Help Drafting

CJ598WK2DQReading2.pdf

Home >English homework help >Help Drafting

A Complexity Framework for Studying Disaster Response Management

Johan Bergstr€om, Christian Uhr and Tove Frykmer Division of Risk Management and Societal Safety, Lund University, PO Box 118, SE-22100 Lund, Sweden. E-mails: [email protected], [email protected], [email protected]

Guided by complexity theory, in this article, we argue that a complex understanding of disaster response management can be achieved by making multiple, transparent and modest interpretations. We suggest an analytical framework in which multiple system interpretations are constructed, all based on explicit analytical choices according to three aspects: (1) system dimension, (2) system scope and (3) system resolution. We apply the framework to a major Swedish forest fire and conclude that direction and coordination as system properties, emerging at a macro level, are the result of inter- play between various patterns of influences. These patterns, we argue, can be con- structed and analysed through a complexity framework allowing for the construction, and contrasting, of multiple system interpretations.

1. The need for a complex understand- ing of disaster response management

In this study, we are interested in how to generate acomplex understanding of disaster response manage- ment. We will introduce an analytical language for studies of how disaster response systems achieve direction and coordination. This language is rooted in complexity theory, allowing for multiple interpreta- tions, based on transparent and clear analytical choices to generate a complex and modest understanding of disaster response. The study begins with outlining what we mean by

direction and coordination in a disaster response con- text. Then, we contrast a complex understanding from a reductionist one to build an analytical framework for studies of disaster response management, which is rooted in complexity theory. The societal response to a large wildfire in Sweden is used to demonstrate how the analytical framework can be applied. Finally, we will end with reflections on some central challenges

related to studying disaster response management from a complexity perspective.

1.1. Achieving direction and coordination in a disaster response system

The meaning of key terms such as coordination, Com- mand & Control and leadership is disparate and some- times also emotive. Scholars who spring from the sociological tradition, focusing on managerial aspects of civil emergencies and disasters, seem to distance themselves from military vocabulary and problem framing. For some scholars, Command & Control can be an expression representing obsolete strictly hierar- chical management paradigms (see, e.g. Quarantelli, 1988; Seddon, 2005). For others, Command & Con- trol science can describe a progressive research field dealing with complexity including self-organization and multi-organizational collaboration (see, e.g. the Com- mand & Control Research Program, CCRP). Terms such as coordination are used in parallel for describing

Journal of Contingencies and Crisis Management Volume 24 Number 3 September 2016

an approach, an activity, a structure or an effect (see Drabek, 2007; Ekman & Uhr, 2015; Quarantelli, 1988; Uhr, 2009). When it comes to other related terms, such as leadership, Stogdill (1974) wrote that ‘there are almost as many different definitions of leadership as there are persons who have attempted to define the concept’ (p. 7). Management as a term is in itself a bottomless barrel of meanings.

This heterogeneous semantic blend does not lead to us arguing for a general standardization of meaning of key terms, because such a condition most likely is impossible due to the uncountable actors using the words. Several attempts have been made in related research fields, for example within risk management (see Kaplan, 1997), though with limited success. Rather, we accede to an approach suggesting that a central aspect of disaster response management is to achieve direction and coordination among available resources in order to meet a range of needs during the acute phase of a disaster. This approach is not intended to provocatively challenge existing paradigms, but rather to bring attention to what needs to be achieved instead of immersing in the details on how to do it. We are interested in providing an analytical framework for studying, in actual cases, multiple inter- pretations of how direction and coordination is achieved and challenges related to achieving direction and coordination.

Ekman and Uhr (2015) advocate the need for direc- tion and coordination to make use of existing capabil- ity. Without direction, the capabilities of the resources will be useless, because they will by definition remain passive. Without coordination, the resources are likely to be in each other’s way and not supporting each other when possible and necessary. We see direction as an internal effect among the resources meaning that the resources engaged in a response are oriented towards formulated goals (Ekman & Uhr, 2015). As emergencies and disasters are phenomena considered to be taking place in dynamic and partially unpre- dictable settings, a goal and the path to it cannot be predetermined in detail. The environment in which the resources act is characterized by uncertainty (Breh- mer, 1992). Directions cannot be predetermined in detail, and there is no single actor grasping what vari- ous needs that might emerge, or what synergies that might occur. Coordination is here seen as an internal effect among the resources meaning that activities and subgoals are adjusted to make most possible use of available resources (Ekman & Uhr, 2015). Disaster scholars such as Comfort suggest that coordination implies ‘aligning one’s actions with those of other rele- vant actors and organisations to achieve a common goal’ (Comfort, 2007, p. 194). Malone and Crowston (1990) define coordination as the act of managing interdependencies between activities performed to

achieve a goal. These definitions harmonize with those suggested by Ekman and Uhr (2015), however, focusing on activities and not effects. Researchers within the field of disaster research

such as Quarantelli (1988), Dynes (1990), Dynes and Aguirre (2008), and Drabek (2007) all advocate a focus on coordination in prescriptive discussions of disaster response management. However, it can be challenging to understand the various aspects and nuances of the term and how to relate it to other terms concerning ‘management’. Coordination (or coordination models) is sometimes described as an alternative to ‘bureaucratic models’, such as the ‘com- mand & control model’ (Drabek & McEntire, 2003). Recently, various models, and studies, of ‘Incident Command Systems’ have addressed the need to make the systems themselves allow for dynamic organiza- tional structures (e.g. Jensen & Waugh, 2014; Rimstad, Nj�a, Rake, & Braut, 2014; Scholtens, Jorritsma, & Hel- sloot, 2014, which all appeared in the same special issue of this journal). We argue that bureaucratic models are tools, in addition to others, for achieving direction and coordination. Some models are func- tional tools and some might be dysfunctional, all depending on the context. In this study, we will pay attention to the systems

of resources that normally are engaged when a society responds to a disaster; a disaster response system that needs direction and coordination at various levels to meet individual, societal and environmental needs. We argue that constructing a complex understanding of such systems is crucial when trying to explore the space of possibilities that any disaster response system has in its attempts to achieve direction and coordination.

