Citizen Observations as Legal Obligations: (Dis)Associations and Representation at the Swedish Land and Environment Court of Appeal

Abstract

There is an increasing international trend in environmental activism to use legal institutions and infrastructures for citizen science (CS) to affect policy and regulation. However, knowledge about observations produced in activist CS and their functions at courts is scarce. To address this, we analyze how citizen observations (COs) reported to an established infrastructure for CS assist in producing legal obligations. Sweden provides an exemplary case due to the integration between CS biodiversity infrastructures, environmental regulation, and the increasing number of legal environmental conflicts. Data was gathered from documentation of legal cases argued in front of the Land and Environment Court of Appeal (LECA) from 2012 to 2020, and interviews with civil servants at Swedish environmental public authorities. Through a qualitative analysis, we find several ways that COs get associated to legal obligations, including through civil servants’ decisions, species spatial-temporality, comparisons between different CS reporting systems, the negotiation of species status on red lists, and interpretations of species behavior. We relate our findings to a broader discussion on what forms of representation matter in environmental regulation and who and what can speak for species and entities making up the world.

Introduction

Citizen science (CS) serves the interests of science, as well as the interests of policymaking and law. It has become an important part of the global initiatives of Open Science, and deemed necessary for monitoring Sustainable Development Goals, installing trust in science and policy, and mitigating alternative facts and social polarization (Fritz et al. 2019; Fraisl et al. 2020, 2023; Kasperowski, Hagen and Rohden 2021, 2023). In parallel to this growing mainstreaming, CS is undertaken by actors who use established CS infrastructures, reporting systems and protocols for activism in areas of biodiversity and climate change (Van Oudheusden and Abe 2021; Ottinger 2022; Turner 2023). Engaged citizens contribute to CS projects not only to advance scientific knowledge on biodiversity but to achieve regulatory change and challenge existing political arrangements that often renders policymakers unable to address environmental threats (cf. Latour and Schultz 2022, 2; Marres 2023, 6). As a result, in Sweden and elsewhere, observations and data derived from CS projects inform the work of environmental regulators, activists, policy, and legal decision-making to a greater degree than before (Kimura and Kinchy 2016; Kasperowski et al. 2023).

However, it is in relation to “proper science” that CS has most often been discussed by social scientists. Studies and evaluations of the quality of citizen observations (COs) and motives of volunteer participants are abundant (see Kosmala et al. 2016; de Sherbinin et al. 2021). Nevertheless, some studies also point to complicated relations of data quality and significance with participants’ security, usage and data rights, inclusiveness, ethics, and dataveillance (Bowser et al. 2020; Guerrini et al. 2018; Benyei et al. 2021; Kasperowski, Hagen and Rohden 2021; Peterson, Kasperowski and van der Wal 2022). Interests in the relations between CS and policy have been mainly in CS as a source for policymaking, an object or instrument of research policy, or in a limited extent as a possible form of direct governance (Schade et al. 2021). We turn to this latter capacity of citizen science to be an instance of direct governance, directing our attention to observations produced in activist-oriented CS projects and their functions in environmental courts.

To investigate this, we look to Sweden, where COs are well integrated with the daily operations of environmental regulation and are increasingly present in conflicts over how humans and other species should be represented to co-exist in the future (Pihl 2020; Jensen, Sandberg and Bengtson 2022; Wennblad 2022). Since 2012, civil servants at Regional Administrative Boards (RABs) and the Swedish Forest Agency (SFA) have relied on technical infrastructures to access biodiversity observations made by members of the public in order to develop and enforce environmental regulation at local and regional government levels (Kasperowski and Hagen 2022).

This integration of COs in Swedish environmental regulatory practices has not gone uncontested, as demonstrated by emergent legal disputes. In this process, landowners, environmentally concerned individuals, and advocacy groups can all make appeals to one of the five environmental courts across Sweden. If no legal decision can be reached in this instance, cases eventually end up at the Land and Environment Court of Appeal (LECA) in Stockholm. In sum, COs are highly integrated in the management of environmental regulation in Sweden through formalized practices and existing infrastructures, offering possibilities for increased understanding of how different stakeholders use established CS infrastructures in legal disputes (Kasperowski and Hagen 2022; cf. Turner 2023).

Purpose and Questions

The purpose of this study is to analyze how citizen observations assist in producing legal obligations. We ask: (1) in what ways are citizen observations associated to legal obligations, and (2) what repercussions might these practices have on notions of representation in science, law and politics? This article echoes calls for critical understandings of how political deliberative processes can take place in cases of environmental justice caused by institutional and regulatory “ignorance,” resulting in conflicts over who and what can speak for other humans, species and entities making up the world (Ottinger 2013; 2022, 3; Braverman 2018; Herzog and Lepenies 2022, 4; Marres 2023, 6; Turner and Wiber 2023). Further, we speak to the different meanings of citizen science as understood by natural scientists, Science and Technology Studies (STS) scholars, and others (Kuchinskaya 2019, 873). For many, the value of citizen science culminates in new digital technologies, such as large portals and platforms, that enable volunteers to join and submit data in the interest of furthering scientific inquiry. For others, the potential in CS is enabling educators, activists and others involved in bottom-up practices, often in the context of the production of local environmental knowledge that may be positioned against institutional scientific knowledge (Lave 2012). In this paper, we aim to defuse this unproductive polarization to show that participation in CS does not necessarily conform to either perspective. We emphasize that environmentally concerned members of the public use and rely upon existing CS infrastructures and technologies (Kasperowski and Kullenberg 2019; Turner 2023). Focusing on the convergence of activist-oriented data collection within existing environmental data infrastructures opens a possibility of understanding how subjects (citizen science activists) meet and create objects of politics in new ways (see Marres 2005).

