Webs of power: forest loss in Guinea

James Fairhead and Melissa Leach

back to issue

THIS paper examines the contrast between the formulation of problems in development policy, and the perspectives of villagers whose views have been subjugated and everyday activities criminalised, within this formulation. We attempt to identify the conditions in which certain demonstrably false ideas about environmental change have come to acquire validity in policy circles, while others, more correct and espoused by inhabitants, have been excluded from consideration and investigation.

Several authors have recently spotlighted the presence of particular off-the-shelf ‘narratives’ current in development institutions, which come to define development problems and justify interventions, particularly in conditions where data is poor, time is short, national agendas are overruled and local consultation impossible (Hoben 1993; Roe 1991, 1995). Narrative construction is the stuff of synthesis overview writing within development agencies and policy research institutes, and of inter-agency analytical alignment in development approaches. Narratives help decision makers to fill confidently the gap between ignorance and expediency.

With the spotlight on the narrative, less attention has been given to the ways that the discursive processes which condition narrative construction also condition the knowledge produced about development problems, including the generation of credible ‘data’ – often in large amounts. Adherents to the environmental degradation view explored in this paper consider that abundant evidence supports their conviction. Furthermore, focus on these narratives encourages analysis to treat the relationship between international and local agendas as one of dislocation, divided by a gulf which the increasing use of development-institutionally acceptable research methods, apparently responsive to local concerns, might help to bridge. Less attention is given to ways that different sections of local society become involved in the discursive processes in which development policy knowledge is produced. Such involvement may have developed over long periods, given that present development concerns frequently build on old debates which have already been incorporated into local political processes.



Environmental issues are pertinent in this respect because of the degree to which environment has become a dominant development concern. They particularly invite critical analysis because of the clarity with which global issues and constituencies, as well as local ones, are involved in defining and responding to the development problem. The analysis summarised in this paper (for a detailed treatment see Fairhead and Leach 1996a) adds to a number of recent historical and social anthropological analyses which have falsified the readings of environmental change that have been driving development policies, revealing major contrasts between external perspectives and locally experienced realities.

Contrasting definitions of the environmental problem contain particular images of local practices and justify contrasting development paradigms, commonly amounting to repression of, as opposed to support to, local techniques and institutions (Behnke and Scoones 1991; Thompson et al. 1986; Tiffen et al. 1993; cf. Leach and Mearns 1996). The case considered in this paper concerns ongoing savannisation of tropical forest which is not, in fact, taking place.



The vegetation of Guinea’s Kissidougou prefecture reflects its position in West Africa’s forest-savanna transition zone, consisting of patches of dense, high semi-deciduous rainforest dispersed in savanna. The forest patches, which surround old and new village sites, have been considered by environmental policy makers for at least a century as the last and endangered relics of a once extensive natural forest cover now destroyed by local farming and fire-setting; a destruction they have continually sought to redress. But the experiences of most of Kissidougou’s Kissi and Kuranko inhabitants, as well as archival and air photographic comparisons, do not support this view. Instead, they show forest islands to be the result of human management, created around villages in savanna by their inhabitants. They also show the woody vegetation cover of savannas to have been increasing during the period when policy makers have believed the opposite (Fairhead and Leach 1996a, 1996b).

West African vegetation maps, which show vegetation zones in more or less horizontal bands, easily lend themselves to interpretation as temporal, as well as spatial transitions. Whether from desertification, Sahelianisation or savannisation of forest, observers have been tempted to see each zone as the anthropogenically degraded derivate of a prior vegetation type. On many maps, the forest-savanna transition zone is marked explicitly as a ‘derived savanna’, or ex-forest zone. And in Guinea, policy makers since the turn of the century have been convinced of this southwards shift, with the conflation of spatial and temporal transitions incorporated into the scientific canon informing national and regional environmental policy.

The first forest reserves established in Kissidougou in 1932 were conceived of as a protective ‘curtain’ to halt the southwards spread of fire and farming induced savannisation. In 1993, the same conflation of spatial with temporal zones provided the logic for a major donor funded environmental rehabilitation project to take 40 Kissidougou farmers on a journey to northern Mali to see the future of their own landscape should protective measures not be undertaken.



Within each vegetation zone, the iconography of spatio-temporal shifts on the vegetation map is complemented by the iconography of ‘divergence from a climax vegetation type’: the notional maximum vegetation which could exist given climatic conditions. This contains the idea of the previous existence of a ‘bigger’ and ‘better’ vegetation type ‘prior to human disturbance’, and closer to the ‘Eden’ which Africa’s environment so often represented in colonial imaginations. In this way, present conditions in each vegetation zone may be envisaged as the anthropogenically degraded derivate of its predecessor. And so in Kissidougou, climate (annual rainfall levels over 1600mm) and the presence of humid forest species and patches are taken as indicative of high forest potential and hence of its past existence.