1.2. A complex understanding

The ultimate aim of this study was to achieve what we call a complex understanding of any disaster man- agement operation. A fundamental principle of com- plexity theory is that it is impossible for any actor to grasp the complete working of the system as a whole. Further, all complexity frameworks have as their start- ing point the holistic principle that the behaviour of the system as a whole cannot be reduced to the func- tioning of the constituent components, but only sought in their nonlinear interactions and relations. Indeed, complexity theory does not offer one coher- ent theory. Using the words of Heylighen, Cilliers and Gershenson: ‘Complexity science is little more than an amalgam of methods, models and metaphors from a variety of disciplines rather than an integrated science’ (Heylighen, Cilliers, & Gershenson, 2007, p. 117). Such disciplines include quantum mechanics (Capra, 1982), biology (Von Bertalanffy, 1950),

A framework for a complex understanding 125

ecosystem theory (Gunderson & Holling, 2002) and even postmodern philosophy (Cilliers, 1998). To describe complexity theory in terms of scientific

facts about how to understand a disaster response system would be self-contradictory. Instead, complex systems are often described in terms of a number of principles that explain why complex problems are so inherently difficult. Such principles typically include nonlinear interactions, openness, emergent behaviour, system dynamics and resilient behaviour emerging from diversity and requisite variety. However, while we also wish to account for such principles in our interpretations of disaster management operations, the complexity framework developed for this study has not been intended to show, or measure, the com- plexity of disaster management operations (below we will refer to this as an ontological approach to com- plexity). Instead, we adopt a social constructivist per- spective, adopted from Cilliers’ approach to complexity (Cilliers, 2005) from which complexity theory implies the need to highlight and analyse multi- ple interpretations of any disaster management opera- tion, make transparent analytical choices and be modest about any claims made regarding the manage- ment of disasters (below referred to as an epistemo- logical approach to complexity).

1.3. Complexity as epistemology or ontology

Studies of disaster response from a complexity per- spective typically approach the notion of complexity as ontology, that is as a characteristic or state of the system. In such studies, the aim is typical to describe how complexity manifests itself in the disaster response system. Comfort (e.g. Comfort, 1994, 1995) utilizes complexity theory in arguing that self-organiza- tion as a phenomenon of disaster response makes the disaster response system complex. Self-organization is in these writings viewed as an emergent property, showing that the system is indeed complex and there- fore poses certain challenges in terms of management strategies. In more recent writings, Comfort argues that system complexity (inherent in the scale and mul- tiple dependencies within a multi-organizational disas- ter response system) give rise to challenges of coordination and suggests solutions in terms of deci- sion support systems that can facilitate coordination among multiple stakeholders as well as incorporating dynamic changes (self-organization) of the disaster response system itself (Comfort, Sungu, Johnson, & Dunn, 2001). With the same focus on how to orga- nize disaster response operations Comfort criticizes the idea of hierarchical organizations, arguing that they are typically unable to adapt to the dynamics of disas- ters (Comfort, 1999; Comfort & Kapucu, 2006). Also Kiel (1995) as well as Atkinson and Moffat (2005)

offer normative approaches, ontologically arguing that since disaster response systems are complex, certain management strategies should be applied.

Previous work has also been conducted focusing on different techniques aiming at mapping complex disas- ter response systems (Abrahamsson, Hassel, & Tehler, 2010; Uhr & Johansson, 2007; Uhr, Johansson, & Fred- holm, 2008) and the connections between different levels of the system (O’Sullivan, Kuziemsky, Toal-Sulli- van, & Corneil, 2013). These studies, using social net- work analysis or system dynamics modelling, have an epistemological take on complexity because they sug- gest a framework for analysing and evaluating emer- gency response. Yet, they are epistemological approaches with ontological commitments because they focus on mapping the complex actor-network and its interactions; that is it focus on ‘showing the complexities’ of the system.

In this study, we distinguish between complexity as epistemology and complexity as ontology to empha- size that our interest is not in assessing whether or not a disaster response system is complex or not, or to what extent, which features of complexity that it shows, or how to best manage the complexities. Our framework rather aims at exploring the space of pos- sibility in working to achieve direction and coordina- tion in any disaster response operation, using complexity theory as the epistemology guiding our analytical choices. A complexity epistemology defies the ability to make final and noncontroversial claims regarding any disaster response operation. It rather embraces multiple perspectives, transparency and modesty about any claims made. As Cilliers puts it:

We cannot have complete knowledge of complex systems; we can only have knowledge in terms of a certain framework. There is no stepping outside of complexity (we are finite beings), thus there is no framework for frameworks. We choose our frame- works (Cilliers, 2005, p. 258).

As Cilliers argues from a complexity perspective there are only normative ways in which to choose frameworks of analysis. Analytical choices are based on normative assumptions of what aspects of disaster management operations that are interesting to inter- pret and analyse. In developing an epistemological framework intended to facilitate the generation of a complex understanding of disaster response manage- ment, we therefore need to incorporate a trans- parency of the analytical choices made, so that our normative positions become open to further discus- sion and critique.

