Synthesis

INTRODUCTION

A core challenge of studying ecosystems and societies as one system is integrating social and ecological theorizing. Although the social and the ecological intertwine in multiple ways, much traditional theory focuses on one or the other (Schlüter et al. 2022). Despite good intentions, attempts for integration are often only partially successful. One reason for this is the apparent incommensurability of methodological and theoretical approaches. However, the mechanism-based approach to theorizing offers an encompassing way to think about social and ecological phenomena by framing them in terms of entities and their interactions. Mechanism-based theorizing has been gaining popularity in science philosophy (Machamer et al. 2000, Craver 2007, Glennan and Illari 2017), including social sciences philosophy (Elster 1989, Hedström and Swedberg 1998, Hedström and Ylikoski 2010) and philosophy of ecology (Pâslaru 2017, González del Solar et al. 2019). It has helped to resolve many traditional problems related to scientific explanation and it is consistent with the way in which social and natural scientists talk about theorizing, causation, and explanation.

Mechanism-based theorizing is based on ideas about scientific explanation. Rather than looking for regularities—like laws of nature or statistical generalizations—or simple statements about causal dependence, scientific explanation should show the cogs and wheels behind the events, that is, the causal processes through which interrelated entities generated the outcome (Elster 1989, Hedström and Ylikoski 2010, Glennan 2017). For mechanism-based theorizing to go beyond ad hoc causal story telling, systematic causal inquiry is key, which involves considering alternative explanations and rigorously searching for evidence that discriminates between alternatives. These ideas have led to increasing attention being given to issues like process tracing (Bennett and Checkel 2014) and explicit theorizing about relations between micro- and macro-scale concepts and processes (Ylikoski 2012).

We talk about micro and macro in terms of scales rather than levels. The metaphor of level is often a source of unnecessary confusion, both in ecology and social sciences. For example, phenomena can be studied at multiple (spatial or temporal) scales, but if this idea is articulated in terms of levels, it suggests a metaphysical commitment to the existence of multiple levels of reality. Although the author might think that she/he is not committed to such things, this might not be clear to the audience. The metaphor of level is especially confusing in discussions of causation and explanation (Ylikoski 2012). In the context of social-ecological systems (SES) research, it is good to remember that scales of phenomena do not usually follow administrative regions.

The potential of mechanism-based theorizing has not gone unnoticed by SES scholars, who have called for this approach to improve the precision and explanatory power of SES theorizing (Sandifer et al. 2015, Carlson et al. 2018, Schlüter et al. 2019b, Rullens et al. 2019). For example, Meyfroidt (2016) argues that understanding causal mechanisms is vital to making sense of why the same cause might produce different effects in land systems and argues for mechanisms-aware policy interventions. Mechanism-based theorizing, especially in the form of middle-range theories, has also been claimed to help produce generalized knowledge in SES (Meyfroidt et al. 2018, Schlüter et al. 2019b). However, calls for mechanism-based explanations and middle-range theorizing have remained quite abstract, as general ideas about causation and explanation do not provide concrete guidance for constructing theories that can integrate social and ecological elements consistently across the micro–macro dimension. The realization of the promise of mechanism-based theorizing requires heuristic guidance. In social theory, Coleman’s diagram has been an important part of such guidance (Coleman 1987, Hedström and Swedberg 1998, Ylikoski 2021). In this paper, we will show how it can help to integrate social and ecological mechanisms in SES theorizing and to conceptualize thorny micro–macro issues that SES research faces.

Our central claim is that an updated version of Coleman’s diagram can guide SES theorizing about the mechanisms underlying social-ecological phenomena. Focusing on social mechanisms improves the understanding of the social dimensions of SES, and the general mechanistic approach helps to conceptualize various ways in which social and ecological processes might intertwine. Although there have been plenty of frameworks for analyzing SES (Binder et al. 2013), the mechanism-based approach guided by the Coleman diagram is distinct as it provides a more general and robust perspective that is not tied too closely to specific theoretical or methodological ideas. Furthermore, it smooths the path toward interdisciplinary integration. Although using the diagram in SES research has been suggested before (Boonstra and de Boer 2014, Biesbroek et al. 2017, Elsawah et al. 2020), this paper provides the first systematic discussion of how ecological processes and social theorizing can be incorporated with its help. We wish to demonstrate that Coleman’s diagram is more than an emblem or a decoration. It is a tool for thinking.

Scholars interested in nature–society interactions have often provided mechanism-based explanations that correspond to one or several of the arrows although they probably did not use Coleman’s diagram. The contribution of the diagram is making explicit this way of thinking in a general model that helps to systematically think about the relations between micro and macro. The Coleman diagram is not in direct competition with earlier approaches, but provides a more general and systematic way to think about many kinds of social-ecological processes.

In our view, the best way to understand the diagram is as a series of questions associated with its arrows (Fig. 1). Using a recently updated version of the diagram (Ylikoski 2019, 2021) as a starting point, we show how ecological causes and mechanisms can be integrated into the diagram familiar to many social scientists. For social scientists, this article provides a way to expand their mechanism-based theorizing to include ecological processes and thus lowers the threshold for them to contribute to social-ecological research. In addition, our analysis and taxonomy of social agents’ attributes makes a novel contribution to understanding of the diagram. For SES researchers, this article provides a systematic introduction to the diagram and its utility for conceptualizing social-ecological processes. We hope to demonstrate that mechanism-based theorizing does not have to remain an abstract metaphor—like cogs and wheels—and that the Coleman diagram can provide useful guidance for consistent application of mechanism-based reasoning.

We have organized this article around the causal questions presented in Fig. 1. In the next section, we describe how the diagram has been used in the social sciences. Then, we dedicate a section to describe and provide examples of various ways in which the diagram can incorporate ecological processes, with a subsection for each of its components. We pay special attention to agent attributes by introducing the concepts of an action situation, mental states, and agent capacities into the diagram to better grasp the interface between agency and social-ecological change. We close this section with an example that comprises the whole diagram. We conclude the article with a discussion about various dimensions of causal complexity in social-ecological phenomena that can be incorporated into theorizing by using the diagram.

THE COLEMAN DIAGRAM

History

The Coleman diagram is a well-known theoretical diagram in sociology. It is often referred to as Coleman’s boat or Coleman’s bathtub. James Coleman first introduced the diagram in 1986 and used it in his subsequent publications (Coleman 1987, 1994). The diagram did not catch on immediately, but after Hedström and Swedberg (1998), it has become a popular representation of the micro–macro challenge in social sciences (for the prehistory of the diagram, see Barbera (2006) and Raub and Voss (2017)). There are various interpretations of the diagram, and people have often adapted it to their own purposes. In this paper, we will employ the version of the diagram presented in Ylikoski (2019, 2021) that makes explicit the way it helps social theorizing.

The Coleman diagram is a rare visual representation in sociological theory. Whereas most theoretical diagrams in sociology are summaries of verbal arguments (Turner 2010, Swedberg 2016), Coleman’s is a cognitive tool for sociological thinking. It can serve as a systematic scheme for articulating social explanations and their presuppositions. The abstract form of the diagram makes it easily adaptable to various situations while, at the same time, raising a sequence of theoretically crucial research questions (Ylikoski 2021). The secret of its intellectual productivity is precisely this ability to raise questions that trigger new lines of thinking about macro phenomena.

A core issue in social theory is the relationship between micro and macro social phenomena (Hedström 2005, Ylikoski 2012). Sociologists are concerned with explaining the emergence, persistence, and change of macro-scale social events, attributes, and processes. Linking these macro facts to micro-scale social events and processes, like individual behavior and social interactions, is a crucial challenge for social theory. The Coleman diagram helps by presenting a series of questions relating macro changes to agents’ behaviors (Ylikoski 2019, 2021). It begins by asking to clarify: what is the macro change (A) that is considered the cause? Then, it proceeds to ask: how does this change affect the relevant agents and their action situations (B)? Once this question (arrow 1) is answered, the next step is to ask: how do changes in B modify agents’ behaviors (arrow 2)? Furthermore, it asks to consider: what consequences do these behavioral changes have for other agents (arrow 3)? Finally, understanding the loop of arrows 2 and 3 for various agents over time makes it possible to see how the macro outcome (D) gets generated (Fig. 1).

Purpose of the diagram

The diagram’s purpose is to help think about the link between two macro phenomena, and it does not represent a fully fledged social theory (Coleman 1987). Thus, it accepts various theoretical ideas, e.g., about human and corporate action (Ylikoski 2021). The research question defines the relevant macro phenomena, and the diagram provides a series of questions about the social mechanisms involved. The diagram is not simplistic or reductionist. Although the questions it raises may seem simple, the answers are often complex and involve multiple causal pathways, heterogeneous agents, and complex institutional background conditions. The diagram is ambitious in the following sense: it is based on the idea that the connection between A and D is not fully explained if questions 1–4 are not answered. This is the idea of a mechanism-based explanation: causal claims about macro phenomena gain strength if we have a better idea about the underlying causal mechanisms (Hedström and Ylikoski 2010). In the case of social sciences, this requires connecting macro events to the micro activities of the agents (Hedström 2005). The diagram shows how to split such explanations into four analytically distinct explanatory subtasks. Naturally, the claim is not that a single study or publication could provide the complete story. The diagram provides a skeleton for an ideal explanatory narrative. Such guidance is valuable, for example, when considering how findings from different studies are connected or when asking whether our explanatory story has missing links. Finally, although the diagram is focused on representing and tracking the consequences (or causes) of specified macro changes, nothing prevents using the diagram multiple times. Thus, when tracking the consequences of multiple macro changes, it is possible to use multiple diagrams with shared elements.