The assumption of anthropogenic degradation of a prior natural forest formation was integral to the first delineation of West African vegetation zones in the early colonial period by the botanist Chevalier. This analysis was transferred directly into contemporary policy since Chevalier was, at the time, the senior most advisor to the French West African colonial administrations responsible for environmental concerns. Subsequently, deductions made from analysis of the botanical composition (‘phyto-sociology’) of vegetation forms in these zones by botanists such as Aubreville, Adam and Schnell reinforced the hypothesis that the forest-savanna mosaic was in temporal transition.

Observing the tree species characteristic of forest patch boundaries, for instance, botanists deduced them to indicate savannised forest (Adam 1948, 1968). They did not consider other possibilities: that this ‘transition woodland’ could represent a stable intermediate form, the establishment of forest in savanna, or the complex outcome of inhabitants’ management strategies.



As Aubreville and Adam, in turn, became senior figures in French West Africa’s forestry administrations, their phytosociology interpreted within the degradation logic became institutionalised as the principal methodology for assessing regional vegetation change, and their publications became key texts in comprehending West African environmental history more generally (Aubreville 1949). Characteristically, these botanists directly observed landscape features and deduced history and people’s impact from them. Their disciplinary position and the social conditions of their fieldwork reinforced their pejorative visions of local farming and fire management practices, rendering it both difficult and seemingly unnecessary to properly verify change with with local people themselves.

It has remained ‘scientifically’ acceptable to interpret vegetation history and anthropogenic impact from snapshot landscape observations, with deductions from plant and other indicators, vegetation surveys, and remotely-sensed imagery now adding to the repertoire. For example, modern observers of Kissidougou often consider the presence of oil palms to indicate that forest has retreated from the area, while the team preparing Guinea’s forestry action plan (République de Guinée 1988) deduced from their air photographic ‘snapshots’ and vegetation surveys that southern Kissidougou was a ‘post-forest’ zone. Similar social distance and pre-conviction as characterised the colonial botanists enables today’s analysts too, to overlook both local people’s environmental experiences and management, and historical methods (eg. oral histories and archive consultation) in comprehending environmental influences and trends.

Historical data sets are nevertheless available today, and their examination produces a very different picture of vegetation change. We compiled a picture of Kissidougou’s vegetation dynamics through elderly people’s oral recollections concerning vegetation use and management, comparative analysis of 1952, 1982 and 1991 air photographs and 1989/92 SPOT satellite data, and landscape descriptions found in archives dating from 1893. Social anthropological fieldwork in Kissi and Kuranko villages throughout 1992 and 1993 provided an understanding of inhabitants’ agro-ecological concepts and techniques and the social conditions of their application in the present, enabling closer enquiry into land use change.



Far from being relics, Kissidougou’s forest islands prove to have been created by local populations. In the majority of villages, elders describe how their ancestors encouraged forest patch formation around settlements which had been founded either in savanna or beside gallery forests. The formation and growth of forest islands around recently established village sites is often visible when 1952 and modern air photographs are compared. Villagers also suggest that woody cover on the upland slopes and plateaux between the forest islands has generally increased during this century, and not declined as has been thought.

In the north and east of the prefecture, grass savannas have become more densely wooded with relatively fire-resistant savanna trees and oil palms. Indeed that oil palms have spread north into savannas, encouraged by villagers, suggests that they may be better seen as outposts of anthropogenic forest advance than as relic indicators of forest retreat. Even more strikingly, in the south and south-east, large expanses of grass and sparse shrub savanna have ceded entirely to forest fallow vegetation: the area is actually a ‘post-savanna’, not a ‘post-forest’ zone.



These southerly savanna-forest transitions are not only evident in air photograph comparison, but are strongly indicated by changes in everyday resource use: for example the introduction of tree felling in agricultural operations, greater availability of preferred fuelwood species, changes in roofing and thatching materials, and changes in termite species associated with particular edible fungi. These demonstrable changes, which reflect long term interactions between the populations of Kissidougou and their forest-savanna vegetation (Fairhead and Leach 1996a), strongly challenge the view of an ongoing southerly shift of vegetation zones, iconography of the ongoing southerly shift of vegetation zones.

Equally, evaluating ‘degradation’ in terms of a vegetation climax is revealed as inadequate when one takes the impact of long term climate history into account. Given that West Africa has experienced both long period, deep climatic fluctuations and changes in climatic variability (Brooks 1986), the history of vegetation form begins to appear as a history of continual transition, rather than of divergence from a single, once extant climax. Recent ecological analysis suggests that such ceaseless transitions depend on multifactor complexes rather than trends in one particular variable; if a transition-causing factor reverts to its pretransition level vegetation may move to another state, but need not return to its original one (Behnke and Scoones 1991; Dublin et al., 1990; Sprugel 1991).

The various forest and savanna forms in the transition zone can be seen as such multiple-determined states dependent on fire, soil, water, seed availability, animal-related, and other conditions. Deflection from one state to another may depend upon a particular, possibly even unique historical conjuncture of ecological and management factors. By altering the balance of interacting factors, people can initiate shifts between states which might be unattainable – or much less likely – through ‘natural’ ecological processes alone. The shifts from savanna to forest in Kissidoguou could be seen in this way (Fairhead and Leach 1996a).



