Contents - Previous - Next

This is the old United Nations University website. Visit the new site at

Appendix : a discussion of environmental and procez criticality

Forms of criticality
Environmental criticality
Implications of procez criticality
Endangerment, criticality, and this case-study

Forms of criticality

Criticality in PROCEZ

Since "criticality" first came into use in the literature on global environmental change, the term has acquired a rather diverse set of meanings, as Kasperson et al. (1990) have demonstrated in a review that traces the use of the term from the early 1970s.~ Arguing that neither purely biophysical nor anthropocentric bases of definition capture what is involved, they initially proposed the following working definition (Kasperson et al., 1990: 16): a continuous portion of the earth's surface, preferably larger than 5,000 km2, constituting a habitat in which human occupation has so changed multiple components of the environment that the quantity and quality of those uses and/or the well-being of the population cannot be sustained, given feasible socio-economic and/or technological responses.

In their revision, however, they focus attention on stages of degradation, being "a decrease in the capacity of the environment as managed to meet its user demands" (Kasperson et al., 1995: 7). They therefore distinguish between "environmental impoverishment," in which the trajectory in the medium to longer term threatens to narrow the range of possibilities for human use, "environmental endangerment," in which the trajectory threatens in the near term to preclude the continuation of current human use systems, and true "environmental criticality" in which this preclusion of continuation of current human use systems is immediate (Kasperson et al., 1995: 25).

On the basis of our earlier, interim assessment of the trajectory in Borneo and the Peninsula (there termed the eastern Sundaland region), Kasperson, Kasperson, and Turner (1995) class this region as among those "endangered" rather than "critical." In the present much fuller review we have explored this assessment as profoundly as we can. We accept that "endangerment" is probably the limit of the effect of present trends. However, we have retained the question of "criticality" in relation to two areas in which current trends are believed to have important global impact: the reduction of biodiversity and the effect of tropical deforestation on global warming. Neither of these was closely considered in our interim statement, and both are excluded from the general discussion of that statement by Kasperson, Kasperson, and Turner. In this Appendix we explore some of the general issues that have, over the life of the PROCEZ project, been raised within and around it. They have had an effect on the strategy we have employed in our analysis.

Types of global environmental change

Turner et al. (1990) separate the types of global environmental change, such as might lead toward criticality, into two classes: "systemic" changes, which are of global domain, however caused, and "cumulative" changes, initially of local domain but which, in combination, have an impact extending from regional to global. Global climatic change is representative of the first, although the worldwide drive for economic development could also be regarded in this category. Deforestation is a prime example of the second, and it is deforestation that is generally seen as the cause of endangerment or criticality in that part of the South-East Asian region discussed here. This approach, which is fairly strongly anthropogenic, to a large degree explains the selection of "case-studies" reported in Kasperson, Kasperson, and Turner (1995). They concern a set of areas in which different aspects of severe human impact on the environment are exhibited, each with a different local set of driving forces. The driving forces in our region are summed up as mainly external though we have had some reason to question this assessment - and the trends bringing about "endangerment" arise principally from deforestation.

As we embarked on our task, we found problems with the def inition of "criticality" initially presented to us. As stated in the Introduction, we saw ourselves as writing a test of criticality in the PROCEZ sense, rather than an account of any demonstrably critical situation. We found it necessary to break the analysis of "criticality" into parts, and even to break the specific question of the effects of deforestation into parts. The downgrading of our test to one of "endangerment" perhaps made this strategy even more appropriate, since it more readily permits relative rather than absolute judgements.

We retain, however, some problems with the attempt to combine the geophysical with the human in a single assessment. In our region, some elements of environmental process are certainly sharpened in their effects by human activity. Other parts of our analysis relate almost wholly to human forces, and the specific nature of the environment other than being part of the hot, wet, and developing tropics is of minimal significance. We therefore feel it necessary to analyse the general issue of criticality, and the track toward it, in order to clarify the argument of our book.