There are several possible entry points to using the diagram. We might start with a macro change (A)—e.g., a new policy or an ecological change—for which we do not know the consequences. In this case, the diagram helps to figure them out by directing our attention to the local micro consequences for the relevant agents and the possible feedback loops (B–C) generated by agents’ actions. We might instead start with the macro consequence (D) and questions on how it was brought about. In another situation, there might be questions about the possible causal relationship between the macro variables A and D—e.g., between indicators that measure aggregate change as it is done in the planetary boundaries and sustainable development goal frameworks. Is it a mere correlation or is there a real causal connection? In this kind of case, following the diagram often helps uncover information about the relevant causal mechanisms that would add credence to the claim about the causal relationship between the macro variables. And if no mechanisms are uncovered, there are good reasons to be skeptical about the causal relationship. Finally, even in cases where we are convinced of the causal connection between the macro-variables A and D, there is still scientific value in understanding how changes in A generate changes in D. Uncovering the underlying mechanisms will also help to understand under which conditions the macro phenomena hold (Ylikoski 2021).

The diagram does not present any substantive hypotheses, nor is it a representation of a causal structure—it is not a poor man’s theory or directed acyclic graph. The idea is that answering the diagram’s questions generates a series of causal hypotheses that can be tested. Furthermore, when all the questions are answered, we have both a much better scientific understanding of the phenomenon and better justificatory support for our causal claims. This is the core idea of the mechanism-based approach to causal inquiry (Hedström and Ylikoski 2010).

The diagram can also be used to illustrate causal scenarios in particular studies. In fact, this is the most common use of the diagram in the literature. However, when used in this way, it should be remembered that the diagram should not be a straitjacket to represent causal scenarios. Often additional arrows and elements are needed. For example, there might be multiple causal pathways, heterogeneous agents, and feedback loops at various scales. The primary purpose of the diagram is to guide theoretical thinking, and it should not be taken as a mandatory format for presenting inquiry results. Causal loop diagrams, network graphs, flow models, and configurations of action-situations will often do better at representing concrete causal scenarios (Banitz et al. 2022).

Asking answerable causal questions with contrasts

Using contrasts and difference-making thinking is often instructive when using the diagram (Ylikoski 2019, 2021). Difference making refers to the idea that it is possible to identify a cause by looking at things that make a difference in the outcome. Let us start with the endpoints. The question of why the system ended up in D is made more precise by considering a relevant alternative state D*. The contrast helps to focus the inquiry: we are now looking for things that make a difference between D and D* rather than focusing on the whole causal process. Similarly, when interested in the causal consequences of A, it makes sense to think about what the sensible comparison case A* would be so that we can focus on figuring out the differences between the two scenarios. Explaining changes in fish population size requires a different contrast from explaining changes in its movement patterns. As this example shows, a natural way of thinking about contrasts is to think about them as changes. So, the relevant question is what changed D* to D, or what are the causal consequences of A* changing to A. This is also the format for basic causal claims, which usually describe how a change makes a difference to the outcome (Woodward 2003).

This sort of contrastive thinking is in line with pragmatics of explanation giving in science and everyday life: we are usually interested in difference makers as our explanations can never cover the whole causal history. More limited, precise, contrastive questions are easier to answer than bigger, non-contrastive ones (Ylikoski 2007). However, bigger questions can be cut down into smaller, more palatable ones by defining contrasts. From the point of view of mechanism-based explanations, the contrastive causal statements are a good starting point because they provide a precise explanandum—i.e., a statement about what is to be explained. This allows asking how the cause brought about the change in a well-defined outcome. It might be that the causal influence is transmitted via multiple causal pathways, but still, a well-formulated explanandum makes it possible to distinguish relevant mechanisms and background conditions from irrelevant ones. An additional benefit of the difference-making interpretation of the diagram is that it provides an excellent way to conceptualize the structural presuppositions of macro–micro and micro–macro claims: they are background conditions for the foreground causal claims (Ylikoski 2021.) The discussion of how contrastive thinking can be used for the arrows follows in the subsection “Putting the arrows to work.”

INCORPORATING THE ECOLOGICAL DIMENSION

Until now, the Coleman diagram has been used in the social sciences, but there is no principled reason why it cannot be used in social-ecological research. To demonstrate this, we will next walk through the diagram step by step to explain how it works, how the ecological dimension can be added, and provide examples from SES studies that have successfully produced mechanism-based explanations. As our discussion will show, there are multiple ways in which ecological and social processes could intertwine from a mechanism-based perspective. The following argument presupposes that social and ecological phenomena belong to the same causal order of the world and can get intertwined. Our discussion does not assume any strict dichotomy between social and ecological processes. Although some processes might be seen as purely ecological or social, most of them include both. To simplify the discussion, in the following, we will call anything that has an important ecological dimension “ecological.” Naturally, more fine-grained conceptual schemes can be introduced, if needed.

There are multiple ways to incorporate ecological events and processes into the diagram. First, they may be implicit background conditions for social processes. It is plausible to assume that ecological factors have this role in ordinary sociological theorizing. They are unrecognized background conditions believed to be present and only become relevant when a change in them interrupts familiar social processes. From the point of view of SES research, this state of affairs is naturally unsatisfactory: ecological events and processes should be parts of explicit theorizing.

The next possibility is that ecological events or processes serve as starting points (A) or as endpoints (D) of the diagram (Fig. 1). In the first case, we are interested in the social and ecological consequences of ecological macro change; in the second case, we are interested in the social and ecological causes of ecological macro changes. Although these possibilities are easy to recognize, it is also possible to consider situations where both A and D are ecological, but mediating causal mechanisms include important social elements. In other words, we have socially mediated ecological macro causation. An alternative possibility is that both A and D are social, but the mediating causal mechanisms include significant ecological elements. In other words, we have ecologically mediated social macro causation (Fig. 2).

Closer consideration of the last two possibilities gives us a better grasp of the situation. The arrows 1–4 raise causal questions whose answers describe causal mechanisms or pathways by which changes at the beginning of the arrow bring about changes at the end of the arrow. The crucial thing is that some or all of these causal pathways can be ecological. In other words, arrows 1, 3, and 4 can incorporate ecological elements in multiple ways. This covers the processes by which the macro change (A) affects agents and their action situations (arrow 1), the processes by which agents’ actions generate consequences to other agents (arrow 3), and how they affect the macro outcome D (arrow 4). Furthermore, the part of the diagram that describes human or corporate action (arrow 2) can be ecological: the action situation of the agent can have ecological elements, the mental states of the agent can be about ecological states or processes, and finally, the consequences of the agent’s action can be ecological.

In Fig. 2, we present examples of studies to illustrate how ecological elements can be incorporated in the different arrows. These examples also showcase how social and ecological can intertwine from this perspective. Depending on the research question, it is possible to consider ecological causes of social change, social causes of ecological change, social mechanisms underlying ecological changes, and ecological mechanisms underlying social change.

Ecological endpoints

Nodes A and D are the starting points to use the diagram. We are interested in the consequences of changes in macro facts A or the causes for the changes in macro facts D. D tells us what we aim to explain, whereas A refers to a suggested cause. The point of the diagram is to figure out their possible causal connection. It is crucial to observe that A and D do not refer to the whole “macro level,” but only to particular facts relevant to the hypothesis under consideration. The relevant macro facts can be pretty diverse, and their scale may vary. Macro changes can be of various types: structural changes, governmental policies, or demographic changes. They could be about revolutions, economic structures, inequalities, market panics, election outcomes, or the disintegration of families. It is impossible to think that all macro facts belong to some unique or comprehensive “social level.” The macro is not a fixed-size scale nor does it refer to “the whole society.” Instead, it is helpful to think of the micro–macro contrast in terms of relative scales (Ylikoski 2012). The macro facts can be of different “sizes:” from local social interactions to whole social systems, depending on the research question.

Once macro changes are conceptualized in this way, it is natural to expand the scope of A to ecological factors. They are large-scale changes connected to human activities, either by affecting agents and their action situations, being affected by the consequences of agents’ activities, or both. Examples could be climate change, lake eutrophication, erosion, biodiversity loss, spatial distribution of organisms, regime shifts, or fish population decline. Likewise, the endpoints (D) can represent outcomes that involve ecological consequences such as changes in harvest rates, entering social-ecological poverty traps, land-use changes, crop domestication, the shaping of cultural landscapes, supply of ecosystem services, changes in patterns of human–nature interaction, etc.