Yet within Guinea, environmental services have been so convinced of the degradation they are combatting as to find it unnecessary to compare their commissioned aerial and satellite images with those from 1952, let alone question their interpretive framework. Even when comparative interpretations are carried out, they are frequently not independent of preconceived ideas of vegetation change. In Kissidougou, the incredulous reactions of forestry staff when presented with 1952 and 1990 air photographs showing increased woody vegetation led them to a sceptical search for ways to render the comparison invalid (the photographs were taken in atypical years, or incomparable seasons).

In other parts of West Africa, similarly surprising results have simply been disbelieved and dismissed: ‘Apparent increases in biomass from pre-disturbance [vegetation] to present were labelled "discrepancy" and such discrepancies were ommitted from further analysis...there seems little possibility that biomass has increased as a result of land use’ (Houghton et al. 1993).



In contrast, justifiable scepticism was cast aside when a comparison of eastern Guinea satellite images taken ten years apart seemed to show significant vegetation degradation, and on which basis major donor funds for a regional environmental rehabilitation programme were secured (Grégoire et al. 1988).

The images of environmental change derived from these ‘scientific’ analyses have been incorporated not only into Guinean environmental institutions, but also into formal sector education and the popular consciousness of state functionaries. They are regularly reproduced in school geography lessons and national university curricula and theses. For those educated within this vision, casual readings of the landscape come to serve as confirmatory evidence; dry season bush fire is taken as proof of a worsening problem, and the conversion to farmland of a few forest islands near the town for urban market-gardening is taken to suggest forest island diminution everywhere.

Such casual landscape readings are often made during the dry season, when external consultants, forestry agents and urban nationals’ visits to villages are concentrated. This is the destructive part of villagers’ normal seasonal cycle, when bush is cleared for farming, fires sweep the savanna and trees are cut for construction or sale. Regeneration during the rainy season, anyway more subtle to observe, escapes attention within this seasonal bias (cf. Chambers 1983).

Interpretations of vegetation degradation are reinforced not only by local observation, but also by the global and regional level analyses with which they are in keeping, and which carry the weight of international authority. Given FAO figures concerning rapid forest loss in West Africa (FAO 1990), for example, it appears inconceivable that Kissidougou should be experiencing anything else. Such figures, so frequently publicised in the more glossy development literature and on the radio, are far more accessible to the environmental administrations and urban public concerned with Kissidougou than are analyses of the locality itself.



Equally, the rhetoric of shared environmental crisis, made so apparent in the 1992 UNCED conference in Rio, appeals far more powerfully to local officials than the statements of the villagers supposedly experiencing these problems. This was made evident in the 1993 ‘Journées de l’Environnement’ conference designed to raise awareness of Kissidougou prefecture’s environmental problems, where both the Prefect and Kissidougou’s urban based environmental NGO’s framed their speeches in terms of global concern with biodiversity loss and the common West African struggle against desertification.

The projection of global and regional concerns onto Kissidougou’s environment has recently heightened, but it is not new; it has informed administrative perceptions since the early colonial period. A concern that deforestation in Kissidougou would damage regional climate and hydrology was apparent in the earliest writings of Chevalier (eg. 1909), and underlay a major watershed rehabilitation programme first outlined in the 1930s, funded in the 1950s following the 1948 Goma inter-African soil conference, and launched again in 1991.



This analysis of environmental change which informs local policy cannot be separated from the financial context in which environmental institutions operate. In Guinea, early colonial administrations first became concerned with the perceived destructiveness of African environmental management because the colonial economy was heavily dependent on ‘threatened’ natural resources; initially wild rubber and then, in Kissidougou, oil palm products and tree crops grown in forest patches (Fairhead and Leach 1995).

In the later colonial and post-colonial periods, more regional and global economic imperatives joined these national ones. In the 1950s, new funding envelopes for regional soil, climate and hydrological conservation became available following the heightened Africa-wide environmental concern epitomised by the 1948 Goma conference. More recently, administrative solvency and development activities have come to rely even more heavily on foreign aid, and have thus become subject to various forms of ‘green conditionality’ (Davies 1992; Davies et al. 1991). This greening of aid, and the specific forms it takes, reflects donors’ needs to satisfy home political constituencies heavily influenced by media images and northern environmental NGOs, as well as their own institutional assessments of African environmental problems.

In Guinea, a large proportion of foreign assistance is now allocated, sectorally and by region, directly to environmental rehabilitation. A new generation of heavily funded environmental projects has emerged, including, in Kissidougou, two component projects of the internationally funded Niger river protection programme. In agricultural and other development activities as well, overt environmental sustainability components are important for attracting future funds. Kissidougou’s prefecture administration, agriculture and forestry services are well aware of the packages which satisfy the donors in this respect: agroforestry programmes, forest conservation and improvement, bush fire control, and rationalisation and reduction of shifting cultivation in favour of intensive wetland rice.



