...] We are doing a good job, we hope that the management of risks will put an end to this problem for local residents.”
42This dissemination of geographical data goes hand in hand with the desire to encourage the reappropriation of the phenomenon of flooding by the inhabitants. In other words, these images, which in this case come from the personal research of the coordinator and Marina, aim to denaturalise the problem of flooding and make it a political problem: the river did not always have its current course, and its transformation, and in particular its embankment, has increased the water pressure and the frequency of floods. Many of these publications have generated comments from local people that will provide new information and support for the CMC. For instance, the picture from 20/11/2015 generated three comments, one of which was from a neighbour sharing another old land registry signed by the public register in 1982. This shows a riverbed with a different form: 'The section between the forest and the old riverbed, marked in red, was not straight either, as can be seen in the following cadastral plan image.' In other words, broadcasts and exchanges on WhatsApp and Facebook groups are gradually enabling the formation of a digital learning community (Fluckiger and Zabban 2023) in areas such as meteorology (see previous section), as well as history and territorial planning, as demonstrated above.
43The digital nature of publications then makes it possible to collect new reactions and testimonies about the floods. The value of digital tools thus lies in their ability to disseminate information, but also in their interactivity. The qualitative research, which complements the analysis of digital exchanges, shows how social media serve to support a mobilisation that multiplies the voices expressed on the subject of flooding. The way people respond and interact within digital spaces then transforms the role of support, in this case the CMC Facebook page, into a data collection tool. The following excerpt shows the desire to collect these reactions from the residents, who criticise the municipality's lack of action and confirm a collective dissatisfaction for which they are the spokespersons:
44“A page was opened on Facebook, people give their opinion [...]they [the residents] tell me on Facebook “nothing has been done, nothing, only promises”, and things like that… After that I said to Marina: “Marina, we're going to take screenshots of all this””
45Bethy, coordinator of CMC of Barrio Luján, Mai, 2016
- 14 Dirección de Aguas.
- 15 Place where historically the river overflows more frequently.
46In addition to amplifying support for the CMC's environmental cause, public platforms such as Facebook and WhatsApp enable the archiving of data, which is later used in administrative and legal documents to support environmental mobilisation. Figure 9 shows a Facebook post, which was then used in letters sent by residents (Figure 10) to the Water Department14 of the Ministry of the Environment, to oppose a real estate project in what they consider to be the critical flood zone15. The photos published by the residents on the committee's Facebook page, are thus collected by the mobilised residents and attached to the administrative mail.
Figure 9. Publication 23 june 2016, on the CMC Facebook page. The publication concerns the agri-food company activities. The neighbours criticise the industry's management of the river. The publication aimed to show the industry's negligence in causing more dangerous floods in the neighbourhood (in this case, due to waste transported by the floods that could have impacted a house). The original image has been modified. Using Illustrator, we translated and cropped the Spanish text and added the red circle to indicate the pallet locations.
- 16 Administrative file at the Ministry of the Environment n°DA, 595-O, Folio 000147-Folio 000151.
Figure 10. Circulation of digital data in other mobilization spaces16.
- 17 Letter to the Human rights defender (Defensoría de los habitantes), July 12, 2017, Archives of the (...)
47This image aims to illustrate how Facebook publications are used in administrative files. As described, the publication in Figure 9 was reused in the letter of complaint to the Water Department about the agri-food company. Similarly, photos shared on WhatsApp were used in a complaint sent to the Defender of Rights17.
48This function of digital technology in the context of mobilisation in Barrio Luján nuances the literature on surveillance studies, which shows that digital traces are a surveillance tool for large private and public actors seeking to understand user behaviour (Gautreau 2016) or ensure security (Hale-Ross 2018). In addition to this work, the case study shows that the generation of traces is also a condition for citizen agency. Digital technology leaves traces that can be dated, localised and controlled. These traces are intentionally left behind to be used later in legal disputes.
- 18 https://www.crhoy.com/ambiente/extrana-coloracion-rojiza-tino-las-aguas-del-rio-ocloro
49The WhatsApp group also shows the same logic of expansion, from emergency management to wider environmental concern. Over the years, Alerta Ocloro shows a reversal of the top-down dynamic that was initially driven by institutions. While the authorities initially created their own WhatsApp groups as channels to facilitate the dissemination of institutional information to residents, to help them manage risks more effectively, many institutional representatives joined Alerta Ocloro, a group created by local residents independently of their efforts. Beyond the problem of floods, this WhatsApp group plays for example an important role in raising awareness of pollution in the river. In July 2018, several residents of Ocloro (non-members of the WhatsApp group) identify the suspicious colour of the river. They inform the institutions18, but also the members of "Alerta Ocloro", in particular Alberto (one of the founding members of the group).
Figure 11. Photo shared by Alberto in “Alerta Ocloro”, July 25, 2018.
50The different municipalities concerned are informed by the members of the group (but not exclusively by them). In August 2018, two representatives of the Ministry of the Environment requested to be added to the Alerta Ocloro, which thus becomes a necessary crossing point, also for institutions. As in the case of police forces, the predominant role of the WhatsApp group chat in the knowledge of the river, and the recognition it enjoys at the local level, pushes the representatives of the Ministry of the Environment to join it. When they joined the group, the ministry's representatives stressed their interest in the approach initiated by the residents:
51“Our aim is to coordinate with the community to ensure that the river is properly monitored for visible contamination. It is very important for us to know in time if there are pollution problems, as there have been in the past, and to be able to act in time. We hope that this channel will allow us to have fluid and honest communication [...] We will provide training so that those who want to actively participate can take samples from the river”.
52José Pedro, official of the Ministry of the Environment (Alerta Ocloro group, 08/13/2018).
