Previous - Next
This is the old United Nations University website. Visit the new site at http://unu.edu
An integrative and systematic approach to environmental criticality must not only view a particular human-environment situation within historical and spatial context but also examine attributes that have been identified as central to such situations. These attributes include sensitivity, fragility, resilience, and vulnerability; the nature of environmental change; threats and disaster; life-support systems; intergenerational export of damage; and societal response systems. From the reviews of these attributes we draw a number of lessons regarding the conceptualization and identification of environmental criticality.
Sensitivity, fragility, resilience, and vulnerability
Key considerations in criticality include the extent to which environmental stresses damage ecosystems and reduce their ability to maintain their basic structures under stress and recover from such damage. Similar considerations of the characteristics of social systems, groups, and individuals add to the set of processes that contribute to differential susceptibility to environmental changes. Indeed, it has long been recognized that environmental hazards are the joint product of stress and exposure on the one hand and fragility and vulnerability on the other. An array of often overlapping or conflicting terms sensitivity, resistance, resilience, marginality, fragility, vulnerability has arisen to address the latter term of this product.
The science of ecology, as Turner and Benjamin (1994) have pointed out, lacks a precise and accepted definition of ecosystem fragility. Ecologists (and some others) use "fragility" to denote a "system at risk," where the notion of "risk" is linked to ecological concepts of stability, instability, and susceptibility (e.g. Yodzis 1980, 545). The chain of linkages characterizing the system provides the ecological base for the notion of fragility. This complex internal structure of the ecological system, with its various pathways and differing sensitivities of species at risk, is receiving detailed analysis in the emerging field of ecological risk assessment (Travis and Morris 1992). Whether the notion of "system" or "land system" will ultimately prove too crude to capture the essential character of change and damage must await the results of these efforts by ecologists and ecological risk analysts.
According to Holling (1986, 296), "stability" is "the propensity of a system to attain or retain an equilibrium condition of steady state or stable oscillation," whereas "resilience" is "the ability of a system to maintain its structure and patterns of behavior in the face of disturbance." Increasingly, however, equilibrium-centred notions appear to be under challenge among ecologists, and linear views of stability and recovery are giving way to notions stressing the cyclic or irregular nature of ecosystem change and the spatio-temporal heterogeneity of landscapes and ecosystems (Turner 1987). Our approach to criticality is consistent with these emerging views.
"Sensitivity" is often used to measure the magnitude of negative impacts of environmental change. Change can have positive and negative impacts on human societies. Here the emphasis is on the degree of stress placed on the land or environmental system and on the identification of the important relationships and processes. Blaikie and Brookfield (1987) use the terms "sensitivity" and "resilience" to describe the quality of land systems. They use "sensitivity" to refer to "the degree to which a given land system undergoes changes due to natural forces, following human interference" and "resilience" to refer to "the ability of land to reproduce its capability after interference, and the measure of the need for human artifice to that end" (Blaikie and Brookfield 1987, 10). For its part, the concept of "fragility" combines two dimensions: the capacity to be wounded by a particular environmental perturbation (either nature- or human-induced), and the ability to maintain structure and essential functions and to recover.
From these discussions, we adopt the following definitions:
- sensitivity (or susceptibility): the degree of ecosystem or ecosystem component change associated with a given degree of human-induced stress;
- resilience: the ability of a particular ecosystem to maintain the basic structure essential to support human uses during perturbations and to recover from such (and especially damaging) changes.
"Fragility" reflects both of these basic properties of human-ecosystem interactions. In conceptions of criticality, we are particularly interested in human-induced stresses that erode irreversibly the ability of the ecosystem to support human productive systems or life itself. Accordingly, we adopt the following definition:
- fragility: the sensitivity of a particular ecosystem to human-induced perturbations and its resilience to such perturbations.
As Turner and Benjamin (1994) point out, these definitions indicate that four quadrants of fragility can be distinguished that represent different risk situations and different "buffering" to stresses (fig. 1.1). Some environments degrade rapidly but maintain themselves under stress and recover; others are sensitive to change and have low capabilities for maintenance and recovery; and so on.
Obviously, these different attributes are highly relevant to analysis of criticality in the face of human-induced environmental stresses. The physical properties of the environmental base matter. Fragile environments degrade more readily under mismanagement and exact higher societal costs for management or for substitution. More robust environments may not degrade so rapidly and may respond to substitutes more economically. It should be noted, however, that less fragile environments are almost invariably those under the most human stress and hence most subject to mismanagement.
