Rigid inst.i.tutions (cultural values and more formal rules).

Lack of stakeholder commitment to share information over the long term.

What inst.i.tutional arrangements are best suited to implementing adaptive management?

Evidence of success The use of 'soft' conceptual and qualitative modelling makes it difficult to communicate outcomes.

The boundaries between adaptive management and background processes can be difficult to distinguish.

Methodologies are needed to gather evidence for and communicate the outcomes of adaptive management to stakeholders.

Ambiguity of definition Multiple, ambiguous definitions make it difficult for resource managers to understand how they can apply this approach.

Is ambiguity a potential strength indicating diversity? Refining the typology of approaches a.s.sociating themselves with this adaptive management will help add clarity.

Complexity, costs and risk Experimentation can be ecologically and economically risky.

Adaptive management is slow and planning costs are high compared to centralised management.

An honest dialogue is needed on the appropriateness of concepts from complexity science such as sub-optimality, uncertainty and diversity.

(Source: based on Medema et al., 2008) From this more bottom-up perspective the key challenges for adaptive management and by implication for integrating adaptation into development planning more generally can be identified: * the need for higher level organisations to be receptive to local viewpoints and undertake learning in response, * the challenges of maintaining local engagement over extended time-spans, and * determining and securing the needed level of technical a.s.sistance and science capacity to ensure the validity and credibility of community-led efforts.

Fernandez-Gimenez et al. (2008) also point to the opportunities that adaptation can open. They note that community-led approaches to adaptive management can be a source of local skill training and employment generation in the establishment of an ecological monitoring workforce. These could in part offset or help to justify the financial costs of adaptation in development.

Coping mechanisms.

The notion of coping has acquired a sizable and well developed literature. It describes the strategies used by those living with rapid onset disasters such as flash floods, and chronic disasters, including drought and food insecurity (Wisner et al., 2004). This matches well the dual interests of adaptation to climate extremes and base-line change. Coping has also been used to explore social change in relation to wider impacts of social violence and personal tragedy (Lee et al., 2009). Despite this wealth of knowledge of direct relevance to climate change adaptation, learning has been limited (Schipper and Pelling, 2006). This makes it important to identify what, if any, are the similarities between coping and adaptation, and what adaptation could usefully take from this literature; and also to make clear the boundaries between these two concepts.

Within the natural disasters and food security literature numerous models for coping have been proposed since the 1970s. These have variously been framed by ent.i.tlements (Sen, 1981), human ecology (Hewitt, 1983), game theory (Uphoff, 1993) and livelihoods a.n.a.lysis (Leach et al., 1997). Across these theoretical realms models tend to be agency focused, the majority operating at the household level and to differentiate coping either by stage or sector of action. Burton et al. (1993) is one of the most encompa.s.sing models, connecting slow cultural change with rapid adjustments. This four-stage model commences with loss absorption where hazard impacts are tolerated, absorbed as part of the ongoing coevolution of socio-ecological systems with no tangible impacts or observed, instrumental adjustments. Stage two, loss acceptance, is reached once the negative effects of a hazard are socially perceived but losses are borne without active mediation. The third stage of loss reduction commences once losses are perceived to be higher than costs for mitigation; this is the focus of most disaster reduction work. A final stage of radical change is reached once hazard impacts can no longer be mitigated and major socio-economic changes are experienced either through impact or attempts to minimise disaster loss. This broad view of coping is useful in identifying coping as simultaneously a long-term (cultural) and short-term (economic) process of realignment to changing environmental conditions. This model also flags the importance of perception on action. The implication of a temporal dimension opens the possibility of tipping points where one stage flips into another through changes in vulnerability or hazard.

Alternative categorisations of coping offer typologies of action; for example, Wisner et al. (2004) identify four kinds of coping action: disaster prevention and loss management (for example, hazard mitigation schemes, early warning systems), diversification of production (for example, the promotion of mixed cropping, livelihood diversification), development of social support networks (for example, informal reciprocity or state welfare) and post-disaster actions to contain loss (for example, opportunistic livelihoods, insurance, novel social organisation). This approach has the advantage of providing technical detail but is restricted to Burton et al.'s stage of loss reduction and possibly radical change. While these models are designed to accommodate action at multiple spatial scales they less easily reveal the trade-offs and interactions of coping interacting across scale. Livelihoods models are one response to this challenge and explicitly situate agents (normally households) within an inst.i.tutional context. Coping responses are located at the interface of actors and inst.i.tutions (Leach et al., 1997).

