World Development, Vol. 23, No. 6, pp. 1023-1035, 1995 Copyright 0 1995 Elsevier Science Ltd Pergamon Printed in Great Britain. All rights reserved 0305-750x/95 $9.50 + 0.00 0305-750X(95)00026-7 False Forest History, Complicit Social Analysis: Rethinking Some West African Environmental Narratives

JAMES FAIRHEAD University of London, U.K. and MELISSA LEACH* University of Sussex, Brighton, U.K.

Summary. - Social science analysis has helped to explain the rapid and recent deforestation supposed to have occurred in Guinea, West Africa. A narrative concerning population growth and the breakdown of past authority and community organization which once maintained “original” forest vegetation guides pol- icy. In two cases, vegetation history sharply contradicts the deforestation analysis and thus exposes the assumptions in its supporting social narrative; assumptions stabilized within regional narratives based more on Western imagination than African realities. For each case and then at the regional level, more appro- priate assumptions are forwarded which better explain demonstrable vegetation change and provide more appropriate policy guidelines.

I. INTRODUCTION explanations of it, our cases pave the way for rethink- ing people-environment relationships in this region. This paper examines social science analyses that We do this by forwarding alternative sets of assump- are being used to explain environmental degradation tions stabilized within narratives which better fit and inform policy responses to it. We focus on two the facts. cases pertinent for exploring the production of applied The specific cases considered here concern social science knowledge about people-environment Guinea’s forest margin zone. They articulate, in dif- relations. They exemplify the type of social analysis ferent ways, the position that local community insti- often brought to bear to explain environmental tutions were once better capable of controlling degradation in Africa, yet it can be demonstrated that environmental resources than they are today, and thus what they explain so successfully has not actually of maintaining a forested environment and resisting taken place. pressure toward its degradation. This articulation Our examples clearly expose a spectrum of enables supposed forest loss to be explained in terms assumptions on which social science analyses - of “institutional breakdown.” An armory of purported whether or not carried out by social scientists as such factors is called to account for such social rupture -tend to draw. These assumptions have strength and credibility in large part because they are linked together, diffused and stabilized within “narratives” *This article is the result of our joint and equal co-author- (Roe, 1991), that is, stories of apparently incontro- ship. It is based on research funded by ESCOR of the vertible logic which provide scripts and justifications Overseas Development Administration, whom we gratefully for development action. But once dissected from the thank. Opinions represented here are, however, the authors’ own, not those of the ODA. Many thanks are also due to our reality they seek to construct, these explanations Guinean co-researchers Dominique Millimouno and Marie reveal instead how the applied social sciences can be Kamano. to Jean-Louis HelliC for field assistance around used to lend weight to popular Western perceptions Ziama, to the villagers we worked with in both case studies, about African society and environment - a mythical and to our collaborating institutions: the Ministere de reality which development interventions are acting to I’Enseignment Superieur; the Direction National des For&s recreate in vain. By stripping away the explained from et de la Chasse, and Projet DERIK, Kissidougou.

1023 1024 WORLD DEVELOPMENT whose results seem so evident in a degraded land- and fire-setting practices, preserving only the belt of scape. These include socioeconomic change and com- forest around their villages to protect their settlements mercialization, increasing mobility, the weakening of from fire and wind, to give necessary shade to tree traditional authority, more individuated farming, the crops, to assist fortifications and hiding, and to pro- new economic and cultural aspirations of the young, vide seclusion for secret ritual activities. They argue new social cleavages, the alienation of local resource that today’s climate would support general forest control to state structures, and the emergence of “anar- cover, and infer from the presence of “relic” forest chic” charcoal, fuelwood and timber businesses to islands that it once did: supply the urban market. The impact of migration is added to these arguments: Eco-ethnic integration once At origin, the forest between Kissidougou and Kankan associated with “forest people” with supposedly for- was.. .a dense, humid, semi-deciduous forest. The trigger est-benign lifestyles has been disrupted by the immi- of degradation is. .the farming system and the fragility of gration or influence of “savanna peoples.” Overlaying climate and soils in tropical regions. Some primary for- all is the specter of population growth, as viewed mations still exist, however, in the form of peri-village forest islands and gallery forests on the banks of water through a Malthusian lens. Foreign observers today courses. These forest islands show the existence of a tend to date such socioenvironmental disruption to the dense forest, which is today replaced in large part by notorious regime of Guinea’s first republic (1958-84) degraded secondary forest. All the stages of degradation under SCkou Tour& imaging the colonial period as are represented: wooded savanna, bush savanna and environmentally friendly, while nationals tend to look grass savanna (Kan II, PIan d’operarion, 1992. pp. to the precolonial period to find “good’ society and 6-7).’ environment. As if to make the point, one scholar forced the social-environmental Eden back to that Deforestation is considered to provoke problems at period documented by the 13th century Arab geogra- several levels, rendering it an urgent policy concern. phers, i.e. a period where his personal moral sympa- At the local level it leads to soil degradation and ren- thies lay (Zerouki, 1993). ders farming less productive and sustainable. At the Social sciences have no monopoly over these regional level - the upper watershed of the Niger social-environmental visions in which a forest past river - deforestation is thought to have caused has become a moral past. They are shared by many irregularities in downstream river flow and in rain- local administrators and school teachers, as much as fall. In addition, it is contributing to global external consultants and university academics. The warming. Something must be done. production of history serves many ends. What will Social analysis has always been instrumental in become clear is that social scientists have been com- explaining this problem and its recent acceleration. In plicit in producing a view of history as one of increas- the early part of this century, the celebrated French ing tension from a harmonious past. Treating this past colonial botanist, Auguste Chevalier, considered as a model and set of objectives for the resolution of greater movement and trade during the postoccupation today’s tensions, they have been forging links period to be responsible for an increase in fire-setting between social and environmental conditions in a way from a previous, less forest-harmful level (Chevalier, that assists in relieving those subjected to their study 1909). He considered that inhabitants conserved the of what little resource control they have. forest islands for cultural reasons, presumably in a sea of otherwise degraded profanity (Chevalier, 1933). In 1948, Adam published the view prevalent in earlier 2. CASE 1: FOREST ISLANDS OF archives that the Mandinka were a “savanna” people KISSIDOUGOU who had migrated southward into the forest zone, and created savanna there (Adam, 1948). In doing so, they (a) The deforestation narrative reportedly forced the original “forest people,” more benign to that resource, south and further into the for- Kissidougou looks degraded. The landscape is est zone. largely savanna, especially open in the dry season More recently, professional social scientists have when fires burn off the grasses and defoliate the few focused on environmental issues in Kissidougou, usu- savanna trees. Nonetheless, rising out of the savanna ally in the pay of international or bilaterally funded and surrounding and hiding each of the prefecture’s environmental programs. One team, responsible for villages, are patches of immense semi-deciduous structuring the European Community-funded Niger humid forest. Scientists and policy makers consider river protection program, illustrate this focus thus: these forest “islands” and the strips of streamside gallery forest to be relics of an original, formerly much Our questions sought to explain the deterioration of the more extensive, dense humid forest cover. Inhabitants environment, viz: erosion and soil impoverishment. have, they suppose, progressively converted forest the drying up of water sources, the origin and nature of into “derived” savanna by their shifting cultivation forest destruction. the origm of perverse use of bush FALSE FOREST HISTORY 1025

fire.. .Parallel to the physical causes of soil erosion, there people with regard to fire” (Green, 1991, p. 20). In a are others of a social, political and religious nature. We typically racialist way, agency for degradation is dif- can suppose a strong relationship between soil erosion, fused into the ambiguity between culture and origin. environmental degradation and the break-up and impov- A study of an area just over the eastern border of erishment of socio-economic structures and relations. Kissidougou, while somewhat cynical of the crisis Environmental management is strongly linked to the state of socio-cultural structures.. .The more a community is in mold of environmental analysis, nevertheless claims equilibrium at the level of social organisation, the health- that: ier is the nature of its relations with the environment. There is a dialectical relationship between social, politi- The degradation of forests - always qualified as cal and religious institutions and ecological equilib- “explosive” - has continued in an accelerating fash- rium. ..In these communities,. ..the existence of the ion...Peasant exploitation is correctly identified as the living is above all justified by a more or less good man- principal factor of destruction, but in general, the mea- agement of what the ancestors have left to them. This sures taken [since colonial times] have only treated the management is inscribed in the collection of laws, con- symptoms, The social reasons for fire setting in hunting crete and abstract, rational and irrational, which, once are. ..closely linked to growing tendencies of commer- disturbed from the exterior, can be the cause of a deteri- cialisation and monetarization in the rural milieu. This oration which manifests itself as much at the level underlines the loss in importance of traditional organisa- of social, religious, political and economic institutions, tions of hunters which, to date, are marked by an anti- as at the level of the environment (Programme commercial character. [Pasture will be threatened by] d’Amenagement des Hautes Bassins du Fleuve Niger, growing immigration of herders into the region, a conse- n.d.. pp. 4-7). quence of the degradation of pastures in the traditional herding regions (e.g., Fouta Djallon). Traditional struc- In a second study, devoted to local fire-setting, the tures which regulate theexploitation ofnatural resources, author aimed to give an inventory of cultural traits most often of pre-Islamic origin, incorporate a series of conservation aspects. Some still operate.. .but a change is which function around the practice of fire. “We have beginning to show itself: a process of social change tried to retrace the transition from a traditional prac- which implies a dissolution of traditional regulative tice to ‘modern’ practice. Our hypothesis was that the structures which are not easily reconcilable with the ‘fire social system’ instituted itself as such, in destroy- commercialisation trends which are more and more ing its host ‘system’, the traditional one” (Zerouki, marked in the region (Stieglitz, 1990, pp. 54. 70. 77). 1993, p. I). In short, the author argues that “moder- nity” is responsible for disrupting the once successful The author, who considers Islam to have disrupted integration of fire control within diffuse sets of intra this “pre-Islamic” tradition, incorporates more agro- and intervillage village social, cultural and political demographic explanations into her explanatory mix: relationships. He finds that “Degradation seems to be recent” and that “it accelerates with the development The period of cultivation being too prolonged or the of an urban network...and population growth.” The fallow period too short, there is too great a loss in the study proposed “solutions to social dysfunctioning” nutritive materials leading finally to an irrevocable degra- (Zerouki, 1993). A coresearcher on this same study dation of the soils. The fallow period is limited to 5-10 expands on the causes of such “dysfunctioning”: years. A tendency for land shortage can be seen (Stieglitz, 1990. p. 71).