Notice that whether we identify a phenomenon with either A or D depends on the research question (Fig. 2). Outcome (D) for one study can be the trigger (A) for another. Conversely, all As can (and should) serve as Ds for other explanations. For instance, forest cover change can be considered A, if our question is about how deforestation affects water tables, but it can also be D, if we want to know how rural–urban migration affects reforestation; naturally, the fact that forest cover is A for one question and D for the other, allows us to connect migration and water tables through forest cover change. There are no unexplainable macro facts, but every explanation must have a starting point.

The diversity of ways in which social and ecological can combine in the diagram can already be seen at this stage. For a start, we have different combinations of ecological and social in the arrows of the boat illustrated in Fig. 2. But in addition, there are combinatorial possibilities within each arrow as one of the endpoints of the arrow, or both, could be ecological. We will return to these possibilities in the subsection “Putting the arrows to work.”

Agency

The bottom half of the diagram, the arrow between B and C, captures the role of theory of action in sociology. B captures the changes for the relevant agents and their situations as generated by the macro change A, and C represents the changes in behavior that are consequences of B. The diagram can be operated with a very liberal notion of agency, so, for example, it does not require a commitment to rational choice theory, which was James Coleman’s preferred theory of action (Ylikoski 2021). For an overview of behavioral theories used in social-ecological research see the work of Schlüter et al. (2017) and Constantino et al. (2021).

The notion of micro is not rigid. Agents do not have to be individual persons; they could also be corporate agents. Contrary to some critics, the diagram does not presuppose strict methodological individualism. Coleman highlighted the importance of non-individual (corporate) agents by arguing that various forms of corporate agency (corporations, states, unions, parties, etc.) have significantly changed our societies over the last two centuries (Coleman 1990). The relevant agents will depend on the case, not on general theoretical principles. Well-understood corporate agents are thus acceptable as bearers of agency without giving up the idea of action-based explanation. In SES research, it is common to encounter corporate agents like neighbor associations as nature stewards, organizations of resource users that engage in collective action, advocacy groups for environmental issues, etc. Whether zooming inside the organization is analytically unhelpful or crucial depends on the research question. The takeaway message is that the diagram should not be understood as an argument for methodological individualism. What matters is agency, not the individuality (or personhood) of agents.

Do the agents have to be individual humans or human corporate agents or is it possible to include other organisms? This is a fascinating question. From the point of view of analyzing micro–macro relationships, the crucial step is zooming into the micro scale. In an ecological model, the agents could be individual organisms, whereas the macro scale might consist of populations or species. Adding the organism scale would be important for understanding the macro dynamics at population scale. However, although organisms are agents, and it is possible to attribute goals and beliefs to them, they are not intentional agents assumed in the usual social science applications of the Coleman diagram. These agents would not only have mental states like beliefs and desires but also have mental states about their own mental states, i.e., metarepresentations (Pettit 1996). Furthermore, the idea of action-based explanation is based on the presumption that we can reliably use human abilities of perspective taking and empathy to better grasp other humans’ reasoning.

In this paper, we will only talk about human agents. Our choice is not based on metaphysical arguments. Although there is a clear psychological difference between a human and, for example, a fish, these differences are empirical matters. However, the empirical differences between humans and other agents matter. When our goal is to understand social-ecological processes, it is important to capture humans as accurately as possible. Furthermore, there are two pragmatic arguments against distributing human attributes too easily. First, although we can understand other humans by simulating their minds with our own (Goldman 2006), it is pretty unclear how reliably this works for other species. Thus, although people might be eager to humanize other species, it is unclear how well this captures their cognitive processes. In other words, mentalizing other species might be highly unreliable. Second, for many purposes, it is unnecessary to assume that non-human agents have a complicated mental life. This is ultimately an empirical matter: if it turns out that the mental states and cognition of non-human and non-organizational agents are indeed crucial for explaining their relevant behaviors, then they could be considered agents. If not, we can employ much simpler assumptions, like behavioral rules. These issues are complicated, and they would require a much more extensive discussion. In this paper, we will stick to the idea that relevant agents are individual humans or corporate agents. In the future, artificial intelligence (AI) agents and robots will be increasingly important parts of society. They will be making choices and decisions, but it remains an open question how their agency should be treated. Basically, the same arguments apply as in the case of organisms: can we reliably mentally simulate them and what are the analytical advantages for postulating extensive mental state to them?

Arrow 2 captures the role of human (or corporate) agency in the processes under analysis. The role of the theory of action is to causally link macro changes to changes in behaviors and their consequences. The node C at the end of the arrow can be conceptualized as choices, behaviors, or actions, depending on the context. For social scientists, changes in behaviors are important targets of explanation and a crucial step in understanding the generation of macro outcomes (D). From the social-ecological point of view, they are equally important. Choices might also be about ecological things, and actions might have ecological consequences—often indirect, unintended ones (Fig. 3).

Agent attributes

Node B can be further dissected in terms of agent attributes. These are the “interfaces” by which arrow 1 (or arrow 3) can influence agents. Agent attributes can be classified into three rough categories: the agent’s (i) action situation, (ii) mental states, and (iii) agent capacities. The categories are rough, because sometimes it might be difficult to decide into which category a certain agent attribute belongs. For example, the notion of interest can describe either an agent’s action situations or their mental states. Earlier sociological literature has not systematically distinguished between these three categories, so we will discuss them in more detail. From the point of view of this paper, it is important that all three can have an interesting ecological dimension. Nature can be a resource or a threat (action situation), the agent can have beliefs, preferences, emotions, and values related to social-ecological affairs (mental states), and the agent capacities might be social-ecological (agent capacities). Note that different theoretical approaches might use slightly distinct conceptualizations of action situations, mental states, and agent capacities. Still, from the more general point of view, they are all agent attributes linking arrows A and C (Table 1).

Action situation

Let us start with the action situation. Elinor Ostrom popularized this term to refer to situations where rules govern interactions (McGinnis and Ostrom 2014). To us, the possibilities of interaction are influenced by institutions and legal obligations, but also by other aspects of social situations like opportunities (e.g., education, housing, and work opportunities), expectations, and rights. The action situation also includes an ecological dimension that can influence action, like the benefits and opportunities provided by surrounding ecosystems, weather conditions, etc. (Schlüter et al. 2019a). Thus, the combination of social and ecological factors makes up the action situation. Elements of the action situation can be conceptualized in many ways, for example, sociologists often talk about opportunity structures or opportunity sets. The crucial thing is that changes in the action situation influence how agents behave. Naturally, this influence usually presupposes cognition by the agent: the agent must observe or learn about changes in the action situation. The implication is that the impact through the action environment might not be immediate: the agents must first recognize the changes and adapt their behavior, if needed. An example of how changes in action situations lead to behavioral change is the case of spatial diversification among small-scale fishers in Baja California, Mexico. Small-scale fishers target different species, depending on fish availability, fishing permits, and vehicles they have access to. When high-priced fish species migrate, the action situation changes. Fishers with access to corresponding licenses and vehicles change their behavior by diversifying spatially, i.e., they follow fish migration; whereas fishers without access to either licenses or vehicles change their behavior by diversifying their local fishing portfolio and targeting less pricey fish (González-Mon et al. 2021).

Mental states

For human behavior, agents’ mental states are crucial for understanding their actions. Folk psychology employs a wide variety of mental concepts: beliefs, goals and desires, motives, emotions, attitudes, values and preferences—we use notions of cognition and mind broadly, which means that these states are not necessarily conscious (Pettit 1996, Goldman 2006, Hutto 2012). Different variants of folk psychology may use different concepts, but in this context, we will not take a stance on them. The crucial point here is that all these folk psychological concepts serve a similar function in social scientific explanations: they mediate between social influences (A) and individual behaviors (C). The changes in the action environment bring about changes in individual actions via changes in agent’s mental states. Despite advances in neurosciences, the intentional concepts employed by folk psychology are indispensable for understanding agents’ behaviors. From our paper’s point of view, the pivotal point is that mental states can be about ecological things. They can be, for example, beliefs and expectations about ecosystems, or goals, desires, and preferences related to them. In other words, the content of mental states can be about socio-ecological states and processes. An example of how mental states about ecosystems link to action is the case of the Tandroy people’s taboo system, according to which they believe that some forest patches are sacred. During the 1980s in the south of Madagascar, the government overruled the customary land tenure system. As collective land rights were no longer acknowledged, the only formal way to claim land was to turn it into farmland, which led to a large-scale cut down. However, the Tandroy people selectively cleared the forest following their taboo system. The “ala kibory,” forest patches that were used as burial places, were spared (Tengö and Heland 2011).

Agent capacities

Finally, agent capacities refer to agents’ attributes that influence their capacity for action, e.g., skills, cognitive abilities, physical capacities, and by extension, available tools and technologies. Consider, e.g., the relevance of literacy and numeracy in the social world. People lacking these skills face a significant disadvantage in modern societies. Agent capacities are also relevant for corporate agents, and for them, the internal social organization of activities is a crucial determinant of the scope of possible actions. Agent capacities are not usually distinguished from other agent attributes in sociological discussions, and their elements are treated as parts of agents’ action situations or mental states. For instance, Hedström’s own preferred model only talks about beliefs, desires, and opportunities, taking the agent capacities for granted (Hedström 2005). However, we argue that distinguishing this additional dimension helps clarify agent attributes and determinants of behavior. For example, it could be possible to analyze the influence of malnutrition on cognition in terms of the resources available in the action situation and relevant mental states, but these conceptual acrobatics would not provide any analytical advantage. Agent capacities are especially useful when considering behaviors that involve direct human–nature interactions. How humans interact with their ecosystems is conditional on agent capacities, which are often extended by available technology. The impact of bottom trawling on marine sediments cannot be explained without the capabilities enabled by drag nets (Jones 1992, Hiddink et al. 2017).