To analyze how COs get associated to legal obligations, we draw from several Swedish environmental controversies between environmentalists and landowners that call into question the legal obligations enforced by Swedish governmental authorities, including County Administrative Boards (CABs), SFA, and the state's environmental court system (cf. Latour 2009, 2015; McGee 2015).

To situate these controversies, the article begins with an account of the organization and makeup of the environmental courts in Sweden. Thereafter, we present our conceptual resources to understand how environmental regulation deals with COs, followed by how the data for the study were collected and curated. We analyze the data to disclose several modes for how COs get associated to or disassociated from legal obligations, specifically through civil servants’ decisions, species presence and locality, comparisons between different CO reporting systems, negotiations over whether a species should be red-listed, and interpretations of species behavior. Lastly, we offer some reflections about who and what can speak for other humans, species and entities making up the world, and discuss how COs are anticipated to change the relations among environmental regulation, representation, and democracy in the Swedish context and beyond.

Environmental Courts of Sweden

The Swedish Environmental Code, established in 1999, gives the environmental courts civil and administrative jurisdiction and extensive enforcement powers. As in the modern legal and liberal tradition, this code applies in principle to all human activities that may damage the environment or human health and is connected to several other Swedish laws and directives from the European Union. The Swedish Environmental Code itself is general; guidelines for implementation and enforcement are issued by public authorities such as any of the local 250 municipalities, the Swedish Environmental Protection Agency (SEPA), CABs and the SFA (Svensk Författningssamling 1998; Bjällås 2010, 179; Naturvårdsverket 2024).

There are five Land and Environment Courts (LECs) in Sweden, located at the Umeå and Östersund district courts in northern Sweden, at the Nacka district court in Stockholm, and at the Växjö and Vänersborg district courts in the south and west of Sweden. A LEC is composed of a panel with one legally trained judge, one environmental technical advisor, and two lay members. By law, all four panel members are equal in the decision-making process (Bjällås 2010, 180). An appeal to a LEC is possible by contesting any decision on environmental matters made by one (or more) of the 458 public authorities that exist in Sweden. ¹

If a settlement cannot be reached, the case travels to the LECA in Stockholm which hears appeals from the five LECs. The LECA is comprised of four judges, any of whom may be replaced by a judge with specific technical training relevant to the appeal. As a rule, decisions by the LECA cannot be appealed, although leave to appeal can be given by the Supreme Court if a case is of special judicial interest. The LECA hears cases as a general meeting rather than in a formal court proceeding. The court normally goes to the site in dispute, and the parties and the people living close to the site are asked to give comments to the court. The court can also ask the responsible local, regional, and central authorities or an independent technical institute or university researcher to comment on the case (Bjällås 2010, 182).

Theoretical and Methodological Considerations

Scholars in STS have found that the epistemologies and ontologies of law and science are networked or co-produced, stabilized and constructed through processes of interaction, translation, and association between actors (Jasanoff 1995, 2006; Latour 2009, 2015; McGee 2015; Cloatre 2018; Kasperowski and Hagen 2022). Influential studies describe nuances between the regimes and practices of science and law. STS scholar Sheila Jasanoff (1995, 2006) shows how the epistemological context changes when scientific findings are used for legal purposes. These social changes create controversies over scientific claims that “must be worked into the particular kinds of propositions, representations, or material objects that the social context of law regards as germane to establishing which party is telling the more plausible story” (Jasanoff 2006, 329). By contrast, in the relational ontology of legal studies inspired by actor-network theory (ANT), social context is not regarded as an explanandum of the epistemological. Rather, through chains of translations and associations, ANT recognizes “all” actors (aka actants), including both humans and non humans, as having agency. Yet agency is not created evenly in actor networks. Citizen observations form into networks, and the strength of network relations waxes and wanes. Stability changes as COs in the network may be assigned agency differently, to be associated to or disassociated from legal obligations (cf. Kasperowski and Hagen 2022, 450).

Though these perspectives differ to a degree, both allow for the exploration of a multitude of objects of knowledge and actors, directing attention to how legal obligations are performed into assemblages of nature and law in environmental controversies (McGee 2015, 61–62). Building on this work, the field of STS has largely been concerned with how epistemological and ontological claims by formal scientific experts are negotiated to meet specific legal criteria. In the context of CS, claims are no longer the exclusive domain of specialized scientists but extend to non-professionals in concerned communities and activist groups (cf. Jasanoff 1995; Bormpoudakis 2019). The technological infrastructure used by activists plays an important role in ascribing COs the qualities of scientific knowledge.

An important difference between science and law is that regulations operate through associations to produce obligations that can legally enforce certain modes of existence (Latour 2009, 2013, 2015; McGee 2015). By contrast, scientific claims, and COs, are held not to serve particular interests because “nature” is not supposed to possess any prescriptive dimension (Latour 2017, 23). This is a defining feature of the “modern” ideal of science (cf. Latour and Schultz 2022, 59–61) that can also be a lever of political ecology. Nonetheless, an environmental legal obligation is intended to represent a view from somewhere, and it can be enforced to serve a particular interest. It is an empirical question in this paper how the non-prescriptive scientific quality of COs is associated to legal obligations. Following the ANT-inspired approach, the technological infrastructure becomes an important actor in the relational ontology of this political ecology, associating observations with scientific qualities that citizens, on their own, would have difficulties mobilizing.

