A CRITICAL REVIEW OF RESILIENCE THEORY AND ITS RELEVANCE FOR SOCIAL WORK

Adrian D van Breda As resilience theory gains traction as a theoretical framework for research in social work, it is important to engage with it critically. This article provides a critical review of resilience theory, drawing on an array of key authors, dating back fifty years. The review addresses three aspects of resilience theory: its definition, the construction of adversity and outcomes, and the nature and scope of resilience processes. The relevance of resilience theory for social work in South Africa is evaluated according to three criteria: the research questions it generates, its contribution to indigenous knowledge and decolonisation, and its contribution to social development

in South Africa will be discussed, with particular attention to current topics of decolonisation, indigenous practice and social development.It is hoped that this will contribute to a common frame of reference for social workers engaging in resilienceinformed research in South Africa.

DEFINING RESILIENCE
Consensus on a definition of resilience has been hard in coming (Southwick, Bonanno, Masten, Panter-Brick & Yehuda, 2014), as different writers put a different spin on the construct.Definitions are exceptionally important to ensure that we talk or write in harmony with each other.The term resilience has, to at least some extent, become an empty word that can be filled with almost any meaning.Some refer to resilience as something intrinsic to the individual, while others refer to it in a more holistic sense.Some refer to resilience as the competencies or capacities of people, while others refer to it as positive functioning in the face of adversity.These multitude of meanings for the same term have led to severe criticisms about the validity of resilience theory (Fletcher & Sarkar, 2013;Kolar, 2011).
Resilience theory has its roots in the study of adversity and an interest in how adverse life experiences impact harmfully on people.Antonovsky (1979) has referred to this as a 'pathogenic' focus, meaning a focus on the origins of illness or (in the social work context) a breakdown in social functioning or wellbeing, which he argues has dominated the social and medical sciences.Key examples of researchers whose work in resilience began with work on vulnerability include Emmy Werner, who conducted longitudinal research over several decades on children born into adverse social conditions in Kauai, Hawaii (Werner & Smith, 1982), Michael Rutter, who studied the intergenerational transmission of poverty and disadvantage (Rutter & Madge, 1976), and Norman Garmezy, who studied the contribution of a range of genetic and environmental risks to the development of schizophrenia (Garmezy, 1971).
These early studies of risk demonstrated that vulnerability contributes to later negative outcomes.'Vulnerability' included a range of factors, including a family history of mental illness (as a proxy for genetic vulnerability), challenges in the prenatal or neonatal period (e.g.maternal malnutrition or smoking, or inadequate neonatal care), problems in the family environment (e.g.marital discord or poor parenting methods), and problems in the broader social environment (e.g.poverty, crime or war) (Garmezy, 1971).'Outcomes' were frequently focused on mental health, as many of the researchers were psychologists or psychiatrists, but Werner's outcomes were more inclusive, including physical, social and intellectual development (Werner, Simonian, Bierman & French, 1967).
These early researchers soon noticed, however, that the relationship between vulnerability and negative outcomes was not universal.While many people have negative outcomes in response to vulnerability, not all do.Some dip and recover, others show little or no deterioration in functioning, and still others appear to achieve higher levels of adaptation than they had before (Masten, 2011).Researchers were thus confronted by exceptional outcomes and needed to generate an empirical and theoretical account for this.Some early researchers used the terms 'invincible' (Dahlin, Cederblad, Antonovsky & Hagnell, 1990;Werner & Smith, 1982) or 'invulnerability' (Anthony, 1987) to refer to those with these exceptional outcomes.Garmezy (1971:114), for example, writes: These are the 'vulnerables' and the 'invulnerables' of a society … Were we to study the forces that move such children to survival and to adaptation, the longrange benefits to our society might be far more significant than our many efforts to construct models of primary prevention designed to curtail the incidence of vulnerability.
This body of research led to one of the frequent definitions of resilience, viz.resilience as an outcome.Outcome definitions include "A stable trajectory of healthy functioning after a highly adverse event" (Bonanno, as cited in Southwick et al., 2014:1) and "Individuals who adapt to extraordinary circumstances, achieving positive and unexpected outcomes in the face of adversity" (Fraser, Richman & Galinsky, 1999:136).Such definitions focus attention on the 'state of being resilient' in the face of adversity, thus on an outcome construction of resilience.