While a successful concept, coping is ultimately misleading as a metaphor for social responses to environmental change at it implies that actors are getting by, doing okay. This can be the case, with agriculturalists, for example, deploying coping mechanisms to get through the low-productivity periods in the annual agricultural cycle (Davies, 1993). But often, acts labelled as coping require the expenditure or conversion of valuable a.s.sets to achieve lower-order outcomes, undermining current capacities and future development options. This ratchet effect (Chambers, 1989) is socially amplified when multiple individuals, households or businesses deploy similar economic strategies selling a.s.sets or changing livelihoods and so undermining market value. Compet.i.tion can turn into collaboration with virtuous magnifier effects through the use of social capital, which can be built up and whose impact can be extended through multiple simultaneous actions. There are, however, limits even to individual and societal stocks of social capital so that continuing environmental stress or repeat shocks can lead to a cascade of failure as social and economic a.s.sets are expended. Figure 2.2 indicates a sequence of coping acts that can lead to collapse as a.s.sets are depleted in the face of unrelieved stress.

Swift (1989) argues that household collapse becomes inevitable once core social and economic a.s.sets are lost and is observed even when macro-economic conditions improve, revealing how individual vulnerability, or capacity to cope, operates with a degree of independence from structural conditions. Households, especially poor households, live with many kinds of risk as well as a desire to fulfil unmet needs and wants. So it is that households have to play off expenditures on immediate household maintenance against investment to recover lost resources or offset antic.i.p.ated risk, and this can make it more difficult to replace savings or productive a.s.sets once they have been expended through coping. The potential for social capital to be undermined through ever more destructive rounds of coping links household collapse to that of collectively held a.s.sets such as social cohesion or notions of community. Commencing with a shift in investment and use from bridging to bonding capital that amplifies cultural difference and compet.i.tive group behaviour (Goodhand et al., 2000), subsequent coping detracts from more fundamental aspects of local social capital through a withdrawal of investment in short-term (health) and long-term (education) social capital, and finally in fragmentation of the most basic social unit the household. As with the economic cascade, cultural contexts will determine the order movement. For Figure 2.2 The coping cascade: coping and erosion of household sustainability (Source: based on Pelling, 2009) example, child sharing is a well-developed coping mechanism in the Caribbean that need not signify approaching household collapse. Here the extended family, not the household, is the basic unit of social organisation (Pelling, 2003b). Broadly, though, as a household approaches collapse subsequent acts are more difficult to reverse.

The delicate balance between the terms coping and climate change adaptation (see Table 2.2), and the negotiation of the intellectual division of labour between them can be found in some early writing on adaptation. Kelly and Adger (2000) define coping capacity as the ability of a unit to respond to an occurrence of harm and to avoid its potential impacts, and adaptive capacity as the ability of a unit to gradually transform its structure, functioning or organisation to survive under hazards threatening its existence. This distinction builds on earlier work; for example, working on food security, Gore (1992, in Davies, 1993) offers a distinction based on the actorinst.i.tution relationships. Coping is the means to survive within the prevailing systems of rules; adaptation is indicated when inst.i.tutions (cultural norms, laws, routine behaviour) and livelihoods change. This distinction is becoming increasingly accepted. Under this rubric an example of coping might be selling cattle during drought, with adaptation signified by migration or a change in livelihood to supplement or replace dependence on livestock. Critics of this division argue that, on the ground, the distinction between coping and adaptation in terms of the depth of consequence for actors is Table 2.2 Distinctions between coping and adaptation.

Coping.

Adaptation.

Source.

The ability of a unit to respond to an occurrence of harm and to avoid its potential impacts The ability of a unit to gradually transform its structure, functioning or organisation to survive under hazards threatening its existence Kelly and Adger (2000) The means to survive within the prevailing systems of rules Change to the inst.i.tutions (cultural norms, laws, routine behaviour) embodied in livelihoods Gore (1992) The range of actions available to respond to the perceived climate change risks in any given policy context Change to the set of available inputs that determine coping capacity Yohe and Tol (2002) The process through which established practices and underlying inst.i.tutions are marshalled when confronted by the impacts of climate change The process through which an actor is able to reflect upon and enact change in those practices and underlying inst.i.tutions that generate root and proximate causes of risk, frame capacity to cope and further rounds of adaptation to climate change Pelling (2010) greatly influenced by the viewpoint of the observer. This blurs the practical utility of the empirical boundaries between coping and adaptation, producing a potential lack of a.n.a.lytical and policy clarity (for example, Saldana-Zorrilla, 2008).