53After several years of existence, the monitoring activities of Alerta Ocloro are no longer limited to issuing flood alerts and maintaining the river. The group's role is much broader: its functions are diversified and respond to a logic of habitation, in the sense of living in a territory, in all its dimensions. This logic of habitation brings together the different aspects of the river: overflowing water, polluted water, space for fauna and flora. This plasticity gives it an added value, compared to other specific applications of participatory science (Lemmens et al. 2021; Liu et al. 2021), whose use is often more specific and socially selective (Wolff 2021). The specificity of "Alerta Ocloro" lies in its ability to combine multiple functions in accordance with the ways of living and experiencing the territory. This polyvalent aspect also helps to explain why the group has survived over time, as it allows residents to go beyond disciplinary segmentations, making connections that are not necessarily recognised by professionals whose perspective is shaped by their training (Flanquart 2016).
54In the Costa Rican context, social media, and in particular instant messaging applications such as WhatsApp, have become a legitimate tool for risk management: mobilised by public institutions to engage citizens and disseminate instructions and information closer to the field, they are also at the heart of the everyday practices of citizens faced with environmental risks. This omnipresence of technology leads to a redefinition of the limits of political action in this field and to a redistribution of agency (Zur 2024) between the different actors involved. The WhatsApp group "Alerta Ocloro" is a paradigmatic example of this discreet empowerment of populations faced with multiple environmental risks (floods, pollution, land use change) and the failure of public authorities to act. Two main results should be highlighted.
- 19 Note in this respect that messages containing a URL are in an extremely small minority (near 1% of (...)
55First, the long-term ethnography of the "Alerta Ocloro" group and its evolutions underlines the interest in studying these barely visible spaces that are nevertheless the bearers of important political reconfigurations. Beyond the technical imaginaries of fluidity carried by public authorities under the umbrella of the "smart city", or the "agilization" promised by digital technology in crisis responses (Delaney and Kitchin 2023), the study of the uses of ordinary digital technologies in urban space makes it possible to grasp the aggregation of infra-political behaviours (Scott 1990), and how they contest and redefine the making of the city, far beyond institutional action. This politicisation from below is possible because the use of the application here is anchored in a well-defined territory, that of the Ocloro watershed; WhatsApp and Facebook are then used as instruments of "placemaking" (Halegoua and Polson 2021), through which residents inquire about the state of their territory, share their knowledge and coordinate their actions on the subject. The socio-technical characteristics of these mainstream social media application (easy access, rapid distribution of short texts and images) are used for the fluid exchange of information, but also, through accumulation and circulation of data, for debating risks, inquiring past states of the river, and ultimately demanding public accountability. This is an interesting result, as WhatsApp has so far often been described in the literature as a space more dedicated to family and friendly sociability (Matassi et al. 2019); not conducive to politicisation, as political content is often considered undesirable in many group chats (Zhu et al. 2024); or associated with the consumption and publication of false information, especially in groups such as Alerta Ocloro that connect individuals with "weak ties" (Kalogeropoulos and Rossini 2023). On the contrary, we observe in this article an emblematic case of virtuous politicisation, based on the sharing of experiences, the cross-checking of information19, with a view to improving the material conditions of local life. Around a core of regular participants, the heterogeneity of the group and its open nature (new members joining and leaving) contribute to its longevity and renewal.
56Secondly, the study of this informal space allows us to better understand the production of lay environmental knowledge and its fluctuating relationship with public authorities. On the one hand, we observe how the WhatsApp group chat, originally created to monitor the river's overflows, tends to aggregate the different aspects of the problem posed by the river in the urban space: flooding, but also pollution, infrastructure works and the modification of its route, and related land issues. The digital platform thus allows residents to network these different political issues, between which they make no distinction, as they are articulated and coexist in their ordinary experience of the urban space; and, by contagion, in the space of the WhatsApp group. In the case we studied, the environmental knowledge produced by the residents is characterised by a holistic tendency that goes against the expert, administrative and epistemic divisions that usually frame these problems (flood risk, biodiversity, land policy pollution). On the other hand, the use of social media is associated here with certain "ways of knowing" (Law and Ruppert 2016) on the part of the inhabitants: by sharing texts, sounds and images, they set up tactics for knowing their territory in a way that is different from the expert framing of the subject. The knowledge they produce mobilises both situated knowledge (Avenue 14 is flooded, the power is down in the neighbourhood, etc.) and experiential knowledge: by habit, residents know that when it rains in one place, the river overflows in another, or that the covering of a rock by the river announces an impending flood. Accumulated and shared, especially through images, this situated and experiential knowledge becomes more widespread and empowers mobilised residents to challenge legitimate political actors on their own ground.
- 20 The success of the Alerta Ocloro group led to the creation of a similar group by residents of anoth (...)
57We showed how the use of social media, in the case of San José, allowed citizens to overtake the neoliberal approach of risk management, producing counter-knowledge through shared experiences, leading to increased accountability for public authorities. These results suggest the importance of informal initiatives based on generic mainstream technologies in terms of risk management, particularly in Latin America where the use of WhatsApp messaging is omnipresent20. Further research is needed to systematise the analysis and, in particular, to shed more light on the conditions for the generalisation of citizen knowledge production in these informal digital spaces.
58Finally, this environmental mobilisation needs to be understood in the context of the Global South, where public institutions are unable to solve complex environmental problems (Araos et al. 2021), especially in metropolitan areas. In this context of constraint, local residents employ resilience practices that combine cooperative relations with public institutions with protest actions. While these processes of empowerment of social movements in the making of the city are well known (Deboulet and Nez 2013; Patsias and Patsias 2006), this article shows that digital technologies make it possible to change the material organisation and timing of mobilisation, by creating both a real-time space for crisis information sharing and a long-term arena allowing the gathering and expansion of lay knowledge and the archiving of data for future mobilisations.