Having recognised these differences in outcomes in the face of adversity, researchers began asking the 'why' question to understand what distinguished those with better outcomes from those with poorer outcomes.This is what Antonovsky (1979) calls the 'salutogenic' question, viz. a question about the origins of health or psychosocial functioning.This salutogenic or resilience question can be phrased as (Van Breda, 2001:14), "Why, when people are exposed to the same stress which causes some to become ill, do some remain healthy?"By asking such a question, researchers are recognising that there are other processes that mediate (i.e. that fall between) adversity and negative outcomes.This way of thinking about resilience is based on a different definition of resilience, viz.resilience as a process.Process definitions include "The capacity to rebound from adversity strengthened and more resourceful" (Walsh, 2006:4), "The potential or manifested capacity of a dynamic system to adapt successfully to disturbances that threaten the function, survival, or development of the system" (Masten, 2015b:187) and "The process of adjusting well to significant adversity" (Theron, 2016:636).Here, resilience centres on the mediating factors or processes that enable positive outcomes in the wake of adversity.This blurring of definitions of resilience emerged as a direct result of the line of scientific enquiry.Van Breda (2017a:227) illustrates this nicely in relation to research by Kobasa (1979) on 'hardiness': In her first studies, [Kobasa] identified samples of executives who had similarly high levels of stressful life events (i.e.adversity) but had different levels of illhealth (i.e.resilience outcomes).She then administered the same tests to both groups and identified dimensions that effectively discriminated between them.These dimensions eventually comprised the construct 'hardiness' (i.e.resilience as a process or protective mechanism), which involves a commitment to invest in one's life situation, a perception of having control over one's circumstances and a belief that change or challenge, rather than stability, is the normal mode of life.Here we see resilience as process and resilience as outcome operating in an integrated fashion.
The process-outcome debate in resilience theory is valid, but creates an unnatural split between process and outcome.Resilience research involves three connected components: adversity, outcomes and mediating factors.It is not possible to think about or research resilience without considering all three components.Nevertheless, the problem with the outcome definition of resilience is that it merely declares the observation of positive outcomes in the face of adversity; it does not explain them.A declaration without an explanation has limited use and for this reason the process definition of resilience is to be preferred.Conceptually, then, resilience is a process that leads to an outcome, and the central focus of resilience research is on the mediating processes (see Figure 1).

FIGURE 1 RESILIENCE AS PROCESS AND OUTCOME
To help distinguish between process and outcome, I join Ungar (2004) in recommending that different terms be used for them, and suggest that 'resilience' is best used as a process definition, and that 'resilient' be reserved for an outcome definition.Thus, one could say that a person or social system is 'resilient' because it evidences good outcomes in the face of adversity.On the other hand, one could say that the 'resilience' of the person or social system is supportive relationships and a hope for the future.
A viable definition of resilience should, therefore, incorporate all three components and focus on the mediating processes.I propose the following definition: The multilevel processes that systems engage in to obtain better-than-expected outcomes in the face or wake of adversity.'Multilevel' means that the resilience processes occur across multiple domains or levels of the social ecology or person-in-environment, rather than only in the individual.'Systems' is used so that the definition can be scaled across different-sized systems, such as cells, individuals, families, organisations and communities, and also non-human systems such as metals, the climate or the economy.'Better-than expected outcomes' will be discussed in the following section.'In the face of' suggests that the adversity is on-going and that resilience processes are at work while the system is still facing the adversity, while 'in the wake of' suggests that the adversity has passed and the system is recovering from the impact of the adversity.

ADVERSITY AND OUTCOMES
Having considered what 'resilience' means, it is necessary to give attention to what is meant by adversity and outcomes.Patterns of adversity can be roughly divided into two categories: chronic and acute (Bonanno & Diminich, 2013).Chronic adversity extends over a considerable period of time and may have a pervasive impact on a person's life.I distinguish between two subcategories of chronic adversity: distal-and proximal-onset.Distal-onset chronic adversity has no clear starting point within the experience of the person (i.e.there is no 'before' the adversity) and may include poverty and family violence, which often extend from birth to adulthood.Proximal-onset chronic adversity, on the other hand, has a defined starting point in the experience of the person, but continues for a significant period of time and may impact on numerous aspects of life, and could include war and natural disasters.Acute adversity, by contrast, has a defined starting point, a relatively brief duration and a limited impact on the whole of life, within a generally well-functioning life context, and may include an accident or assault (Bonanno & Diminich, 2013).