According to inquiry on the one hand from elders.. .nnd on the other by IFAN in 1968, the whole region was cov- This is the position on demographic change held by ered with forest about 99 years ago. corresponding to the most analysts. Ponsart-Dureau, for example, an agron- Samorian period. War chiefs used fire for better visibility omy student advising a nearby project, considers that: and for encampments. The introduction of the locomo- tive during the colonial period had a serious impact on the around 194.5. the forest, according to the elders, reached vegetation. Since independence, there has been demysti- a limit 30 km north of Kissidougou town. Today, its fication of sacred forests and of islands considered once northern limit is found at the level of Gueckedou- as cult places, the installation of wood mtlls, and brick- Macenta, thus having retreated about 100 km... mnkmg Nomadic farmmg and herding, uncontrolled Demographic growth forces the villagers to exploit their hush fire. forest tire, and runaway demography. aggra- land completely, and to practice deforestanon whtch dis- vate the process of vjegetation degradation already begun equilibriates the natural milieu (Ponsnrt-Dureau. 1986. (Fobna ef trl.. I Y93.p. 19). pp. 9-10 and 60).

Other recent expert vtews have drawn on conven- Thus in different ways, each of these analyses tional social analysis to assert once again that the contributes to a narrative now as prominent in Kuranko people (who speak a Mandinka dialect) are a Kissidougou’s education and administrative circles as savanna people and brought bush lire practices with it is in social science analyses. Once Kissidougou had them when they pushed the Kissi further south. “As an extensive forest cover, maintained under low pop- forest people. the K~ssi arc not as careless as savanna ulation densities and by a functional social order 1026 WORLD DEVELOPMENT whose regulations controlled and limited people’s vation of forest islands in favor of “the community.” inherently degrading land and vegetation use. The breakdown of such organized resource management (b) The counternarrative under internal and external pressures, combined with population growth, has led to the deforestation Examining how vegetation has actually changed in aparently so evident in the landscape today. Observers Kissidougou is a necessary first step in evaluating invariably consider degradation as a recent, ongoing these social science analyses. Fortunately, a number and aggravating problem. The social and economic of historical data sources make this possible - changes are, like “runaway demography,” always seen sources ignored or deemed unnecessary by social ana- to be accelerating out from a “zero point” (the arche- lysts convinced of the degradation they were explain- type “tradition” so dear to Malinowskian social ing.* Aerial photographs exist for Kissidougou which anthropologists and the object of description in old clearly show the state of the vegetation in 1952-53. ethnographies). A host of indicators is drawn upon to These provide incontrovertible evidence that during support ideas concerning recent and ongoing degrada- this recent, supposedly most degrading period, the tion, such as rainfall decline since the mid-1950s the vegetation pattern and area of forest and savanna have drying-up of certain water sources, and more. in fact remained relatively stable. Changes which have Policy implications have followed logically from occurred do not involve forest loss; rather there are the assumptions contained and stabilized within this large areas where forest cover has increased, and narrative, and have changed little since its first elabo- where savannas have become more, not less, woody. ration in the early colonial period. The first policy Forest islands have formed and enlarged, and in many emphasis is on the reduction of upland farming - areas, savannas evident in the 1950s have ceded to seen as inherently forest and soil degrading and secondary forest vegetation. becoming more so under greater individualization and To examine vegetation change further back, we population growth - in favor of swamp farming. reviewed descriptions and maps of Kissidougou’s What upland use must remain needs to be rationalized landscape made during the early French military occu- and intensified (e.g., through “model” agroforestry pation (roughly the 1890s to 1910). as well as indica- systems, reorganization of tenure and fallow systems). tors of past vegetation that emerge from oral history Second, policies have focused on bush-fire control and accounts of everyday life in the youth of today’s through externally imposed prohibitions, regulations elderly people. These sources make clear that what and practices (e.g., early-burning). Third, policies was true for 1952-94 is equally true for 1893-1952. have attempted to control deforestation both through Moreover, villagers suggest, quite contrary to policy prohibitions on the felling of a list of protected tree interpretations, that they established forest islands species (largely those forest species commercially around their settlements, and that it is their work valuable for timber and most representative of the which encourages the formation of secondary forest “original” forest cover) and through the reservation of thicket in savanna. In 27 of the 38 villages we investi- certain forest patches. Fourth, there are attempts at gated, elders recounted how their ancestors had forest reconstitution through tree planting in village founded settlements in savanna and gradually encour- territories. Uniting these policies is their recourse to aged the growth of forest around them. technology “packages” well established in the region Earlier documentary sources from the 178Os-1860s such as inland valley swamp development and tree do not suggest extensive forest cover; indeed they sug- planting from nurseries. Uniting them, too, is their gest the opposite. Both Harrison, traveling to Kissi attempt to establish or reestablish control and organi- areas (c. 1780, see Hair, 1962), and, as we shall see in zation in resource management; although with the next case, Seymour (1859/60) in Toma country changes in development philosophy, there have been south-east of Kissi, describe short grass savannas and changes in the levels deemed appropriate. Thus in an absolute scarcity of trees in places which now sup- Guinea’s colonial and first republic periods, the degra- port extensive dense humid forest. Sims (1859/60), dation narrative justified removing the villagers’ (dys- speaking