Environmental criticality

A more restricted approach

Although we have differed only in degree from the design proposed by the PROCEZ core-group in presenting our material, our view is that the usefulness of the concepts of "environmental criticality" and the stages leading toward it may suffer if they are not also defined more precisely in strictly environmental terms. We write principally of "criticality" in order to avoid confusion. We agree that a state called "environmental criticality" can arise only in the presence of people, for it is, like resource degradation, a concept that can have meaning only in relation to resident human occupation of a part of the biosphere. However, environmental criticality (and likewise "impoverishment" and "endangerment") should be reserved to mean criticality of the biophysical environment, and not simply that of the people who live in it. A less specific condition of criticality can be said to exist if the health, livelihood, and welfare of the occupiers are at risk, but this is environmental criticality only if a hazard arises from causes that are either embedded within the natural environment or arise from changes that have taken place in that environment. This wider meaning can shade too readily into endangerments of another kind, those that arise from the lifestyle, activities, or poverty of people, from malfunction of the economy, or - at another level - from proximity to wastedisposal outlets or defective industrial plants; the criticality facing the people who died and suffered at Bhopal and around Chernobyl is an example of the latter category. Although these hazards exist in a broadly defined natural "environment," they cannot properly be regarded as environmental. Those resulting from purely human cause may make environmental criticality worse, but they are not themselves necessarily part of that criticality. Problems arise immediately if we fail to make this distinction.

Identifying environmental impoverishment, endangerment, and criticality

Even with this restriction, these terms are not easy to define and, except in extreme cases, nor are the conditions easy to recognize in practice. "Criticality" does not properly describe an environment in a short-lived state of crisis brought about by some unusual or recurrent event, such as a hurricane or earthquake. An environment that is merely badly managed is not necessarily in a critical or even impoverished condition. Nor should these terms be used to describe generally degraded but still productive landscapes, of which there are a great many in the world. It is important to decide whether an environment that is particularly vulnerable to damage is necessarily likely to become "critical." If it is being adequately managed, so that the risk of damage is minimized, it may become severely degraded only under exceptional natural conditions.

A related group of concepts

Such terms as "sensitivity," "resilience" or "ability to recover," and "vulnerability" or "fragility" are additionally in need of sharper definition, and this is necessary if "criticality" itself is to be properly defined. These terms are older in scientific use than "criticality" and, moreover, they generally refer specifically to the physical environment. A discussion of the meanings and the interrelationships will pave the way for progress.

In writing of the problems of land degradation some years ago, Blaikie and Brookfield (1987:10) proposed that "sensitivity" should mean "the degree to which a given land system undergoes changes due to natural forces, following human interference."2 This was not wholly adequate, especially in the wider sense introduced by PROCEZ. Human activity does more than just release or accelerate natural forces; it also introduces new elements and, in modern times, chemical elements to which the land may vary significantly in sensitivity. Moreover, what we seek to describe is sensitivity only to adverse environmental change; deposition by "warping" of riverine or estuarine silt within managed embanked basins is a change immediately brought about by the natural causes of flood and tide, but we would not use the term "sensitivity" in this context.

Sensitivity and fragility

The distinction between "sensitivity," a general term, and the more specific term "fragility" is important. Jodha (1990: 2) has usefully defined a "fragile" resource as one "which cannot tolerate the degree of disturbance implied by the intensity of use associated with specific usage." Jodha refers specifically to unstable mountain slopes in the Himalayas and to rain-fed, semi-arid regions in northern India. His apparent meaning, in more precise terms, is that these environments have, under human use, very high sensitivity, respectively, to the erosivity of heavy rain and to drought. What makes them "fragile" is a combination of this high sensitivity with the intensive use to which they are subjected. The net effect is to make present forms of usage unsustainable.

Within the South-East Asian region, the area that most readily comes to mind in such a context is the uplands of Java, where a large literature alleges high rates of anthropogenic erosion, implying an environment that is "fragile" in Jodha's specific context (e.g. Donner, 1987; Seckler, 1987; Palte, 1989; Hardjono, 1991). Nibbering (1991a, 1991b), however, shows that even massive erosion can be brought partly under control by management, as it is to a great degree similarly controlled by agricultural terracing in the Nepal Himalayas (Ives and Messerli, 1989). Moreover, high erosion rates reflect an abundance of weathered material available to be eroded, and, once this is reduced, the rate will in any case diminish. If the land becomes agriculturally unusable, then it was indeed "fragile," but in Java this does not seem to be the case except over small areas. We have seen that severe erosion has also taken place in parts of our region. However, use of the term "fragility" may have been too strong; the quality of what remains is also important, and so is the ability of the land, and its plant cover, to be restored.