In the service of scientific work, COs depend upon immutables in the “background” that form a “residual realism” (Chilvers and Kearnes 2020, 349). These immutables exist beyond the reach of most actors to alter and bestow observations with scientific qualities that allow them to move in different contexts (cf. Lammes 2017, 1020). For instance, the reporting protocols in Artportalen, the Swedish Species Information System, move without issue between citizens and the twelve ² groups of species that can be reported in the system. However, as COs move from the citizen science portal to the reporting protocols recognized by municipalities and courts, such “immutables” no longer become self-evident or assist in COs being associated with legal obligations (cf. Dugdale 1999; de Laet and Mol 2000; Johnson 2008; Lammes 2017; Benyei et al. 2021. See also Star and Griesemer 1989).

In tandem with post-humanist and relational ontological theoretical frameworks, we reject an essentialism of what COs or environmental regulation are or should be (Callon 2007; Levi and Valverde 2008, 808). COs can produce different temporal and spatial effects due to associations and disassociations between actors in unresolved cases in environmental regulation (Dugdale 1999, 118; Caliskan and Callon 2010, 9; Latour 2013; Lammes 2017, 1019; Ottinger 2022, 2). In this article, we describe the associative practices taking place, as well as practices which disassociate COs from their potential legal obligations. That is, we view COs as assemblages (Ong and Collier 2007) of associations and disassociations between multiple actors. Our study connects to a growing body of legal studies, which critically discuss and explore regulatory and legal processes in analyses of more-than-human networks (see Callon 1986; Latour 1996, 2015, 2017; McGee 2015; Reeves and Peters 2021; Buocz and Eisenberger 2023; Turner 2023; Turner and Wiber 2023).

Empirical Data

To understand the roles of COs in environmental regulation, one needs access to empirical data that can show how observations are only as real and stable as their ability to contain and sustain associations among different actors (Murdoch 1997, 331). Hence, we collated a dataset of 42 legal cases argued in front of the LECA in Sweden from 2012 to 2020 that rely on COs reported to Artportalen and Svensk fågeltaxering (the Swedish Bird Taxation) (cf. Johnson 2008; Law and Singleton 2005, 347; Herzog and Lepenies 2022, 5). We found these cases by putting together a text corpus of the approximately 3,000 cases ³ the LECA processed during 2012 to 2020 and then searched for the names of databases that utilize COs ⁴ Only Artportalen and Svensk fågeltaxering resulted in hits in the database.

Searching for the names of biodiversity databases allowed us to ensure that COs were present in the court cases we analyzed. The number of cases involving COs increased over the sample period. For instance, most cases involving observations from Artportalen occurred between 2017 to 2020. This may be because several environmental organizations active in Sweden endorse the use of citizen science and portals like Artportalen to affect environmental regulation (cf. Greenpeace 2021). Even though the 42 cases we found are a sliver of the total number of court cases tried by the LECA during this time, this sample encapsulates controversial legal decisions in respect to COs and, as such, point to issues that the Swedish environmental authorities and the lower levels of the Swedish Environmental Court System were unable to resolve. The 42 cases relate to different environmental controversies, including housing construction, logging and deforestation, landfills, groundwater and run-off diversion through damming or drainage, creation of rock-quarries, localization of wind turbines, and more.

To analyze this data, we followed a triangulation approach, which was performed by all authors independently analyzing the content of each case in full, with the aim of identifying how COs are associated with legal obligations. This approach yielded a preliminary coding of such associations and these were then compared and grouped by the authors, resulting in four modes by which COs were associated to legal obligations.

To complement our court case data, we also performed 18 semi-structured in-depth interviews with civil servants at Swedish environmental public authorities. Interviews were conducted during the spring and autumn of 2019 at the SFA and at CABs, with individuals who had been involved in the type of court cases we analyze. The purpose was to understand the predicaments facing civil servants as they evaluate and determine how to associate COs to legal obligations in preparation for court appearances. The interviews were also coded and analyzed by each of the authors independently following a triangulation approach.

Qualitative Approach and Some Limitations

The 42 LECA court cases and the interviews with civil servants have been subject to qualitative content analyses with the aim of attaining theoretical saturation following our perspective on COs being associated or disassociated to legal obligations (Sebele-Mpofu 2020). However, the data analyzed in this article cannot fully account for and represent the state of COs in the Swedish court system. Our inclusion criteria do not guarantee that all instances in which COs mattered were identified, as cases do not necessarily name databases or data origins in court records. To get a complete sample, one would need to meticulously go through all 3,000+ cases manually.

Our ambition is to chart and explore COs that prompt disputes, regardless of whether COs are successful in becoming legal obligations in the individual cases. In other words, this study does not analyze successful or unsuccessful associative processes. Instead, it analyzes how observations become subject to associations and disassociations in controversial cases. We have not gone into detail concerning what is at stake in each case (the exact location of a wind turbine, deforestation, establishment of a rock quarry, etc.); our methodological considerations were based on empirical material that could yield modes for associating COs to legal obligations that would be relatively independent of individual cases. Generalizations were thus not based on representation, but on comparisons of data from interviews and legal documents, with the aim of identifying modes that transcend individual cases.

So, although few countries have developed environmental courts or similar juridical institutions and technical infrastructures for CS as Sweden, the results presented here contribute to discussions on the national and local political consequences of international ambitions to turn CS into global infrastructures for biodiversity monitoring (cf. Turner 2023). Aggregated data from Artportalen already has global reach as it contributes COs to the Global Biodiversity Information Facility (GBIF). As we show, many actors and their agency get redistributed as they are mobilized into political ecologies and conflicts over how COs are associated with legal obligations. This has implications for issues of political representation beyond the Swedish context.