Recognising these patterns is important, because they suggest different resilience pathways and because they may be differently prevalent in different communities (Bonanno & Diminich, 2013).Acute and proximal-onset chronic adversities permit one to think of resilience as a 'bouncing back' to a previous (pre-trauma) level of functioning, while distal-onset chronic adversity does not, since there is no 'before', and patterns of sustained coping or stress resistance may be more likely.With chronic adversity, resilience involves coping in the face of the adversity (while it is ongoing), while resilience to acute adversity involves recovering in the wake of adversity (after it has ended).Bonanno and Diminich (2013) coined the term 'emergent resilience' to describe the response to chronic adversity, which may show much longer periods of uneven outcomes and a gradual improvement over time, and 'minimal-impact resilience' as the acute adversity resilience pathway, characterised by a mild decline in functioning in response to the adversity and a rapid recovery.
Given the tremendous disparities in socioeconomic and development status, it may be argued that those living in the Global North are more likely to experience acute adversity and proximal-onset chronic adversity related to natural disasters, while those in the Global South are more likely to experience distal-onset chronic adversity due to poverty and rolling conflict, and proximal-onset chronic adversity due to war and natural disasters, and combinations of all three.In light of this, patterns of resilienceand thus how one defines outcomes, identifies resilience processes and constructs resilience pathwaysmay vary considerably between the Global North and South.
Further definitional issues emerge in the way outcomes are constructed, from which three main points emerge.First, we read earlier how Garmezy (1971:114) divided the world into "the 'vulnerables' and the 'invulnerables'".This approach of bifurcating the world into distinct groups is common in resilience research; Masten (2011:500) divides the world into "maladaptive and … adaptive groups".There is an intuitive appeal and often a usefulness in this division (though the terminology is unfortunate).A widely used instance in youth research is the dichotomous variable NEET (Not in Employment, Education or Training) (DHET, 2017), which enables a quick and impactful division of youths into NEET (which is considered a negative outcome, and which also increases vulnerability to further negative outcomes, such as crime) and not NEET.
However, in general, researchers should avoid such a simplistic (one might say unrealistic) division of the world.Probably no-one is either resilient or not resilient.It would be better to talk about a range of outcomes extending from more negative to more positive (Condly, 2006).Statistically, researchers will achieve more nuanced and powerful results if they measure outcomes on a continuous scale than dichotomously, because there is far greater sensitivity in measurement between a slightly more resilient outcome and a slightly less resilient outcome.
Second, researchers are advised to be specific about what outcomes they are talking about to avoid inferring the general from the specific.Masten (2011:499), for example, writes about adaptive and maladaptive children, as if describing the whole of their psychosocial functioning, but in fact she defines adaptation according to three discrete fractions of life, viz."academic achievement, friendship, and rule-abiding (vs.antisocial) behavior".A child who, for example, was doing poorly academically, a loner and acting out would be categorised as 'maladaptive', even though the child might have strong self-esteem, be excelling in creative activities and have a loving relationship with her or his parents.Instead, it would be more accurate and respectful of people's dignity to define the resilient outcomes in relation to the specific variables used, such as higher or lower academic achievement or friendships.
And third, the question can be asked: What constitutes a good outcome?Resilience research does require a differentiation in outcomes between 'good' and 'bad', or 'better' and 'worse'.The actual outcomes need to be relevant to the study at hand and the social context and developmental stage of the participants.The decision on which outcomes to select may be based on qualitative research with a sample of the participants themselves to determine what is most important to them, and how they define their own outcomes.The normative value of outcomes remains a challenge, however.What, for example, can be considered a 'good' outcome for a young person living in a poor, drug-ridden community, or an older woman raising several foster children in a remote and chronically under-resourced rural community?Some researchers are using the term "better-than-expected" rather than 'good' outcomes (Luthar, Crossman & Small, 2015:3;Rutter, 2012;Ungar, Ghazinour & Richter, 2013).This seems to provide better contextualisation for what 'good' means within the range of outcomes across the population under study.Thus, for example, the young person above who remains 'drug free' may be considered to have a better-than-expected outcome, and the woman who manages to continue loving her foster children, caring for herself and being hopeful for the future may also have a better-than-expected outcome in the face of adversity, even if these outcomes might not conform to societal notions of 'success'.