Resilience, or ability to recover

Following Holling(1978: 11), "resilience" is the "property that allows a system to absorb and utilize (or even benefit from) change." Blaikie and Brookfield (1987) use it rather differently, meaning essentially the ability of a land system to recover spontaneously after interference, and especially after damaging interference. Ives and Messerli (1989: 241) describe resilience as "the generalized ability of a system to cope with the unexpected," but they are referring to people-environment systems, and thus classify the rapid repair of landslide damage in the middle hills of Nepal as an example of "resilience." This seems, once more, to confuse two rather different issues, although the Ives and Messerli definition can remain useful in its application to natural systems alone.

A distinction should be drawn between the response to interference or shocks of different sorts: physical, chemical, and biological. For example, landforms built from very competent rock may have low sensitivity to physical damage, and recover well from such damage as does take place, yet some soils may readily become impoverished, and biodiversity may recover only very slowly and poorly after interference. In this sense these land systems, taken as a whole, are not resilient. Landforms very prone to slope failure, on the other hand, clearly exhibit high sensitivity to this form of damage, yet in smaller landslips the soils might even be improved by the pedoturbation (turning over of the soil), and a good vegetation cover might quickly become established (Humphreys and Brookfield, 1991). Such sites are "resilient" even in Holling's rather specific sense, yet remain at high risk of new physical failure, sensitive to any severe rainfall event. Clearly, a useful approach might be to draw up matrices of sensitivity to different forms of damage and of resilience or ability to recover following shocks of different types.

Taken together, however, sensitivity and ability to recover can be used to define the natural conditions that might, more or less readily, lead to something like "criticality" under interference or shock. All forests are equally sensitive to interference by chain-saw, but their soils differ widely in sensitivity to exposure, and the forests themselves differ in their ability to recover. Where soils are particularly sensitive to erosion and impoverishment, so that forest cannot recover after clearance, the outcome of even a single natural or humangenerated shock might be a degraded and almost useless landscape, very vulnerable to further damage by erosion and fire. In other forest environments, however, only massive and repeated interference would yield such a result. As we have seen, this may be the case in at least large parts of the region discussed in this book.

Buffering capacity, and a redefinition of "fragility"

Natural differences in sensitivity and ability to recover are fundamental and interrelated. They depend on climate, geology and topography, and available energy, as well as on the existing stock of biota. Together, these determine the "buffering capacity" of the biosphere and its inequalities from place to place and, to a lesser degree, also through time. When we speak of a "fragile environment" we should mean, essentially, one in which the collective buffering capacity in face of all or most forms of interference is limited. Contrary to Jodha's (1990) definition, this does not mean that such an environment can sustain only light use, for such light use might itself cause severe damage. Some of the most "fragile" environments demand intensive use of conservation practices in their management, which are difficult to sustain except under high productivity or with large expenditure of resources. Their condition may become "critical" if these inputs are reduced or withdrawn. By contrast, environments in which the buffering capacity is high can sustain more casual systems of management under a wide range of intensities, at least under normal conditions, and for a long time. To use degrees of "fragility" scaled in relation to the total buffering capacity of a system would follow logically. Moreover, such an approach would acknowledge the role of good management of resources in intensive use.