of the area just to the southeast (between functional, incapable) “control” over resources in Beyla and Kerouane) writes that: “There are no trees; favor of the state. In bush tire, upland use, timber- the whole country is prairie; for firewood the people felling and forest reservation policies, government have to substitute cow dung, and a kind of moss which administrations took over resource tenure and regu- grows abundantly in that country.” This picture of lated local use through permits, fines and at times mil- less, not more, forest cover in the 19th century is sup- itary repression. More recently, emphasis has shifted ported by several sets of early oral history data. All the somewhat toward patching-up, reconstituting or above villages claiming foundation in savanna were replacing broken community control over resources: established during or before the 19th century. Several “gestion de terroir villageois” approaches provide a village foundation stories in the south refer to conflicts context for village-level planning of bush fire, upland triggered by the scarcity of construction wood, seem- and forest use, “participatory” tree planting, and reser- ingly bizarre given the present forest and thicket veg- FALSE FOREST HISTORY 1027 etation, and in certain areas savanna grasses are said tenance of long-term productivity is in many cases to have changed from those associated with drier cli- built into short-term production patterns; whether car- mates to those associated with wetter ones. ried out for oneself, one’s household or one’s com- It appears, therefore, that social science analyses in pound, these improvements frequently interact with Kissidougou have been providing explanations for others - spatially or temporally - so that the com- forest loss which has not actually been taking place. In bined effect on resource enrichment is greater than the doing so, they have supported a vegetation-change sum of their parts. Thus, the fires set in the early and narrative quite at odds with -even the reverse of - mid-dry season by hunters to clear small hunting more demonstrable environmental “facts.” This casts grounds, and by others to protect property and fallows, into question the relationships between society, create barriers to more devastating later fires; and the demography and environment valorized in these small tree crop plantations which people make and analyses. As we suggest now, there are other ways of protect behind their kitchen gardens add to the cre- conceiving of these relationships - countemarra- ation of the village forest island. For much “resource tives, if you will - which better fit and explain vege- management” there is no need for village or higher tation history as demonstrated.3 level management structures to “regulate degrading The first reconception involves recognizing that pressures.” Nevertheless, village authorities do inter- local land use can be vegetation-enriching as well as vene in certain vegetation-influencing activities - degrading. It can (and often does) serve to increase the e.g., in managing early-burning around the village, in proportion of useful vegetation forms and species in protecting palm trees, in imposing cattle-tethering the landscape according to prevailing local values and dates, and in coordinating the fallow rotations of farm- productivity criteria. This has often meant increasing ers’ contiguous plots in some Kissi areas. Village and the prevalence of forest forms in a once more savanna higher level organizations also exert control over landscape. Thus, for example, villagers have encour- external factors which influence the agricultural envi- aged the formation of forest islands around their vil- ronment, such as in negotiating with prospective Fula lages for protection, convenient shelter for tree crops (Peuhl) pastoralist settlers or representatives of the and sources of gathering products, and the conceal- forest service. ment of ritual activities. They have achieved this both In this context, socioeconomic change has been through everyday use of village margin land (for articulated in shifts in landscape enrichment priorities instance, in the thatch and fence-grass collection and and in the composition of a continued resource man- cattle-tethering which reduce flammable grasses, and agement constellation. Villagers have, for example, in the household waste deposition which fertilizes the adapted forest island quality to suit changing socio- forest successions beginning to develop), as well as economic conditions and commercial signals - man- through deliberately applied techniques (such as aging them as fortresses during precolonial warfare, planting forest-initiating trees and cultivating the mar- extending them for coffee planting when this became gins to create soil conditions suitable for tree estab- profitable, and abandoning coffee in favor of fruit tree lishment). In addition, on the slopes and plateaux and gathering-product enrichment as prices fell again. between forest islands, local farming and fire use prac- Urban employment opportunities, youth emigration tices tend to maintain existing woody cover, and and more individual economic opportunities have upgrade soils and vegetation from savanna to forest contributed to changes in farming organization, but conditions. Much farming is concentrated on land that today’s smaller farm-households use and improve fal- farmers have improved, whether by long-term alter- lows as large compound ones did earlier, and modem ations to edaphic quality through habitation, garden- women’s individualized, commercial food cropping is ing and gardening-like cultivation; or by shorter-term concentrated in the forms of upland gardening that fallow improvement through intensive cattle-grazing, upgrade soils and vegetation (Leach and Fairhead, seed-source protection, the multiplication of savanna 1995). Village-level authorities have played a contin- trees from suckers, or distributing forest-initiating uing, though shifting role within this historically flex- creepers. These forms of knowledge and practice are ible and diverse management constellation. There found among all of Kissidougou’s ethno-linguistic have been many social and economic changes, and groups. There seems little basis for distinguishing there are many new social and economic problems, between “forest” and “savanna” people. but these changes are rendered visible in the landscape A related reconception concerns the character of largely through changing land use and management natural resource management “organization.” priorities, not through organizational “breakdown” Environmental management in this region seems to and vegetation degradation. depend - and always depended - less on commu- Explaining demonstrable vegetation change also nity-level authorities and sociocultural organizations suggests relationships between demographic and (which might be “threatened” by social change), than environmental change very different from the “rapid on the sum of a much more diffuse set of relations; a population growth-deforestation” relationship upheld constellation more than a structure. Indeed, the main- by the policy narrative. Despite the problems of recon- 1028 WORLD DEVELOPMENT strutting precolonial populations, evidence certainly Viewing people-environment relations in terms does not support the idea of dramatic population of landscape enrichment-through-use by a diverse growth or even steady one-way increase. Comparing resource management constellation responding to census data suggests that Kissidougou’s rural popula- changing incentives thus better explains (provides a tion has increased by only 70% since 1917. Growth counternarrative which better fits) demonstrable veg- pockets have been concentrated around Kissidougou etation and population history. Policies conceived town and major road axes, and in many areas popula- within the degradation narrative have sometimes tion has remained almost stagnant. Precolonial evi- undermined these relations, as well as created more dence suggests that certain areas had early 19th general problems for villagers. In removing local con- century rural populations significantly higher than trol over resources, they have sometimes interfered today, and suffered radical depopulation during late with local management of them. In the north, for 19th century wars. Indeed oral accounts, explorers’ example, external fire control and prohibition pre- reports, early 16th-18th century documents which vented villagers operating their sequenced manage- mention the region, and broader regional history and ment-throngh-use strategies,4 forcing clandestine archaeology combine to suggest that Kissidougou had coping strategies and rendering village and plantation relatively high farming populations from the 16th cen- protection more difficult. Removal of local resource tury and long before. There is clearly as little evidence tenure has reduced villagers’ abilities to profit from for dramatic population increase in the present century past enrichment activities (e.g., in selling their forest from a low precolonial baseline as there is for dra- island trees for timber) and their incentives for further matic forest loss. landscape enrichment. The implementation of repres- In this context, Kissidougou’s forest increase sive environmental policies has in effect taxed rural trends might be supposed to relate to population stag- populations for supposedly harmful activities which nation or decline. This reversed argument, however, were, in fact, benign or beneficial. More recent still depends on the assumption that local land use approaches, which focus on decentralizing resource tends to convert forest and forest fallows to savanna, control by establishing village-level organization and and thus that more people means more forest loss. A management plans, actually risk undermining the counternarrative better fits evidence of local land-use existing flexible, diverse constellation of resource practices and vegetation history: from an earlier situ- management relations. When initiated by state agen- ation of greater savanna extent, there has been a cies with considerable foreign support and presence broadly positive relationship between the peopling of and predefined ideas about environmental dynamics, this region and its forest cover. First, as settlements are real decentralization can be undermined. Finally, but associated with the formation of forest islands, more by no means least, the investment in “redressing” villages mean more forest islands. This relationship Kissidougou’s supposed environmental degradation, has been modified by changes in population distribu- an investment reaching unprecedented levels amid tion and settlement patterns, with greater multiplica- current aid donor concerns, carries heavy opportunity tion of settlements and forest islands during the 19th costs in terms of other more pressing rural develop- century when dispersed settlement was a survival ment problems left unaddressed. strategy, than in the 20th when much population Vegetation history and its counternarrative of land- growth has been accommodated through the expan- scape enrichment entail different policy implications, sion of existing settlements, and indeed some consol- emphasizing support to proven local practices and idation linked to depopulation. Still, new settlements determinants of change. There are clearly many tech- and forest formation have more recently been associ- niques and land uses that serve to increase forest ated with the movement of village sites. Second, cover, and which could provide an effective basis for greater population density assists the control of fire, external support. In working with the local ecology of both by providing the necessary labor and by creating fire, soils, vegetation successions and animal dynam- the demand, filling the landscape with more places ics, these “integrated vegetation management” prac- (upgraded fallows, plantations, settlement sites) tices are more locally appropriate, integrated with the which people need to protect. In certain cases, the den- social matrix and thus more cost-effective in terms of sity of such protected sites of denser vegetation easily labor than are the forestry “packages” generally pro- enables the entire exclusion of fire from the territory. posed by outside agencies. Given that farming in the The districts where upland savannas have recently region is not inevitably degrading, environmental ceded to dense forest fallow vegetation correlate policy may look to support as well as to “rationalize broadly with the areas where population has grown. and regulate” agriculture, specifically to support those By contrast, low population densities make fire pre- upland farming practices which improve soils and fal- vention impossible, and are a major factor in the per- low vegetation rather than concentrate technical effort sistence of running fire in the north and of the exclusively on swamps. Fundamentally, rather than particular “living with fire” management strategies increase external intervention in the organization