During Kissidougou’s ‘Journées de l’Environnement’ the prefecture’s no. 2 administrator stated explicitly that: ‘Donors are interested principally in environmental projects, so we must solicit their aid to ensure the development of the prefecture.’ He suggested that other localities learn from the example of the Niger protection project zones where schools, water and other infrastructural developments were provided in exchange for local participation in environmental protection. The emergence of local, urban based environmental NGOs such as Kissidougou’s Friends of Nature Society has also been encouraged by recent donor interest, not only in environmental issues but also in the claimed capacity of NGOs to achieve ‘participatory’ development. In short, presenting a degrading or threatened environment has become an imperative to gain access to donors’ funds. In this respect, our own findings were often considered subversive: threatening to the prefecture’s future financial and development interests, and to the continued employment and material privileges of environmental project administrators and extension workers.

Considering the environment as degrading and threatened is equally crucial to the solvency of state environmental institutions when they do not receive donor support. Since their inception, Francophone West African forestry services have derived revenues from the sale of permits and licences for timber and wildlife exploitation, and fines for what became enviornmental crimes in breaking state environmental laws. In Guinea, setting bush fires actually carried the death penalty during the 1970s (Law 08/AN/72 of 14 September 1972). Environmental services have been able to gain such revenues only by taking control over the management of natural resources (e.g. fire and trees), and this through deeming villagers to be incapable and destructive resource custodians.



Revenues are thus ensured by a reading of the landscape as degraded and degrading; of forest islands as disappearing relics in an increasingly grassy savanna, not as created in an increasingly woody one. The importance to forestry staff of informal receipts gained while applying policies of repression only accentuates the imperative for this environmental reading, while the antagonistic relationship thus engendered between forestry agents and villagers bars communication about villagers’ own environmental experiences. Thus at local and national, as well as international levels, the economic structures within which environmental agencies operate frame the ways that information is derived.

The attitudes of forestry staff depend not only on their financial and educational status as forestry service members, but also on their socio cultural positions. They share with many other formally educated, urban based Guineans a particular vision of villagers’ resource management capabilities. This image – of the rural farmer as environmental destroyer, and rural farming and forestry techniques as backward, in need of modernisation – conforms with and helps justify urban intellectuals’ self-distinction as modern and progressive. Such distinctions were reinforced under Sekou Touré’s 1958-84 regime when the urbanised were politically and economically privileged, and their vision of a highly mechanised, capital intensive technical future dominated rural development approaches (Rivière 1976).



As greater attention has come to be paid to environmental rather than agricultural issues, the environmental component of this degrading view of village capabilities has become dominant. Generalised notions about ‘man’s destructive impact on the environment’, projected locally, have entered the numerous processes through urban educated people who understand themselves as relatively more ‘civilised’ or ‘globalised’ (cf. Bledsoe 1990). And just as urban circles benefitted from the agricultural modernisation which wrested resource control from villagers, so they have become the main beneficiaries of environmental control, keeping the moral high ground while gaining from policies such as those removing timber cutting rights from ‘irresponsible’ villagers.

Images of forest loss in Kissidougou are also reinforced as part of processes of ethnic distinction, which depend on colonial portrayals and their subsequent incorporation into local political discourse. From the earliest, colonial constructions of ethnic difference among Kissidougou’s populations rested partly on stereotypes concerning their environmental behaviour. The Kissi were seen as ‘forest people’ like other groups further south (Toma, Guerze), with cultural proclivities towards forest conservation. Kissi were seen as more forest loving largely because of the centrality to local life of the ‘forêt sacrée’ (or ‘secret society’) initiation ceremonies which they held in their forest islands.

Kissi ‘sacred’ forests were upheld as veritable islands of nature conservation amid secular destruction around them (Aubreville 1939), just as they are by today’s policy makers. This was despite the reality that only part of a forest island would be devoted to the ‘sacred’ forest institution, medicines and powers, and that the ‘sacred’ technologies would usually be installed only after the forest island had itself been formed.



As ‘forest people’, Kissi were contrasted ethnically with the more northerly ‘savanna people’ of Maninka origin, including Kissidougou’s Kuranko populations. In the context of historical and ongoing southwards Maninka migration, their fire-setting in savanna farming, honey collecting, and hunting was considered responsible for southwards savannisation (Adam 1948). Where Kuranko lived within forest islands, this was perceived as learning from the Kissi (Administrateur de Cercle 1913), as it is by modern environmental projects. Yet Kuranko themselves associate prosperity and fortune with inhabiting a forest island (haraye ye tu le ro), and where Maninka immigrants have moved into Kissi villages, they have usually been incorporated into Kissi society and land management. That the majority of supposedly ‘forest’ Kissi families can trace descent from a Maninka family of savanna origin clearly undermines such arguments of ethnically driven environmental degradation.

During the First Republic, Sekou Touré’s state regime encouraged villages to move out of the ‘mystified obscurity’ of their forest islands into ‘the open’; into the ‘clarity’ and ‘modernity’ upheld by the regime’s cultural demystification policy, and into the roadside world more accessible to its demands (Rivière, 1969). This policy drew on and reinforced ethnic stereotypes, deepening their construction in terms of forest. Maninka self-representations often draw on the ideal of social clarity, of openness and simplicity in language and expression, and draw an explicit contrast between their clear ‘savanna language’ (kan gbe) and the secrecy and obscurity of the forest culture and languages which they find difficult to learn.