RESILIENCE PROCESSES
I have argued that the heart of resilience research is the mediating processes (also referred to as resilience processes or protective resources) which enable people to achieve better-than-expected outcomes in the face or wake of adversity.Much of the earlier research on resilience was focused on identifying individual factorstypically intrapsychic factorsthat distinguished those with better-than-expected outcomes from those with expected or poor outcomes (Van Breda, 2001).These include constructs such as hardiness (Kobasa, 1979), sense of coherence (Antonovsky, 1979), self-efficacy (Bandura, 1982) and, more recently, grit (Duckworth, Peterson, Matthews & Kelly, 2007).These classic constructs have been augmented with a wide array of other individualised variables, notably intelligence, problem-solving skills, emotional regulation, motivation to succeed, faith and hope (Masten, 2015a:148).
These kinds of individual resilience processes remain prominent in resilience studies, but they have come under much criticism.Most recently this has taken the form of a critique of resilience theory being in the service of a neoliberal agenda (Joseph, 2013), though this view in turn has its critics (Schmidt, 2015).Garrett (2016Garrett ( :1918) ) particularly critiques "the way in which societal and political context is minimized in that 'resilience' is usually individualised", and the way resilience "gels with the politics of neo-liberalisation".Neoliberalism is associated with the decentralising of responsibility for personal growth and development to the level of the individual.The individual is thus responsible for improving her or his life context, with little or no support from the state, permitting the dismantling of the welfare state.If the individual is responsible for her or his own well-being, the state is free to disregard adverse social systems and dynamics, such as poverty, racism, lack of access to resources and poor-quality education.Individuals are, in effect, made responsible for dealing with collective challenges that should be dealt with by collective structures, such as the state.
The emphasis of a large amount of resilience literature on how individuals transcend social risks does appear to support this alignment of resilience theory with neoliberalism.While Harrison (2013:99) is careful to point out that "it is not the [resilience] concept itself that is problematic, but the ways in which it may lend itself to use in academic and policy thinking", she also points out the toll of resilience (Masten, 2015a).Resilience to chronic, particularly distal-onset chronic adversity, draws for extended periods on a well of resources that is not limitless.Eventually, these wells run dry through resilience exercised in response to adversity after adversity.Women, in particular, are vulnerable to such exhaustion, due to the gendered structure of caretaking.While individuals and families in such contexts do demonstrate resilience, this comes at a cost to those individuals and families.Resilience research or theory which valorises such resilience, without considering the social structures that create or perpetuate adversity, has colluded with a neoliberal agenda.
However, there have been a number of developments within resilience theory over the past decade or so that provide the building blocks for addressing these concerns.To at least some extent, the critiques above are based on a narrow reading of resilience theory.Contemporary resilience theory increasingly gives attention to understanding resilience within broader social systems, and to engaging with issues of power and social justice.
While considerable attention has been directed to individualised resilience processes, even the earliest resilience studies emphasised the centrality of relationships.Large longitudinal studies have been consistent in showing the importance of an early caregiving relationship for developmental outcomes through childhood and into adulthood (Werner, 2013:94): those who by midlife showed better-than-expected outcomes "relied on sources of support within the family and community that increased their competence and efficacy, decreased the number of stressful life events they subsequently encountered, and opened up new opportunities for them."Not only early childhood relationships, but also relationships in people's current life, including adult life, are important mediating factors in the face or wake of adversity (Van den Berg, George, Du Plessis, Botha, Basson, De Villiers & Makola, 2013;Ebersöhn, 2012).This evidence, which is abundant in resilience research, locates resilience processes not so much within individuals, but within networks of social relationships with family, friends, school, colleagues and neighbourhoods (Hartling, 2008).Even some individual resilience processes, such as sustained self-esteem, have been shown to be more relational than intrapsychic; here one thinks of Cooley's looking-glass self as the relational source of self and self-worth (Stephan & Maiano, 2007).Hartling (2008) is, however, careful to point out that social support is not equivalent to connectionsthe former tends to be unidirectional (one person supported by another), while the latter are mutual, bidirectional.Relationship-centred resilience aligns well with African ubuntu values, which emphasise social connections as the crucible of personhood.Some authors use the term 'interdependence' or 'interconnectedness' as Western synonyms for ubuntu (Theron & Phasha, 2015).Ubuntu, together with a broader connectedness with one's cultural heritage, is an important source of resilience.