of used there. resource management within villages. the more FALSE FOREST HISTORY 1029 important priority is to create the enabling policy and conservation project (Baum and Weimer, 1992). The socioeconomic conditions in which local resource assumptions it forwards are stabilized within a narra- management constellations can act effectively. This tive not dissimilar to Kissidougou’s, involving grow- implies a shift on the part of environmental agencies ing populations of immigrant and indigenous farmers away from direction (through repression or organiza- who have lost “traditional” values and organizational tional restructuring as in assisted “community con- forms, and who are seeking and de-wooding forested trol”) toward recognizing and supporting the diverse land. institutions which are actually engaged in resource As in Kissidougou, a strong contrast is drawn management, and toward a more responsive role in between a forest people, the indigenous Toma providing requested services at the village level. (Loma), and a savanna people, in this case the Konianke (Mandinka), whose immigration and savanna ways threaten the forest. Thus we read of the 3. CASE 2: THE ZIAMA FOREST RESERVE Toma that they are “largely fixed in their customary conceptions and habitual mode of life” (Baum and (a) The deforestation narrative Weimer, 1992). The authors explain that the Toma “historical and social evolution as a people in a forest Traveling south from Kissidougou, one enters the environment...favours a tendency to contemplation Upper Guinean forest region. Within Guinea this and sobriety.” These attitudes supported a lifestyle region is populous, and there are only two significant and traditional society which existed in harmony with intact forest blocks, the northern-most of which is the forest. The peripheral geographical situation of the Ziama.s Covering an area of about 120,ooO hectares, Toma in terms of communication, and the largely Ziama was designated a colonial forest reserve in uncommercialized nature of past economy, supported 1932, made an international biosphere reserve within these tendencies, so the argument goes. the “Man and the Biosphere” program in 1980, and is Nevertheless, it is maintained that the Toma have now the subject of a major World Bank financed con- lost their forest ways: “the forest has largely lost its servation project. Policy narratives concerning Ziama customary importance, in favour of an essentially reproduce those of Kissidougou to a significant extent, agricultural use of space. This evidences, without with one major scale exception: changes in the status doubt, profound changes in economic orientation, of a major forest block are at stake, and the conserva- especially among the Toma, ancient hunters and gath- tion concern is partly global. erers” (Baum and Weimer, 1992). The authors are The Ziama forest is considered to be under consid- surprised to find that women manage the principal erable threat as an important relic of a once much crop, rice, and this serves to reinforce the idea that the greater forest cover. As Table 1, drawn from an IUCN Toma have only just learned to farm; it “reflects, with- report on Ziama, indicates, forest cover in this part of out doubt the historical agricultural experience of a Guinea is now only 20% of what it was “at origin,” migratory farming, on small areas, only partially and the report emphasizes that the forest is regressing cleared.” This view builds on colonial perceptions that rapidly. Apart from the loss of biodiversity (of con- Toma had “a very primitive agriculture, quite anar- siderable international concern) this reduction is said chic, centred on pluvial rice based on forest clear- to be causing a drying-out of the local and regional ings.. .Those of the north have practically destroyed climate, evident in drier water sources and courses, the cover of trees, those of the south, in the valleys and thereby increasing forest loss in a vicious cycle that peneplanes, are still crushed by the forest” (Port&es, threatens regional agriculture. 1965, pp. 688 and 726). Regional studies and administrative perceptions are Changes in the Toma agricultural economy are based on social analysis of this deforestation and linked to the opening-up of the area to commerce and encroachment on the remaining Ziama reserve. The markets and to the need to feed growing populations. most detailed and explicit version of the “analysis” is Both these trends are linked to the immigration and found in a socioeconomic study commissioned by a influence of Mandinka people from the north - immi- gration which is also central in explaining the area’s demographic evolution. The authors present a picture of a long-term, very gradual peopling of the Ziama region through the immigration of Toma people and Period Area (hectares) then brusque changes as Mandinka began to immi- grate, now represented by second or third generation At origin I .930.000 C. 1958 I .3oo.ooo migrants. It is said that there were two villages present C. 1980 I .075,000 in the reserve when it was designated in 1932, Boo and 19X6 397.ooo Kpanya, having 542 and 370 persons at that time. Boo, which now has a population of some 1,600 is said to Source: RCpubliqur de Guin&z (1990). have had a population of 500 when it was founded, 1030 WORLD DEVELOPMENT giving the impression that while the forest might have nas. The Ziama mountain massif, now considered the been lightly inhabited for long periods by forest heart of the primary forest, was either bare rock or people, it is only since the mid-19th century that it covered “with cane grass and scarcely any tree but the has been under a threat which is ever increasing. palm” (Seymour, 1859/60). From the top of the mas- Immigration into the region is reported to have risen sif, Seymour describes the plain as “covered with by four to sevenfold in 60 years. This rapid population small bushes and grass, and it gives the country the growth is seen to have created severe land pressure in appearance of an old farm, with palms standing scat- the areas neighboring the reserve. Assumptions about tered all over it.” The ascription of 19th century iden- carrying capacity under shifting cultivation are used to tity as “forest people,” however dubious in itself, argue that population:land ratios are now “fully satu- seems highly inappropriate for these savanna- rated,” and this largely accounts for farmers encroach- dwelling Toma of Ziama. ing on the reserved land for farming. The region had large populations, by all accounts This narrative-concerning