Many Kissi perceived Sekou Touré’s regime as Maninka-biased, and considered the attempts it made to evict them from their forests and suppress ‘sacred forest’ schools as attempts to disempower the institution which had hitherto defended the Kissi from Maninka domination, whether cultural or military. The political conditions from 1958-1984 therefore reinforced the significance of forest symbolism in Kissidougou’s local and ethnically charged political discourse.



In this context, both the present privileging of the forest, and the view that it is threatened as portrayed by the forestry service, coincide with the broader politico-ethnic interests of urban Kissi; interests heightened in the run-up to multi-party electoral processes beginning in December 1993. Sharing one forest – where the forest islands of neighbouring villages have come to touch each other – is one of the strongest metaphors of Kissi political solidarity, linked as it used to be to alliance in warfare and forest initiation. Accepting the idea that the Kissi region could (until even recently) have been united in one forest provides a politically appealing vision of unity, as does blaming Maninka immigration for forest loss. These views are most often voiced within the politically influential urban Kissi community, but can also be heard in villages when rural Kissi use environmental issues to make politico-ethnic points.

Distinctions between urban institutional and rural villagers’ perceptions of environmental change also derive from different valuations of vegetation quality. For urban observers and the forestry service, high value is accorded to large forest trees, whether for recent global reasons or for the commercial gains to be made from timber exploitation, which has recently become big business in Kissidougou.



Villagers do not share this valuation, not least because the forestry laws designed to regulate timber exploitation (preserve the environment) deny them all but an insignificant royalty from trees cut by outsiders in their forest islands. Their values are conditioned, instead, by the importance of different vegetation types and species in agriculture, gathering, settlement and tree crop protection and cultural practices, and in which lower bush fallow vegetation is frequently more useful than high forest (Leach and Fairhead 1993). The large trees of forest islands are, in fact, more the ‘fortuitous’ consequence of villagers’ environmental management for other reasons than a deliberately encouraged feature. While the felling of these trees may be of little consequence to villagers (or to forest area in the long term), to urban and official observers it epitomises, and thus reinforces their conviction of, environmental destruction.



The image of environmental degradation in Kissidougou is supported by apparently successful explanations for it in terms of local land use practices and their changing socio economic, demographic and institutional contexts. Just as the prevalent socio-cultural, institutional and financial structures lead certain readings of and methods for investigating environmental change to dominate, while excluding others from consideration, so these same structures influence the methods and theories brought to bear in understanding why the environment has changed.

Policy makers’ thinking has long been dominated by the view that local land use encourages savannisation and reduces savanna tree cover and soil quality. These apparent processes of degradation are readily observable in the short term: in, for example, the clearing and burning of wooded lands for farming, and the setting of fire by hunters and herders. But less attention is paid to processes of regeneration and the impact of local practices on them. In villagers’ experience, their land use has, in the long run, maintained or enhanced woody vegetation cover and soil quality. The logic of local cultivation practices which encourage the advance of forest in this region has been documented both by ourselves (Fairhead and Leach 1996b) and in neighbouring Cote d’Ivoire (Blanc-Pamard and Spichiger 1973). Villagers tend to consider themselves as improving once less productive lands, rather than reducing the productivity of once ‘naturally’ productive ones.

Nevertheless, the contrasting external image of local land use as inevitably degrading is combined with particular theories about the impact of demographic and social change to account for the long term degradation which policy makers believe to have taken place. Discussions in development circles of the links between population and environment, poverty and environment, and social organisation and environmental management have set terms of debate which guide causal interpretations by development personnel, consultants, and national institutions.

Given that it is explanations of supposed environmental degradation which are being sought, and given the prevailing intellectual, social and fiscal structures which condition causal analysis, all but the dominant strands of thinking within these debates tend to be suppressed at the project level. Thus it is Malthusian views of the relationship between population and environment, the deduction that impoverishment forces villagers to draw down their natural resources, and the notion of a ‘tragedy of the commons’, which are used to explain increasing environmental degradation in Kissidougou.



Environmental degradation is attributed to assumed demographic trends by policy makers who believe that, since local land use is degrading, more people must mean more degradation, principally through extra upland use. An image of low pre-colonial population densities is commonly linked to the supposed existence of extensive forest cover then, and rapid population growth during this century (and now refugee settlements) are held to account for forest decline. Short fallows and long cultivation periods on savanna uplands are often taken as evidence of modern population pressure. That local farmers use intensive cultivation practices for positive ecological and economic reasons, unrelated to population pressure, is not considered. Nor does the possibility that population growth could enable environmental improvement receive attention. Yet in Kissidougou where there are more villages, there are more forest islands, and more people can mean that there is more intensive, soil and vegetation enhancing savanna cultivation and more generalised fire control.

Socio-economic theories to explain supposed recent environmental degradation attribute it partly to modern poverty, forcing villagers to sacrifice sustainable long term resource management in favour of short term uses assumed to be degrading. Recent environmental degradation is also explained through the idea that modern resource use is disorganised and individualistic; a vision shared by many local administrators as much as external consultants and university academics.