Relationships are part of a larger set called the 'social environment'.Some resilience researchers are drawing on the foundational social work concept of the person-inenvironment (PIE) to construct a more holistic picture of resilience processes.Van Breda (2017b), for example, categorises resilience processes into those that are individual or personal (the P in PIE, such as spirituality and optimism), those that are in the social environment (the E in PIE), which includes by social relationships (with family and friends, for example) and the environment (such as community safety and family financial security), and those that are interactional (the I in PIE, referring to processes the link person and environment, such as team work and empathy).The value of a PIE approach to resilience is to foreground the interactions between people and their social environmentsthus the resilience processes lie not just in the individual or in the environment, but in the way these transact.For example, while relationships are central to resilience processes, fostering positive relationships requires individuals to develop a set of social skills to elicit helpful and supportive responses from others, and for others (such as families and teachers) to develop understanding of these individuals and the challenges they are facing and a willingness to engage with them.Simply having individuals and others in the same location is insufficient; there has to be interaction between them.Lerner, Agans, Arbeit, Chase, Weiner, Schmid & Warren (2013:293) denote this focus on the interaction as "individual↔context relations".Transactional ecological theory (Felner & DeVries, 2013) extends this still further by attending to the transactions between systems in the social environment (e.g. between a child's parents and school).
Ungar's social ecologies of resilience takes this a step further.Ungar (2012:15) argues that while the resilience of individuals is a result of a combination of personal and environmental factors, "resilience is as, or more, dependent on the capacity of the individual's physical and social ecology to potentiate positive development under stress than the capacity of individuals to exercise personal agency during their recovery from risk exposure."Within the social ecologies of resilience, then, emphasis is given to family relations, social structures, services (e.g.welfare, health and education) and culture as the central and most powerful resilience resources (Bottrell, 2009).As a result, resilience-building interventions focus not on the individual, but on the social environment.
These new developments, which give far greater weight to the social environment than earlier psychologised resilience theory, are now evolving into what Wright, Masten & Narayan, (2013:30) refer to as a "fourth wave" of resilience research, which "is focused on multilevel dynamics and the many processes linking genes, neurobiological adaptation, brain development, behavior, and context at multiple levels".This wave has been made possible by a combination of rapid developments in genetic and brain science and statistical methods.Research on gene-environment interaction (Rutter, 2013) shows that certain genetic features, in interaction with environmental conditions (such as recurrent negative life events or child abuse), contribute to the development of psychopathology.By extension, it is anticipated that genes may also interact with positive environments to bring about more positive outcomes.Research on the biology of resilience is, however, still in its infancy, making this a cutting edge of the resilience field.
While one part of the resilience field is moving towards the super-micro, other parts are moving towards the critical macro.Bottrell's work with young people from a poor community in Australia, for example, raises the notion of 'resistance' within resilience theory.She asks (2009:335), "How much adversity should resilient individuals endure before social arrangements rather than individuals are targeted for intervention?"In the context of structural inequality, resistance to adversity is more appropriate than resilience.Bottrell (2007:599) defines resistance as "practices which express opposition to rules and norms in specific contexts, and which contain critiques of social relations, from the lived experience of marginalisation."In her work, resistance is in the interests of developing identity, rather than in political conscientisation or social change.Resistance is thus an exercise of agency in adverse social contexts.
Hart extends resistance as identity work to resilience as "changing the odds" (Hart, Gagnon, Eryigit-Madzwamuse, Cameron, Aranda, Rathbone & Heaver, 2016:7).She defines resilience as "overcoming adversity, whilst also potentially changing, or even dramatically transforming, (aspects of) that adversity" (3).In so doing, Hart et al. introduce a social justice discourse into resilience theory and seek to reduce or eliminate the vulnerability itself (Prilleltensky & Prilleltensky, 2005).In this way, the mediating processes of resilience are targeted not at accommodating to the adversity, but at challenging the adversity.This is a form of empowerment that is radically different from the empowerment inherent in resilience as 'helping individuals rise above their circumstances'.Hart et al. note that this changes the nature and role of research from one of knowledge generation to one of activism.