Discussions of the differential susceptibility of social groups and individuals to losses from environmental changes have used diverse nomenclature, often paralleling discussions of ecosystems. Although "vulnerability" frequently refers to the differential susceptibility of both social and biophysical systems, it appears to be emerging as the most common term in the former context (Chambers 1989; Dow 1992; Downing 1991; Gleick 1990; Green 1990; Liverman 1990; Susman, O'Keefe, and Wisner 1983). At an abstract level, the social and ecological dimensions share concerns with the level of impact of a change and the ability to cope with the stress and to continue functioning. Throughout this volume, "vulnerability" refers mainly to the social dimension.
Vulnerability in this sense is a product of three dimensions: exposure, resistance (the ability to withstand impacts), and resilience (the ability to maintain basic structures and to recover from losses). These dimensions, slightly different from those applied above to ecosystems, incorporate the costs of recovery and the role of social relations and decision-making at a variety of social levels, from nation to household, in determining conditions of exposure as well as exposure itself. The distinction is an important one in relation to human activities leading to environmental change as well as the variety of adjustments or coping measures adopted in the face of environmental changes. For instance, the knowledge and resources that people bring to a new farming enterprise will affect the outcomes of that enterprise in that environment. Successful adjustments of farming techniques throughout a season can mitigate the impacts of unpredictable climatic conditions. The level of support that society gives to these farmers will also make a difference in the losses they experience in the case of poor harvest due to drought or declining land capability.
Green (1990, 14), in reference to flood hazards, expresses vulnerability as a relationship between changing conditions and normal use patterns. He defines vulnerability as a function of "susceptibility (the extent to which the presence of water will affect inputs or outputs of an activity); dependency (the degree to which an activity requires a particular good as an input to function normally); and transferability (the ability of an activity to respond by deferring demand, using substitutes or relocating)." Expanded to consider a broader set of relationships between nature and society, these characteristics become important dimensions in discussing the consequences of environmental changes.
Researchers have identified a broad variety of factors, including social relations (particularly race, ethnicity, class, and gender), institutional characteristics, demographic attributes (such as age and reproductive status), individual decision-making and perception, types of technology employed, and political-economic relations that may well contribute to vulnerability (Cannon 1994; Downing 1991; Liverman 1990). The scale of analysis differs among these factors; the appropriate scale for the analysis of vulnerability depends to a great extent on the type of event and the question of concern. The vulnerability of a country to declining agricultural productivity in a region will differ from that of farming households in the region.
The character of change
Discussions of environmental criticality often express particular concern with irreversible change. Our Common Future, for example, opens with a world of environmental stresses, the result of increasing interconnectedness in which "many regions face risks of irreversible damage to the human environment that threaten the basis for human progress" (WCED 1987, 27). The book usually specifies particular irreversible losses deemed as critical, such as species or particular vegetation systems that are unlikely to reappear, once lost or changed, regardless of socio-economic or technical response. Irreversibility is also linked to concerns over sustainable development through definitions that speak of societies that satisfy their needs without reducing the opportunities for future generations to satisfy theirs. Concerns with irreversible change or sustainability are often vague in specifying exactly what should be preserved for future generations, but quality of life, raw materials needed for innovation, and the store of capital assets and natural resources are commonly cited.
Rates of change are recognized as a second key aspect. Indeed, the definition of particular environmental changes as problems often rests less upon the magnitude than on the acceleration of change, as in concerns about exponential increases in Amazonian deforestation or the rapidity with which new chemical substances are introduced into the environment. In particular, the judged relationship between relative rates of environmental change and the estimated time needed for society to respond through prevention, adaptation, or adjustment has gained recognition as an important variable.
It is increasingly apparent that many environmental changes may be discontinuous. Threshold effects, non-linear relationships, surprises, and chaotic readjustments may thus be typical of society-environment relationships. Not only does this possibility have far-reaching implications for the construction of knowledge bases and assessment methodologies, but it suggests that human institutions and management strategies that assume linear change will face intrinsic difficulties in anticipating and interpreting environmental change. It also suggests that causality will often be elusive and that miscalculations and surprises will flourish. The recognition that essential parts of needed knowledge are unknowable and that threat situations involve moral and ethical issues has led Funtowicz and Ravetz (forthcoming) to argue for a "second order" science.