Yohe and Tol (2002) offer a nuance on the distinction between coping and adaptation described above. They see adaptive capacity as describing the set of available inputs that determine coping capacity which itself is manifest in the range of actions available to responding to perceived climate change risks in any given policy context. Adaptive capacity is determined by underlying social factors: resources, inst.i.tutions, social capital, human capital, risk spreading, information management and awareness. Their availability is context specific and path dependent. Coping capacity is defined by the range of practical measures that can be taken to reduce risk. The range, feasibility and efficiency of these measures is determined by adaptive capacity. This logic reveals some insightful outcomes in the relationships between inputs and actions (adaptation and coping). Enhanced investment in the 'weakest link' component of adaptive capacity has the advantage of raising coping capacity across the board or at least until the next weakest link emerges to limit coping. By the same token investing in one component in isolation need not increase coping capacity. Adding to the resource base may, for example, have no effect on coping capacity if inst.i.tutional processes or decision-making structures block implementation.

The distinction being made by these authors reflects other attempts to disentangle distinct relationships between actors and their environment. This helps provide some depth to the more narrowly focused challenge of coping/adaptation in climate change. The interest of Freire (1969) was to make transparent the potential role of education in society much like the climate change problem, his concern was to see development as a process that contained what the poor knew and what they imagined they could do with knowledge. The distinction between 'adapted man' (that is, someone who has learnt to live with the current system) and 'critical consciousness' has parallels with coping and adaptation. Adapted man corresponds with coping where successive rounds of coping, that is, of accommodating one's life to live with hazard, describe well the ratchet effect undermining a.s.sets and human wellbeing. Critical consciousness the ability to see one's position in society as a function of social structures as a prerequisite to seeking ways of making change in those structures has great parallels with the inst.i.tutional dimensions of adaptation described above. The difference is that climate change has to date been driven predominantly by a concern for maintaining efficiency in the output of economic systems and livelihoods rather than in the balance of power between actors or as embodied in inst.i.tutions. Thus the current modes of defining adaptation go only halfway to meet Freire; they acknowledge the action to change inst.i.tutions but do not emphasise the potential for emanc.i.p.ation that this could bring nor indeed that this could be a parallel and even motivating goal for climate change adaptation.

The systems worldview that has had a great influence on recent thinking about human responses to climate change also recognises the potential for more profound change (for example, Flood and Romm, 1996; Pelling et al., 2007b). Argyris and Schon (1996) identified three kinds of learning, termed first, second and third loop learning. Only the first two are encompa.s.sed routinely in the distinction between coping and adaptation in climate change literature. First loop learning corresponds with coping learning to improve what you already do. Second loop learning corresponds with adaptation learning to change the mechanisms used to meet your goals. Third loop learning learning that results from a change in the underlying values that determine goals and actions is less clearly expressed within current adaptation theory.

The lack of emphasis in climate change literature on adaptation as critical consciousness or third loop learning is likely a reflection on the difficulty of making clear empirical a.s.sociations between climate change related impacts and social change of this order. Chapters 5 and 8 aim to provide one step forward in opening this discussion. There is also the possibility that the climate change community which has its eyes tightly focused on the IPCC process, and which in turn is a product of negotiated content between science and governments has not found a.n.a.lysis of power as part of adaptation to be a priority. It risks alienating the political and technical decision-makers for whom the IPCC endeavour is designed to support.

Another area where coping is still a predominant term, and one where further development could prove insightful for work on climate change, is the psychological literature. This work views coping as an interior action determined by the interaction of cognitive and emotional process, but acknowledging interaction with socialised values, access to information and socialhistorical context. Individual ability to cope with stress a.s.sociated with catastrophe has been described as psychological resilience (Walsh, 2002). This literature is most developed in the USA, with Hurricane Katrina stimulating many studies including Lee et al. (2009) who identified psychological resilience as an outcome of survivors' perseverance, ability to work through emerging difficulties and ability to maintain an optimistic view of recovery. Amongst this group those who suffered human loss were least able to cope, with property loss having only a minor impact on capacity for psychological recovery. Other hurricane events in the US have shown that survivors who reported more resource loss also reported higher levels of active and risk-reducing behaviour (Benight et al., 1999). This has important implications for the appropriateness of mainstream methodologies for measuring disaster impact and for disaster response and recovery efforts which predominantly focus on economic and physical rather than social and psychological aspects.