Results and Analysis: Modes of Association

In this section we analyze four modes in which citizen observations are associated to and disassociated from legal obligations. Our analytic account starts at the desks of civil servants at environmental public authorities, and details how COs are associated to legal obligations through the kinds of validation, environmental laws and conventions, and protocols offered by Artportalen. Our analysis proceeds through litigation and lawsuits challenging civil servants’ decisions in front of the LECA in Sweden. We describe how the predicaments associated with civil servants’ decisions result in (1) species, mostly birds, being disassociated from the localities of their observation; (2) COs reported to different reporting systems affecting obligations; (3) species being disassociated from obligations despite being red-listed, and (4) species behavior being associated to obligations. These modes exemplify how COs can be assigned agency differently, to be associated to or disassociated from legal obligations.

Producing Predicaments Through Civil Servants’ Efforts

The work of civil servants within Swedish environmental authorities at municipal and regional levels is greatly influenced by COs, predominantly those reported to the Swedish Species Information System, Artportalen. This online platform launched in 2000 has developed from a reporting system for amateur ornithologists into an obligatory passage point for Swedish civil servants overseeing public decision-making on environmental regulation (Knape et al. 2021; Peterson, Kasperowski and van der Wal 2022). Artportalen is now one of the largest biodiversity observation data centers in the world, collecting observations by the public and is extensively used by civil servants at environmental public authorities in Sweden and abroad. Artportalen is the largest national multi-taxon biological recording CS online platform contributing to the GBIF. GBIF was established by the OECD in 1999 to aggregate and provide biodiversity data about all types of life on Earth (Jönsson et al. 2023; Global Biodiversity Information Facility 2024).

All observations in Artportalen undergo multiple stages of validation before they reach civil servants. The first validation is made by the observer following system-mandated inputs. Then, other observers validate the observation. Last, one of eighteen species validation committees evaluate observations of interest (e.g., red-listed or very rare species). Through this process, Artportalen involves a network of experts, both scientists and laypeople, to decide which species to red-list. The database currently contains more than 100 million observations on different species, mostly birds, used to maintain environmental laws in monitoring sustainable development goals set by the Swedish Parliament (Sveriges Miljömål 1999), by municipalities, CABs, and the SFA, which are the main public authorities responsible for implementing environmental regulation.

COs, therefore, get associated to legal obligations under regulations administered by municipal and regional civil servants who evaluate and determine COs at the local and regional level. A civil servant at the CAB or SFA manages many cases and requests concerning the environment, including giving advice on larger interventions by companies or individual forest owners, the creation of natural reserves, conducting environmental impact assessments for large infrastructural projects, issuing building permits, etc. A civil servant at a CAB or SFA divides her duties between supervision, consultation, and the issuing of permits. These responsibilities demand that she determines which COs are relevant and authoritative (Kasperowski and Hagen 2022, 452).

However, as the number of COs in Artportalen and Svensk fågeltaxering continues to grow and become more important with respect to environmental regulation, civil servants experience their duties as becoming increasingly difficult and pressing. For instance, civil servants must make decisions based on increasing numbers and complexity of COs, interpreting when laws should be enforced. ⁵

In this “new role,” civil servants have a say in determining which observations become legal obligations for landowners to adhere to. However, due to their increase in number and complexity, observations become more susceptible for questioning and become subject to disassociation by civil servants. For instance, civil servants increasingly deal with COs reported by non-governmental organizations (NGOs) and activists who support biodiversity conservation. Many such observations are made in a bid to stop logging in places of high conservation value, having resulted in several successful lawsuits against Swedish public authorities for not fulfilling national or international environmental laws and conventions. This puts civil servants under increased pressure. In some cases, “activist” observations intended to be associated with potential obligations might be disassociated already at the civil servant's desk. If a civil servant notices that someone has recently registered an account on Artportalen and reports many badly tagged observations to stop logging in a certain area, this will evoke suspicion and will be investigated, especially in relation to red-listed species. ⁶ The civil servant acts as a frontline for preventing some COs from being associated to legal obligations.

Civil servants at the CABs and SFA also need to deal with COs that are purposefully vague. On occasion, obligatory inputs such as the location of a CO are not specific enough. Observations of large birds of prey, for instance, are not always accompanied by clear geographical coordinates, a strategy to keep the location of nesting places secret:

In the case of the eagle-owl, there is no information on nests in the material received from the County Administrative Board or Artdatabanken. The reason for this is probably that the people who work with eagle-owls…are extremely secretive and do not report the location (Mark- och miljödomstolen vid Växjö tingsrätt 2019-09-20).

Although civil servants could disassociate these observations from obligations directly, they have to follow up on observations of red-listed species, often by contacting the observer or an evaluation committee member, to get more specific information to validate and establish trust in the observation (Kasperowski and Hagen 2022). This relationship between civil servant and observer or committee member becomes paramount for associating or disassociating COs to obligations. Hence, civil servants tend to trust COs made by long-term observers who make recurrent observations and are members of reputable organizations (Mark- och miljödomstolen vid Växjö tingsrätt 2018–04–13; Mark- och miljööverdomstolen 2020–12–09).

A member of the…bird club has for a long time inventoried and ringed Siberian Jays, so the material is therefore considered a credible source for the species’ observations. In the concerned areas, it is a matter of recurring observations over several years and also of nests, which are reported in Artportalen (Mark- och miljödomstolen vid Östersunds tingsrätt 2019-08-27a).