Further, Cilliers’ approach to complexity embraces the principle of multiple perspectives that ‘different descriptions will decompose the system in different

126 Johan Bergstr€om, Christian Uhr and Tove Frykmer

ways. Different descriptions may also have different degrees of complexity’ (ibid, p. 257). Consequently, our analytical framework has been developed to account for multiple interpretations of one disaster response operation. This in order to generate knowl- edge characterized by epistemological pluralism (Healy, 2003) and to avoid the reductionist reasoning of com- ing up with one correct version of the system and its behaviour.

The final point that we want to stress and include in our efforts to develop a framework able to facilitate the generation of a complex understanding of disaster management operation is the modesty that Cilliers argues for:

A modest position should not be a weak position, but a responsible one. Such a position will be devel- oped by examining three arguments: the argument that modest positions lead to relativism, the argu- ment that modest positions are subject to the per- formative contradiction and the argument that modest positions are vague (Cilliers, 2005, p. 256).

This appeal to the modesty of the analyst is rooted in the complexity theory view of knowledge as funda- mentally provisional, depending on time, space and framework of analysis. The way in which we wish to account for this modesty in our analytical framework is again in making our analytical choices transparent and in making clear that additional interpretations will always be possible and should always be appreciated in further debate.

To summarize our view on complexity, the above review suggest that a framework developed with the intention to generate a complex understanding of dis- aster response needs to:

• Allow for a discussion on direction and coordina- tion as emergent phenomena from the various inter- actions and relations in the disaster response.

• Allow for multiple interpretations of such interac- tions and relations within the same analysis.

• Highlight the potential conflicts, overlaps or even interactions between such perspectives.

• Make explicit the analytical choices made for con- structing the multiple interpretations.

• Be modest in recognizing that additional interpreta- tions will always be possible and should always be appreciated.

2. Building a complex understanding: multiple interpretations based on three system aspects

To address the above-mentioned criteria for building a complex understanding, our analytical framework is

based on constructing multiple system interpretations, and for each system, interpretation makes explicit the analytical choices according to three aspects: (1) sys- tem dimension, (2) system scope and (3) system reso- lution. Each of these three aspects is outlined below.

2.1. Aspect one: dimension

The first analytical choice to make explicit in a disas- ter response analysis rooted in a complexity episte- mology is what we call the system dimension. A system dimension is the theoretical perspective con- structing the analysis of relations and interactions between system actors and elements. To do so, the dimension will include both the actor(s) and the rela- tionships studied. One example of a system dimension is an informal network, where the actors are individu- als and the relations connecting them to each other are built on trust. Another dimension could be bureaucracies describing formal roles and formal power relations. Approaching the system as held together by interacting functions, (a functionalist approach), is yet another example of a system dimen- sion. In conjunction with an analysis, one may include principally different elements, for example individuals using computers that are interacting through commu- nication relations. When studying direction and coordination during

disaster response, relations connecting the actors are of particular interest if they enable influence, such as communication relations, formal power relations, trust relations etcetera. In fact, influence could be seen as a relation between actors in itself (see Fredholm & Uhr, 2007).

2.2. Aspect two: scope

The aspect of scope refers to the analytical drawing of the boundary around what actors and relationships that are to be included in the interpretation. If con- structing an interpretation based on the dimension bureaucracies, describing roles and formal power rela- tions, it is necessary to create an outer boundary for what roles to include in the interpretation (not what kind of roles or the relationships between them as in the case of the aspect dimension above). The system scope can be based on relevance, for example by including the bureaucratic structures perceived to be relevant for the response operation (and consequently argue why others are not). It can also be based on spatial demarcations, for example by including only bureaucracies describing roles and formal power relations on an accident scene, at the county adminis- trative board or similar. Typically, system scope will be defined in terms of both relevance and spatial demarcations.

A framework for a complex understanding 127

2.3. Aspect three: resolution

The system resolution has to do with a level of detail. As we choose larger scopes of analysis, it becomes necessary to simplify the elements by clustering them. Sometimes, it is relevant to interpret single roles and their relations within one organization, but sometimes it is necessary to choose a lower system resolution by clustering roles. Informal networks based on trust can be analysed from a group perspective, and technical system dimensions can be described as interlinked clusters of communication devices and so on. There is a logical link between system resolution

and system scope. A wide system scope often sacri- fices a detailed system resolution, similar to a wide angle lens. A narrow system scope allows detailed sys- tem descriptions, similar to a macro lens, but sacri- fices a comprehensive analysis. Within a complexity epistemology, it therefore makes sense to vary the system resolution in different interpretations in order to compare and contrast them with each other.

2.4. Grasping the interpretations

The aim of us introducing this framework is to facilitate the generating of a complex understanding of the space of possibility for achieving direction and coordination in disaster response. To us, a complex understanding implies the interpretation of a diversity of perspective: a variety of dimensions, scopes and resolutions. Such a diversity of interpretations not only generates a multi- tude of interpretations, but also accounts of how the interpretations themselves interact, complement each other or contradict each other. Conflicting interpreta- tions add to the complex understanding and the analyti- cal modesty that we try to stimulate. Using the multiple interpretations generated through

the framework of transparently choosing aspects of study ultimately make us able to construct a convincing description of how direction and coordination are achieved in disaster response operations. Also, we aim at constructing different challenges related to achieving direction and coordination and ultimately learning out- put in terms of advice concerning how to meet the challenges.