Psychology has begun to offer some insight into the factors leading to individual wellbeing and empowerment post-disaster, although the link to material coping actions is as yet less well defined. Psychological traits a.s.sociated with coping following Hurricane Katrina included a heightened sense of control over one's destiny and of personal growth. These in turn were attributed to survivors who were problem-focused, accepting of loss, optimistic and held a religious worldview (Linley and Joseph, 2004). In the general population talking, staying informed and praying enabled coping, emerging as predictors of decreased psychological stress during post-disaster relocation (Spence et al., 2007), with spirituality particularly significant for older African American Katrina evacuees (Lawson and Thomas, 2007). In a comparison of psychological resilience pre-and post-Katrina, Kessler et al. (2006) found reduced thoughts of suicide after the disaster in survivors expressing faith in their ability to rebuild their life and a realisation of inner strength. This is important in providing an empirical link for adaptation, between internal processes of belief, ident.i.ty and self-worth and external actions, in this sad case ill.u.s.trated through suicide rates. Outside the US, following the 2003 earthquake in Guatemala, feelings of self-control and self-a.s.surance were also found a.s.sociated with adaptation outcomes of 'successful survivors' who reconceptualised the crises as opportunities for acquiring new skills (Vazquez et al., 2005). This work provides one approach for promoting a progressive response to climate change through acknowledging the interplay of social and psychological root causes (Moos, 2002), but this has yet to be systematically applied (Zamani et al., 2006). It provides an initial evidence base to begin a characterisation of specific psychological orientations a.s.sociated with adaptation and linking interior and exterior expressions of adaptation, taking us closer to gaining some leverage on the ways in which individuals and social collectives might move between different cognitive, emotional and potentially intellectual states; the latter opening scope for the study of shifts between 'adaptive man' and critical consciousness or first, second and third loop learning.

In order to incorporate deeper levels of change while retaining close links to the existing literature adaptation to climate change is defined here as: The process through which an actor is able to reflect upon and enact change in root and proximate causes of risk.

This formulation sees coping as the range of actions currently being enacted in response to a specific hazard context. These are made possible by existing coping capacity (which may extend beyond the range of coping acts observed at any one time). Adaptation describes the process of reflection and potentially of material change in the structures, values and behaviours that constrain coping capacity and its translation into action. Coping then is an expression of past rounds of adaptation. Both adaptation and coping will unfold simultaneously and continuously in shaping humanenvironment relations, they will interact and on the ground they may be hard to separate as reflection and application occur hand-in-hand. Still, from an a.n.a.lytical perspective and for policy formulation there is a value in distinguishing these two components of humanenvironment relations.

The coproduction of vulnerability/security by coping and adaptation brings the possibility that adapting to climate change can undermine as well as strengthen capacities and actions directed at coping with contemporary climate related risks. Coping may be limited for longer-term gain or a result of ignorance or injustice in the implementation of adaptation. This can be seen in the loss of income accepted by low-income families who are able to provide an education for their children. This is an adaptive action that constrains contemporary coping capacity, but with the aim of providing future gains that will provide the means for better family wellbeing including capacity to cope with uncertainty and shocks a.s.sociated with the climate change. More likely, the immediacy of political life will produce a tendency for coping that distracts from or undermines the critical reflection and long-term view of adaptation. The danger is that coping is felt to be sufficient so that the potentially difficult questions and changes in development that adaptation might bring are temporarily evaded. At the scale of large social systems, this tension is ill.u.s.trated by the trade-off between short-term social disruption and the long-term easing of socio-ecological friction proposed by Handmer and Dovers (1996) (see below).

Adaptation as a contemporary development concern.