Typically, issues concerning scattered observations and secret positions are resolved by the civil servants, relying on trust established in the process of validating observations. In other words, civil servants try to align the scientific associations of COs to meet legal obligations. When someone questions these associations, COs can end up at the LECA. In these circumstances, civil servants’ evaluations can then appear as support for different actor's claims or be deconstructed:

With regard to the red-listed bird species reported to Artportalen, the municipality notes that the observations are not coordinated, which, together with the fact that the reliability of Artportalen can be questioned, means that the observations must be considered with caution. (Mark- och miljööverdomstolen 2019-0314)

Thus, information provided to the courts by the municipalities and regional public authorities can uphold the associations made by civil servants or disassociate COs through eventual arguments or contradicting observations. Associations or disassociations made by a civil servant can be used in the LECA by different actors for their own interests. In this way, civil servants at the CABs or SFA either uphold associations or disassociate COs as a first stage on their path to producing legal obligations.

Disassociating Species and Locality

When a stakeholder files a suit against a decision made by a civil servant, a species locality can serve as a mode of associating COs to legal obligations. In the case below where activists react to logging plans by a forestry company, locality becomes integral in a struggle to prevent deforestation:

Perhaps the most important reason, both locally and globally, for a species to disappear is destroyed habitats, i.e., habitats disappear and/or are fragmented. Artdatabanken states that the Siberian Jay gradually disappears, especially at the edge of its habitat. If deforestation is permitted in cases such as these, the County Administrative Board will see great difficulties in slowing down the development [toward extinction] (Mark- och miljödomstolen vid Östersunds tingsrätt 2017-10-06a)

Essentially, once an organism is observed, the place, and not just the species, becomes subject to potential obligations. An observation of a red-listed species serves to preserve a geographical space, as well as living beings. It is therefore important that associative relations between the location and the species observed can be upheld. Or, for an opposing actor such as a forestry company, for that association to come apart:

The investigation does not show that the proposed deforestation would affect the Siberian Jay's conservation status regionally, or that its southern distribution limit risks being moved in the way proposed by the Swedish Forest Agency and the Swedish Environmental Protection Agency. The investigation also does not show that the continuous ecological function of the habitat cannot be maintained in the area if deforestation is allowed. (Mark- och miljööverdomstolen 2020-12-09)

Here, COs risk being disassociated from locality in two ways. The first way is to disconnect the observations from their geographical position, without disassociating a species’ red listing. That is, the argument moves the organism into a generalized, unthreatened space—a variant of biodiversity offsetting, often linked to green grabbing (Bormpoudakis 2019, 549). The second way focuses on the land itself and not the organism. It disassociates the ontological and ecological features of the landscape from the COs to make possible changes to the land. Such disassociations imply that land-use changes will have no significant ecological impact on a species.

Locality can also be already disassociated from COs by the regional and local environmental regulators. Civil servants must deal with ever-increasing numbers of COs that trigger obligations, particularly those reported by activists to protect land or water from being used in ways not compatible with biodiversity and the species’ protection ordinance. To adhere to the species’ protection ordinance by the letter, civil servants felt they would have to “close down Sweden,” for instance by denying applications for logging from landowners. The species to be protected are clearly listed in the Endangered Species Protection Ordinance, and these are targeted by activists in their reports to Artportalen. So how should this be done?

Too many rejects (or accepts) of logging applications or other development plans can also result in litigations against decisions made by civil servants at CABs. These complicated relations between law and COs was repeatedly voiced by civil servants in interviews for this study. Artportalen is a reporting tool for science, but laws turn COs into something complicated. As a result, civil servants are sometimes incentivized to disassociate locality from COs. Finding ways to do so is a way to keep Sweden “open.” ⁷ In other words, civil servants recognize that to associate COs with obligations in every possible instance, for instance in relation to red-listed species, would pose a risk of inviting landowners and forest companies to launch litigation. On the other hand, too liberal interpretations of the Endangered Species Protection Ordinance risk disassociating COs from legal obligations, which would result in opposition and legal action from activists. This creates a highly politicized ecology in which associations or disassociations become subject to contestation by several stakeholders who would like to see COs produce obligations fulfilling their specific interests.

Pitting CO Reporting Systems Against Each Other

Different reporting systems for COs can lead to different obligations. When COs from Artportalen are questioned as evidence in court cases related to birds, they are sometimes compared to COs reported to Svensk fågeltaxering. Observations in Artportalen and Svensk fågeltaxering rely on different CS practices for observing and classifying species and biodiversity. Artportalen is described by its management at the Swedish Agricultural University as a meeting place for anyone interested in nature to easily keep track of their own and others’ observations, and exchange knowledge with each other. The portal provides information on different species, whereabouts, times of observation and, importantly, species behavior (e.g., seasonal variations, foraging, nesting, courting etc.).

On the other hand, observations in Svensk fågeltaxering are made by volunteers who follow predetermined routes during specified time frames. Often, these observations are made by the same volunteers on an ongoing basis over several years. As a result, they produce series of long-term comparable observations, sometimes dating back to the 1970s, which can be searched to determine changes in the numbers of birds over time. COs from Svensk fågeltaxering are thus associated with the rigorous and disciplined way of attaining them. Unlike in Artportalen, observations reported in Svensk fågeltaxering do not contain descriptions of bird behavior.