Environmental criticality defined in terms of imminent breakdown

It should follow from this discussion that "endangerment" or "criticality" ought to mean that there is a condition of, respectively, threatened or imminent crisis/breakdown in the natural or humanassisted buffering capacity of a fragile system. Such a crisis could be demonstrated in several ways. One is by catastrophic failure to meet production goals that were earlier attained, coupled with severe degradation of soil and water, as in the Amu Darya and Aral Sea case in central Asia (Kotlyakov, 1991), which Kasperson, Kasperson, and Turner (1995) conclude to be a region truly in a state of

"criticality. " Another is by a major increase in erosion, salinization, or acidity, coupled with rapid pauperization of the biota, as in a forest region that quickly becomes transformed into established lowdiversity grassland after clearance and burning, open to further degradation by exposure and erosion. There are such conditions in our region. The essence of "environmental criticality" therefore lies in environmental changes that threaten greatly to reduce the lifesupport capacity of a sizeable region on a long-term basis. Properly, such a loaded term as "criticality" should be reserved for situations in which problems of an unusual order of severity are currently or imminently encountered.

This differs from the PROCEZ core-group definition in its emphasis on the state of the environment itself, irrespective of the degree of human pressure on, and transformation of, that environment. Moreover, as Kasperson, Kasperson, and Turner (1995) note, in all but the Amu Darya-Aral Sea case studied by PROCEZ, there has generally been an improvement in aggregate human welfare. But this does not mean that there is no critical situation in the biophysical environment. There are elements of the biophysical environment of Borneo and the Peninsula that are indeed in a critical condition, especially from a greatly increased hazard of fire. Defining criticality in terms of the biophysical environment is to expose the fact that the need for remedial action is urgent, whatever the state of human welfare at the present time.

One could perhaps also write of a "post-critical" system in which no remedial action that is within even possible bounds of cost is feasible: a region of "badlands" or a totally decertified area would be examples. Whether or not the destruction of a large part of the oldgrowth rain forests of Malesia constitutes a condition of "postcriticality" has been discussed in this book. More commonly, some form of repair, or at least adaptation of usage, is possible in most environmentally critical regions. In a seemingly totally decertified region, exclusion of all grazing fauna would lead to some restoration of vegetation. Even in the Amu Darya and Aral Sea case, an obvious example of true environmental criticality, a set of quite simple remedial actions could undo some of the damage, although total repair would require major change in the use of land and water, at a cost clearly prohibitive under present economic and political conditions in the region. Repair of criticality, however, demands a proper understanding of causation, and this requires a major intellectual wrench abandoning the common assumption that a self-evident proximate cause is also the whole cause of the problem (Brookfield, in press).

Causation of environmental criticality

We have argued that environmental criticality is a property of the biophysical environment under use, but of the natural environment none the less. Its impacts on people and livelihood are symptomatic, not central. Given the enormous scale of modern human interference, the causes of environmental criticality might well lie squarely in that interference, but this is by placing the buffering capacity under excessive stress, by breaking all the unwritten and often unknown rules for the management of fragility. Beyond the proximate cause in most, if not almost all, cases is a set of underlying causes or conditions. The burden of the present argument is that these conditions vary through space and time, and that it is either exposure of, or pressure on, the most sensitive and least resilient systems that triggers environmental criticality and its approach.

There are some cases in which the biophysical share in the causation of criticality is greater than in others, and we have already discussed one such in the Himalayas. Another is the sustained decline of rainfall in the Sahel region from the 1960s to at least the 1980s. It might be true that the explanation of the 1973 famine was to be found in the "economic and social situation prevailing in the 1960s, before the drought" (Garcia, 1981: 191). The subsequent chronic criticality of the Sahel region cannot, however, be so unicausally explained. There was a long-term secular or cyclic decline (by as much as a quarter) in the rainfall of the most sensitive areas, as well as social and political failures that made its impact far worse and made efforts at management almost impossible.

The underlying biophysical causes of criticality should not be overemphasized, but they cannot safely be ignored. Criticality is brought about by human pressures of a level inappropriate to the conditions, and by unwise management in conjunction with natural sensitivity and low resilience. Sometimes, the natural conditions themselves are of major significance. Moreover, this consideration gains force from the modern context in which environmental criticality is most often discussed, that of the unknowns of global climatic change. For, as Parry (1990) has warned, those parts of the agricultural world where production is marginal or that currently exhibit environmental crit icality are those that are most likely to experience much greater problems of adaptation in a warmer world.