Interactions of agent attributes

Naturally, the interactions between the three dimensions are crucially important. Learning, an agent capacity, allows acquiring updated belief, a mental state, about their opportunity space, i.e., action situation. This knowledge might lead agents to develop new skills, and in doing so, expand their capacities, and better skills or tools for making observations, or preserving them can help agents further learn about their environment. Other interactions are possible that go beyond the scope of our discussion, but the pivotal thing is to keep an eye on all three dimensions and their interactions as they are often the critical factors that explain things like the differential success of agents, e.g., to adapt to crises. Social-ecological systems scholars have articulated frameworks that account for more than one of these dimensions and relationships among them (Raymond et al. 2018, Gillette et al. 2022), and they can be accommodated in mechanism-based theorizing. However, it is good to bear in mind that, for many purposes, it is unnecessary to analyze agency in a complicated way.

Putting the arrows to work

We now have an idea of what the nodes in the diagram represent—and how they can capture the ecological dimension. The arrows are slightly more difficult. Nodes represent the starting points (arrow tails) and endpoints (arrowheads). The best way to understand the arrow shafts is as explanatory relationships based on causal dependencies. A natural way to interpret them is in terms of difference making, as discussed in the subsection “Asking answerable questions with contrasts.” In the case of arrow 1, e.g., a change in A brings about a change in B. In the same way, the whole captures the idea that the consequences of changes in A are transmitted along the arrow up to the macro change D—the final explanandum. Answering the questions captured in the arrows helps investigate each of the sections. It must be born in mind that the mechanisms that are the answers to A–B and C–D questions presuppose background conditions on which relationships are contingent (institutional, cultural, ecological, etc.). If the background conditions were different, the dependence would be different or even absent. The unarticulated background conditions in ordinary social scientific uses of the diagram hide implicit ecological assumptions. Acknowledging these assumptions explicitly is a way to make arrows more ecological. However, there is no reason to stop there. Ecological macro starting points, ecological causal pathways, and ecological endpoints, like agents’ action situations are also possible. In other words, there are many ways in which an arrow could be ecological (Fig. 4).

The downward arrow shaft

Arrow 1 directs causal inquiry toward the dependency relationships between the macro change (A) and agent attributes that influence action (B). A benefit of the notion of scale when using the diagram is that there is no need to conceive arrow 1 as somewhat mysterious “downward causation.” The down-to-earth idea of large-scale social or ecological factors influencing smaller-scale facts about agents and their action situations is enough (Ylikoski 2012). In other words, it is a question of how large-scale changes bring about smaller-scale changes. To incorporate the ecological dimension in this arrow, there are three possibilities: the starting point (A), the arrow shaft (or part of it), and the arrowhead (B). As the starting point has already been discussed in the subsection “Ecological endpoints,” we will focus here on the latter two.

The question associated with the downward arrow asks “How does macro change make a difference for micro-scale processes?” (Fig. 1). The idea is to trace the consequences of a larger-scale change on the scale of agents. Following difference-making thinking, this is equivalent to asking how changes from A to A* lead to changes from B to B*. For instance: How climate change makes a difference in herders’ income from the arid north of Kenya? The answer can look something like this: climate change makes a difference in rain patterns as it increases drought frequency, which in turn makes a difference in primary production and thus fodder availability, which makes a difference in herd survival, which in turn makes a difference in herders’ income (Ng’ang’a et al. 2016a, b). This example shows a chain of difference making compatible with mechanisms-based theorizing: the answer to a question about mechanisms will tell us why and how a change in A leads to a change in B.

Although the explanation-seeking question associated with an arrow might be simple, the answer is often complex. As in the previous example, the causal link between A and B may be mediated rather than direct, which implies that it should be considered a causal chain. It is also possible to find multiple causal pathways (mechanisms), which may interact. For example, a legislative change can influence agents both by changing their legal rights and opportunities and by influencing their values. If we zoom in on any mechanism, we will find finer-grain steps, however, the investigative context determines how detailed analysis of the causal pathway is needed. Sometimes it is enough to recognize the causal dependence between A and B, but in other cases, paying more attention to alternative pathways and their details is crucial. For example, when the causal pathways are counteracting or modifying each other’s influence, the more detailed focus makes sense. Similarly, the details matter more if the causal effect is very context sensitive. From the point of view of SES, indirect and interacting causal pathways are an important locus of intertwinedness (or coupling) of social and ecological processes. One or more of the links in the causal chain might be ecological, and this holds for cases in which multiple causal pathways interact.

The downward arrowhead

We have already discussed the relevance of ecological elements for thinking about agent attributes or node B (“Agency” subsection). However, we like to highlight the relevance of ecological opportunities for action situations. Changes in opportunities can lead to changes in courses of action because they affect the available options for making choices. This holds both for social and ecological opportunities. Often, the available opportunity space is determined jointly by social and ecological situational elements. For example, fishers’ decisions on what, when, for how long, and how far to fish depend on ecological elements such as the distribution, diversity, and abundance of fish, as well as on social elements such as fish prices, institutional constraints, and alternative sources of income (Holland 2008, van Putten et al. 2012, Boonstra and Hentati-Sundberg 2016, González-Mon et al. 2021).

Opportunities provided by ecosystems are relevant to agent heterogeneity. How the distribution of ecological opportunities among agents intersects with societal opportunities—like those associated with gender, socioeconomic status, age, and occupation—is critical to understanding the differential impacts of ecological changes on agents. For example, farmers having the same social opportunities will experience the consequences of a drought very differently depending on whether their crops are water demanding or drought resistant, and whether they have access to a well or not. Similarly, opportunities provided by social structures, like institutions, might mediate between agents and opportunities provided by ecosystems, which modifies the extent of agents’ action situations, and thus their response to macro changes. For example, food-processing companies often expand their ecological opportunities by relying on multiple suppliers who, in turn, rely on multiple ecosystems. The consequences of environmental shocks caused by climate change are experienced as drastically different by the agents in the large processing company compared with the small-scale farm that relies only on local ecosystems (Davis et al. 2021).

Although differences in opportunities are a major source of agent heterogeneity, the differences in action situations are not the only source of heterogeneity; agents differ also in their mental states and capacities.

Micro-scale interdependencies

When Coleman originally proposed the diagram, one of his main points was that social scientific research pays a lot of attention to the questions associated with arrows 1 and 2, whereas the theoretically more challenging final steps of the diagram get much less attention. For example, this is the main point of his criticism of Max Weber’s (1905) thesis about protestant ethics and the development of modern capitalism (Coleman 1986). His purpose was to highlight the importance of the micro–macro transition: how do changes in individual behaviors bring about macro changes? It is common to assume that the macro order is just an aggregate effect of micro facts about individuals, but Coleman opposed this reductionist assumption: the transition from micro to macro is more than mere aggregation, especially in those cases that involve structural change. However, Coleman’s discussion of the final arrow remained sketchy. For this reason, the improved diagram used in this paper incorporates an additional feedback arrow 3 that helps to conceptualize the dynamic element of social change (Ylikoski 2021).

The idea behind arrow 3 is that theoretically interesting processes are rarely simple one-shot causal chains that connect A and D. In most interesting cases, the B is not merely influenced by changes in A, but also by the consequences of agents’ actions, that is, by C. In other words, agents’ actions influence their own and other agents’ later action situations, mental states, and agent capacities. Furthermore, this feedback loop might influence how the causal influence of A is transmitted to B, for example, modifying the causal impact either directly or by influencing, e.g., institutional or ecological background conditions. The backward loop provided by arrow 3 describes this interdependence between agents. Now the unfolding social process can be analyzed as a looping process, and D can be understood as the macro-scale product of this process (Ylikoski 2019, 2021).

Schelling’s (1978) segregation model illustrates this idea nicely. The initial setting can be interpreted as a macro change that allows families to move freely to new neighborhoods. In other words, after this initial macro change, the agents who are dissatisfied with the social composition of their neighborhood are free to move. Their movements, however, change the composition of both the neighborhood that they are leaving and the neighborhood where they are arriving. A consequence of these moves is that some agents that were satisfied earlier are now facing an unsatisfactory situation. Arrow 3 captures this feedback loop. The feedback is based on the interdependence between the agents: what an agent does affects what others can do and what they want to do. The explanandum (D) is only the final outcome of this process, so producing it could require multiple rounds of feedback.