Hence, these different reporting systems provide unique datasets that get used in Swedish environmental courts to associate or disassociate COs to legal obligations. This typically happens by introducing temporal and spatial uncertainties in observations that have reached the LECA:

Data in Artdatabanken…can be questioned, as they differ from the Svensk fågeltaxering data. The Siberian Jay is not threatened and has a positive trend. A possible explanation for the claims in Artdatabanken may be that the species has natural fluctuations between years, and if an inventory is made in a certain year, the wrong conclusions can be drawn. (Mark- och miljödomstolen vid Östersunds tingsrätt 2017-10-06b)

Elsewhere:

For about forty years, volunteer and knowledgeable ornithologists have been counting birds in several hundred locations around Sweden, both during summer and winter. The bird count is performed in the same way every year. Comparisons can thus be made over the years. Unlike the reporting to Artportalen and Artdatabanken, the Svensk fågeltaxering monitoring assessments are based on a scientific method. Greater emphasis should therefore be placed on data from Svensk fågeltaxering. (Mark- och miljööverdomstolen 2019-08-27)

By comparing reporting systems, COs in Artportalen get disassociated from potential obligations through COs in Svensk fågeltaxering. These aspects include their temporality in relation to when the CO gets made, in terms of phenology as well as scientific method. Additionally, long-term aggregated data from Svensk fågeltaxering also produces ranges rather than localized observations, which further acts as a resource to disassociate COs in Artportalen, by generalizing time and space. The temporality and spatiality associated with COs in Svensk fågeltaxering are typically used to question place-bound observations in Artportalen, disassociating their local value (e.g., the home range of the Siberian Jay) to beyond or outside a local place and into an “aggregated space,” often referred to as the nation or the species’ natural global distribution area (Mark- och miljödomstolen vid Östersunds tingsrätt 2019-08-27a).

The distribution area for The Siberian Jay is far from exact. The conservation status of the species is good. The Siberian Jay is described as viable both globally and nationally, which is the highest degree on the scale. The population has been stable since the 1990s and has increased since at least 1998. In 2015, the largest number of Siberian Jays ever was measured in Sweden. Local deforestation would not affect the conservation status of the Siberian Jay. (Mark- och miljödomstolen vid Östersunds tingsrätt 2019-08-27a)

On the other hand, COs in Artportalen can also be used to disassociate COs in Svensk fågeltaxering that generalize time and space. This generality is challenged because it cannot encompass the richness of observations made in a particular place and, therefore, can be of little use for local environmental regulation:

Data from the Svensk fågeltaxering are reliable in terms of developments at the national level, but they cannot be used for regional assessments. This is due to a relatively low number of observations and a low probability of observations of Siberian Jays in the southern and south-eastern parts of the county. (Mark- och miljödomstolen vid Östersunds tingsrätt 2017-10-06b)

Observations in Artportalen—associated with a specific place, species behavior in that place, and the trusted individual who performed the observations—can function to disassociate COs from Svensk fågeltaxering as irrelevant for local environmental regulation. COs from both reporting systems have the potential to disassociate each other's observations in the process of environmental regulation. What COs from Artportalen do in this context is to localize time and space, whereas Svensk fågeltaxering COs generalize them. Hence, COs from both reporting systems can be complementary for scientific reasons, for example they are sometimes harmonized to produce habitat suitability models (see Bradter et al. 2018; Henckel et al. 2020). However, in an environmental regulation context, when COs are used for legal purposes, COs are at the heart of political conflicts about particular versus universal ecologies (cf. Jasanoff 2006, 329).

Disassociating COs from Obligations Through Red Lists

Swedish environmental regulators and stakeholders place much value on the conservation status of species (Kasperowski and Hagen 2022, 10). The relation between how COs and red lists get associated to conservation status are therefore important in an environmental regulatory context. At times, conservation status associated to COs and red lists align:

According to Artportalen, colonies of Sand Martins have been continuously observed in the quarry….Given that the species are classified as endangered on the Swedish Red List, the quarry constitutes a site of importance for the species’ conservation status, at least regionally. (Mark- och miljödomstolen vid Vänersborgs tingsrätt 2017-05-08)

When such alignment occurs, obligations can come into effect. However, the potential for alignments can be frustrated because red lists also can be used to disassociate COs from obligations:

The grouse (like the Capercaillie) is a highly fluctuating species with large ups and downs over time, similar to those reported by SOF/Birdlife, which is common for obvious reasons. The grouse is not included in the Swedish red list, and the trends that SOF/Birdlife refers to [Svensk fågeltaxering 2016] apply mainly to southern Sweden, not the northern parts where the grouse is common. Venues with six to ten roosters are common in large parts of Norrland. (Mark- och miljödomstolen vid Umeå tingsrätt 2017-06-16)

Here, the grouse's absence from the Swedish red list is used to disassociate claims about grouse made by the Swedish ornithological association (Sveriges ornitologiska förening (SOF)). Hence, the value of a local observation can be disassociated from legal obligations depending upon the species inclusion and its status on a red list. Nonetheless, even when a species’ conservation status has been codified on a red list, its status can be disassociated from the red list by COs. In 2015, COs were used to claim that the Siberian Jay was still under threat even though its red list status changed from “near threatened” to “viable:”

According to a red list assessment for Siberian Jay from Artdatabanken in 2015, it is included in the category viable. Previously, it was judged to be in the category near threatened. The species is declining in number, but now at such a slow pace that it has returned to the category viable. It is thus not on the red list. The European Commission's advice states that the fact that a species is not endangered, i.e., does not face any direct risk of extinction, does not automatically mean that it has a favorable conservation status. The obligations of the Member States are something more than just avoiding extinction. (Mark- och miljödomstolen vid Växjö tingsrätt 2018-09-25; Mark- och miljödomstolen vid Östersunds tingsrätt 2019-08-27a)

In this case, COs expose the interpretive flexibility of the Jay's conservation status as represented on the red list. Thus, COs are used to challenge the agency of lists as a means of creating a simple act of representation. Red lists and COs both get used to disassociate a species’ conservation status. Such disassociations affect the potential for COs to produce legal obligations.