In a slightly different direction, but also within the broader family of social justice, Ungar (2004) draws on social constructionist theory to suggest that even apparently antisocial activities may be evidence of resilience processes contributing to resilient outcomes, when carefully examined from an emic perspective.Terms like adversity, resilience and better-than-expected outcomes are social constructions, and may vary from context to context.While there may be some universally accepted norms, such as the value of a loving family or sufficient food, much of what defines the resilience pathways of children (and indeed all people) is locally defined.Troubled youths, for example, frequently report 'antisocial behaviours' (such as drug use, self-harm and gang membership) as resilience processes, because they result in experiences of meaning, belonging and power.Ungar (2011:8) has termed these "hidden resilience", by which he means "functional but culturally nonnormative substitute adaptations".Recognising the hidden value of such behaviours contributes to social justice by redefining marginalised and social excluded individuals as people endeavouring to overcome adversity (Bottrell, 2007).
It should be evident that one of the key debates in resilience theory that is relevant for social work is between agency and structure.Agency reflects the power that individuals exercise over their lives and social environment, while structure reflects the macro systems that constrain the choices and opportunities of individuals.The tension between agency and structure is almost as old as the discipline of sociology (Romanos, 2014).Van Breda (2016c) uses the transition of young South Africans out of the child care system into young adulthood as a case example to argue that, in the real lives of individuals, both agency and structure are important.Focusing on agency without structure can lead to the deeper oppression of people by unjust social systems, while focusing on structure without agency can lead to people's disempowerment and marginalisation.Rather, both agency and structure, and the interactions between them, are necessary for resilience and social development.

RELEVANCE OF RESILIENCE THEORY FOR SOCIAL WORK IN SOUTH AFRICA
In drawing towards a conclusion, the question of the relevance of resilience theory for South African social work needs to be asked.It appears that resilience theory has been strongly critiqued, particularly for its neoliberal tendencies.Can such a theory contribute to social work thinking and practice in South Africa, and other countries in the Global South?Admitting my bias, I think the answer is yes, for three main reasons.
First, resilience theory guides the kinds of research questions we ask.Based on my experience of reading numerous social work research proposals, it appears that social workers are all too often interested in asking pathogenic questions, about how adversity or vulnerability impacts negatively on the lives of people.While these are important questions, they short-change social work, which is a profession and academic discipline concerned with facilitating change.The global definition of social work (IFSW, 2014) starts with the words, "Social work is a practice-based profession and an academic discipline that promotes social change and development, social cohesion, and the empowerment and liberation of people."In other words, social work is a changeoriented discipline.
Understanding vulnerability (or risk or adversity) is an important stage of the assessment phase in social work practice.A significant proportion of a rigorous assessment is understanding the challenges clients face, and the history and context of those challenges.But there is more to a client than just their challenges, and thus a good assessment addresses not just the problem, but also the person of the client (Van Breda, 2015a); it is the whole person that is of interest to the social worker.The clients' coping resources, strengths and attempts to deal with their challenges is an important part of understanding a whole person within their social environment.If social work is genuinely concerned with the dignity and rights of people, then attention should be given to the full spectrum of human experience, both negative and positive, both vulnerability and strength.This can be illustrated by linking to some of the sisters of resilience (solution-focused therapy, narrative therapy and the strengths perspective), though to a large extent these four sisters grew up unaware of each other, based on the general lack of cross-references between them (except for narrative therapy and the solution-focused approach).In various ways these practice models decentre and downplay the value of assessing people's problems.Solution-focused therapy (De Shazer & Dolan, 2007:2) is perhaps clearest about this point: "The solution is not necessarily directly related to the problem".This model argues that understanding problems does not clearly lead to an intervention.Instead, an intervention is best informed by the client's view of what life would be like in the absence of the problem.In a similar way the strengths perspective (particularly in its earliest forms (Weick et al., 1989:353)) eschews any focus on client problems: "The assessment focuses exclusively on the client's capabilities and aspirations in all life domains".White's (1989) narrative therapy is somewhat different, in that serious attention is given to 'the problem' through his use of externalising conversations, though he indicates that frequently he does not have externalising conversations with his clients (White, 2001).The heart of narrative therapy is in the identification and mobilisation of unique outcomes -those instances where the client's life was not dominated by 'the problem'.
In a similar way resilience theory has a major concern with the mediating processes that enable systems to achieve better-than-expected outcomes in the face or wake of adversity.The adversity is most certainly central in resilience theory, in contrast to solution-focused therapy and the strengths perspective.Indeed, one cannot talk about resilience in the absence of adversity.However, the heart of resilience theory is the resilience processes that mediate adversity and outcomes.In line with the practice models mentioned above, understanding these resilience processes positions one better to inform policy and practice than understanding the adversity only.
