The challenges

First, crises and disasters can dramatically affect people’s lives and their physical and psychosocial health in multiple ways. In our opinion, the theory and language of trauma, for example, have moved our understanding forward and are highly relevant. However, they also have limitations: they tend to focus on cohorts of people, which tends to draw research onto diagnoses rather than psychosocial processes. This allows the perception that everyone experiences events in similar ways and creates expectations that negative effects are inevitable, overshadowing instances when growth occurs.

Second, primary stressors are powerful and inherent in the events themselves (e.g. floodwater, infections, injuries, death). Secondary stressors originate in people’s life circumstances before an extreme event (e.g. illness, poverty), in pre-disaster societal arrangements (e.g. inefficient work circumstances, bureaucracy) or in inadequate responses to that extreme event (e.g. defective governmental responses). ^{Reference Williams, Ntontis, Alfadhli, Drury and Amlôt2} When improving responses to these challenges, we need an overarching paradigm that encourages attending to secondary stressors and their complex psychosocial impacts, which can be as powerful as primary stressors and, importantly, many are remediable. ^{Reference Williams, Ntontis, Alfadhli, Drury and Amlôt2}

Third, too often the effects of crises and disasters last much longer than we previously thought. ^{Reference Dückers, Stroebe, Baliatsas, Spreeuwenberg, Brüning and Stroebe1} The model that we seek must increase recognition of the importance of exploring what happens over time. ^{Reference Jordan, Shannon, Browne, Carroll, Maguire and Kerrigan3} Trajectories of becoming unwell and recovery, and what powers them differentially, are important in deciding how best to intervene. ^{Reference Oppo, Forresi, Barbieri and Koenen4}

Fourth, common approaches struggle to articulate a basic matter: namely, that crises interrupt complex flowing series of interconnected social processes. This risks relegating the sense-making experiences of people who have been swept up by disasters, including recognising their collective and relational natures.

Fifth, the construct of resilience is based on a theory that describes essentially invisible forces that are posited to account for differential sensitivity to the influential variables. Large amounts of literature treat resilience as a property of the person. Attempts to identify and quantify the associated factors have led to people who appear to cope well with an extreme event being described as resilient, despite this sometimes being situation-specific. This approach can take us on a circular path in which resilience is both cause and outcome. ^{Reference Ntontis, Drury, Amlôt, Rubin and Williams5} Moreover, this risks attributing responsibility for resilience to individual persons and away from social systems and organisations, and simplified perspectives often find their way into policy. ^{Reference Ntontis, Drury, Amlôt, Rubin and Williams5} Nonetheless, social and societal constructs of resilience are proving helpful; Norris et al treat resilience as a process through which people and communities use their resources effectively to return to a positive trajectory of functioning, allowing for integration of findings from multiple perspectives and levels of analysis. ^{Reference Norris, Stevens, Pfefferbaum, Wyche and Pfefferbaum6} Social identities, for example, have been associated with increased well-being, collective efficacy and social support. ^{Reference Muldoon7} Trauma may result from valued identities being undermined, and whether one is traumatised by a crisis can be a function of whether adequate resources exist to maintain a positive sense of self, maintain or develop new social connections, feel valued and supported and deal collectively with stressors. ^{Reference Muldoon7}

In summary, all the themes we identify require continued study and need to be brought together. Three issues appear crucial to articulating any sustainable wider theoretical framework. First, we need to think in terms of unfolding, changeable and interruptible processes. Second, we should recognise that processes weave together complex relations between events that unfold at different scales and through different media. Since there is never just one process but always complexity, we require process thinking to be relational and consider the togetherness of things. Third, we need ways to grasp how people’s experiences play out over time, and how those experiences and their consequences may be different for people in different societal locations.

Liminality

The concept of liminality is a centrepiece of processual and relational social and psychological theories that take systems and structures seriously as incompletions in process, organising space and unfolding through time. Liminality was introduced in 1967 by Turner, ^{Reference Turner8} who realised the psychosocial value of addressing the processual nature of human existence and espoused the importance of capturing the fluidity, movement and indeterminacy that structure seems to lack.

We propose that liminality might serve as the basis for a broader paradigm because it refers directly to scenes of change and meets our challenges. A limen is a threshold or margin but, in addition to this spatial meaning, it indicates a temporal threshold in which the connection between past and future is suspended. Suspensions are often introduced by extreme events through which we find ourselves no longer who we were but not yet who we may become. The concept is valuable in manifold ways. First, rather than separating and isolating relevant factors and variables, liminality has a social orientation and serves to connect the otherwise disconnected and to draw attention to hidden relations. Second, liminality works well in circumstances in which norms and roles are unsettled and re-evaluation of identity is required. Third, liminality concerns the space-time of relations and was born out of a concern with the dynamics of transition, foregrounding movement and passage. Fourth, liminality views transitions as transformations: not just how one thing relates to another, but also how one thing becomes another. Fifth, transformations need not be for the worse but can be for the better, or both; we need to understand better the conditions under which a transformation can be productive for people involved. This entails ethics: liminal paradigms must recognise cultural matters and be based on co-production. Sixth, a liminality paradigm can shed new light on the psychosocial genesis and consequences of disasters, crises and other disturbing events. But under what conditions does a liminal event yield trauma, positive transformation or change? We have more to learn.

The way forward

Researching and practising mental healthcare requires models that allow better integration of findings from a rich variety of disciplines and concepts, while recognising the many transitions involved. We are interested in further developing an adequately processual, relational and experiential psychosocial theory that encompasses interactively the contributions of many disciplines. Could liminality be a helpful vehicle for conducting processual research on emergent circumstances and disasters of all kinds? Suggesting new paradigms does not necessarily mean that they gain purchase: often prevalent models remain dominant. Nevertheless, the fields of disaster psychology and psychiatry have expanded rapidly over the past 40 years, and finding a way to bring together the diversity of our knowledge and practice and the disciplines involved is now due.

We illustrate this by recognising that a complex relationship can be identified between the liminal situations created during crises in which secondary stressors, and psychosocial processes in relation to people’s identities, can facilitate or hinder coping, adaptation and recovery. We are looking for an approach that (a) deals well with common characteristics of human systems that include ambivalence and uncertainty, and is able to recognise interruptions in the flow of events and relationships and the need for narrative reconstruction; (b) views resilience as a process that relies on sets of resources that people develop within social systems, and a concept of mutual interdependence rather than a static property of individual people; (c) examines stressors and risk factors as they unfold within social systems; and (d) examines the psychosocial processes that can facilitate or hinder coping and development. Studies of the impact of wars and other disasters on psychosocial and physical health have led to an emphasis away from the idea that symptoms must have a single organic source because the experiential component of the symptoms is evident. A processual liminality theory would encourage reframing this away from a deterministic cause-and-effect framework.

We think that a new integrated research and practice lens for disaster healthcare is required. We need to learn how to examine the conditions in which possibilities play out and better integrate insights from different disciplines and levels of analysis. We think that liminality meets the tests we have set.