In many versions of this narrative, a picture of people in greater ‘harmony’ with their forested environment is projected onto the pre-colonial period; a harmony maintained either by efficacious traditional authority (Green 1991; Stiegelitz 1990) or, in more sophisticated terms, by the integration of fire control within intra and inter-village social, cultural and political relationships (Zerouki 1993). An armoury of factors is held to have ruptured this controlled harmony, including socio-economic change, the weakening of traditional authority, new economic and cultural aspirations and social divisions, and the alienation of local resource control to state structures.

The logical policy implication is that resource use can be rendered sustainable by improving forms of ‘regulation’, ‘authority’ and ‘organisation’, whether by greater state control (e.g. over timber cutting and fire) or, in recent policy emphasis, by ‘rebuilding’ community institutions. These dominant social and demographic explanations for degradation, and the idea of degradation itself, seem to be mutually sustaining. From within this complex the actual history of peoples environmental use and the complex influences on it fail to receive serious attention.



The institutional and financial structures in which social science is applied to environmental problems in Guinea strongly support such uncritical explanations of degradation. Studies are commissioned by donor agencies and projects who need (or at least, must be seen to have sought) socio-economic information to help them tackle the environmental problems integral to their institutional survival in more ‘appropriate’ and ‘participatory’ ways. The environmental problem is thus built into the very terms of reference of consultants who have neither the time nor the social position to investigate village natural resource management and its changes on any other terms.

This problem is not necessarily solved when consultants are Guinean, or even working in their own areas; indeed it can be compounded by the urban intellectual images which such local consultants bring to bear. Furthermore, as the dominant social and demographic explanations of environmental degradation are the stuff of academic debate, consultancy reports phrased in their terms gain easy acceptance and credibility.

The interface between environmental-development agencies and villagers, which has developed over more than half a century and often in antagonistic ways, renders the communication of local environmental experiences highly problematic. Villagers, faced by questions about deforestation and environmental change, have learned to confirm what they know the questioners expect to hear. This is not only through fear, politeness and the awareness that the truth will be met with incredulity, but also through the desire to maintain good relations with authoritative outsiders who may bring as yet unknown benefits; a school, road or advantageous recognition to the village, for example. In such discussions, the historical ecology that villagers portray is as politically inflected as in their oral histories concerning settlement foundation, where images of initial vacancy (high forest, empty savanna, or abundant wild animals) often justify the firstcomer status of current residents (cf. Dupré 1991; Hill 1984).



Like the prefecture administration, many village authorities realise the benefits which can accompany community participation in environmental rehabilitation, and in this context may publicly agree to the ‘urgent need’ to plant trees, establish village environmental management committees and so on. Nevertheless, acceptance is not without anxiety: over losing land to ‘project’ trees; over losing control over management of local ecologies to outsiders ignorant of their specificities; and over the unknown future demands that apparently generous projects of unknown origin and intent, huge financial resources and foreign interests may later exact. Everyday forms of resistance thus frequently underlie overt participation: letting project tree nurseries and plantations burn in the dry season, for example, and ensuring that necessary fires are set in ways contrary to agreed project procedures.

It has been surprising to us how little the personal lifetime experiences of development workers from the prefecture influence the way that Kissidougou’s environment has come to be perceived. This may be because personal environmental histories have too limited a spatial coverage to challenge a generality, or because unbroken personal histories are themselves rare: state officials are frequently transferred and are in preference posted to areas with which they are unfamiliar, so they have frequently been away from their childhood village environments for long periods. Such people almost invariably justify their perceptions of historical deforestation with examples drawn from roadsides and urban peripheries, with which they have more continual familiarity but which in Kissidougou are the proverbial exceptions to the rule.



Scientific challenge to the dominant analysis in Guinea is also rare. This is partly because the scientific information and ecological theory which questions the derived savanna model, and which often proves to support the farmers’ explanations which we have investigated, is dispersed among different disciplines and their specialist academic journals. These are largely inaccessible to policy makers and national academic institutions. Information from each discipline alone (e.g. botany, hydrology, soil science, demography and climate history) is insufficient to shift thinking in a sufficiently fundamental way; lack of inter-disciplinary criticism seems, indeed, to promote consistency.

In any case, little such discussion enters the information bulletins of multinational organisations (e.g. FAO), NGOs, development journals and the media; the sources on which most development personnel rely for environmental science information. Fundamentally, the precepts basic to local science which challenge conventional savannisation wisdom are not easily apprehended by researchers ill-disposed either to listen or to understand.

This environmental case illustrates in a particularly striking way how development problems and policies are constituted within diverse, seemingly disparate relations. The vision of environmental degradation in Kissidougou to which so many people are drawn for different reasons has, for a hundred years now, been sustained within their scientific, social, political, institutional and financial relationships. These relationships have evolved, showing the operation of power in a Foucauldian sense (Foucault 1976).