Making Species Behavior Count

As we have shown, COs that are part of cases argued in front of the LECA are associated to additional information that can be provided by reporters of observations in different reporting systems. For instance, the protocol for reporting observations in Artportalen includes options to provide additional “interesting information,” including details about species behavior and cyclical or seasonal biological events, such as migrations, egg laying and flowering (Artportalen: rapportsystem för växter, djur och svampar 2016, 7). Though reporting such details is sometimes not regarded as obligatory by reporters, COs with associations to foraging, courting, breeding and nesting behaviors can be very important for producing legal obligations. The possibility of COs to turn into obligations can depend upon the behaviors with which a species, most commonly birds, can be associated.

However, since species behavior is typically not obligatory to record in the reporting systems for COs, this association often remains loosely connected or overlooked within environmental regulation. Hence, civil servants follow up on observations that include species behavior by contacting observers for more specific information. If this association does not get clearly established, it can be viewed with skepticism:

Artdatabanken states that the established nesting of Corn Crake is few and that it is very uncertain whether the population is self-reproducing on the Swedish mainland (Mark- och miljödomstolen vid Nacka tingsrätt 2012-05-09).

To disassociate behavioral observations from legal obligations, it is often claimed in litigations that only few observations have been reported to Artportalen, or that observations are made far apart in time and therefore only have historical and not current value (quantity, frequency, timing):

The observation was made in 2010 and refers to one individual. Since then, no observations have been made in the area in question and no individuals of Great Crested Newts have been located by the inventories carried out by the company. (Mark- och miljödomstolen vid Växjö tingsrätt 2015-02-13)

Such disassociations are met by opposing stakeholders’ efforts to make or strengthen associations between species behavior and legal obligations. Often, environmental groups’ lawsuits claim that responsible parties, including public authorities like the SFA or CAB or consultants acting for these authorities, have not considered the full range (phenology) of species’ behavior in their decision-making. To facilitate such contestations, they perform their own inventories—taking advantage of the possibilities enabled by Artportalen to achieve regulatory changes and to associate species behavior to their observations (Mark- och miljödomstolen vid Växjö tingsrätt 2018-09-25; cf. Mark- och miljödomstolen vid Vänersborgs tingsrätt 2014-10-02; Mark- och miljööverdomstolen 2015-12-08; Mark- och miljödomstolen vid Vänersborgs tingsrätt 2015-07-02; Mark- och miljödomstolen vid Nacka tingsrätt 2019-03-14):

The Ospreys in the company area have been studied in…a substandard study that covers far too little time, only a couple of days.…Also, according to this study, the birds in their search for food would fly northwest…and return the same way. Based on a number of [our] own observations…and from ornithologists…this is seriously incorrect. There are a number of reported observations (spread over the breeding season in Artportalen) of Ospreys. (Mark- och miljödomstolen vid Vänersborgs tingsrätt 2017-06-30)

By signaling the Osprey's behavior through their own observations reported to Artportalen, environmental groups build associations that question the previous observations and data (Mark- och miljödomstolen vid Vänersborgs tingsrätt 2015-10-23).

In a case related to green-spotted toads, this red-listed species was observed along the perimeter of an area scheduled for development. Stakeholders associated the toad's behavior to legal obligations by arguing that the “green-spotted toad found at…shore meadows ca 700 m to the south also will move and be present in the environment that is under consideration for planning” (Mark- och miljödomstolen vid Växjö tingsrätt 2013-02-28.; cf. also 2014-12-11; Mark- och miljödomstolen vid Östersunds tingsrätt 2017-04-12; Mark- och miljödomstolen vid Umeå tingsrätt 2017-06-16, Mark- och miljödomstolen vid Vänersborgs tingsrätt 2019-06-05). Here, the toad's spatial behavior is used to intervene in the planning process and COs become a tool to leverage possible future associations between COs and legal obligations. In a case involving an inventory of mushrooms, new COs were reported to Artportalen to supplement as well as question a preexisting survey:

The company's compilation shows that an inventory of mushrooms was made during a very dry year. No new inventory was carried out, despite a recommendation to this effect. According to Artdatabanken, a unique fungal flora has been identified with a number of endangered species [in the area]. (Mark- och miljödomstolen vid Nacka tingsrätt 2011-11-30)

In this way, environmental activists strengthen existing associations and provide additional data to challenge previous observations, for example when records are separated by long periods of time, (Mark- och miljööverdomstolen 2019-01-22) or when observations are considered old (Mark- och miljööverdomstolen 2019-08-27). This practice points out how COs’ associations to current or future number, status, presence or behavior of a species can aid in producing legal obligations (Mark- och miljööverdomstolen 2018-11-15).

Concluding Discussion: Conflicts Over Political Representation

In this paper we set out to understand how citizen observations assist in producing legal obligations. We accomplished this by finding four modes that observations can be associated or disassociated to legal obligations: through data on species locality, through COs in different reporting systems, through a species’ relation to its red-list status, and through species behavior.