The preceding discussion on the antecedents of adaptation reveals the framing behind contemporary understandings of adaptation. This is not always explicitly acknowledged in the climate change literature but can be felt, for example, in the pervasive influence of systems thinking. Systems theory has had a far-reaching influence with its promise of providing a mechanism to integrate the social and natural. It is used in cybernetics, adaptive management and to a lesser extent in coevolution as well as in contemporary adaptation studies, particularly through work on resilience (Folke, 2006). The aspect of adaptation given prominence in each application reflects fashions in social scientific research as much as the underlying use of systems theory in each case. Cybernetics, developed at a time when positivism was seen as providing new scope for generalisable theory, sought to apply a value neutral, technical epistemology. It is reductive, opening scope for mathematical modelling of behaviour but not able to incorporate the significance of competing values and power asymmetries in shaping action. Adaptive management acknowledges the role of difference in access to information and decision-making capacity in shaping adaptive processes and outcomes, but does not have power as a focus of a.n.a.lysis; like cybernetics the focus is on technical aspects but in this case with a view to informing policy learning. Coevolution orients adaptation less towards the search for ways in which to manage risk and change and is more interested in adaptation as a process, a state of living with uncertainty. It stands back from technical and management a.n.a.lysis to examine the bigger picture of historical change where contesting values are included as a driver for change alongside knowledge, technology, organisational forms and the natural environment. Coping is the outlier in offering a legacy for adaptation that is grounded not in systems theory but in development studies. Connections between nature and society are context specific and hard to generalise from, although a common language has been developed through work on vulnerability (partly originated as a critique of the cybernetic school) that acknowledges both the roots of coping in political-economy but also the influence of values and social viewpoint in shaping decisions and options for adaptation. These four approaches highlight a tension in understandings of adaptation which persists today. This is between policy friendly but reductive a.n.a.lysis on the one hand, and holistic, value sensitive and critical but potentially unwieldy work on the other.

The antecedents also offer guidance on the qualities that promote adaptive capacity. These include parsimony (that the best adaptive choice is that which expends least resource); flexibility; diversity; monitoring to facilitate appropriate change (as distinct from managing to maintain stasis); learning as a facet of policy systems and organisations as well as individuals; and a realisation that observed adaptation, while a positive attribute, is also a sign of stress and a play-off that can signify approaching collapse and reduced wellbeing. These ideas have been taken up by resilience thinking and have a strong influence on contemporary framings of adaptation (see below). They also set adaptation apart from other logics for a.s.sessing development, perhaps most important that of economic maximisation, a cornerstone of economic globalisation. This argues economies should invest in what they do best, leading to a concentration of a.s.sets and closing off options for diversity and flexibility in the productive sectors (Pelling and Uitto, 2001).

The aim of this section is to examine the contemporary conceptualisation of adaptation in detail. We review a typology of adaptation, discuss the influence of resilience on the conceptualisation of adaptation and the significance of social thresholds as tipping points for adaptive change, and compare economic and ethical frameworks for evaluating adaptive choices. This sets the context for the proposal of the three adaptation pathways resilience, transition and transformation that are then developed throughout the remaining chapters.

A typology of adaptation.

Following the technocentric bias of its antecedents, much of the early work on adaptation was theorised as a technical act of adjusting economic or other functions to a changing external environment. This bias has gradually been eroded. An important literature in this regard has been that focusing on adaptation in developing country contexts (Adger et al., 2003; Nelson et al., 2007) including urban (Satterthwaite et al., 2009) and rural (Tanner and Mitch.e.l.l, 2008) contexts. Contributions have also been made from work demonstrating the need to include values, feelings and emotions in decision-making (O'Brien, 2009).

As summarised in Chapter 1, a sizable and fundamental literature on adaptation is directed towards differentiating adaptations (see Smit et al., 2000; Smit and Pilifosova, 2001). Table 2.3 presents a typology of adaptation to be taken forward in this framework, and also distinguishes between the impacts of different adaptive actions. These include actions that respond to perceived positive as well as negative impacts of climate change; those that are felt directly (heat events), indirectly (the price of food or water) or through perturbations in socio-ecological systems (political instability). They are acts unfolding within many sectors (urban planning, water management, agriculture development, transport planning and so on) and using a range of vehicles (technical innovation, legislative reform, market adjustment, professional training, behavioural change). They describe both the nature of an adaptive action and its scope of impact.

Table 2.3 A typology of adaptation.

Criteria.

Options.

Nature of Adaptive Action.

