Is resilience a normative concept?

Abstract

In this paper, we engage with the question of the normative content of the resilience concept. The issues are approached in two consecutive steps. First, we proceed from a narrow construal of the resilience concept – as the ability of a system to absorb a disturbance – and show that under an analysis of normative concepts as evaluative concepts resilience comes out as descriptive. In the second part of the paper, we argue that (1) for systems of interest (primarily social systems or system with a social component) we seem to have options with respect to how they are described and (2) that this matters for what is to be taken as a sign of resilience as opposed to a sign of the lack of resilience for such systems. We discuss the implications of this for how the concept should be applied in practice and suggest that users of the resilience concept face a choice between versions of the concept that are either ontologically or normatively charged.

Keywords:

• Resilience
• normativity
• concepts

Notes

1. Lance Gunderson pointed out, in a recent lecture at Lund University, that the concept is descriptive and took that to imply that it is not a good thing in itself. Otherwise this stance is to a considerable extent implicit in the approach, rather than explicitly defended. The idea seems to be that if resilience is descriptive then resilience theorists are merely engaged in the objective reporting on the facts. So normative conclusions from that process say, to build more resilience, are not believed to be tinted by any particular perspective.

2. The underlying intuition here appears to be rather widely shared, c.f. Carpenter et al. (2001).

3. The basic conceptual distinction is common, although the terms used differ. Schrader-Frechette and McCoy, for instance, prefer the terms ‘dynamic balance’ and ‘persistence’ (Shrader-Frechette & McCoy, 1993), Orians (1975) uses the term ‘inertia’ as a label for what Holling calls ‘resilience’ and Grimm and Wissel (1997) opts for ‘domain of attraction.’ This is mostly a matter of terminological preference and the underlying concepts are more or less the same. More important here is the theoretical shift in ecology (and also economics) – spearheaded by ecologists like Holling – from an emphasis on stability and the dynamics of ecosystems close to equilibrium to a focus on dynamics far from equilibrium. An excellent overview of this shift can be found in Zebrowski (2013).

4. This has proved to be difficult with the resilience concept. Stability can easily be operationalised as the speed of return to the reference state, or the distance (with respect to some variable) between the state that is returned to and the reference state (see e.g. Newton, 2016).

5. Resilience theorists do explicitly talk of identities; Walker and Salt write that resilience is ‘the capacity of a system to absorb disturbance and reorganize so as to retain essentially the same function, structure, and feedbacks – to have the same identity’ (Walker & Salt, 2012, p. 3).

6. If one is mainly concerned with highly idealised models where disturbances are stipulated to work on only very few variables, all disturbances are, for trivial reasons, qualitatively similar from within the system. For a real system, like a coastal community, however, there is often reason to differentiate between different kinds of disturbances.

7. Note that under alternative descriptions of the system, the tennis ball can be considered stable; for example, if I is not thought of as structural integrity, but its shape. Moreover, what makes the tennis ball resilient under one assignment of I is what makes it stable under another.

8. Notably we do not mean to suggest here that resilience can only be deployed meaningfully to systems that are fundamentally static. Indeed, in most of the literature on resilience the emphasis is on flexibility and change (see e.g. Folke et al., 2010). On a purely conceptual level, it makes no sense to talk of resilience if there is no continuity at all (c.f. Cumming & Collier, 2005). What constitutes continuity for a given system, and how that is best determined given some set of (e.g. epistemological) constraints, is a different question.

9. Again, a good overview can be found in Brand and Jax (2007).

10. We are here taking a stance on these concepts that may appear to be strict to the point of misrepresenting them. The concept of resilience is commonly part of a larger system of ideas and within that larger system of ideas, flexibility in the long run in a very natural or obvious way comes out as something good or desirable. It is by being flexible that complex adaptive systems survive, so from the perspective of the systems themselves (or from within them), being flexible (or having adaptive capacity etc.) is a good thing. This is certainly the case and, without getting too far ahead of ourselves, this is a central part of the point of this paper. The commitment to this particular ontology is not always how the applicability of the concept of resilience is judged, and sometimes this commitment is explicitly downplayed. This is a source of confusion.

11. See also Folke et al. (2010).

12. Garmezy considers resilience to be ‘Functional adequacy … (the maintenance of competent functioning despite interfering emotionality) … as the benchmark of resilient behaviour under stress’ (as quoted in (Olsson, Bond, Burns, Vella-Brodrick, & Sawyer, 2003, p. 3). See also (Rutter, 1993) for a discussion of different definitions of the resilience concept in psychology.

13. Developing this concept is beyond the scope of this particular paper and what comes out as admissible or not to some extent depends on what fundamental philosophy of science one is inclined to subscribe to. The social constructivist and the naive realist may have rather different intuitions regarding how most systems are. That does not, however, change the fundamental tenets of this paper but merely offers a different way of discussing the matters so that the reader can feel free at this point to attach his or her favourite epistemology.

14. Another possibility here is that there are several admissible descriptions but they are all equivalent with respect to how resilience is construed. For simplicity, we file that category under systems monism.

15. System individuation matters here, although we shall not pursue that particular issue further. In brief, it is perhaps possible to argue that it is not a matter of several descriptions of one system, but rather several descriptions of several systems that just happen to overlap in certain respects.

16. The idea here would be that social-ecological systems are complex adaptive systems, and complex adaptive systems are a natural kind (for an introduction see Bird & Tobin, 2017): a set of entities that share some essential properties. In other words, expanding the resilience framework to apply not only to ecosystems but also to social systems and social-ecological systems involves making a substantive claim about the nature of these systems, and the features they share. Whether or not resilience theorists indeed are of this view is a little uncertain; but it is in no way an implausible reading. Although resilience theorists sometimes emphasise the provisional nature of their framework – it is merely to be viewed as a heuristic device, or a metaphor (see e.g. Carpenter et al., 2001, p. 766; Holling & Gunderson, 2002, p. 49) – the claim that social or social-ecological system just are complex adaptive systems, or self-organising systems, is virtually endemic (see e.g. Walker & Salt, 2012).

17. There is a general epistemological argument to made for pluralism. Representations, such as scientific theories, concepts, and models, are inherently partial (see Kellert, Longino, & Waters, 2006).

18. With pluralism follows epistemological modesty; the good pluralist acknowledges the limits under which he or she operates. It is notable that pluralism with respect to the models deployed by resilience theorists has been advocated before, see e.g. Cumming & Collier, 2005.