a last remaining block significantly larger, not smaller than today. Thus tak- of “pristine” natural forest, threatened by recent ing the enclave villages as an example, Anderson (in socioeconomic change and population pressure - 1874) considered Kpanya as “very large” (when his provides a powerful justification for conservation. account described 2,500 people as small) and It also entails guidelines for conservation policy. Seymour (in 1859/60) estimated Boo to have 3,600 “Original” forest is easily defined as a global or inhabitants. In addition, as the elders of the villages regional heritage, and its conservation by global and describe, these large villages had many smaller depen- regional guardians a moral imperative. The narrative dent settlements which no longer exist. The region enables reserve administrations to list deforestation was evidently highly agricultural. Seymour and problems concerning climate and water as if they had Anderson describe large savanna farms of rice, maize never happened before, and to justify the urgency of and cassava stretching as far as the eye could see, and conservation using arguments about their irreversibil- the short fallows necessary to sustain large popula- ity. Within earlier, colonially derived approaches, the tions. It was also commercially prosperous. Seymour reservation of such forest, often as part of the state’s noted 50 looms and five blacksmiths in Boo, and domain, was acceptably justified with minimal regard found some women wearing jewelery worth $20-30 at for local interests in using reserved land and resources. that time. A little further north, at Kuankan, people In Ziama as elsewhere, “policing and patrolling” walked several miles from the mountains to the plain approaches characterized early forest conservation. to sell firewood. As Seymour noted, “Firewood is More recently, emphasis has been placed on the need scarce about this large city, but they have a good mar- to gain the participation, acceptance and support of ket, and it would do a person good to see the activity local populations if conservation is to be sustainable. of the little boys, who are the principal traders in this Since local resistance to and failure to respect the line.” Both enclave settlements had daily and weekly reserve are seen in terms of land shortage and eco- markets, as did all the major towns, distant some eight nomic pressure, the presumed policy needs are for to 10 kilometers one from another, and these traded in socioeconomic development and agricultural intensi- foodstuffs, livestock, cash crops (such as cotton and fication in the marginal area around the forest, accom- kola), and artisanal goods of every description. The panied by restricting of land tenure as necessary, to region was not economically or geographically mar- reduce current and future pressures on the reserve. ginal, but central to busy and long-established forest- savanna trade routes. Thus in the mid- 19th century the Ziama area clearly (b) The counternarrative did not fit the images which today’s policy narratives construct for it. Unsurprisingly, then, its subsequent Once again, examination of historical data showing history also overturns the conventional narrative’s how vegetation, population and society have changed image of unilineal population increase and forest in this region reveal the extent to which the assump- destruction. The story which explains how this region tions stabilized within this narrative are ill-founded. became a “primary forest” reserve within only 130 In the Ziama case, detailed descriptions come from years of being heavily populated savanna turns, the published writings of several highly educated instead, on the wars which affected the area during America-Liberians who visited what is today the for- 1870-1910 (Fairhead and Leach, 1994). Sustained est reserve in the mid- 19th century (Anderson, 1870; military conflict first with Mandinka groups and then Starr, 1912; Seymour, 1859/60). What they saw and with the colonizing French caused major depopulation described in no way conformed to the enduring image and economic devastation. It is this, not the extension of sparse Toma hunter-gatherer populations living in of persistently low precolonial population densities, harmony with an isolated high forest. The two which explains the region’s sparse populations at the “enclave” villages, now situated within the forests turn of the century. On the abandoned settlements, covering the wide Diani river plain, then lay in savan- lields and fallows, forest grew. By 1932 the French FALSE FOREST HISTORY 1031 colonial administration recognized secondary forest Kissidougou and Ziama, are examples of a broader worthy of reservation. That the forest grew so fast sug- narrative. This broader narrative contains and stabi- gests that earlier intensive farming and savanna main- lizes assumptions which have been applied in the spe- tenance did not cause irreversible damage to forest cific cases, but which are also written into national, vegetation potential; indeed it may indicate the posi- regional and international policy documents. tive legacy of previous local management practices, as Thus it was lamented in work incorporated into in Kissidougou. By the early 198Os, conservationists Guinea’s agricultural development policy strategy were failing to distinguish Ziama’s forest regrowth that: “The north of forest Guinea (Beyla, Kissidougou from primary forest. Populations since 1932 have not and Gueckedou) is no longer a pre-forest region, but grown by the 400-700% suggested in the socioeco- an ‘ex-forest’ or ‘post forest’ region!” Stating the nar- nomic study. Using the study’s own statistics, in the rative in its perhaps most succinct and pure form, it 41 villages in the vicinity of the reserve populations was asserted that: have increased by only 80% since 1932, or 120% if recent influxes of Liberian and Sierra Leonean This degradation of the natural environment...is the refugees are taken into account. result of an evolution of rural societies little adapted to It is clear that the stabilized assumptions which the rapid structural, demographic and economic changes social science researchers are using to understand the this century, and above all, these last years. ..The prob- nature and change of people-environment interactions lem today is the recession of traditional control of the orderly exploitation of space and its resources, which has in Ziama are completely at odds with a more demon- not managed to follow or adapt to the recent and very strable