However, the Schelling example does not fully capture the potential of the diagram. In many interesting cases, the result of the process is a structural change, e.g., the rules of the game change (Hernes 1976). In the Schelling model, the rules for the movement of agents remain the same, only the distribution of the agents changes. In structural change, the structural and institutional rules also change. Espeland and colleagues’ (2016) study of the effects of U.S. News law school rankings is a social scientific example of this. The authors show how introducing school rankings produced long-standing changes in American legal education: how students were admitted, how the schools used their resources, to whom the schools targeted their marketing, and how the career services advised fresh graduates. These changes, in turn, changed U.S. legal education structurally, e.g., by changing the composition of the student body and increasing similarity and status competition between the schools. These changes can be captured by the revised Coleman diagram (Ylikoski 2019, 2021) by focusing on how the usability of the ranking as a cognitive tool changed applicants’ behaviors, which in turn affected the deans, admissions officers, employers, and university administrators, which then affected how the students thought about the rankings.

The endpoint of arrow 3 is purposefully ambiguous. The behavioral consequences could affect other agents directly or via institutional and structural conditions at any scale. This is an important empirical question and the arrow’s purpose is to remind the researcher that it is pivotal to look for these feedback loops as they are often crucial for understanding social change. Arrow 3 highlights how thinking with the diagram is a natural way of approaching the complexity of social processes: heterogeneity and interdependence of agents, sensitivity to the background conditions, and feedback processes. Although only the last is explicitly represented in the diagram, the questions about others are its natural corollaries (Ylikoski 2021).

The same line of reasoning applies to social-ecological processes. Let us start with the beginning of arrow 3: the agent behaviors (C) have ecological consequences. Some of these may have consequences for the ecological aspects of the action situations the agents will face (B) later on. These consequences might be indirect as they might be transmitted by the background conditions of arrow 1, rather than directly influencing the action situations. In other words, the scale of the endpoint of arrow 3 can be variable, so paying attention to it makes a lot of sense. For instance, the consequences of deforestation might affect agents through changes in local micro-climatic conditions, regional forest connectivity, or national market prices for arable land. Next, the mechanisms supporting arrow 3 might be ecological. Here, the situation is analogical to arrow 1: any elements of the arrow might be ecological (or social). Thus, the general lessons from Fig. 4 also apply here.

From micro-scale interdependencies to macro outcomes

Once the social process generated by arrows 2 and 3 loops is understood, it is easy to understand how the outcome (D) is generated. We only need to zoom out to the macro scale so that we can see what the larger-scale consequences of the micro-scale behaviors and loops are. Careful attention to the interdependencies captured by the loops can be expected to be particularly useful to theorizing environmental governance transformations as it is a phenomenon that involves multiple interdependencies of many actors across scales and often involves structural change.

To illustrate the relationship between micro and macro, it is best to show a full example of how the diagram works as a series of analytical questions. We follow up on the example provided in the previous section inspired by the work by Ng’ang’a and colleagues (2016a, b) on migration and climate change among herders in the north of Kenya. Although the authors did not use Coleman’s diagram, their reasoning is analogous to it and thus serves as an illustration. The main question of their research was: “What are the consequences of climate change on rural–urban migration from pastoralist systems in the north of Kenya?” The hypothesis was that, as climate change increases, migration will accelerate until entire communities of herders move to urban settlements. Let us first look at some important background facts and then break the main question down into smaller ones following Coleman’s diagram.

The important background facts about herding are the following. Livestock serves multiple purposes for the herders. It allows them to obtain nutrients from the scarce vegetation, as draft power, and provides a commodity that can be sold to buy maize. Families seek to assemble a herd that secures constant access to milk. Socioeconomic status is tightly linked to the size of the family’s herd, and wealthier families often hire families that do not own a herd to help them with herding. Now, we can move to the first question (arrow 1): How does climate change affect herders? The authors’ answer is that climate change increases the duration and frequency of droughts. During droughts, fodder and water become scarce, leading to cattle deaths. Herds are decimated, but families with smaller herds lose the capacity to self-sustain. These are dramatic changes in herders’ action situations. The next question (arrow 2) is, how does herder families’ behavior change due to reduced herd size? According to the authors, families whose herd is not big enough to sustain the household but can still gather resources to send a family member, often a young male, to a nearby city to get a job and supplement the family income. The follow-up question (arrow 3) is the following: What were the consequences of migration for the communities? The authors observe multiple consequences. First, women get the burden of raising children and becoming household heads. They seek support from neighbors and relatives to endure the period before migrants can send their remittances. Second, the remittances allow families to recover from drought impacts and, in time, replace drought-sensitive livestock with drought-resistant (camels and goats) livestock. Third, earlier migrants make migration easier for other herders as they can more easily navigate the city and find a job. This leads to the final question (arrow 4): What were the overall consequences of remittances, gendered migration, and livestock substitution on migration rates? Although the droughts keep occurring, migration rates slow down after the first peak. The decline in migration is due to two principal reasons. First, the number of young males available for migration shrinks. Second, the drought-resistant livestock acquired with the help of remittances prevent previously vulnerable families from migrating. This mechanism-based explanation debunks the original hypothesis that assumed a linear relationship between climate change and migration.

DEALING WITH CAUSAL COMPLEXITY

The Coleman diagram provides a series of simple questions that need to be answered if the macro fact of interest is to be satisfactorily explained (Fig. 1). However, as we have seen, the answers to these questions can be complex. It is good to consider what kinds of causal complexity the diagram can incorporate. From the point of view of SES research, it is interesting to observe the various forms of intertwinedness of the social and ecological that these forms of causal complexity incorporate (Folke et al. 2016, Schlüter et al. 2020). Due to space limitations, it is impossible to discuss these dimensions of complexity in detail. However, we have collected eight different ways the diagram can accommodate complexity (Table 2).

The first three sources of causal complexity are related to individual arrows of the diagram. The idea is that although the question associated with each of the arrows is simple, the answer to it might be complex in three different ways. The agency provides two major additional sources of causal complexity: heterogeneity and interdependency. Finally, it is possible to use the diagram iteratively to capture broader changes in the overall causal configuration related to structural changes and different rates of change.

CONCLUSION

The integration of social and ecological theorizing is a core challenge for SES research, and many frameworks have been proposed to advance this goal. Binder and colleagues (2013) present a useful classification of such attempts. What they find is that the majority of the approaches fail to deal with micro–macro interactions of the social system and also fail to address interactions between different scales of ecological processes. Furthermore, many of them do not succeed in analyzing the interactions between social and ecological systems. Especially rare were frameworks that allowed theorizing about various feedback loops in social-ecological systems. The ecological Coleman diagram is not in direct competition with any of these frameworks as they have different purposes or operate at a different level of abstraction. For instance, the purpose of the Management and Transition Framework and the Human–Environment System Framework is the analysis of settings aimed at producing desired outcomes through management, whereas the purpose of the Coleman diagram is to guide causal reasoning and mechanism-based theorizing (Pahl-Wostl et al. 2010, Scholz et al. 2011). The Social-Ecological Systems framework is robust, comprehensive, and successful at explaining specific phenomena, namely sustainable and unsustainable outcomes of common-pool resources (Ostrom 2007, 2009). Similarly, other frameworks, such as the Turner Vulnerability Framework and the Sustainable Livelihood Approach, focus on specific types of outcomes and settings (Scoones 1998, Turner et al. 2003). The contribution of Coleman’s diagram is complementary as it provides intellectual tools to address these issues in a consistent manner. Thus, it should be regarded as a meta-tool for conceptualizing social-ecological processes. For instance, systems dynamics and causal loop diagrams, popular analytical tools in SES research (e.g., Zamora-Maldonado et al. 2021), frame feedbacks in terms of aggregate variables but often miss the role of agents and interactions underlying the causal relationship between variables; this could be mitigated by using Coleman’s diagram. This more abstract mechanism-focused perspective also helps to think about social-ecological processes in a more balanced manner and also to see the limitations of anthropocentric approaches that are dominated, e.g., by governance concerns. Naturally, many analytical strategies and methodological frameworks are compatible with the Coleman diagram and will be useful to answer the questions contained in its arrows, e.g., the social-ecological network approach, or the network of action situations approach, and agent-based modeling, to mention a few (Bodin et al. 2019, Kimmich et al. 2023).

Social-ecological systems researchers, along with researchers in many other fields in social and biological sciences, have called for clear-headed mechanism-based theorizing. This paper has shown how the Coleman diagram, familiarly used for mechanism-based theorizing in the social sciences, can be expanded to incorporate ecological and social-ecological processes. This expands the usability of the diagram and hopefully makes it a valuable addition to the theoretical toolbox of SES research. We also hope it makes it easier to build bridges between traditional social science researchers and social-ecological researchers. Although the diagram is not a magic formula for producing better SES theories and explanations, it provides a valuable thinking tool for SES researchers. It can be used to construct social-ecological explanations and to explore the causal consequences of policy changes and macro-scale ecological change. Furthermore, it can help to conceive possibilities for the intertwinedness (or coupling) of social and ecological systems while providing a methodical way of thinking about how macro outcomes and causes are related to micro processes.