3. Application of the analytical frame- work on the recent Swedish wildfire

In this section, we use the analytical framework intro- duced above to understand how direction and coordi- nation were achieved during a major wildfire in Sweden 2014. Two of the authors behind this article were present as observers and researchers in the joint command centre for 4 days. Our analysis is to be seen as a demonstration of how to use the theoretical

framework and not as a thorough analysis of the disas- ter response system during the event. The data under- lying the demonstration of the framework comprise field notes, observations, public reports and 24 inter- views, of which five were on site and 19 after the operation ended. The data were collected for an anal- ysis of direction and coordination during the handling of the wildfire and will now serve the purpose of showing multiple interpretations in our analytical framework. Examples from interviews have been translated from Swedish by the authors.

The extensive wildfire broke out on the 31 July 2014 in the county of V€astmanland in Sweden. The emergency response operation was officially ended on the 11 September. During the response, a wide range of societal resources were engaged, such as fire and rescue service organizations from different parts of Sweden, the police, the County Administrative Board, central agencies, the Swedish Armed Forces, volunteer organizations, unorganized volunteers, private organi- zations and Italian and French water bombing aircrafts with crew.

Below, we will introduce six different interpreta- tions of how direction and coordination were achieved during different stages of the event. Following the interpretations, we discuss shortly how such inter- pretations together contribute to a complex under- standing of the disaster response system. We should already here acknowledge that our aim with the example is not to generate such a complex under- standing, but only to show the principle path towards it with the selected interpretations.

3.1. Interpretation one – a system of two sepa- rate formal organizations

The Swedish system for civil safety is, as most of the Swedish political system, based on a decentralized principle of local responsibility. As is regulated in the Swedish Civil Protection Act, municipalities are responsible for providing fire and rescue services within their geographical borders. In the case of the wildfire of V€astmanland in 2014, the fire actually started close to, and soon crossed, the borders of two different municipalities, meaning that two different fire and rescue service organizations were involved in the response.

In an interpretation of the formal system of the response operation during the first days of the wild- fire, we take the formal roles as actors and the formal power relations between them as our dimension of study. As our scope, we combine relevance and spatial demarcation by focusing on the senior commanding staff within the relevant fire and rescue service organi- zations present at the command posts rather than all individual fire fighters involved. This enables us to

128 Johan Bergstr€om, Christian Uhr and Tove Frykmer

interpret the more strategic challenges of achieving direction and coordination in the response operation. The resolution chosen for this interpretation is at the individual level. A conceptual illustration of this inter- pretation is shown in Figure 1.

When an incident affects several municipalities, as in our case, the Swedish fire and rescue service orga- nizations should cooperate and coordinate their resources in the response operation. Despite this, the formal fire and rescue service organizations involved in the wildfire did not achieve this but rather remained as two separate organizations with separate incident commands (Ministry of Justice, 2015), which is illustrated in the figure above with the dashed line in between the formal organizations. This can also be exemplified with examples from our interviews:

We had one incident command in XX [location] and one in XX [location], I think. And it was unclear: who is the incident commander here and who is the boss here?

And then I realized that the work was still per- formed on two different sides.

It was really strange with two separate incident response operations, I thought that was very strange. I didn’t understand it.

During Monday the 4 August, the fire spread rapidly and the response operation expanded in terms of actors and resources, such as the Swedish Armed Forces, volunteer organizations and forest property owners (Ministry of Justice, 2015). The actors of this

formal system realized that the response system lacked the ability to achieve acceptable direction and coordination among the resources to meet the needs and therefore asked the County Administrative Board to assume responsibility of the handling of the wildfire (County Administrative Board of V€astmanland, 2014). This was executed on Tuesday the 5th and at the same time the County Administrative Board appointed a highly experienced commander to be the new inci- dent commander of the operation (Ministry of Justice, 2015).

3.2. Interpretation two – a formal, centralized system

From the moment that the County Administrative Board of V€astmanland assumed responsibility for the disaster response operation, we need a new interpre- tation of the formal system. The aspect resolution (in- dividual level) remains the same but we need to change the dimension and scope of the interpretation. The dimension now consists of formal roles and for- mal relations between them and the scope takes the form of the joint command centre that was estab- lished at a conference facility in the village of Ramn€as. For a conceptual illustration of interpretation two, see Figure 2. The previously active fire and rescue service organi-

zations were still active, but the command structure now changed. In addition to the fire and rescue ser- vice, other organizations and authorities such as the police and the Swedish Armed Forces were included in the system, as well as County Administrative Board representatives. These actors are represented with

Figure 1. Conceptual illustration of interpretation one.

Figure 2. Conceptual illustration of interpretation two.

A framework for a complex understanding 129

formal roles horizontally attached to the incident commander and with their respective organizational structure beneath. The appointed incident commander expressed in one of our interviews that the greatest challenge during the start-up of the new formal cen- tralized organization was to ‘form one response oper- ation with all aspects included’. As the formal, centralized organization was formed

an earlier sense of chaos was changed into a sense of calm focus, as expressed in one of our interviews:

On the Wednesday a structure had been created, there were staff on site as well as in the incident command, we have a staff room, we have positions, we have titles, we have assignments. [. . .] Since that day, it looks like home.

During observations on site, we could note that also actors not formally involved in the chain of com- mand ranging from this newly appointed commander expressed that he was also their commander. The interpretation of the formal command structure is not able to capture such accounts, and since most obser- vation reports and early evaluations of the operation remain at the level of formal roles and responsibilities, we need additional interpretations of how direction and coordination were achieved in the disaster response operation, leading to interpretation three.

3.3. Interpretation three – a system of trust

The commander had great freedom in selecting the fire and rescue service staff acting in the new joint command centre organization. Several appointments reflect relations of trust between the actors rather than formal relations and can be seen in our interview with the commander:

Then I began to pick from the group of people I know and trust in order to man these functions.

Further down in the now extended fire and rescue service organization similar patterns of trust can be distinguished:

Then, in the relationship with XX [the commander] and XX [the chief of staff] there was a more per- sonal trust which made it very simple.