Second, resilience theory celebrates local and indigenous knowledge, and through that can make a contribution to decolonising social work theory and practice.Although some resilience research uses quantitative designs, which may make it vulnerable to an etic approach, much resilience research is qualitative, which is more amenable to an emic approach, allowing the voice and experience of participants to come to the fore.Through such approaches, local knowledge and practice about what helps people navigate hard times has a good chance of emerging.
In addition to this, there is a growing interest among resilience researchers in cultural expressions of resilience.Theron has been particularly attentive to cultural manifestations of resilience in South Africa.She and her colleagues note that many of the resilience processes identified in international studies manifest locally also, including attachment, mastery and meaning-making (Theron & Phasha, 2015:54), but observe that local research also draws attention to processes that are embedded in indigenous world views and values, notably "interdependence, spirituality, and duty to kin".In addition, attachments are not primarily to parents, as in most Western societies, but rather to the extended family (Theron & Theron, 2011).Other research points to the centrality of "strong women, an ethic of human and spiritual care, and an education-facilitated forward-focus" (Theron, 2016:660).These processes appear, however, to have both promotive and inhibiting effects on young people.They facilitate better-than-expected outcomes by linking people both to a collective heritage and to present family and community networks, but also become tainted by experiences of ubuntu not being lived out (when they are maltreated or sexually assaulted, for example) (Theron & Phasha, 2015).
Research on cultural resilience processes in South Africa still needs further development.Relatively little has been written on the subject, particularly by social workers.Cultural practice has been identified by social work academics as a priority area for the decolonisation of social work (ASASWEI, 2017), and could be advanced by investigating cultural resilience processes.However, critical engagement with this is also required.First, some cultural resilience processes, such as traditional spirituality and practices, are more highly valued by older persons than by the youth, and can become points of conflict rather than resilience for younger people (Cook & White, 2006).Second, Van Breda (2016b) observes that some apparently cultural resilience processes (notably 'passive acceptance', which Theron and Theron (2011) refer to as 'equanimity' or 'stoicism') are probably less an expression of culture and more of generations of socio-political oppression and poverty.Engaging in such critique, however, is also an investment in decoloniality.
Finally, resilience theory provides a useful framework for research that bridges the micro-macro divide, which could make important contributions to deepening social development theory (Patel, 2015).The developmental social welfare approach is the national approach to welfare in South Africa, informing and guiding all social welfare and social work activities.Patel identifies five key themes that have shaped this welfare approach in South Africa, one of which is "reconciling the micro-macro divide in developmental social welfare theory and practice" (Section 3.3).This is accomplished by recognising the need for micro, meso and macro social work services, ensuring that all social workers are competent in working at all three levels, and promoting integrated service delivery across the levels by all service providers.
Resilience theory, as has been shown in this article, has been torn at times by the tension between agency and structure, which is sometimes referred to in the sociological literature as micro and macro.As argued by Van Breda (2016c), resilience theory, particularly resilience informed by person-in-environment, transactional ecological and social ecological theories, is well placed to guide researchers in bridging this divide in their research by identifying a full spectrum of resilience resources that contribute to human flourishing.The emerging literature on social justice in resilience provides particularly interesting opportunities for new research in this area.

CONCLUSION
In conclusion, this article has explained resilience theory and elucidated some of the key points of tension and emergent thinking.Resilience theory is not without its critics, though the theory itself remains resilient in the face of the criticismsa testimony to its intuitive appeal and the usefulness of the theory in understanding the human experience of adversity and in informing policy and practice.At the same time, the critique has been useful in encouraging resilience theorists and researchers to examine the taken-forgranted assumptions within their work, which has led to greater recognition of the importance of the social environment and of social justice, among other things.In this way, resilience theory continues grow and develop.
In the final section I argued that resilience theory is of great relevance in South Africa at a time when it continues to translate social development theory into practice, grapple with rolling poverty and underdevelopment, and embrace the opportunity and challenge of decolonising social work thinking and practice.Resilience theory, intelligently and critically applied, can help to open up new understandings of how people in the resource-constrained environment of South Africa work for their growth and development, and how social structures of inequality and opportunity can be mobilised to cultivate a society that cherishes social flourishing.