Hence not only the simple magnitude but other aspects of change must enter into concepts of environmental criticality. The reversibility of changes, their rates, and their possibly discontinuous and nonlinear character deserve particular attention, whereas past and possible future trajectories of change as well as magnitudes of change warrant consideration.
Threats and disaster
Notions of criticality embody assessments, often implicit, of losses and harm arising from alterations in nature-society relationships. Inevitably' such notions cannot be restricted to disruptions in productive systems or economic loss but involve broader effects on human values about nature. This is consistent with the definition of a "hazard" as a "threat to people and what they value" (Kales, Hohenemser, and Kasperson 1985, 21).
Societal capacities to respond and subjective human interpretation of events are important considerations in the regional context. Quarantelli (1985, 1987), for example, views disasters as a subset of studies addressing social crises (where crisis is assumed to refer to an unstable condition, a crucial turning point in the life-course of an individual, family, or nation). Thus disasters, whether the result of natural or of social events, are primarily social phenomena and can be assessed only in terms of society's capabilities to respond. Other researchers seek an intermediate ground and a more interactive definition. For instance, natural hazards are frequently described in terms of their duration, scope of impact, cause, and speed of onset all measures that affect society's ability to predict, control, and give warnings (Wenger 1978). The ecologic-symbolic approach of Kroll-Smith and Couch (1990) conceptualizes disasters in terms of society-environment relationships.
Sociological research offers a further perspective on the effects of environmental change. An accepted list of needed social functions might include: (1) production-distribution-consumption; (2) socialization; (3) social participation; (4) social control; and (5) mutual support. One well-known definition of disaster is "an event, concentrated in time and space, in which a society or a relatively self-sufficient subdivision of a society undergoes severe danger and incurs such losses to its members and physical appurtenances that the social structure is disrupted and the fulfillment of all or some of essential functions of the society is prevented" (Fritz 1961, 655).
These perspectives introduce several considerations useful in conceptualizing environmental criticality. First, both changes to the environment and changes to society require assessment, particularly for establishing causal relationships. The degree of environmental change or social disruption will vary among events and among communities. Although some of the contributing factors will be environmental, others will be social, so that attributing a disaster to either environmental or social causes alone may be problematic. Relatively minor environmental changes can easily become disasters under unusually adverse social and economic circumstances. In short, environmental hazards are interactive phenomena.
Life-support systems
Among the varied threats to the environment, the most central (and most relevant to notions of criticality) is endangerment of the ability to sustain human life over the long term. Although no general agreement exists on this subject, discussions have treated such matters as basic functions, basic needs, and human well-being. These may be clustered under the general heading of "life-support systems," so long as that term is understood broadly to incorporate the environmental resources that sustain the economy as well as those - such as water and air - that support life in an immediate sense.
Some previous investigations of criticality have sought to assess the effects of environmental degradation on environmental life-support systems. A primary focus of the Soviet red-zone maps has been the long-term and potentially irreversible impacts on what are described as "life-support capabilities" (Masher and Sdasyuk 1991). A concern with "life-support" functions of the environment was also central to the arguments of Our Common Future (WCED 1987) and pervaded much of the rhetoric at the Earth Summit in Rio de Janeiro. Lacking a rigorous definition and means of measurement, however, this notion has thus far proven difficult to use in rigorous assessment, yet the concept strikes at an essential quality of human-environment interactions.
E. P. Odum (1989, 13) defines the "life-support environment" as "that part of the earth that provides physiological necessities of life, namely, food and other energy mineral nutrients, air and water," and a "life-support system" as the "functional term for the environment, organisms, processes, and resources interacting to provide these physical necessities." He sees these systems as under stress from pollution, poor management practices, and population pressure, with early warnings apparent in the erosion of prime agricultural soils and the death of trees in industrialized countries.
Social scientists have adapted and expanded Odum's fundamentally physiological definition of life-support systems. The "necessities of life" to which Odum refers have received extensive treatment in the social sciences under such umbrella terms as "basic human needs" - generally the likes of health, food, water, education, shelter, and sanitation (Haq and Burki 1980; World Bank 1980). Non-material attributes such as participation, political rights, cultural identity, and a sense of purpose in life are sometimes cited as basic needs (Stewart 1985; Streeten 1980). For others, basic needs are important largely because of their contributions to more fundamental aspects of human life, health, and education, and these goals are viewed as the primary objectives (Stewart 1985). The World Commission on Environment and Development argues that sustainable development requires meeting basic needs, however defined, and offering all an opportunity to achieve their aspirations for a better life (WCED 1987, 44).
