This book is about causal thinking and use of causal idiom in general, with the aim to provide the basic understanding required to explore the diversity of causal reasoning about social-ecological systems (SES) in particular.Footnote 1 Three questions stand in focus: (i) What are the meanings of different causal expressions, (ii) what is sufficient evidencefor inferences from observations to causal relations, and (iii) how to handle the diversity of causal relations in SES?

As a starter, let’s have a brief look at three excerpts from SES research. The first is from (Hruska, 2017):

A social–ecological system (SES) is a combination of social and ecological actors and processes that influence each other in profound ways. The SES framework is not a research methodology or a checklist to identify problems. It is a conceptual framework designed to keep both the social and ecological components of a system in focus so that the interactions between them can be scrutinised for drivers of change and causes of specific outcomes. Resilience, adaptability, and transformability have been identified as the three related attributes of SESs that determine their future trajectories. Identifying feedbacks between social and ecological components of the system at multiple scales is a key to SES-based analysis.

...[T]he SES framework facilitates identification of cross-system feedbacks to explain otherwise puzzling outcomes. While information intensive and logistically challenging in the management context, the SES framework can help overcome intractable challenges to working rangelands such as rangelandconversion and climate change. The primary benefit of the SES framework is the improved ability to prevent or correct social policies that cause negative ecological outcomes, and to achieve ecological objectives in ways that support, rather than hurt, rangeland users. (op. cit. p. 263)

Several expressions indicating causal relations are here used: ‘interactions’, ‘drivers of change’, ‘causes’, ‘feedbacks’, ‘prevent’, ‘correct’ and ‘achieve objectives’. The authors do not give any precise definitions of these expressions, apparently thinking that they are sufficiently common and well understood in ordinary language use or within the SES community.

Later in the chapter we find a diagram showing the components of rangeland social-ecological systems, but it is no diagram in the ordinary sense of this word; It is no more than a display of a number of concepts ordered in five or six groups. It does not suggest anything about causal relations.

The caption says a lot more, but almost nothing of what is said in that text is displayed in any way in the figure (Fig. 1.1).

Fig. 1.1
A block diagram. The regional global ecological system with its local ecosystem and regional global social system with its local system interact with natural resource management. Livestock grazing involves spatial and temporal patterns. The demanded ecosystem services are provided to the social system.

Generalized diagram of a rangeland social–ecological system. Humans and the environment interact in countless ways outside of natural resource management, but the interactions are most directly planned, manipulated, and monitored in natural resource management activities. Local, regional, and global social processes can all shape natural resource use and management activities. While resource policy may be set at large geographical scales (e.g., national), management activities occur within a single ecosystem. Livestock grazing differs from other types of natural resource use in that it is indirect; rather than directly manipulating a rangeland ecosystem, livestock operators devote their primary attention to managing livestock, and the livestock interact directly with the rangeland (adapted from Hruska, 2017, 266)

Full size image

The second excerpt is a case study of green turtle fisheries in Nicaragua. Here is the summary:

2.3 Robustness Summary. The Nicaraguan green turtle fishery does not represent a robust system of CPR governance. Persistent poverty, lack of alternative employment opportunities, and a high population growth rate (initially within communities, but more recently through in-migration) continue to be the main drivers for the commercialization of this fishery under a domestic subsistence use exception to endangered species protection of green turtles. Instead of protecting the species from exploitation, the ratification of the Convention on International Trade in Endangered Species (CITES) in 1977 by the national government, and subsequent closure of the legal international market for green turtle meat, has merely led to a shift in focus by turtle fishers from responding to international market demands to creating and satisfying a domestic demand for green turtles. Although de jure rules limiting the harvesting of green turtles exist at all government levels, including at the territorial, municipal, and community level in the RAAS and the RAAN, there is no overarching coordination of those rules, and no monitoring or enforcement, but for the collection of harvesting data that is being conducted by a researcher formerly involved with an international NGO. In essence, the fishery is de facto operated year-round without any restrictions. Prior limiting factors, such as the special skills required to navigate sailing dories and harpoon turtles, have been eliminated through the increased use of motor boats and turtle nets. The literature mentions three factors that provide evidenceof the long-term unsustainability of the fishery: (1) actual capture rates are believed to be significantly higher than reported; (2) a majority of the animals captured are large, sexually immature juveniles and adult turtles from the Tortuguero natal nesting site, which effectively removes the base for a future breeding population; and (3) recent declines in capture rates in regions with previous turtle abundance (Lagueux et al. 2014). Given many Miskitos reliance on green turtles as a sole source of cash revenues, a turtle population collapse could have significant social-ecological consequences (Brady et al., 2015).