Degree of collaboration.

Degree of focus.

Degree of forethought.

Phasing.

individual or collective.

purposeful or incidental.

spontaneous or planned.

proactive or reactive.

Scope of Impact..

Target.

Timescale.

Future wellbeing.

Social consequences.

Developmental orientation.

proximate, intermediary or root causes of risk.

immediate or delayed.

climate-proofing or maladaptation.

regressive or progressive.

autonomous or integrated.

Adaptation is purposeful when directed towards a recognised hazard or opportunity (retro-fitting of a building) and incidental when undertaken in response to some other pressure that has consequences for exposure, susceptibility or adaptive capacity (economic opportunities leading to migration out of a flood-p.r.o.ne location). Proactive adaptation is that which takes place before a risk manifests into hazard (disaster risk reduction); reactive adaptation takes place during or after an event (disaster reconstruction). The scope of adaptive action can be distinguished between that which seeks to change material a.s.sets or practices set against less direct inst.i.tutional change (see Pelling and High, 2005). This is reflected in the potential targets of climate change which may be proximate (crop variety), intermediary (local decision-making systems) or root causes (politicaleconomic structures and development visions). Timescale acknowledges that adaptation can have immediate (changing built forms) or delayed (investing in health and education) benefits. The impacts of adaptation on the future wellbeing of others are indicated by acts that could be termed as climate proofing (the integration of mitigation) or maladaptation (adaptation that increases vulnerability); socially regressive or progressive depending on redistributive consequences, and autonomous to (isolated and contained) or integrated in (undertaken with awareness of and aiming at synergies with the actions of others) development.

Resilience and adaptation.

Resilience is popularly understood as the degree of elasticity in a system, its ability to rebound or bounce back after experiencing some stress or shock. It is indicated by the degree of flexibility and persistence of particular functions. That resilience is not simply synonymous with adaptation has been well demonstrated by Walker et al. (2006a) who argue that adaptation can undermine resilience when adaptation in one location or sector undermines resilience elsewhere, where management focus on a known risk distracts attention from emergent hazard and vulnerability, and that increased efficiency in adaptation (through risk management, for example) can lead to inst.i.tutional or infrastructural inertia and loss of resilient flexibility.

Resilience has been contrasted both with stability and vulnerability. Stability, according to Holling (1973), is an attribute of systems that return to a state of equilibrium after a disturbance. This compares with resilient systems that might be quite unstable and undergo ongoing fluctuation but still persist. Stability is more desirable in circ.u.mstances where environmental perturbations are mild; resilience is most useful as an attribute of systems living with extremes of impact and unpredictability. Within the disaster risk community, resilience has been interpreted as the opposite of vulnerability. The more resilient, the less vulnerable. But this belies the complexity of the conceptual relationship between these terms which have also been constructed as nested with vulnerability being shaped by resilience (Manyena, 2006) which for some in turn incorporates adaptive capacity (Gallopin, 2006). Stability and vulnerability provide useful bounding concepts for resilience. They suggest that resilience is about the potential for flexibility to reduce vulnerability and allow specific functions to persist. What it does not tell us is how these functions are identified or who decides (Lebel et al., 2006). This requires a more critical engagement with social processes shaping resilience (see Chapters 3, 6, 7 and 8).

Working with the idea of resilience, and especially efforts that seek to measure it are made difficult because of its multifaceted character. The processes and pressures determining resilience for a unit of a.s.sessment change with spatial, temporal and social scale a community may be resilient to climate change a.s.sociated hurricane risk (through early warning and evacuation, for example) but less resilient to the long-term inflections of climate change with the local and global economy. The subjects of a.n.a.lysis are also wide, bringing diversity but also fragmentation to the study of resilience. Cutter et al. (2008) identify studies attributing resilience and related metrics to ecological systems (biodiversity), social systems (social networks), economic systems (wealth generation), inst.i.tutional systems (partic.i.p.ation), infrastructure systems (design standards) and community competence (risk perception) (Folke, 2006; Paton and Johnston, 2006; Rose, 2004; Perrow, 1999; Vale and Campanella, 2005).