When we reported at nine there were a lot of rotations and stuff going on in there, and one per- son heard that I was from Gothenburg and then he pointed at me and said that you will be responsible for the north-eastern sector, then he asked me who I was.

A senior commanding officer in the fire and rescue service came up with the expression ‘the Ramn€as spirit’, which reflected the high level of trust in the response organization:

Number one in this success story [achieving direc- tion and coordination in Ramn€as] is the County Board governor who gave confidence to XX [the commander] and in a way to me. Number two was XX [the commander], he gave confidence, he allowed for a great arena to act on. I think I came up with the expression ‘the Ramn€as spirit’ and I think that is a product of all that, where it started with the County Board governor.

Additional examples of trust can be seen in how actors not formally subordinate to the commander or the fire and rescue service, such as the police, were nevertheless perceived as so:

Yes, I experience that it is that way, that he [the commander] is the one people turn to, he gets that role at least.

We [the fire and rescue service] own the response operation. [. . .] When the police or the Armed Forces want to do something they ask me.

Also members of the police acting at this joint com- mand centre expressed to us, in informal conversa- tions, that they perceived themselves as subordinate to the commander in charge of the fire and rescue service operation. Several observations on site also justify seeing the organization of the joint command centre as based on relations of trust, rather than of only formal delegations. One observation involves a fire and rescue service employee, being interested in what was going on in Ramn€as managed to hitchhike with a military helicopter to the joint command cen- tre. There, he happened to know the person respon- sible of operational control, who soon had his friend appointed head of fire and rescue operations for a particular geographical sector.

Thus, our interpretation three entails a new dimen- sion of humans as actors and relations of trust between them. We keep the scope as the joint com- mand centre in Ramn€as, and the resolution the individ- uals acting there. For a conceptual illustration of interpretation three, see Figure 3.

This also becomes an interpretation of how the for- mal joint command organization is dynamically orga- nized and re-organized partially based on relationships of trust. However, this does not make the organiza- tion informal. Using the interpretations together, we start to build a more complex understanding of the

130 Johan Bergstr€om, Christian Uhr and Tove Frykmer

disaster response operations and how direction and coordination were achieved.

3.4. Interpretation four – a functional system

The response operation was managed from one room at the Ramn€as conference centre, which can be called the incident command room. In the room relevant actors such as the fire and rescue service’s operational senior staff, air space coordinators and the Swedish Transport Administration were gathered. Interviews with staff from the incident command room revealed that relations in this room were mainly based on func- tions rather than formal relations and that one would turn to the relevant function in order to achieve direction and coordination:

I didn’t ask any commanding fire and rescue officer when we wanted water bombing, instead I went straight to the helicopter responsible.

We sat so close to each other in Ramn€as so it was better to go out to the military with what we needed, the police was in the same room.

In this room there were many actors, for example the Swedish Transport Administration is an actor that I usually never use that way. [. . .] It probably built on that we were in the same room, it became natural that one ‘here’s a task that I can solve’, that is the Swedish Transport Administration voluntarily took over a task so that we [the fire and rescue service] didn’t need to handle that.

This functional system is our interpretation four. As dimension, we have functions and functional relations between them. The room sets a spatial demarcation for the scope of the system, and as resolution, we take the individual level. For a conceptual illustration of interpretation four, see Figure 4.

3.5. Interpretation five – a socio-technical net- work of forest property owners and their tools

Disasters involve actors who are not engaged in any formal command organization (Drabek et al., 2003), but still achieve direction and coordination of resources to meet certain needs. During observations on site, we could note that local forest property own- ers came together to try to protect their property. As they perceived a need (a threat to their property) and had the tools to act, they were able to achieve direction and coordination among their own resources to meet the need of protecting their properties from fire by, for example creating firebreaks in the forest. The composition of interpretation five can be under- stood as a result of self-organization. However, one must bear in mind that we emphasize interpretations and thus leave ontological considerations aside. Exam- ples from our interviews also justify this interpreta- tion:

There were property owners, forest property own- ers, they had interests but they didn’t express it that way, but they wanted to help.

There were a lot of civilians who wanted to help, mostly people living there who had resources they believed we needed, like tractors and so on.

Our interpretation five is a socio-technical network of forest property owners and their tools forming the dimension. The forest property owners were held together by agreements, probably based on previous relationships and a common incentive to protect their property. Thus, the relations in this interpretation con- sist of these agreements as well as ownership linking the owners to their tools. The scope of this interpreta- tion is made up of the spatial demarcation of forest property owners and their resources in the specific areas threatened by the fire. The resolution is at the

Figure 3. Conceptual illustration of interpretation three.

Figure 4. Conceptual illustration of interpretation four.

A framework for a complex understanding 131

level of individuals and their tools. In Figure 5, we show a conceptual illustration of interpretation five.

3.6. Interpretation six – a system of communication

Due to the large area covered by the wildfire, the operation was separated in geographical sectors to divide the workload into smaller segments, something that is commonly done in fire and rescue service operations. From our functional system, orders were sent to one or several of these geographical sectors, where fire and rescue service staff and other resources such as the Swedish Armed Forces and organized/self-organized volunteers assisted in per- forming the tasks. Due to perceived unclear formal responsibilities, it was not at all times clear for the fire and rescue service personnel whether they formally ‘owned’ the resources in the sector and whether they consequently could give formal orders to them. Nev- ertheless, the actors in the geographical sector prag- matically worked together to achieve direction and coordination to solve the needs, which can be noted in our interviews:

I had around six units and military there. On Wed- nesday there were volunteers [forest property own- ers] there, on Thursday we became friends and on Friday we hugged when I was going home. I had an awesome cooperation with the volunteers.