Others would argue for "well-being" or "quality of life" as the appropriate term for the health of nature-society relationships. Appraisals addressing these concepts are unavoidably subjective (at least in part), embedded as they are in cultural values and perspectives. If endangerment is viewed as threats to human well-being or quality of life, both material and value-related attributes will need to be considered. Income, health, nutrition, and literacy are common such indicators. Economic growth measures, such as per capita GNP or GDP, are sometimes used as indicators of human well-being because of their general relationship with other desirable factors. They also are often criticized, however, because they fail to address issues or negative impacts of economic growth and do not accurately reflect such considerations as the importance of informal and unpaid work and costs not accounted for by market transactions. broader and more integrative measure is the well-established "international human suffering index" (Population Crisis Committee 1987, 1992), which taps 10 measures: life expectancy, daily calorie supply, clean drinking water, infant immunization, secondary school enrollment, GNP per capita, rate of inflation, communications technology, political freedoms, and civil rights. The "human development index" (HDI) developed by the United Nations Development Programme combines indicators of life expectancy, educational attainment, and income to produce a composite measure of human development at both national and regional scales. Although the HDI has drawn fire for its ranking of countries, it is a useful tool, whose strength lies in its broad-based perspective: "Human development concerns all activities - from production processes, to institutional changes, to policy dialogues" (UNDP 1992, 13).
Assessments of these sorts need to address not only what is threatened by environmental degradation but also who. The most sensitive, vulnerable, marginal, or fragile typically experience impacts well before others and suffer greater harm. It is not unusual that risks are concentrated in particular regions or social groups. Notions of endangerment must discriminate and assess effects at different scales, with attention to variations by ethnicity, class, and gender. It is often these most vulnerable groups who are driven to erode their own (and their children's and grandchildren's) environmental life-support systems.
Basic measures of needs and quality of life assess the life-support capacities of the nature-society relationship from the social side. Beginning on the physical side, researchers have used various measures to indicate the degree of change in the environment or a consequent decline in productivity in order to identify areas whose life-support systems are threatened. The rangeland classification scheme of the US Bureau of Land Management ranks the quality of rangelands according to the degree of departure of the present vegetation from the ecological potential of the area (BLM 1979). The variables measured include species composition, production, and ground cover. Similarly, classifications of desertification have been used to identify and to map the most degraded areas (Berry and Ford 1977; Reining 1978; UNSO 1992). Dregne (1983, 5), for example, maps global desertification according to four categories - very severe, severe, moderate, and slight - based on four indicators: plant cover as a measure of rangeland condition; erosion (by type of erosion rather than measures of soil loss); amount of salinization or waterlogging; and the percentage reduction in crop yields. According to this definition and these indicators, very severe instances of desertification are those in which the ecological changes reduce agricultural or livestock production, a reduction likely to translate into a decline in human wealth or well-being.
Although human welfare and the environment are fundamentally related, the relationship is complex. Declines in human well-being often occur for reasons entirely unrelated to environmental degradation. Taken alone, falling indices of human welfare would be as inadequate to underpin a judgement of environmental criticality as would be purely physical measures of change. The exclusive use of the latter would fly in the face of the fact that societies typically tolerate some level of environmental degradation, and that even serious ecological or environmental degradation, as we shall see repeatedly in this volume, often is accompanied by improvements in human well-being and standard of living. The concept of life-support systems, by focusing attention on those aspects of the environment critical to sustaining human occupancy, avoids the extreme geocentric position of equating large or rapid ecological transformation of any sort with significant environmental degradation.
Environmental degradation can undermine life-support systems in a variety of ways. Consider three different regional situations with different resource bases and productive systems. The first, and the most simple, involves the mining of a finite resource, such as coal or oil. Criticality is reached when the resource becomes scarce as a result of resource depletion, limited knowledge of the resource base, market constraints, or technological limitations. In the second case, criticality occurs after the rate of exploitation of an otherwise renewable resource (e.g. soils, groundwater, fisheries) exceeds the rate of renewal. For instance, from the perspective of resource analysis, the section of the sustainable-yield curve for fisheries resources where the population has been depleted beyond its ability to reproduce itself is the "critical zone" (Rees 1985). The third case reaches criticality when the demand for disposal exceeds the assimilative capacity of the environment. For example, toxic emissions may result in high levels of air pollution and adverse health effects, and sewage disposal into a river may increase the biological oxygen demand to the point that flora and fauna die.