This paper also contributes to a better understanding of the diagram itself. First, the paper demonstrates the heuristic usefulness of the updated version of the Coleman diagram (Ylikoski 2019, 2021). The new version allows an easy incorporation of ecological mechanisms, which was not obvious in earlier interpretations of the diagram. Furthermore, the paper expands the diagram’s conceptual underpinnings by distinguishing three kinds of agent attributes: the action situation, mental states, and agent capacities. These concepts provide a better grasp of the different sorts of agent attributes that have not been discussed in social science literature. The paper also contributes to social theory. It shows that ecological mechanisms and phenomena can be incorporated into social theorizing. When both social and ecological processes can be theorized in terms of causal mechanisms, there is one less reason to keep up arbitrary boundaries between the fields. This makes mechanisms-based theorizing a viable avenue for developing more ambitious interdisciplinary theories about major challenges facing both people and ecosystems.

LITERATURE CITED

Banitz, T., T. Hertz, L.-G. Johansson, E. Lindkvist, R. Martínez-Peña, S. Radosavljevic, M. Schlüter, K. Wennberg, P. K. Ylikoski, and V. Grimm. 2022. Visualization of causation in social-ecological systems. Ecology and Society 27(1): 31. https://doi.org/10.5751/ES-13030-270131

Barbera, F. 2006. A star is born? The authors, principles and objectives of analytical sociology. Revista de Sociologia 80: 31. https://doi.org/10.5565/rev/papers/v80n0.1768

Bennett, A., and J. T. Checkel, editors. 2014. Process tracing: from metaphor to analytic tool. Cambridge University Press, Cambridge, UK. https://doi.org/10.1017/CBO9781139858472

Biesbroek, R., J. Dupuis, and A. Wellstead. 2017. Explaining through causal mechanisms: resilience and governance of social–ecological systems. Current Opinion in Environmental Sustainability 28:64-70. https://doi.org/10.1016/j.cosust.2017.08.007

Binder, C. R., J. Hinkel, P. W. G. Bots, and C. Pahl-Wostl. 2013. Comparison of frameworks for analyzing social-ecological systems. Ecology and Society 18(4): 26. http://dx.doi.org/10.5751/ES-05551-180426

Bodin, Ö., S. M. Alexander, J. Baggio, M. L. Barnes, R. Berardo, G. S. Cumming, L. E. Dee, A. P. Fischer, M. Fischer, M. Mancilla Garcia, and A. M. Guerrero. 2019. Improving network approaches to the study of complex social–ecological interdependencies. Nature Sustainability 2(7):551-559. https://doi.org/10.1038/s41893-019-0308-0

Bodin, Ö., G. Robins, R. R. J. McAllister, A. M. Guerrero, B. Crona, M. Tengö, and M. Lubell. 2016. Theorizing benefits and constraints in collaborative environmental governance: a transdisciplinary social-ecological network approach for empirical investigations. Ecology and Society 21(1):40. https://doi.org/10.5751/ES-08368-210140

Boonstra, W. J., and F. W. de Boer. 2014. The historical dynamics of social–ecological traps. AMBIO 43(3):260-274. https://doi.org/10.1007/s13280-013-0419-1

Boonstra, W. J., and J. Hentati-Sundberg. 2016. Classifying fishers’ behaviour. An invitation to fishing styles. Fish and Fisheries 17(1):78-100. https://doi.org/10.1111/faf.12092

Carlson, A., J. Zaehringer, R. Garrett, R. Felipe Bicudo Silva, P. Furumo, A. Raya Rey, A. Torres, M. Gon Chung, Y. Li, and J. Liu. 2018. Toward rigorous telecoupling causal attribution: a systematic review and typology. Sustainability 10(12): 4426. https://doi.org/10.3390/su10124426

Chaudhary, S., A. McGregor, D. Houston, and N. Chettri. 2018. Environmental justice and ecosystem services: a disaggregated analysis of community access to forest benefits in Nepal. Ecosystem Services 29:99-115. https://doi.org/10.1016/j.ecoser.2017.10.020

Coleman, J. S. 1986. Social theory, social research, and a theory of action. American Journal of Sociology 91(6):1309-1335. https://doi.org/10.1086/228423

Coleman, J. S. 1987. Microfoundations and macrosocial behavior. The Micro-Macro Link 153: 173.

Coleman, J. S. 1990. Foundations of social theory. Harvard University Press, London, UK.

Coleman, J. S. 1994. Foundations of social theory. Harvard University Press, Cambridge, Massachusetts, USA.

Constantino, S., M. Schlüter, E. U. Weber, and N. Wijermans. 2021. Cognition and behavior in context: a framework and theories to explain natural resource use decisions in social-ecological systems. Sustainability Science 16(5):1651-1671. https://doi.org/10.1007/s11625-021-00989-w

Corsi, S., L. V. Marchisio, and L. Orsi. 2017. Connecting smallholder farmers to local markets: drivers of collective action, land tenure and food security in East Chad. Land Use Policy 68:39-47. https://doi.org/10.1016/j.landusepol.2017.07.025

Craver, C. F. 2007. Explaining the brain: mechanisms and the mosaic unity of neuroscience. Clarendon Press, Oxford, UK. https://doi.org/10.1093/acprof:oso/9780199299317.001.0001

Crona, B. I., T. van Holt, M. Petersson, T. M. Daw, and E. Buchary. 2015. Using social–ecological syndromes to understand impacts of international seafood trade on small-scale fisheries. Global Environmental Change 35:162-175. https://doi.org/10.1016/j.gloenvcha.2015.07.006

Cumming, G. S. 2018. A review of social dilemmas and social–ecological traps in conservation and natural resource management. Conservation Letters 11(1): e12376. https://doi.org/10.1111/conl.12376

Davidson, K. 2014. A typology to categorize the ideologies of actors in the sustainable development debate: a political economy typology of sustainability. Sustainable Development 22(1):1-14. https://doi.org/10.1002/sd.520

Davis, K. F., S. Downs, and J. A. Gephart. 2021. Towards food supply chain resilience to environmental shocks. Nature Food 2(1):54-65. https://doi.org/10.1038/s43016-020-00196-3

Delgado-Serrano, M. del M., E. Oteros-Rozas, P. Vanwildemeersch, C. Ortíz-Guerrero, S. London, and R. Escalante. 2015. Local perceptions on social-ecological dynamics in Latin America in three community-based natural resource management systems. Ecology and Society 20(4): 24. http://doi.org/10.5751/ES-07965-200424

Elsawah, S., T. Filatova, A. J. Jakeman, A. J. Kettner, M. L. Zellner, I. N. Athanasiadis, S. H. Hamilton, R. L. Axtell, D. G. Brown, J. M. Gilligan, M. A. Janssen, D. T. Robinson, J. Rozenberg, I. I. T. Ullah, and S. J. Lade. 2020. Eight grand challenges in socio-environmental systems modeling. Socio-Environmental Systems Modelling 2:16226-16226. https://doi.org/10.18174/sesmo.2020a16226

Elster, J. 1989. Nuts and bolts for the social sciences. Cambridge University Press, Cambridge, UK. https://doi.org/10.1017/CBO9780511812255

Epstein, G., J. Pittman, S. M. Alexander, S. Berdej, T. Dyck, U. Kreitmair, K. J. Rathwell, S. Villamayor-Tomas, J. Vogt, and D. Armitage. 2015. Institutional fit and the sustainability of social–ecological systems. Open Issue 14:3-40. https://doi.org/10.1016/j.cosust.2015.03.005

Espeland, W. N., M. Sauder, and W. Espeland. 2016. Engines of anxiety: academic rankings, reputation, and accountability. Russell Sage Foundation, New York, New York, USA.

Fazey, I., J. A. Fazey, J. Fischer, K. Sherren, J. Warren, R. F. Noss, and S. R. Dovers. 2007. Adaptive capacity and learning to learn as leverage for social–ecological resilience. Frontiers in Ecology and the Environment 5(7):375-380. https://doi.org/10.1890/1540-9295(2007)5[375:ACALTL]2.0.CO;2

Folke, C., R. Biggs, A. V. Norström, B. Reyers, and J. Rockström. 2016. Social-ecological resilience and biosphere-based sustainability science. Ecology and Society 21(3): 41. https://doi.org/10.5751/ES-08748-210341

Gelcich, S., T. P. Hughes, P. Olsson, C. Folke, O. Defeo, M. Fernandez, S. Foale, L. H. Gunderson, C. Rodriguez-Sickert, M. Scheffer, R. S. Steneck, and J. C. Castilla. 2010. Navigating transformations in governance of Chilean marine coastal resources. Proceedings of the National Academy of Sciences 107(39):16794-16799. https://doi.org/10.1073/pnas.1012021107

Gillette, M. B., M. Arias Schreiber, and N. Siegrist. 2022. Sweden’s good fishers: skills and values in a transforming social-ecological field. Kritisk Etnografi: Swedish Journal of Anthropology 5(1):37-59. https://doi.org/10.33063/diva-490449

Glennan, S. 2017. The new mechanical philosophy. Oxford University Press, Oxford, UK. https://doi.org/10.1093/oso/9780198779711.001.0001

Glennan, S., and P. M. Illari. 2017. The Routledge handbook of mechanisms and mechanical philosophy. Routledge, London, UK. https://doi.org/10.4324/9781315731544

Goldman, A. I. 2006. Simulating minds: the philosophy, psychology, and neuroscience of mindreading. Oxford University Press on Demand, New York, New York, USA. https://doi.org/10.1093/0195138929.001.0001