Out there were a lot of military helping out, they managed themselves and just reported how much people they had and then we [fire and rescue ser- vice] split them up.

The organised volunteers and the military were fan- tastic, they just delivered.

One fire and rescue service commander responsible for a geographical sector mentioned how the cooper- ation around firebreaks took place:

There were forest machines, and after the machines came bulldozers and after was a military tank with a water tank and after that came two men putting out small fires that could appear after the machines.

Our interpretation six is made by looking at the relationships between the fire and rescue service as a group and other group of actors in the sector, which then also includes our socio-technical network of property owners and their tools, that is interpretation five. The dimension of this interpretation consists of humans as actors connected by relations of communi- cation rather than, for example formal relations. The scope includes spatial demarcation of the geographical sector and the resolution is based on interacting groups of actors, representing the fire and rescue services, forest property owners, the Swedish armed forces and organized volunteers. This interpretation is shown as a conceptual illustration in Figure 6 with the fire and rescue service as the hub of communication.

Not everything was unproblematic in our system of communication. For instance, the initiatives of the property owners described in interpretation five were often performed without the knowledge of the fire and rescue service staff and the fire and rescue ser- vice did not have any influence on the direction of the resources. When the forest property owners did not get response from their official contact attempts with the fire and rescue service (Ministry of Justice, 2015), they probably decided to act on own initiative. Not knowing when or where the local property owners were acting was looked upon with concern for the safety of the property owners. When one of the authors joined for a control drive in the area, fire- breaks not ordered by members of the fire and rescue service were discovered. Prior to the occasion, there

Figure 5. Conceptual illustration of interpretation five.

132 Johan Bergstr€om, Christian Uhr and Tove Frykmer

was also an incident when the joint command centre needed to interact with local and self-directing actors by actually rescuing them from dangerous situations. In addition, the forest property owners were at times perceived as difficult to control by the fire and rescue services:

So when I get out there, around ten farmers [forest property owners] with these fertilizers stand there, and when you call for more people and so on, then there is a lot going on. And these farmers were really difficult to control.

3.7. Grasping the interpretations

In Table 1, a summary of the six interpretations is given.

How was direction and coordination achieved in this Swedish fire disaster? What are the needs that should be met by achieving direction and coordination among the resources? What defines a resource? Are local incentives a problem or a necessary resource?

Table 1 suggests that the answers are to be found in a complex understanding allowing multiple perspec- tives. Based on our observations, we conclude that different interpretations are possible, but that single actors are trapped in their respective rationalities and thus have difficulties in grasping multiple perspectives allowing a complex understanding. This may not only affect the quality of evaluation and learning processes after an emergency/crisis/disaster, it also becomes an operational problem during a response, for example to the wildfire case described in this study. When working to achieve direction and coordination among a conglomerate of different resources (fire are rescue service, visiting volunteers, military, local property owners, etc.), not only the existence of various goals must be acknowledged. One must also realize that direction and coordination as system properties, emerging at a macro level, are the result of interplay between various patterns of influences visible only through multiple interpretations. In this short introduction to how our analytical

framework can be used, we have included different perspectives generating not only different answers to

Figure 6. Conceptual illustration of interpretation six.

Table 1. Summary of the Six Interpretations

Interpretation Dimension Scope Resolution

1. A system of two separate formal organizations

Formal roles and power relations Fire and rescue service command in two municipalities

Individuals

2. A formal, centralized system Formal roles and power relations The joint command centre Individuals 3. A system of trust Humans and trust relations The joint command centre Individuals 4. A functional system Functions and functional relations The incident command room Individuals 5. A socio-technical network of forest property owners and their tools

Forest property owners with tools and agreement/ownership relations

Forest property owners and their tools in specific areas

Individuals and tools

6. A system of communication Humans and communication relations The geographical sector Groups

A framework for a complex understanding 133

the questions asked above, but also additional ques- tions, such as Are self-organized local actors with resources a coordination problem to consider or a necessary resource to appreciate? Do direction and coordination emerge through means of formal struc- tures or structures of trust? What happens when these interpretations coexist and interact? What hap- pens when actors change their views of each other or even themselves? The complex understanding is shown when we are modest enough to construct this pluralism of possible interpretations.

4. Discussion of implications for disaster studies and response management

The first critical question to be raised concerning our suggested framework, and our demonstration of how it can be used, is whether it allows for an understand- ing of complex systems or a complex understanding of systems. The complex understanding implies the construction of multiple interpretations, and a mod- esty about the ‘correctness’ of any such interpretation by acknowledging that interpretations are provisional in time as well as space; acknowledging that there will always be room for additional interpretations (the more ontological assumption that no actor, or researcher can fully grasp the complexity of the whole system). Not only will there always be room for addi- tional interpretations, interpretations will also only seem valid in constrained temporal and spatial scales. This is not only because the dynamics of the disasters changes the response operation in a tactical sense, but also because actors’ views of other actors will change and actors’ self-images might change too as a disaster unfolds. Further, this modesty is also achieved by being clear about the analytical choices made to con- struct different interpretations. This is in our frame- work operationalized as the different aspects of every interpretation. We should also note that one aspect, in which our

view of complexity theory differs from the way that it is typically operationalized in the field of disaster research, is our view of the principle of emergence. In the disaster literature, emergence is typically made synonymous with self-organization and ‘ad hoc solu- tions’ (e.g. Boersma, Comfort, Groenendaal, & Wol- bers, 2014; Comfort, 1994). To us, seeing emergence as the core system argument that the (emergent) behaviour of the whole cannot be reduced to the behaviour of the constituent components but only understood in terms of interactions and relations (Dekker, Bergstr€om, Amer-W�ahlin, & Cilliers, 2013), self-organization does not emerge as much as it has emerging consequences in terms of additional chal- lenges to the system’s abilities to achieve direction and coordination. Again, this is a matter of perspective

(involving dimension, scope and resolution of the sys- tem), but our focus is rather on direction and coordi- nation as emergent phenomena, emerging from the interactions between self-organization, communication, use of various resources, formal command structures, regulations, weather and other aspects perceived rele- vant to the one studying a disaster response system.