In looking at these problems as rates of replacement and issues of assimilation, Daly (1990) finds the basis to begin to operationalize notions of sustainable development using three guidelines: (1) rates of harvest should not exceed regeneration rates (sustained yield); (2) rates of waste emissions should not exceed the rates of the assimilative capacity of the environment; and (3) for quasi-sustainable use of non-renewable resources, the harvest of these resources should be paired with compensating investment in a renewable substitute. Gaps in these relationships suggest differing situations in which trajectories in human-environment relations may lead to criticality or some similar condition.
It is possible, however, to deplete resources within an area without major social consequences as long as the deficit can be made up from elsewhere. Hence, the external linkages of an area with other regions are of no small importance. Most regions are sustained to some degree by inputs from elsewhere. For a region whose economic base is substantially disconnected from local physical resources, even substantial transformation of the environment may pose little threat to life support, as long as pollution sinks are not seriously overburdened. For a regional economy that depends closely on regional natural resources, on the other hand, rates of depletion considerably in excess of rates of natural replenishment may progress to criticality if the costs of resource substitution through imports become so high as to make the linkage unsustainable. A region would enter a state of criticality if environmental change undermined the productive activities that sustain its population to the point that the costs of substitution for essential inputs from outside can no longer be sustained and no feasible societal responses exist that are capable of mitigating the ongoing degradation or sustaining the same level or quality of habitation (as indicated by size of population and level of human wellbeing).
Generally missing from these treatments of life-support systems and assessments of well-being are the spiritual dimensions of human-environment interactions. However much ecological crises are depicted and characterized in material terms, they are doubtless also spiritual, religious, and cultural in their roots. The ecocentric perspective to which O'Riordan (1976) refers finds its expression in ideas of stewardship, of human spirituality and development, of responsibility to future generations, and of respect for other species. Although basic-needs and life-support-systems approaches may inevitably need to focus on material conditions, it bears emphasizing that spiritual dimensions may be no less important, no matter how difficult they are to conceptualize and measure.
The literature on life-support systems, important as it is, strains definitional integrity by opening the interpretation of criticality and endangerment to cascading sets of issues that take us far afield from environment-based human needs. Many attributes cited as basic needs are virtually unmeasurable at this time, which questions the practical usefulness of such broader approaches. Life-support systems are a strong reminder, on the other hand, of what might be ignored in narrow interpretations of criticality and endangerment. Larger systemic relationships should not be lost from view in conceptualizing and analysing criticality.
The future
No approach to endangerment or criticality can escape issues of time orientation and connections between the past, present, and future. Future generations will experience many of the impacts of current environmental changes. An important dimension of environmental loss is what it forebodes for the future (Weiss 1989).
These impacts, alas, are murky, for we know little of the future's abilities, values, and needs. Present actions, even ones that appear exploitative, may - unintentionally or counter-intuitively as well as deliberately - add to rather than subtract from the wealth, resources, and options available to the future. We do, however, have some indications of potential degradation to be exported to the future by present actions, the costs of current interventions, and the uncertainties that characterize choices. In ecological terms, the impacts of current activities often have long latency periods. Adequate knowledge is often lacking to assess the implications of present actions and the levels of costs to export to future generations. But it is clear that present actions may draw down the "capital" held in nature, reducing the resources and the options available to future generations in meeting their needs and aspirations. Impositions on future generations occur in at least three major ways: (1) depletion of resources; (2) degradation of environmental quality; and (3) discriminatory access to the environmental resources and benefits enjoyed by previous generations (Weiss 1990). In proposing an innovative set of principles to guide intergenerational equity, Weiss argues that
Every generation receives a natural and cultural legacy in trust from its ancestors and holds it in trust for its descendants. This trust imposes upon each generation the obligation to conserve the environment and natural and cultural resources for future generations. The trust also gives each generation the right to use and benefit from the natural and cultural legacy of its ancestors. These rights and obligations, which may be called planetary rights obligations, form the corpus of a proposed new doctrine of intergenerational equity in international environmental law. (Weiss 1990, 7)
Notions of security in the face of an uncertain world and future enter into assessments of endangerment. Threats occur in terms of both increased risk and reduced options for human responses. Environmental degradation and resource depletion reduce our abilities to cope with foreseen or unforeseen changes, increase our losses, place unacceptable burdens on future generations, and can exacerbate tensions among nation-states. Nature's capital includes both proven and potential assets, and it is enhanced by the human creation and maintenance of "landesque capital." Landesque capital refers to "any investment in land with an anticipated life well beyond that of the present crop, or crop cycle" and "involves substantial 'saving' of labour and other inputs for future production." It includes, for example, terracing, irrigation, and drainage (Blaikie and Brookfield 1987, 9). Depleting either potential or actual assets undermines the flexibility to respond to future challenges or surprises. Biodiversity, for example, may hold yet unrecognized potential for medical and agricultural development. The diversity and buffering capacity in the environment are a source of security for unknowable future impacts or demands.