Although these four modes of association are not exhaustive for the Swedish context, we argue that COs are likely being associated to legal obligations in other countries through similar modes, especially considering trends in aggregating biodiversity data at national and international scales (Peterson, Kasperowski and van der Wal 2022). Moreover, our findings points to that COs can assist in producing legal obligations through multiple pathways. COs get associated with legal obligations through networks in which human and nonhuman agencies are distributed in many ways. In the context of environmental regulation, COs’ associations can be strengthened, weakened, or come undone through disassociation. Through this process, COs assist in producing legal obligations; for whom and what they benefit or disadvantage are empirical questions. These results are consistent with studies finding that observations change as they are performed across different temporalities and spatialities, as actors assemble “eclectic articulations” of observations, and move beyond pre-given formats for citizen science participation (see Buchanan 2017, 521; Hyysalo, Pollock and Williams 2019; Benyei et al. 2021).

Additionally, as evidenced in the data, the use of existing technical infrastructures and biodiversity observation databases are examples of how volunteer participation in CS often occupies a middle ground position (cf. Benyei et al. 2021, 759), contributing to science but also performing political action in relation to ecological challenges. The increasing pressure on civil servants, the connections between species and locality, opposing reporting systems, red lists, and species behavior are the practices of association and disassociation that matter most.

These four modes of association point to issues of political representation. When uses of a particular area of land or water become contested, in Sweden and elsewhere, various actors may respond by producing and using COs. From an environmental activist perspective, the justification for such actions is commonly that economic elites have captured political action, or that public authorities and governments fail to tackle pressing environmental challenges. For activists, court rooms, national legal systems and international conventions offer alternative routes for advancing political decisions and regulation of pressing environmental challenges. The potential for COs to become political instruments, for instance when environmental activists aim to trigger legal obligations by reporting multiple sightings of red-listed species to existing CS infrastructures, has already led to conflicts given the increasing interest in CS among national public authorities, governments, multilateral organizations, and NGOs. Facilitating a participatory, active citizenship for the public to engage in environmental monitoring can mobilize “activists” and “data collectors” simultaneously.

We confirm a central tenet of STS: science as much as CS are inherently political and, for this reason, there are grounds for skepticism about imaginaries that CS will achieve an impartial democratization of science (cf. Macq, Tancoigne and Strasser 2020), where everyone is represented. Most CS projects show a strong bias whereby participants identify as male, and most COs reported to biodiversity databases come from highly educated, upper-middle class, middle-aged or older, white participants (Purcell, Garibay and Dickinson 2012; Paleco et al. 2021). For almost two decades, Artportalen has exhibited such imbalances in age and gender, featuring a male-dominated birding culture now reproduced in the digital (Jönsson et al. 2023). Such inequalities have spurred ethical discussions on the necessity of including more diverse groups of participants in biodiversity observations (Mor Barak 2020).

The positive narratives of broader inclusion in CS tell of increasing participants’ knowledge (Brossard, Lewenstein and Bonney 2005; Jordan et al. 2011), science literacy (Crall et al. 2013; Bonney et al. 2016), engagement in conservation activities, environmental advocacy, and positive attitudes toward nature (Johnson et al. 2014; Lewandowski and Oberhauser 2017; McKinley et al. 2017; Sharma et al. 2019). However, CS also tends to reproduce imbalances in gender, class, education and distribution of projects between the Global North and South (Benyei et al. 2021). Serious consideration must be given to the question of who and what can speak for others as COs start to inform policy and law.

For some CS participants—in the context of rampant biodiversity decline and climate change—it becomes necessary to mobilize established infrastructures, national environmental regulations, international conventions, and certain modes of association to change or question the role of public authorities (Turner 2023; Turner and Wiber 2023). In other words, they recognize that COs have the potential to change unproductive governance structures. At the same time, our analysis demonstrates that achieving representation through COs also has the capacity to put some mechanisms of representative democracy out of play (cf. Chilvers and Kearnes 2020, 358). Certain actors can be accused of sidestepping democratic means of representation (e.g., voting) by producing COs that can be associated with legal obligations and that matter for the environmental regulatory context.

These practices directly point to what forms of representation matter and thus who and what can speak for other species and entities making up the world. The study of COs and CS in legal and political representation and deliberation is a growing field (see Ottinger 2013; 2022; Kimura 2016; Braverman 2018; Pearse 2020; Herzog and Lepenies 2022; Kasperowski et al. 2023; Turner and Wiber 2023). Previous studies have found that when existing environmental governance and regulatory structures are found to be weak, distrusted, or insufficient by environmentally concerned communities and stakeholders, COs are seen as an alternative political modus operandi (see Marres 2005; Gloppen and St Clair 2012; Ottinger 2013, 2022; Latour 2017). Moreover, the politics of scientific and democratic representations have been connected to efforts by environmental groups and communities devoted to decentering humans from the destructive legacies associated with the Anthropocene and species extinction events (Kramm 2020; Reeves and Peters 2021). With this study, we highlight where representation gets realized/decided through certain modes of associative and disassociative practices. Such practices occur at key sites for contestation and include COs’ connections to the space-times of species presence and behavior and classifications of what species merit protection. How such modes may affect the political processes of representative democracies is well beyond this paper, but certainly a question for the future, as different environments might be acted upon in ways that some actors may view as undemocratic, and others as a political necessity. If we want to understand who or what can speak for entities that make up the world, we could benefit from further understanding the associative and dissociative modes that produce legal obligations in the courthouse (cf. Braverman 2018; Turner and Wiber 2023).