One of the first critical engagements with resilience from the perspective of environmental risk management came from Handmer and Dovers' (1996) proposal of a three-way cla.s.sification of resilience. This insightful framework has echoes of Burton et al.'s (1993) cla.s.sification for coping and still offers a great deal. It highlights both the contested and context specific character of adaptation that this book argues for, and is worth describing in some detail. The three-way cla.s.sification presented resilience as: (1) resistance and maintenance; (2) change at the margins; and (3) openness and adaptability.

Resistance and maintenance is commonplace, particularly within authoritarian political contexts where access to information is controlled. It is characterised by resistance to change; actors may deny a risk exists with resources being invested to maintain the status quo and support existing authorities in power. When risk is undeniable these systems typically delay action through a call for greater scientific research before action is possible. Vulnerability can be held at bay by resource expenditure; for example, in food aid or through containing local hazard risk through hard engineering 'solutions'. But this can generate additional risks for other places and times through global flows of energy, resources and waste. This type of resilience offers an easy path for risk management, there is little threat to the status quo and considerable stress could be absorbed. However, when overcome the system would be threatened with almost complete collapse Diamond's (2005) thesis on the collapse of ancient civilisations reminds us of this possibility.

Change at the margins is perhaps the most common response to environmental threat. Risk is acknowledged and adaptations undertaken, but limited to those that do not threaten core attributes of the dominant system. They respond to symptoms, not root causes. Advocates argue that this form of resilience offers an incremental reform, but it is as or more likely to delay more major reforms by offering a false sense of security. Preference for near-term stability over radical reform for the wellbeing of future generations provides a strong incentive for this form of resilience. This approach is well ill.u.s.trated by the Hyogo Framework for Action on Disaster Risk Management, which sets forth an international agenda agreed by nations for managing disaster risks including those a.s.sociated with climatic extremes. Not surprisingly given the vested interests of dominant voices in the international community for the status quo, the framework is limited. It calls for the integration of risk management policy into development frameworks, the increasing of local capacity for risk reduction and response, and for new systems of disaster risk identification and information management (ISDR, 2005).

Social systems displaying openness and adaptability tackle the root causes of risk, are flexible and prepared to change direction rather than resist change in the face of uncertainty. That this mode of resilience is so rare is testament to the huge inertia the results from personal and collective investment in the status quo. Large fixed capital investments make change difficult as do investments in soft infrastructure preferences for certain types of education or cultural values making shifts painful in industrial societies. Dangers also lie with this form of resilience: instability will lead to some ineffective decisions and maladaptation would need to be prepared for within individual sectors as a cost of wider systems flexibility. These are both worries that decision-makers have cited in making it difficult for them to commit to adaptive management strategies, as described above.

Handmer and Dovers prefigure their account by a caution that while the three cla.s.sifications are designed to cover the full range of policy responses to the adaptation challenge, most actors will operate in only a small part of this range. This points to a central dilemma for progressive adaptation that the comfort zone for adaptive action is relatively small because both those with power and the marginalised are wary of the instability they fear from significant social change (see Chapter 5). Resilience then has the possibility of both identifying the scope for flexibility within the socially accepted bounds of stability but also making transparent for all social observers the range of choices foregone. Mapping the characteristics of social systems that are more or less amenable to these three forms of resilience is a key foundation for the a.n.a.lytical framework development in this book which places emphasis on the processes through which systems undertake or resist adaptive change.

More contemporary work on resilience and its relationships with vulnerability and adaptation have also applied critical reasoning. This has focused on the advantages of inclusive governance. This, it is argued, facilitates better flexibility and provides additional benefit from the decentralisation of power. On the down side, greater partic.i.p.ation can lead to loose inst.i.tutional arrangements that may be captured and distorted by existing vested interests (Adger et al., 2005b; Plummer and Armitage, 2007). Still, the balance of argument (and existing centrality of inst.i.tutional arrangements) calls for a greater emphasis to be placed on the inclusion of local and lay voices and of diverse stakeholders in shaping agendas for resilience through adaptation and adaptive management (Nelson et al., 2007). This is needed both to raise the political and policy profile of our current sustainability crisis and to search for fair and legitimate responses. Greater inclusiveness in decision-making can help to add richness and value to governance systems in contrast to the current dominant approaches which tend to emphasise management control. When inevitable failures occur and disasters materialise this approach risks the undermining of legitimacy and public engagement in collective efforts to change practices and reduce risk. This takes us back to Handmer and Dovers' (1996) a.n.a.lysis of the problem of resilience and shows just how little distance has been travelled in the intervening years.

