The search for a unifying framework to explain resilience is a worthy goal, but Kalisch et al. have chosen an overly reductionistic explanation for positive adaptation under adversity. Research on resilience is showing that the protective processes that account for people's survival are not just individual, but are as likely to be social and ecological aspects of the individual's life (Masten Reference Masten2011; Theron et al. Reference Theron, Liebenberg and Ungar2015; Ungar et al. Reference Ungar, Liebenberg, Armstrong, Dudding and van de Vijver2012). By its very nature, resilience implies the presence of a stressor, and the nature of that stressor will influence how each protective process affects outcomes related to well-being. Whereas a positive appraisal style may be adaptive in some contexts where there is lower exposure to risk, it may lead to worse mental health outcomes, or be ineffective, in contexts where it produces faulty cognitions about the threats posed and the solutions required to resolve life's challenges (Seligman Reference Seligman2011). In this brief discussion, I will define resilience in social ecological terms, discuss the differential impact social ecologies have on resilience processes, and discuss the dangers that a focus on positive appraisal style may have for clinical processes and social policies that unintentionally place responsibility for change solely on the individual.

Whether we look at the work of the epigeneticist (Jaffee et al. Reference Jaffee, Caspi, Moffitt, Polo-Tomas and Taylor2007), the neuropsychiatrist (Perry Reference Perry2009), the medical anthropologist (Panter-Brick & Eggerman Reference Panter-Brick, Eggerman and Ungar2012), or the clinical social worker (Sanders et al. Reference Sanders, Munford, Liebenberg and Ungar2014), there is general agreement: When individuals are exposed to significant amounts of adversity, the amount of variance in positive developmental outcomes that is accounted for by the person's environment will be equal to or greater than that accounted for by individual variables (Abramson et al. Reference Abramson, Park, Stehling-Ariza and Redlener2010; Cicchetti Reference Cicchetti, Miklowitz and Cicchetti2010; Kassis et al. Reference Kassis, Artz and Moldenhauer2013; Tol et al. Reference Tol, Barbui, Galappatti, Silove, Betancourt, Souza, Golaz and van Ommeren2011). For example, Romanian orphans did much better developmentally when they were adopted by well-resourced families in Britain (Beckett et al. Reference Beckett, Maughan, Rutter, Castle, Colvert, Groothues, Kreppner, Stevens, O'connor and Sonuga-Barke2006); young children who are burdened with risk factors that predict delinquency are more likely to avoid early school dropout and criminality when their environments (homes and schools) are loaded with resources to socialize them well (Brame et al. Reference Brame, Nagin and Tremblay2001).

In both cases – neglected orphans and potential delinquents – the locus of control for changes to life trajectory is external to the child, though a positive disposition and willingness to take advantage of these resources can help individuals develop in positive ways. For this reason, resilience is increasingly being thought of as the capacity of systems to adapt, rather than the capacity of individuals to overcome challenges. A social ecological model of resilience suggests that human systems help individuals navigate to resources, and negotiate for resources to be provided in meaningful ways (Ungar Reference Ungar2011).

Even if we make the assumption that it is possible to reduce resilience to a cognitive process, the environment/individual interaction (with the environment likely accounting for more of the variance in outcomes than individual factors such as cognitive style, intelligence, or personality) will influence the degree of successful coping that is achieved. For example, in a nonresponsive environment (e.g., chronic parental neglect) or a consistently oppressive environment (e.g., persistent racism), a positive appraisal may actually have the unintended consequence of undermining the experiencing of one's personal efficacy (Hopkins et al. Reference Hopkins, Taylor, D'Antoine, Zubrick and Ungar2012; Obradović et al. Reference Obradović, Bush, Stamperdahl, Adler and Boyce2010). After all, children who are racially marginalized but who make an unrealistically positive appraisal of their capacity to resist oppression may actually be ill-prepared to cope with the challenges they face. Instead, an accurate appraisal of the world as grim and hopeless may cue children to emotionally withdraw in order to protect themselves from exposure to threats to their mental health (APA Task Force 2008).

In one sense, Kalisch et al. are correct. We require a transdiagnostic approach to resilience that identifies the processes (plural, not singular) that are associated with resilience. Unfortunately, a model that reduces resilience to purely internal cognitive processes overlooks the need for a differential “diagnosis” of resilience (Ungar Reference Ungar2015) that accounts for the complex patterns of environmental influence along with cognitive processes. In this sense, an experiment that models resilience with a rodent should never forget that the subject would do quite fine if the handler stopped imposing stressors in the first place. Rather than reducing our understanding of resilience, our models should be more systemic and ecological. Different factors matter more or less in different contexts.

A tentative model first published by Ungar (Reference Ungar2011) accounted for these complex aspects of individuals' interactions with their social ecologies. To summarize:

In the context of above normal exposure to adversity (∑A>average A for a population):

$${R_{1\comma 2\comma 3 \ldots }} \;= {f\lpar P_{\rm SC}\comma {E} \rpar \; \over \lpar {O_{Av}}\comma \ {O_{Ac}} \rpar \lpar {M} \rpar }$$

In the equation, resilience processes over time (R _{1, 2, 3…}) will vary by the interaction between a person (P) and his or her environment (E), while keeping in mind the person's strengths (S) and challenges (C). Processes are mediated by the opportunities (O) that are available (AV) and accessible (AC) for adaptive coping. They also are mediated by the socially constructed meaning systems (M) that shape appraisals of the risks and resources that individuals experience (e.g., whether they experience these resources as useful). This social ecological interpretation of resilience includes positive appraisal style as part of the denominator (M), but keeps it fully contextualized.

Although Kalisch et al. certainly expand the theory of appraisal style and generously note the complexity of appraisals that are related to resilience, their model's locus of control remains internal brain functions, when we know that changes to the environment can trigger brain functions that, quite literally, shape cognitions and resilience processes. This is more than a question of which comes first, trigger or cognition. Arguably, where there is significant risk exposure, it is the environment that accounts for the majority of the change in individuals' appraisals. Positive reappraisal is preferred, but unrealistic unless one is privileged enough to have opportunities (such as family supports) to justify optimism.