We have distinguished between complexity as epis- temology and complexity as ontology in order to emphasize that we do not intend to argue for ‘how’ complex a particular disaster response operation was or for how to best direct and coordinate complex dis- aster management operations. Instead, we highlight how some critical, and in the disaster management lit- erature typically ignored, points from complexity the- ory can inform our development of a complex understanding. However, of course, there is a connec- tion here between complexity as ontology and com- plexity as epistemology. Hence; generating a complex understanding is not simply an academic exercise but can feed back to practitioners and system designers in several ways. First and foremost, just showing the potential coexistence of multiple interpretations of one single disaster response operation could be a ped- agogical tool for teaching some central challenges to efforts for achieving direction and coordination and serve as a start of discussion for how to address some of them. Second, how different interpretations coexist and influence effort to achieve direction and coordina- tion could be built in to scenarios used in disaster response training and simulation. Third, our concep- tual tool for understanding various interpretations of recourses engaged in response operations can support efficient response in naturalistic settings, that is in real emergency situations. One example is when arranging meetings for composing situational pictures. Instead of trying to achieve a one dimensional (homogeneous) understanding of resources engaged in a response operation, participants can be encouraged to try to achieve a heterogeneous holistic understanding, consti- tuted by different coexisting interpretations.

It might be possible that some readers interpret our idea on complex understanding as an alternative to existing response doctrines, such as ICS/NIMS. That is not our intension. The aim has been to pro- mote an approach that enables further understanding and reveals the need for being humble when con- structing representations of reality.

Our six interpretations of the wildfire disaster response operation are only brief introductions to what could be subject for further analysis. For example, to deepen the interpretation of a functional system models such as the functional resonance analysis method (FRAM) aimed at analysing functional relations in terms of how different functions ‘resonate’ with each other (Hollnagel, 2012) could be useful. When going

134 Johan Bergstr€om, Christian Uhr and Tove Frykmer

ahead with deeper analysis of the disaster response operation manifested as interlinked social networks, other analysis methods, including snowballing (Uhr, 2009), can be employed. The intention of our (deliber- ately broad) framework is to make it work for events of multiple scales and to allow it to include multiple methods of analysis for the multiple interpretations made. Bigger scale events than the Swedish forest fire that we have analysed would much likely require a great number of system interpretations in order to generate what could be claimed to be a complex understanding.

Perhaps, the most important implication of the framework that we suggest is that of being modest in the claims we make, we hope that also people involved in disaster response (or evaluations of disas- ter response) will approach each other with humble- ness. That is what a complex understanding allows by defying simple statements and solutions often revert- ing to heroes and villains.

References Abrahamsson, M., Hassel, H. and Tehler, H. (2010), ‘Towards a System-Oriented Framework for Analysing and Evaluating Emergency Response’, Journal of Contingencies and Crisis Management, Volume 18, Number 1, pp. 14–25.

Atkinson, S.R. and Moffat, J. (2005), The Agile Organization: From Informal Networks to Complex Effects and Agility. CCRP, Washington, DC.

Boersma, K., Comfort, L., Groenendaal, J. and Wolbers, J. (2014), ‘Editorial: Incident Command Systems: A Dynamic Tension Among Goals, Rules and Practice’, Journal of Contingencies and Crisis Management, Volume 22, Number 1, pp. 1–4.

Brehmer, B. (1992), ‘Dynamic Decision Making: Human Control of Complex Systems’, Acta Psychologica, Volume 81, Number 3, pp. 211–241.

Capra, F. (1982), The Turning Point. Simon & Schuster, New York. Cilliers, P. (1998), Complexity and Postmodernism: Understanding Complex Systems. Routledge, London.

Cilliers, P. (2005), ‘Complexity, Deconstruction and Relativism’, Theory, Culture & Society, Volume 22, Number 5, pp. 255–267.

Comfort, L.K. (1994), ‘Self-Organization in Complex Systems’, Journal of Public Administration Research and Theory, Volume 4, Number 3, pp. 393–410.

Comfort, L.K. (1995), Self Organization in Disaster Response and Recovery: The Maharashtra, India Earthquake of September 30, 1993. FMHI Publications, Natural Hazards Research and Applications Information Center, University of Colorado, Paper 42.

Comfort, L.K. (1999), Shared Risk: Complex Systems in Seismic Response. Pergamon, New York.

Comfort, L.K. (2007), ‘Crisis Management in Hindsight: Cognition, Communication, Coordination and Control’, Public Administration Review, Volume 67, pp. 189–197.

Comfort, L.K., Sungu, Y., Johnson, D. and Dunn, M. (2001), ‘Complex Systems in Crisis: Anticipation and Resilience in Dynamic Environments’, Journal of Contingencies and Crisis Management, Volume 9, Number 3, pp. 144–158.

Comfort, L.K. and Kapucu, N. (2006), ‘Inter-organizational Coordination in Extreme Events: The World Trade Center Attacks, September 11, 2001’, Natural Hazards, Volume 39, Number 2, pp. 309–327.