Implications of procez criticality

One interpretation of PROCEZ

Writing within a simulated "history" to 1996 of research on the human dimensions of global environmental change, based on work principally in or from the United States - completed by 1991, Rockwell and Moss (1992:14) gave a prominent place to PROCEZ. They described its aims and hypotheses in these terms:

The project focuses on whether and how environmental degradation leads to deterioration in human health and well-being. Because human occupation inevitably leads to some degree of environmental change, the project has sought to understand the causes and effects of degradation, perceptions of its occurrence, and responses to it.

In distinction from the PROCEZ core-group, Rockwell and Moss were, it seems, implying that degradation itself is the mark of criticality. In this sense, major degradation of land cover does, or should, constitute a state of affairs that might be linked with decline in the quality of human life. This, therefore, returns us to a central aspect of the PROCEZ inquiries, which is the connection between adverse environmental change and declining quality of life support.

Interpretation in the context of PROCEZ criticality

We therefore had to approach our problem from two sides. On the one hand, we sought what in this case-study might constitute environmental endangerment approaching criticality, and asked how this affects human welfare. On the other, we had to inquire into declining life-support capacity itself, where that can be identified, and inquire how far this might be related to environmental degradation, broadly or narrowly conceived. The question of perception of, and response to, endangerment also became important. However, what should seemingly be excluded from "criticality" are environmental changes that cannot be regarded as causing degradation of lifesupport capacity, and deterioration of human health and well-being that cannot be explained in terms of environmental degradation. These are complex issues.

Our response to this set of challenges has been to place emphasis on changes, positive and negative, in the state of human welfare. While doing this, however, we linked the argument at all stages to changes in the environment, and specifically discussed evidence of degradation, whatever its effects. To do this more effectively, we looked beyond the immediate scene and also examined change in the historical past. This enabled us to see that trajectories can be reversed, and that variation, as well as secular change, is important. We tried to be rigorous in making linkages between human change and environmental change, and also attempted to separate proximate causes from underlying causes of change. In this way we hoped to avoid simplistic conclusions.

Social breakdown, conflict, and environmental criticality

Rockwell and Moss (1992: 14) went beyond the aims of PROCEZ, and (without consultation with the project principals) also hazarded a guess at its conclusions. Projecting notionally into the future, they went on:

The project has explored such tentative generalizations as the hypothesis that environmental degradation leads to socioeconomic decline only within areas that are dependent on primary production and that are not substantially engaged in trade with other areas; in more economically open systems, environmental degradation apparently has to reach catastrophic proportions before it precipitates socioeconomic decline.

This conclusion certainly does not emerge from our case-study, nor does it seem to emerge from any of the others in PROCEZ. We have to presume that it is a hypothesis of Rockwell and Moss, and regret that they attributed it to this group of scholars in so prominent a place as the journal Environment. However, in being thus bold, they were linking our project to wider concerns prevailing among some social scientists who have considered the consequences of climatic and other forms of change. Their views are both important and influential. It is therefore worth while to digress and examine this "finding" more closely.

The idea that human-induced environmental pressures, when on a major scale, may have consequences that will so disrupt society as to affect national and even international security has been voiced by a number of prominent writers (e.g. Gleick,1989; Mathews, 1989; Myers, 1989b). Most agree that, although the greatest such problems threaten the developing countries, they have potential impact worldwide, not least through the effects of conflict both within and between countries, and even between developing and developed worlds as a whole.3 Homer-Dixon (1991) reviewed the arguments and some of the evidence, focusing on conflict that involves the possibility of violence. Environmental change in the poorer countries will, arguably, lead to increased scarcity, thence to conflict arising from relatively greater deprivation in some groups than in others. The consequence will be refugee flows that will cause the better off to defend their resources, and hence their ethnic or social identity, and to problems of distributive justice that may be resolved, in the short term, by repression. One can see more evidence of this in the growing West European resistance to refugees from poorer regions lying east and south than in the developing world itself. Assuming that the theory is correct, however, it would seem to follow that the early identification of the most critical zones and most vulnerable people is in everyone's interest.