Adaptation thresholds.

Acts of adaptation are stimulated by the crossing of risk, hazard and/or vulnerability thresholds. Each threshold is socially constructed, a product of intervening properties including identification, information and communication systems, political and cultural context and the relative, perceived importance of other risks, hazards and vulnerabilities that compete for attention. The existence of social thresholds explains the 'lumpiness' of human experience, where history does not unfold as a gradual story but in fits and starts. Forward looking adaptation, or the impacts of climate change resulting from a lack of sufficient adaptation, may be catalysts for the breaching of thresholds.

Risk is ever present in society. The level of risk that is accepted by different social actors determines the first threshold (see discussion on coping) and is shaped by whose values and visions for the future count in society (Adger et al., 2009a). For any social group the level of acceptable risk can change as scientific innovation, media interest and public education influence awareness amongst the public and decision-makers. Communication between science, decision-makers, the media and the public is determined by norms of trust. Trust is built over time by the everyday performance of scientific or government bodies but is easy to lose (Slovic, 1999). Where there is a confidence gap in advisory bodies, the government or science, popular regard for new risk announcements will be greeted with scepticism (Kasperson et al., 2005), with a preference for self-reliance or fatalism amongst those at risk and potentially resistance to any coordinated adaptation. It is here that dedicated intermediary organisations or individuals that can translate climate science into the language of target audiences (such as agricultural extension agencies) play a significant role in shaping people's willingness to reduce risk (Huq, 2008). Indeed part of the challenge facing adaptation to climate change is the need to communicate without confusing, and the science community that has championed climate change research thus far has not found this easy (Hulme, 2009).

Climate change is felt locally through many environmental indicators. Figure 2.3 represents how just one say precipitation is influenced by climate change and how this is related to the timing and scope of coping and adaptation. In this case climate change produces reduced hazardousness at minimum extremes (drought) but increased hazardousness at maximum values (flood). In this way new hazard thresholds challenge existing hazard management strategies which are breached until the changing hazard threshold is recognised (E1) and responded to (E2). The distance between these two points reflects the risk acceptance and communication thresholds described above. A final threshold that determines adaptation comes from changing vulnerability profiles.

Demographic and economic change in particular influence the likelihood of adaptation. This is often not integrated into accounts of adaptive capacity and action (see Figure 2.3) but is particularly important in rapidly changing contexts such as rapid urbanisation, economic restructuring or where social tensions might lead to armed violence. The vulnerability threshold suggests there is a critical ma.s.s of a.s.sets or people at risk and of risk management capacity that are needed for adaptation to be likely. This also has consequences for the kind of adaptation undertaken. Thus a small coastal settlement may undertake independent, spontaneous adaptations to protect livelihoods in the face of sea level rise, but should this area be subject to investment by the corporate tourism sector and subsequent high levels of labour in-migration adaptation may become more coordinated, collective and planned.

Figure 2.3, although stylised, is useful in demonstrating several other attributes of adaptation (Fussel, 2007). It shows the disproportionate ability of extreme over average climatic conditions to stimulate adaptation, the need to consider natural climatic variability and anthropocentric climate change together in planning adaptations, and the continuous process of review and response needed of adaptation to climate change as hazard thresholds change (E3). The fuzzyness inherent Figure 2.3 Adaptation thresholds.

(Source: based on Fussel, 2007).

in labelling adaptation as reactive or proactive is revealed with the decision to adapt being both a reaction to the preceding extreme event and a proactive antic.i.p.ation of future risk. A reactive motivation can lead to a proactive adaptation. This said, the time needed to make decisions to adapt to climate change and complete adaptive measures such as major infrastructure works or the reform of housing stock is often several years if not decades so that incremental adaptation may be dangerous and costly (Reeder et al., 2009). In contrast, planning over extended timeframes opens decision-making to uncertainty. As the limits of scientific knowledge are reached so decisions are based increasingly on value judgements. These in turn are shaped by the structures and norms of governance systems and culturalhistorical expressions of acceptable risk that inform and legitimate adaptation (Paavola and Adger, 2006). Ultimately this directs scrutiny to questions about who it is that determines the principles upon which adaptive choices are made as much as the nature of the decisions themselves.

Evaluating adaptive choices: economics and ethics.