The intellectual, social, political and financial structures which sustain the vision of environmental degradation in Kissidougou form a constellation in which each element reinforces the others in a sticky web. It is too simple to suggest that mistakes are made in Kissidougou because erroneous information is uncritically inherited, although this has sometimes been the case. It is more that the same basic analysis is perpetually reconstituted over and over again within prevailing institutional, financial and explanatory climates. Nor is it necessarily the case that particular people or institutions are pursuing conscious and direct personal interests in using information for political or economic ends; rather, all are subject to and are the vehicles of the same conjuncture of intellectual, institutional and economic structures.

The degradation vision has evolved over time and in ways which mean that today it cannot be attributed only to donor agencies and their narratives. It is partly the product of a long history of interaction with and incorporation into local social and political processes, and is thus today partly sustained by them. This is not to say that villagers’ everyday ecological practice is influenced by the deforestation reasoning, but merely that their ecological reasoning is subjugated in much political interaction.

Views of degradation in Kissidougou are not sustained on the basis of ignorance, but through the continual production of supportive knowledge. ‘What has taken place... is the production of effective instruments for the formation and accumulation of knowledge – methods of observation, techniques of registration, procedures for investigation and research, apparatuses of control. All this means that power, when it is exercised through these subtle mechanisms, cannot but evolve, organise and put into circulation a knowledge, or rather apparatuses of knowledge’ (Foucault 1976: 102).

Those who are convinced of degradation do not lack data to support their convictions, and it is within this methodologically supported certainty that alternative methods and data sets have been disqualified as inadequate, naive, or unscientific.



The disjuncture between locally lived reality and the degradation discourse therefore has to be considered as a political as much as a methodological one. Local environmental experience and history are, as we have seen, not easily accessible across farmers’ interface with environmental agencies and urban intellectuals. Recent attempts to overcome problems of information transfer at the interface through taking a more participatory approach to research and development (eg. PRA) – as current in Kissidougou as elsewhere – are not the unproblematic answer they may first appear, without serious attention to altering the intellectual, institutional and financial structures which are implicated in the production of knowledge and of confidence in it.



* This paper is the result of our joint and equal co-authorship. It draws on research funded by ESCOR of the Overseas Development Administration (ODA) to whom we are grateful; opinions represented here are, however, our own and not those of the ODA. Many thanks are also due to our co-researchers in Guinea, Dominique Millimouno and Marie Kamano; to the villagers, administrators and project staff. Revised version of a paper published in R. Grillo and R.L. Stirratt (eds), Discourses of Development. Berg Press, Oxford, 1997.




Adam, J. G., 1948, ‘Les reliques boisées et les essences des savanes dans la zone préforestière en Guinée francaise’, Bull. Soc. Bot. Fr. 98, 22-26.

Adam, J. G., 1968, ‘Flore et végétation de la lisière de la forêt dense en Guinée’, Bulletin IFAN Series A, 30, 3, 920-952.

Administrateur de Cercle 1913, Rapport sur Kissidougou 1913, Guinean National Archives, Conakry, 2D175.

Aubreville, A., 1939, ‘Forêts reliques en AOF’, Revue de Botanique Appliqée et Agronomie Tropicale, 1939, 215.

Aubreville, A., 1949, Climats, forêts et désertification de l’Afrique tropicale, Paris.

Behnke, R. and I. Scoones, 1991, ‘Rethinking rangeland ecology: implications for rangeland management in Africa’, ODI/IIED Issues Paper no 33.

Blanc-Pamard, C. and R. Spichiger, 1973, ‘Contact forêt-savane et recru forestier en Côte d’Ivoire’, L’Espace Géographique 3, 199-206.

Bledsoe, C. 1990, ‘No success without struggle: social mobility and hardship for foster children in Sierra Leone’, Man (N.S.) 25, 70-88.

Boserup, E. 1965, The Conditions of Agricultural Growth: agrarian change under population pressure. London: Earthscan.

Brooks, G.E., 1986, ‘A provisional historical schema for Western Africa based on seven climatic periods,’ Cahiers d’Etudes Africaines 101-102, XXVI-1-2: 43-62.

Chambers, R. 1983, Rural Development: putting the last first. London: Longman.

Chevalier, A., 1909, Rapport sur les nouvelles recherches sur les plantes à caoutchouc de la Guinée française, Senegalese National Archives, 1G276.

Darling, P.J. 1993, Updating some African population myths. Paper presented at the First World Optimum Population Congress, 8-11 August 1993.

Davies, S., 1992, ‘Green conditionality and food security: winners and losers from the greening of aid’, Journal of International Development 4(2).

Davies, S. and M. Leach, 1991, ‘Globalism versus villagism: food security and the environment at national and international levels’, IDS Bulletin 22(3).

Dublin, H., A. Sinclair and J. McGlade, 1990, ‘Elephants and fire as causes of multiple stable states in the Serengeti-Mara woodlands’, Journal of Animal Ecology 59, 1147-1164.

Dupré, G., 1991, Les arbres, le fourré et le jardin: les plantes dans la société de Aribinda, Burkina Faso, in G. Dupré (ed), Savoirs pay-sans et dévéloppement, Karthala-ORSTOM, Paris, 181-194.