The core causal claim is that there are ‘three factors that provide evidence of the long-term unsustainability of the fishery’. This formulation tells us that plausibly there are three causes of unsustainability, although the word ‘cause’ is not used. It is followed by a short account of the causal mechanisms.

The third excerpt is the abstract of a paper about the governance of coastal fisheries in Chile:

Here we explore social, political, and ecological aspects of a transformation in governance of Chile’s coastal marine resources, from 1980 to today. Critical elements in the initial preparatory phase of the transformation were (i) recognition of the depletion of resource stocks, (ii) scientific knowledge on the ecology and resilience of targeted species and their role in ecosystem dynamics, and (iii) demonstration-scale experimental trials, building on smaller-scale scientific experiments, which identified new management pathways. The trials improved cooperation among scientists and fishers, integrating knowledge and establishing trust. Political turbulence and resource stock collapse provided a window of opportunity that triggered the transformation, supported by new enabling legislation. Essential elements to navigate this transformation were the ability to network knowledge from the local level to influence the decision-making processes at the national level, and a preexisting social network of fishers that provided political leverage through a national confederation of artisanal fishing collectives. The resultant governance scheme includes a revolutionary national system of marine tenure that allocates user rights and responsibilities to fisher collectives. Although fine tuning is necessary to build resilience of this new regime, this transformation has improved the sustainability of the interconnected social–ecological system. Our analysis of how this transformation unfolded provides insights into how the Chilean system could be further developed and identifies generalised pathways for improved governance of marine resources around the world (Gelcich et al., 2010).

In all three examples it is clear that a main goal of the research is to arrive at knowledge about system’s dynamics and what to do to change things, in the first case to improve management of rangelands, in the second case to arrive at a more sustainable turtle fishery and in the third case to explain the change in governance of coastal fisheries. In order to know what to do one needs causal knowledge, reliable knowledge of the form, ‘If we do X, then Y will probably occur’.

Causal knowledge is thus often wanted because one wants to understand why and how SES change and to obtain guidance for future actions. However, in neither case is the reported research particularly illuminating about the criteria for counting something as a cause of something else. And with one exception, there is no indication of what kind of evidence the authors require for inferences to causes of observed states of affairs.

This, we believe, is a rather common feature of much empirical science; causal notions are often used without much reflection and questions about evidencefor causal relations are often not discussed. The intense discussion among philosophers about the concept of causation has had little influence on empirical scientists. This we will try to remedy to some extent.

Broadly speaking there are three types of questions addressed by SES researchers:

  1. 1.

    What are the causes of a particular undesirable state of affairs?

  2. 2.

    What are the effects of a possible intervention in a social-ecological system?

  3. 3.

    What is the causal structure of a SES?

Answers to questions 1 and 2 are causal explanations, while answers to the third one are constitutive explanations, i.e., explanations that consist of descriptions of the parts making up the system. This distinction will be elaborated upon in Chap. 8.

The rest of this book is divided into three parts. The Part I, Chaps. 2, 3, and 4, is a general analysis and discussion of the use of causal idiom in ordinary language. It does not result in an explicit definition of the terms ‘cause’ and ‘effect’, but it gives their meaning in particular contexts. This is a necessary prerequisite for a discussion about causal inference in science, which is the topic of the Part II, Chaps. 5–8. The Part III, Chap. 9, is about the diversity of causal reasoning in SES research. It discusses how this diversity results from the diverse backgrounds, interests, epistemological stances and scientific norms of researchers and practitioners in this multi-disciplinary field and where and how during a research process it manifests. It also provides suggestions about how to navigate this often confusing diversity.

While the chapters do have, of course, a sequence, and although we suggest to follow this sequence while reading, there is not necessarily a strict linear flow from simple to complex, or basic to advance, as a textbook on, e.g., statistics would be organised. Rather, there are multiple aspects to consider, when trying to understand what ‘causation’ can mean, and what the challenges are when trying to understand causation. Therefore, to some degree, each chapter has its own main focus, which is relevant but also strongly connected to most of the topics of the other chapters. To make this very clear, we added, after each chapter’s abstract, a bullet point list summarising the main lessons to be learned in this chapter, plus a brief summary why and how the chapter’s topics are relevant regarding the overall purpose of this book, which is being better prepared to ask and answer questions about causation, both in general and for SES in particular.