County Administrative Board of V€astmanland (2014), The Acting of the County Administrative Board. The First Week of the Wildfire 31st of July to 7th of August 2014. (L€ansstyrelsen V€astmanlands l€an. Beskrivning av l€ansstyrelsens agerande. Skogsbrandens f€orsta vecka 31 juli – 7 augusti 2014).

Dekker, S., Bergstr€om, J., Amer-W�ahlin, I. and Cilliers, P. (2013), ‘Complicated, Complex, and Compliant: Best Practice in Obstetrics’, Cognition, Technology & Work, Volume 15, Number 2, pp. 185–195.

Drabek, T.E. (2007), ‘Community Processes: Coordination’, in Rodr�ıgues, H., Quarantelli, E.L. and Dynes, R.R. (eds), Handbook of Disaster Research. Springer, New York, pp. 217–233.

Drabek, T.E. and McEntire, D.A. (2003), ‘Emergent Phenomena and the Sociology of Disaster: Lessons, Trends and Opportunities from the Research Literature’, Disaster Prevention and Management, Volume 12, Number 2, pp. 97–112.

Dynes, R. (1990), Community Emergency Planning: False Assumptions and Inappropriate Analogies. Disaster Research Center, Newark, DE.

Dynes, R. and Aguirre, B.E. (2008), Organizational Adaptation to Crisis: Mechanisms of Coordination and Structural Change Crisis Management,, Vol 2. SAGE Publications, Thousand Oaks, CA, pp. 320–324.

Ekman, O. and Uhr, C. (2015), Global Model for Direction and Coordination in Multi-Actor Crisis Management, 20th ICCRTS.

Fredholm, L. and Uhr, C. (2007), Ledning som inflytanderelationer mellan beslutsfattare. LUCRAM-rapport 1013.

Gunderson, L.H. and Holling, C.S. (2002), Panarchy: Understanding Transformations in Human and Natural Systems. Island Press, Washington, DC.

Healy, S. (2003), ‘Epistemological Pluralism and the ‘Politics of Choice’’, Futures, Volume 35, Number 7, pp. 689–701.

Heylighen, F., Cilliers, P. and Gershenson, C. (2007), ‘Complexity and Philosophy’, in Bogg, J. and Geyer, R. (eds), Complexity, Science and Society. Radcliffe Publishing, Oxford, pp. 117–134.

Hollnagel, E. (2012), FRAM – The Functional Resonance Analysis Method: Modelling Complex Socio-technical Systems. Ashgate, Farnham, UK.

Jensen, J. and Waugh, W.L. (2014), ‘The United States’ Experience with the Incident Command System: What We Think We Know and What We Need to Know More About’, Journal of Contingencies and Crisis Management, Volume 22, Number 1, pp. 5–17.

Kaplan, S. (1997), ‘The Words of Risk Analysis’, Risk Analysis, Volume 17, pp. 407–417.

Kiel, L.D. (1995), ‘Chaos Theory and Disaster Response Management: Lessons for Managing Periods of Extreme Instability’, in Koehler, G.A. (ed), What Disaster Management Can Learn from Chaos Theory. The California Research Bureau, California State Library, Sacramento, CA.

Malone, T.W. and Crowston, K. (1990), What Is Coordination Theory and How Can It Help Design Cooperative Work Systems? Proceedings of the Conference on Computer Supported Cooperative Work, Los Angeles.

Ministry of Justice (2015), Sj€oqvist, A. and Str€omberg, C. Wildfire Accident Investigation Report. (Justitiedepartementet. Rapport fr�an skogsbrandutredningen).

O’Sullivan, T.L., Kuziemsky, C.E., Toal-Sullivan, D. and Corneil, W. (2013), ‘Unraveling the Complexities of Disaster Management: A Framework for Critical Social Infrastructure to Promote Population Health and Resilience’, Social Science & Medicine, Volume 93, pp. 238–246.

Quarantelli, E.L. (1988), ‘Disaster Crisis Management: A Summary of Research Findings’, Journal of Management Studies, Volume 25, Number 4, pp. 373–385.

Rimstad, R., Nj�a, O., Rake, E.L. and Braut, G.S. (2014), ‘Incident Command and Information Flows in a Large-Scale Emergency Operation’, Journal of Contingencies and Crisis Management, Volume 22, Number 1, pp. 29–38.

Scholtens, A., Jorritsma, J. and Helsloot, I. (2014), ‘On the Need for a Paradigm Shift in the Dutch Command and Information System for the Acute Phase of Disasters’, Journal of Contingencies and Crisis Management, Volume 22, Number 1, pp. 39–51.

Seddon, J., (2005). Freedom from Command & Control - Rethinking Management for Lean Service. Productivity Press. NY, US.

Stogdill, R.M. (1974), Handbook of Leadership: A Survey of the Literature. Free Press, New York.

Uhr, C. (2009), Multi-Organizational Emergency Response Management. A Frame- Work for Further Development. Department of Fire Safety Engineering and Systems Safety, Lund University, Lund, Sweden.

Uhr, C. and Johansson, H. (2007), ‘Mapping an Emergency Management Network’, International Journal of Emergency Management, Volume 4, Number 1, pp. 104–118.

Uhr, C., Johansson, H. and Fredholm, L. (2008), ‘Analysing Emergency Response Systems’, Journal of Contingencies and Crisis Management, Volume 16, Number 2, pp. 80–90.

Von Bertalanffy, L. (1950), ‘An Outline of General System Theory’, British Journal for the Philosophy of Science, Volume 1, pp. 139–164.

A framework for a complex understanding 135

Copyright of Journal of Contingencies & Crisis Management is the property of Wiley- Blackwell and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.