FAO, 1990, Interim report on forest resources assessment 1990. Committee on Forestry, tenth session, 24-28 September 1990, Rome.

Fairhead, J. and M. Leach, 1993, Degrading people? The misuse of history in Guinea’s environmental policy. Paper presented at the African Studies Association meeting, Boston, 4-7 December 1993.

Fairhead, J., and M. Leach, 1995, ‘Reading forest history backwards: the interaction of policy and local land use in Guinea, 1893-1993’, Environment and History 1(1), March 1995.

Fairhead, J. and M. Leach, 1996a, Misreading the African Landscape: society and ecology in a forest-savanna mosaic. African Studies Series. Cambridge and New York: Cambridge University Press.

Fairhead, J. and M. Leach, 1996b,‘Enriching landscapes: social history and the management of transition ecology in Guinea’s forest- savanna mosaic’, Africa 66(1), 16-46.

Fairhead, J. and M. Leach, 1998, Reframing Deforestation: Global analyses and local realities – studies in West Africa. London: Routledge.

Fairhead, J. and M. Leach with D. Millimouno and M. Kamano, 1992a, Forests of the past? archival, oral historical and demographic evidence for Kissidougou Prefecture’s vegetation history, COLA Working Paper 1, Kissidougou.

Fairhead, J. and M. Leach with D. Millimouno and M. Kamano, 1992b, ‘Managed productivity: the technical knowledge used in local natural resource management in Kissidougou Prefecture, COLA Working Paper 3, Kissidougou.

Foucault, M., 1976, ‘Two lectures’, in C. Gordon (ed), Power/Knowledge: selected interviews and other writings 1972-1977, 78-108. The Harvester Press.

Green, W. 1991, Lutte contre les feux de brousse. Report for project DERIK, Dévéloppement Rural Intégré de Kissidougou.

Grégoire, J.M., S. Flasse et J.P. Malingreau, 1988, Evaluation de l’action des feux de brousse, de novembre 1987 à février 1988, dans la région frontalière Guinée-Sierra Leone, Projet Régional FED-CILSS-CCR ‘Surveillance des Ressources Naturelles Renouvables au Sahel-Volet Guinée’, ISPRA: EEC

Guinée Service Forestier, Rapport annuel de fin d’année 1932, ANS 2G32(70).

Hill, M. 1984, ‘Where to begin? the place of the hunter founders in Mende histories’, Anthropos 79, 653-656.

Hoben, A. 1993, The political economy of land tenure and environmental policy in Ethiopia. Draft summary of paper presented at African Studies Association meeting, Boston, December 1993.

Holling, C. S. 1973, ‘Resilience and stability of ecological systems’, Annual Reveiw of Ecology and Systematics 4, 1-23.

Houghton, R.A., J.D. Unruh and P.A. Lefebre, 1993, ‘Current land cover in the tropics and its potential for sequestering carbon’, Global Biogeographical Cycles 7(2), 305-320.

Leach, M. and R. Mearns (eds), 1996, The Lie of the Land: challenging received wisdom on the African environment. London: James Currey Publishers Ltd. and Heinemann.

Leach, M. and J. Fairhead, 1993, ‘Whose social forestry and why? people, trees and managed continuity in Guinea’s forest-savanna mosaic’, Zeitschrift fur Wirtschaftsgeographie 37(2), 86-101.

Noy-Meir, I., 1982, Stability of plant-herbivore models and possible applications to savanna, in B. Huntley and B. H. Walker (eds), Ecology of Tropical Savannas. 591-609.

Ostrom, E., 1990, Governing the Commons: the evolution of institutions for collective action. Cambridge: Cambridge University Press.

République de Guinée, 1988, Politique forestière et plan d’action, TFAP 1988. Conakry.

Rivière, C., 1969, ‘Fétichisme et démystification: l’exemple guinéen’, Afrique documents 102(3), 131-168.

Rivière, C., 1976, ‘Bourgeoisies du tracteur’, Revue Française d’Etudes Africaines 123, 74-101.

Roe, E., 1991, ‘"Development narratives" or making the best of blueprint development’, World Development 19 (4), 287-300.

Roe, E., 1995, ‘Except-Africa: postscript to a special section on development narratives’, World Development 23 (6), 1065-1070.

Sprugel, D.G., 1991, ‘Disturbance, equilibrium, and environmental variability: what is "natural" vegetation in a changing environment?’ Biological Conservation 58, 1-18.

Stiegelitz, F.V., 1990, Exploitation forestière rurale et réhabilitation des forêts: Premiers résultats d’un projet de recherche inter-disciplinaire en Haute-Guinée, Berlin.

Thompson, M., T. Hatley and M. Warburton 1986, Uncertainty on a Himalayan Scale. London: Ethnographica.

Tiffen, M., M. Mortimore and F. Gichuki, 1993, More People, Less Erosion: environmental recovery in Kenya. Chichester: Wiley.

Zerouki, B., 1993, Etude relative au feu aupres des populations des bassins versants types du Haut Niger, Programme d’Amenagement des Bassins Versants Types du Haut Niger, Conakry.