When we come to seek evidence in the contemporary Southern world, however, there are problems.4 Where environmental stress can plausibly be associated with conflict, it is more as trigger than as underlying cause. In the Horn of Africa and the Sudan, to cite the starkest of examples, distress induced by drought and land-resource degradation has been palpable for most of the past 20 years, during which long-running civil wars and large-scale refugee movements have afflicted at least three countries. But although there has certainly been an interplay of environmental stress with conflict, it is not at all clear which is the basic cause; for example, famine in 1972-1974 may have been the proximate trigger of the Ethiopian revolution, but the seeds of conflict were already present.

Conflict and social breakdown may arise specifically from pressures and disasters as proximate causes. However, the underlying causes lie elsewhere. The contemporary problems of the underdeveloped countries are due most fundamentally to underdevelopment itself, and only then to population and economic pressure on the resources available for use and on the degradation of those resources. Cases in which degradation affects human health and welfare can certainly be identified, but in complex association with other causes. In no case can social breakdown yet be unequivocally attributed to degradation, or even to environmental catastrophe alone. Criticality, whether strictly environmental or as defined by PROCEZ, cannot be given so deterministic a role in human affairs.

Endangerment, criticality, and this case-study

What has been relevant?

The problems of interpretation discussed in this Appendix have significance for the manner in which we presented and analysed environmental and societal change in our region. There is heavy pressure on natural resources (forest resources in particular), but also on some soils and increasingly on energy sources. For the species-rich natural forests themselves this impact is certainly critical, and for large areas may already be post-critical. For other elements of the biosphere the impact is, as we saw, far more patchy and uncertain.

The extent to which resource depletion threatens human occupancy and well-being is problematical and selective, especially since this region is "frontier" to a wider and more populous area in which economic development and social improvement are advancing apace. In so far as there is endangerment or criticality in these latter respects, we found that it flows as much from the development strategy itself, and its differential impact on the region's people and industries, as from environmental change.

Two widely publicized aspects of the situation in this region were, however, important. One is the effect of deforestation on biodiversity and on global climate, both matters of wide extra-regional concern. The other is alleged criticality for the forest-dwelling people and their livelihood. A third, of less widespread international concern, is the sustainability of new activities made possible by forest depletion and degradation. All these are areas in which a strictly environmental criticality, as we define it, has potential importance, and we gave some prominence to the identification, or refutation, of such environmental criticality in our analysis. However, the dominant force is economic development, together with population growth in what is still, in the main, a sparsely peopled region. In so far as the state of economy and of welfare are indicators, we could not ignore the evident fact that their main determinants lie outside the environmental sphere. In writing a report on a PROCEZ case-study, we were therefore concerned with the interplay between rapid but uneven economic and social change, the biophysical environment that it affects, the changes taking place in that environment, and societal consequences, both within the region and beyond it.


1. The first part of this Appendix draws substantially on ideas presented in a paper by Brookfield, written in 1992 for publication in a collection edited by N. F. Glazovsky intended to be published by the Institute of Geography, Russian Academy of Sciences. Up to the time of completion of the present book, this publication has still not appeared.

2. In their revisions, Kasperson, Kasperson, and Turner (1995) draw quite heavily on some of the definitions proposed by Blaikie and Brookfield (1987), albeit with some modifications, most of which Brookfield would accept. The present discussion, however, still stands in its own right and only necessary changes have been made.

3. We perhaps witnessed such a conflict in verbal form in the preparations for the United Nations Conference on Environment and Development in 1992. In a preliminary meeting of the developing countries at Kuala Lumpur in April, the Prime Minister of Malaysia was a prominent advocate of a developing country view that was hostile to the developed lands regarding the distribution of responsibility for global environmental protection.

4. Here we draw on discussion at a workshop on "Environmental Change, Economic Decline, and Civil Strife," organized by the American Academy of Arts and Sciences and the University of Toronto, and held at the Institute of Strategic and International Studies in Kuala Lumpur, Malaysia, in November 1991. Brookfield's arguments, voiced at the meeting but unpublished, are reflected here.

Contents - Previous - Next