Adding the dimension of the future complicates many of the issues surrounding criticality. Restricting the scope of impacts on human societies to the current generation fails to give adequate weight to accumulating deterioration and its ultimate impacts on the human condition (in future generations). Yet in many past cases in which hindsight reveals that degradation of essential resources had pointed inexorably toward an environmental catastrophe, that catastrophe was averted through economic or technological change - not necessarily undertaken for environmental reasons. Resources were substituted and environmental degradation was displaced or alleviated (e.g. Hagerstrand and Lohm 1990; Pfister and Messerli 1990; Riebsame 1990; Whitmore et al. 1990). The lessons are that seemingly severe degradation may ultimately prove to be ephemeral because of future adaptations and changes. On the other hand, even changes that do not directly threaten current or near-term uses may none the less be of great importance because of the costs that they are likely to exact from future users of the environment.
Response systems
Environmental threats, as we have already seen, are interactive phenomena. Similarly, environmental change can be assessed only by examination of the complex of interactive contributing factors at work in a particular landscape and culture. These factors include natural variability, human-induced stresses, the sensitivity and resilience of the ecosystem, the vulnerability of the population, and, not least, the goals and capacities of response and management systems.
These response systems are embedded in political economies and cultures and take on goals and structures consistent with those larger contexts. The state commonly lends its power to dominant groups and classes to support their accumulation while marginalizing others (Blaikie 1985; Blaikie and Brookfield 1987). Thus, scarce resources are typically not allocated to intervene in peripheral regions headed for criticality; indeed, conscious exploitation and extraction of resources often continue unabated in the face of warnings. In other cases, environmental degradation is less overtly driven by state policy but rather reflects a willingness to exploit nature's assets for short-term gain or to focus political resources on other societal priorities. In any event, political will and priorities are often the determining factors for intervention to deflect an approaching condition of environmental criticality, as the subsequent chapters on Amazonia and the Aral Sea will attest.
But capabilities and resources are also important. Delayed and ineffective responses to emerging environmental criticality across many capitalist and socialist societies suggest that the roots of inadequate response cut across, or extend beyond, political economies and state behaviour. Meadows and associates (1992) note that delayed response over lengthy periods toward accumulating environmental criticality is commonplace, not exceptional. It is apparent, as Blaikie and Brookfield (1987) argue, that grand theory will not provide satisfying explanations of particular regional situations, a recurrent observation from scholars of nature-society relationships. The response systems operating under both opportunities and constraints need detailed analysis conducted in regional context.
Such analysis will certainly assess political relations that characterize decision-making structures at various scales. But it will also need to address the adequacy of various knowledge systems - how alerting occurs, what catalysts for action occur, how problems and data are constructed, how interventions and management strategies evolve, and how social learning occurs (Glantz 1988; Lindblom and Cohen 1979; Ravetz 1986). Clearly the degree to which emerging environmental degradation in any situation is critical depends upon the resources and capabilities that can be brought to bear and the political will to allocate them to ameliorate damage. As Argent and O'Riordan argue in chapter 8, sea-level rise poses an extraordinary risk to the Netherlands, but the very high response capability of that society shrinks the degree to which criticality is likely to emerge.
Summary
From the foregoing discussion and the larger literature on which it draws, we derive several lessons for a sound approach to environmental criticality:
1. Environmental criticality is an interactive phenomenon that relates to types and rates of environmental change, the fragility of the ecosystem, the vulnerability of the population affected, and response capabilities.