González del Solar, R., L. Marone, and J. Lopez de Casenave. 2019. Mechanismic approaches to explanation in ecology. Pages 555-573 in M. R. Matthews, editor. Mario Bunge: a centenary festschrift. Springer International Publishing, Cham, Switzerland. https://doi.org/10.1007/978-3-030-16673-1_31

Gonzalez-Mon, B., Ö. Bodin, E. Lindkvist, T. H. Frawley, A. Giron-Nava, X. Basurto, M. Nenadovic, and M. Schlüter. 2021. Spatial diversification as a mechanism to adapt to environmental changes in small-scale fisheries. Environmental Science and Policy 116:246-257. https://doi.org/10.1016/j.envsci.2020.11.006

Hedström, P. 2005. Dissecting the social: on the principles of analytical sociology. Cambridge University Press, Cambridge, UK. https://doi.org/10.1017/CBO9780511488801

Hedström, P., and R. Swedberg. 1998. Social mechanisms: an analytical approach to social theory. Cambridge University Press, Cambridge, UK. https://doi.org/10.1017/CBO9780511663901

Hedström, P., and P. Ylikoski. 2010. Causal mechanisms in the social sciences. Annual Review of Sociology 36:49-67. https://doi.org/10.1146/annurev.soc.012809.102632

Hernes, G. 1976. Structural change in social processes. American Journal of Sociology 82(3):513-547. https://doi.org/10.1086/226352

Heslin, A., N. D. Deckard, R. Oakes, and A. Montero-Colbert. 2019. Displacement and resettlement: understanding the role of climate change in contemporary migration. Pages 237-260 in R. Mechler, L. Bouwer, T. Schinko, S. Surminski, and J. Linnerooth-Bayer, editors. Loss and damage from climate change. climate risk management, policy and governance. Springer, Cham, Switzerland. https://doi.org/10.1007/978-3-319-72026-5

Hiddink, J. G., S. Jennings, M. Sciberras, C. L. Szostek, K. M. Hughes, N. Ellis, A. D. Rijnsdorp, R. A. McConnaughey, T. Mazor, R. Hilborn, J. S. Collie, C. R. Pitcher, R. O. Amoroso, A. M. Parma, P. Suuronen, and M. J. Kaiser. 2017. Global analysis of depletion and recovery of seabed biota after bottom trawling disturbance. Proceedings of the National Academy of Sciences 114(31):8301-8306. https://doi.org/10.1073/pnas.1618858114

Holland, D. S. 2008. Are fishermen rational? A fishing expedition. Marine Resource Economics 23(3):325-344. https://doi.org/10.1086/mre.23.3.42629621

Homer-Dixon, T., B. Walker, R. Biggs, A.-S. Crépin, C. Folke, E. F. Lambin, G. D. Peterson, J. Rockström, M. Scheffer, W. Steffen, and M. Troell. 2015. Synchronous failure: the emerging causal architecture of global crisis. Ecology and Society 20(3): 6. https://doi.org/10.5751/ES-07681-200306

Hutto, D. D. 2012. Folk psychological narratives: the sociocultural basis of understanding reasons. MIT Press, Cambridge, Massachusetts, USA. https://doi.org/10.7551/mitpress/7525.001.0001

Jones, J. B. 1992. Environmental impact of trawling on the seabed: a review. New Zealand Journal of Marine and Freshwater Research 26(1):59-67. https://doi.org/10.1080/00288330.1992.9516500

Kaaronen, R. O. 2017. Affording sustainability: adopting a theory of affordances as a guiding heuristic for environmental policy. Frontiers in Psychology 8: 1974. https://doi.org/10.3389/fpsyg.2017.01974

Kimmich, C., Baldwin, E., Kellner, E., Oberlack, C. and Villamayor-Tomas, S., 2023. Networks of action situations: a systematic review of empirical research. Sustainability Science 18(1):11-26. https://doi.org/10.1007/s11625-022-01121-2

Lade, S. J., S. Niiranen, J. Hentati-Sundberg, T. Blenckner, W. J. Boonstra, K. Orach, M. F. Quaas, H. Österblom, and M. Schlüter. 2015. An empirical model of the Baltic Sea reveals the importance of social dynamics for ecological regime shifts. Proceedings of the National Academy of Sciences 112(35):11120-11125. https://doi.org/10.1073/pnas.1504954112

Lambin, E. F., and P. Meyfroidt. 2010. Land use transitions: socio-ecological feedback versus socio-economic change. Land Use Policy 27(2):108-118. https://doi.org/10.1016/j.landusepol.2009.09.003

Lansing, J. S., S. Thurner, N. N. Chung, A. Coudurier-Curveur, Ç. Karakaş, K. A. Fesenmyer, and L. Y. Chew. 2017. Adaptive self-organization of Bali’s ancient rice terraces. Proceedings of the National Academy of Sciences 114(25):6504-6509. https://doi.org/10.1073/pnas.1605369114

Machamer, P., L. Darden, and C. F. Craver. 2000. Thinking about mechanisms. Philosophy of Science 67(1):1-25. https://doi.org/10.1086/392759

Martin, R., and M. Schlüter. 2015. Combining system dynamics and agent-based modeling to analyze social-ecological interactions—an example from modeling restoration of a shallow lake. Frontiers in Environmental Science 3: 66. https://doi.org/10.3389/fenvs.2015.00066

Masterson, V. A., J. P. Enqvist, R. C. Stedman, and M. Tengö. 2019. Sense of place in social–ecological systems: from theory to empirics. Sustainability Science 14(3):555-564. https://doi.org/10.1007/s11625-019-00695-8

McGinnis, M. D., and E. Ostrom. 2014. Social-ecological system framework: initial changes and continuing challenges. Ecology and Society 19(2): 30. https://doi.org/10.5751/ES-06387-190230

Meyfroidt, P. 2016. Approaches and terminology for causal analysis in land systems science. Journal of Land Use Science 11(5):501-522. https://doi.org/10.1080/1747423X.2015.1117530

Meyfroidt, P., R. Roy Chowdhury, A. de Bremond, E. C. Ellis, K.-H. Erb, T. Filatova, R. D. Garrett, J. M. Grove, A. Heinimann, T. Kuemmerle, C. A. Kull, E. F. Lambin, Y. Landon, Y. le Polain de Waroux, P. Messerli, D. Müller, J. Ø. Nielsen, G. D. Peterson, V. Rodriguez García, M. Schlüter, B. L. Turner, and P. H. Verburg. 2018. Middle-range theories of land system change. Global Environmental Change 53:52-67. https://doi.org/10.1016/j.gloenvcha.2018.08.006

Milfont, T. L., L. Evans, C. G. Sibley, J. Ries, and A. Cunningham. 2014. Proximity to coast is linked to climate change belief. PLoS One 9(7): e103180. https://doi.org/10.1371/journal.pone.0103180

Ng’ang’a, S. K., E. H. Bulte, K. E. Giller, J. M. McIntire, and M. C. Rufino. 2016a. Migration and self-protection against climate change: a case study of Samburu County, Kenya. World Development 84:55-68. https://doi.org/10.1016/j.worlddev.2016.04.002

Ng’ang’a, S. K., E. H. Bulte, K. E. Giller, N. N. Ndiwa, S. C. Kifugo, J. M. McIntire, M. Herrero, and M. C. Rufino. 2016b. Livestock wealth and social capital as insurance against climate risk: a case study of Samburu County in Kenya. Agricultural Systems 146:44-54. https://doi.org/10.1016/j.agsy.2016.04.004

Olsson, P., L. H. Gunderson, S. R. Carpenter, P. Ryan, L. Lebel, C. Folke, and C. S. Holling. 2006. Shooting the rapids: navigating transitions to adaptive governance of social-ecological systems. Ecology and Society 11(1): 18. https://doi.org/10.5751/ES-01595-110118

Orach, K., A. Duit, and M. Schlüter. 2020. Sustainable natural resource governance under interest group competition in policy-making. Nature Human Behaviour 4(9):898-909. https://doi.org/10.1038/s41562-020-0885-y

Ostrom, E. 2007. A diagnostic approach for going beyond panaceas. Proceedings of the National Academy of Sciences of the United States of America 104:15181-15187. http://dx. doi.org/10.1073/pnas.0702288104

Ostrom, E. 2009. A general framework for analyzing sustainability of social-ecological systems. Science 325:419-422. http://dx.doi.org/10.1126/science.1172133

Pahl-Wostl, C., G. Holtz, B. Kastens, and C. Knieper. 2010. Analyzing complex water governance regimes: the management and transition framework. Environmental Science and Policy 13:571-581. http://dx.doi.org/10.1016/j. envsci.2010.08.006

Pâslaru, V. 2017. Mechanisms in ecology 1. Pages 348-361 in S. Glennan and P. Illari, editors. The Routledge handbook of mechanisms and mechanical philosophy. Routledge, New York, New York, USA. https://doi.org/10.4324/9781315731544-26

Pettit, P. 1996. The common mind: an essay on psychology, society, and politics. Oxford University Press, Oxford, UK. https://doi.org/10.1093/0195106458.001.0001