2. Criticality must be assessed within the particular landscape and culture of occurrence and must be placed in historical and spatial context.
3. Human-environment trajectories appear particularly likely to lead to criticality in situations that have some combination of
• economies of high sensitivity and low resilience to environmental change;
• human societies with high social and economic vulnerability;
• economies strongly dependent upon local environmental resources;
• frontier areas exposed to new forms of use; and
• close linkage with, and dependent position vis-á-vis, global markets or distant political authority.
4. Non-linear and discontinuous environmental change holds a high potential for exacerbating societal diagnosis and delayed responses.
5. Criticality refers to situations in which emerging environmental degradation threatens to overtax the resources of the environment, leading to a loss of life-support capability.
6. The concept of a critical region alone does not adequately capture the range of identifiable situations. Additional categories identifying different kinds and degrees of criticality are required.
7. Much of the change inflicted by human pressures on the environment may impose costs on future generations that need to be included in approaches to endangerment and criticality. On the other hand, many of the perceived environmental threats of today may be dispelled in the near future.
Regions and the regional approach
We began by observing that environmental problems are not distributed uniformly across the earth and that environmentally critical situations are often concentrated in regions. We suggest also that the regional level offers advantages for the study of environmental criticality. The substantial variation in setting, process, impact, and ability to respond across the earth's surface makes a meaningful global aggregate assessment problematic if not impossible. Findings derived from local or micro-scale studies, on the other hand, may thwart generalization because of their particularity. Examples of criticality in small (local or micro-scale) areas of the world are not difficult to find: places suffering from severe erosion or soil impoverishment, from desertification, from pollution, or from resource depletion. In these cases, however, abandonment or relocation, or a change to a different livelihood system, may be costly for the local population but are not severely disruptive at the societal level and do not pose the difficult questions that regional criticality does. The persistence of the region as a middle level for geographic study seems to reflect "the recognition that it offers a meeting ground for the local and global poles of empirical research" (Meyer et al. 1992, 273). In such research, "vertical synthesis," or the close study of the interaction of processes within a region, can be usefully supplemented by "horizontal synthesis." The potential insights offered by such comparative study of regions are apparent in studies by geographers and others.
To use a regional approach, even a comparative one, to understanding criticality is to run the risk of ignoring significant aspects of the problem that transcend the region's boundaries. Many of the causes of regional environmental degradation cannot be understood without attention to processes occurring globally and in other regions, nor can the pool of responses available be gauged accurately. A cautionary tale in this regard is provided by the UN Food and Agriculture Organization's treatment of "critical zones" and "critical countries" (FAO 1984). This study assumes, unrealistically, that there is no movement of surplus food between regions in the country and it accordingly focuses on imbalances between population and food production from closed and static land resources. It estimates levels of inputs requisite to bring available food and population into balance and then maps "critical zones" in relation to estimated levels of needed inputs. These assumptions of spatially closed systems involve many problems (as this volume will make clear) and point to the need to consider regional linkages with other areas and global markets in any sub-global treatment of endangerment and criticality. The identification of the region as a reasonably homogeneous unit should also not be allowed to obscure the importance of differences across sub-areas and social groups within it. Finally, it is possible that disjunctures in the driving forces of change are identifiable at different spatial scales, with forces evident at the global level perhaps obscured at the regional and vice versa.
We define a "region" for the purposes of this investigation as a continuous portion of the earth's surface, characterized by a rough match between a distinct physical environment and a system or set of systems of human use. The term "region" as generally used implies a middle-level scale of study, spatially continuous and not smaller than, say, a large island or a megacity and its hinterland, but not larger than a subcontinent. It has long been recognized that the delineation of regions involves a degree of subjectivity because sharp boundaries in natural and social phenomena do not generally exist. The fuzziness of boundaries poses especially difficult problems where both natural and social processes are important. Different factors to be taken into account include national and subnational boundaries, climatic zones, topography, the economic base, and so on. Rarely do these features co-vary across space. Some degree of heterogeneity is unavoidable, but for our purposes some potential regional units exhibit too much diversity to be usable. The natural and social conditions and processes of transformation operating within the Amazon Basin, for example, make it a workable region for the assessment of criticality in a way that nation-states like India, Nigeria, and Chile - though smaller in area would not because of the scale of their internal variety. (We recognize the variability within Amazonia as well, but contend, on a comparative basis, that it is less than in the states noted.)