Pitt, A. N., C. A. Schultz, and J. J. Vaske. 2019. Engaging youth in public lands monitoring: opportunities for enhancing ecological literacy and environmental stewardship. Environmental Education Research 25(9):1386-1399. https://doi.org/10.1080/13504622.2019.1649368

Prell, C., M. Reed, L. Racin, and K. Hubacek. 2010. Competing structure, competing views: the role of formal and informal social structures in shaping stakeholder perceptions. Ecology and Society 15(4): 34. https://doi.org/10.5751/ES-03652-150434

Raub, W., and T. Voss. 2017. Micro-macro models in sociology: antecedents of Coleman’s diagram. Pages 11-36 in B. Jann and W. Przepiorka, editors. Social dilemmas, institutions, and the evolution of cooperation. De Gruyter Oldenbourg, Berlin, Germany and Boston, Massachusetts, USA. https://doi.org/10.1515/9783110472974-002

Raymond, C. M., M. Giusti, and S. Barthel. 2018. An embodied perspective on the co-production of cultural ecosystem services: toward embodied ecosystems. Journal of Environmental Planning and Management 61(5-6):778-799. https://doi.org/10.1080/09640568.2017.1312300

Romero, P., and O. Melo. 2021. Can a territorial use right for fisheries management make a difference for fishing communities? Marine Policy 124: 104359. https://doi.org/10.1016/j.marpol.2020.104359

Rudel, T. K., O. T. Coomes, E. Moran, F. Achard, A. Angelsen, J. Xu, and E. Lambin. 2005. Forest transitions: towards a global understanding of land use change. Global Environmental Change 15(1):23-31. https://doi.org/10.1016/j.gloenvcha.2004.11.001

Rullens, V., A. M. Lohrer, M. Townsend, and C. A. Pilditch. 2019. Ecological mechanisms underpinning ecosystem service bundles in marine environments—a case study for shellfish. Frontiers in Marine Science 6: 409. https://doi.org/10.3389/fmars.2019.00409

Sandifer, P. A., A. E. Sutton-Grier, and B. P. Ward. 2015. Exploring connections among nature, biodiversity, ecosystem services, and human health and well-being: opportunities to enhance health and biodiversity conservation. Ecosystem Services 12:1-15. https://doi.org/10.1016/j.ecoser.2014.12.007

Schelling, T. C. 1978. Micromotives and macrobehavior. Norton, New York, New York, USA.

Schill, C., T. Lindahl, and A.-S. Crépin. 2015. Collective action and the risk of ecosystem regime shifts: insights from a laboratory experiment. Ecology and Society 20(1): 48. https://doi.org/10.5751/ES-07318-200148

Schlüter, M., A. Baeza, G. Dressler, K. Frank, J. Groeneveld, W. Jager, M. A. Janssen, R. R. McAllister, B. Müller, and K. Orach. 2017. A framework for mapping and comparing behavioural theories in models of social-ecological systems. Ecological Economics 131:21-35. https://doi.org/10.1016/j.ecolecon.2016.08.008

Schlüter, M., G. Caniglia, K. Orach, Ö., Bodin, N. Magliocca, P. Meyfroidt, and B. Reyers. 2022. Why care about theories? Innovative ways of theorizing in sustainability science. Current Opinion in Environmental Sustainability 54: 101154. https://doi.org/10.1016/j.cosust.2022.101154

Schlüter, M., L. J. Haider, S. J. Lade, E. Lindkvist, R. Martin, K. Orach, N. Wijermans, and C. Folke. 2019a. Capturing emergent phenomena in social-ecological systems: an analytical framework. Ecology and Society 24(3): 11. https://doi.org/10.5751/ES-11012-240311

Schlüter, M., T. Hertz, and M. Mancilla García. 2020. Social-ecological intertwinedness: an attempt at a clarification. SSRN Electronic Journal 2020: 3727968. https://doi.org/10.2139/ssrn.3727968

Schlüter, M., K. Orach, E. Lindkvist, R. Martin, N. Wijermans, Ö. Bodin, and W. J. Boonstra. 2019b. Toward a methodology for explaining and theorizing about social-ecological phenomena. Current Opinion in Environmental Sustainability 39:44-53. https://doi.org/10.1016/j.cosust.2019.06.011

Scholz, R. W., C. R. Binder, and D. J. Lang. 2011. The HES framework. Pages 453-462 in R. W. Scholz. Environmental literacy in science and society: from knowledge to decisions. Cambridge University Press, Cambridge, UK. https://doi.org/10.1017/CBO9780511921520.020

Schurman, R. A. 1996. Snails, southern hake and sustainability: neoliberalism and natural resource exports in Chile. World Development 24(11):1695-1709. https://doi.org/10.1016/0305-750X(96)00069-1

Scoones, I. 1998. Sustainable rural livelihoods: a framework for analysis. IDS Working Paper 72. Institute of Development Studies, University of Sussex, Brighton, UK.

Swedberg, R. 2016. Can you visualize theory? On the use of visual thinking in theory pictures, theorizing diagrams, and visual sketches. Sociological Theory 34(3):250-275. https://doi.org/10.1177/0735275116664380

Tengö, M., and J. V. Heland. 2011. Adaptive capacity of local indigenous institutions: the case of the taboo forests of southern Madagascar. Pages 37-74 in E. Boyd and C. Folke, editors. Adapting institutions. Cambridge University Press, Cambridge, UK. https://doi.org/10.1017/CBO9781139017237.006

Traynor, J. 2017. A History of Almond Pollination in California. Bee World 94(3):69–79. https://doi.org/10.1080/0005772X.2017.1353273

Turner, B. L., R. E. Kasperson, P. Matson, J. J. McCarthy, R. W. Corell, L. Christensen, N. Eckley, J. X. Kasperson, A. Luers, M. L. Martello, C. Polsky, A. Pulsipher, and A. Schiller. 2003. A framework for vulnerability analysis in sustainability science. Proceedings of the National Academy of Sciences 100(14):8074-8079. https://doi.org/10.1073/pnas.1231335100

Turner, C. 2010. Investigating sociological theory. SAGE, London, UK. https://doi.org/10.4135/9781446251898

van Aelst, K., and N. Holvoet. 2016. Intersections of gender and marital status in accessing climate change adaptation: evidence from rural Tanzania. World Development 79:40-50. https://doi.org/10.1016/j.worlddev.2015.11.003

van Putten, I. E., S. Kulmala, O. Thébaud, N. Dowling, K. G. Hamon, T. Hutton, and S. Pascoe. 2012. Theories and behavioural drivers underlying fleet dynamics models: theories and behavioural drivers. Fish and Fisheries 13(2):216-235. https://doi.org/10.1111/j.1467-2979.2011.00430.x

Ward, P., and S. Hindmarsh. 2007. An overview of historical changes in the fishing gear and practices of pelagic longliners, with particular reference to Japan’s Pacific fleet. Reviews in Fish Biology and Fisheries 17(4):501-516. https://doi.org/10.1007/s11160-007-9051-0

Weber, M. 1905. Die protestantische Ethik und der Geist des Kapitalismus. J. C. B. Mohr, Tübingen, Germany.

Westley, F. R., O. Tjörnbo, L. Schultz, P. Olsson, C. Folke, B. Crona, and Ö. Bodin. 2013. A theory of transformative agency in linked social-ecological systems. Ecology and Society 18(3): 27. https://doi.org/10.5751/ES-05072-180327

Wijermans, N., W. J. Boonstra, K. Orach, J. Hentati-Sundberg, and M. Schlüter. 2020. Behavioural diversity in fishing—towards a next generation of fishery models. Fish and Fisheries 21(5):872-890. http://doi.org/10.1111/faf.12466

Woodward, J. 2003. Making things happen. Oxford University Press, Oxford, UK. https://doi.org/10.1093/0195155270.001.0001

Ylikoski, P. 2007. The idea of contrastive explanandum. Pages 27-42 in J. Renn and L. Patton, editors. Rethinking explanation. Springer, Dordrecht, The Netherlands. https://doi.org/10.1007/978-1-4020-5581-2_3

Ylikoski, P. 2012. Micro, macro, and mechanisms. Pages 21-45 in H. Kincaid, editor. The Oxford handbook of philosophy of the social sciences. Oxford University Press, Oxford, UK. https://doi.org/10.1093/oxfordhb/9780195392753.013.0002

Ylikoski, P. 2019. Mechanism-based theorizing and generalization from case studies. Studies in History and Philosophy of Science Part A 78:14–22. https://doi.org/10.1016/j.shpsa.2018.11.009

Ylikoski, P. 2021. Understanding the Coleman boat. Pages 49-63 in G. Manzo, editor. Research handbook on analytical sociology. Edward Elgar Publishing, Northampton, Massachusetts, USA. https://doi.org/10.4337/9781789906851.00008

Zamora-Maldonado, H. C., V. S. Avila-Foucat, V. G. Sánchez-Sotomayor, and R. Lee. 2021. Social-ecological resilience modeling: water stress effects in the bighorn sheep management system in Baja California Sur, Mexico. Ecological Complexity 45: 100884. https://doi.org/10.1016/j.ecocom.2020.100884