counternarrative centering on warfare, depop- rapid change in the rural world. This management ulation, forest regeneration and land alienation. The becomes insufficient given a brutal increase in popula- latter narrative better encompasses the experience and tion [and] a progressive loss in the power of traditional attitudes of today’s Toma inhabitants, whose promi- control, due to the destructuration of rural society, the nent display on village houses of portraits of ancestors new amplitude of migration and the push towards who were killed or who fled during the wars testifies agrarian individualism and the monetization of the local that these past events are not forgotten. It is largely this economy (RCpublique de Guinte, 1989, p. 8). mismatch of narratives which underlies the failure of the reserve administration to build any constructive This narrative is the script of international donors, relationship with local inhabitants. Instead, their rela- and one could till shelves with its versions across tionships are tense and have at times erupted into vio- Africa and beyond. Focusing on the population lence. Development activities around the reserve have component, a recent World Bank policy review seemed inadequate to calm this conflict and prevent argues that: “encroachment” on land within it. Achieving sustain- able conservation, let alone of a participatory nature, .traditional farming and livestock husbandry practices, tnditional dependency on wood for energy and for build- remains a distant and unlikely goal, and much invest- ing material, traditional land tenure arrangements and ment has been wasted in the effort. traditional burdens on rural women worked well when When today’s inhabitants “encroach” they are population densities were low and population grew attempting both to reclaim ancestral lands, and to slowly. With the shock of extremely rapid population reestablish control over a once peopled and prosper- growth.. .these practices could not evolve fast enough. ous, now ex-social domain politically alienated from Thus they became the major source of forest destruction them. Recognizing this suggests alternative, poten- and degradation of the rural environment (Cleaver, 1992, tially more fruitful guidelines for policy. If policy p. 67). makers are to engage sensitively and productively with local communities, then local inhabitants’ histor- This, it is argued, leads to vicious spirals of short- ical experiences need to be incorporated into policy ening fallows, land depletion, yield declines, and sub- dialogues and negotiations. Moreover, historically sequent migration to marginal lands and forests. grounded claims to land and political authority need to Environmental crisis results less from the overall be recognized and seriously addressed through con- effect of population pressure on resource availability, servation arrangements which, for example, cede as a classic Malthusian position would have it, than tenurial control to local landholders, within the con- from the multiple effects of population pressure on the text of leasing or management agreements which fully institutions seeking to control resource access and use. recognize the value their lands now have for others. In exemplifying how inapplicable current regional narratives can be to local situations, the Kissidougou 4. THE REGIONAL NARRATIVE AND and Ziama cases invite a more fundamental examina- ITS ALTERNATIVE tion of the origins and purposes of the regional narra- tive itself. More than empirical evidence, such The specific narratives, concerning vegetation narratives depend on - and expose - the field of change and its social causes used to support policy in Western imagination concerning African society; in 1032 WORLD DEVELOPMENT particular, they show that stereotypes born of the colo- facts of vegetation history, so at a more general level nial era are alive and well in the applied social sci- we follow the spirit of the articles in this special sec- ences. Whether they are used to justify policies of tion to suggest other assumptions and a stabilizing external repression, or policies of social reorganiza- narrative which better reflect realities surrounding tion and “participatory” development, the narratives African environments. The parameters of this coun- justify and make imperative a role for the outsider in temarrative accord with recent developments in eco- the control of rural resources. The broader assump- logical and social theory and, significantly, they do not tions which the regional narratives contain can be perpetuate the imperative for outside intervention in summarized as follows: local resource control. (a) that African vegetation was once “original,” While the old narratives held within them a view of consisting of a climax vegetation, i.e. the ultimate ecology which had to explain the disappearance of a stage of plant succession which can exist under natural climax vegetation, newer strands of ecological given ecological conditions. Prevailing ecological theory reject the idea of a single environmental maxi- conditions are unchanging, so that what could exist mum. When climate historians suggest that Africa has today (e.g., humid forest in the forest and preforest experienced both long-period, deep climatic fluctua- region) did recently exist. Against this most natural tions and changes in climatic variability, the history of vegetation one can judge levels of “degradation”; vegetation begins to be seen as a history of continual (b) that African society can be seen, at origin, in transition, rather than of divergence from a single, terms of a traditional “functional order.” Such order once-extant climax. Recent theory suggests that such was once harmoniously integrated with “natural” repeated transitions are likely to be between particular vegetation (e.g., as epitomized in the idea of a “for- “stable” vegetation states, each determined by a multi- est people” and a “savanna people”). African farm- factor complex, rather than by trends in any particular ing, land and resource-use practices degrade or are variable. If the transition-causing factor reverts to its at best benign to the original vegetation. pretransition level, vegetation may move to another Degradation is thus limited only by functional state, but need not return to its initial one (Sprugel, social organization (regulation and authority). 1991; Scoones, 1994). Given the multiplicity of inter- From environmental degradation one can diagnose acting factors influencing each state, shifts between the social ills of organizational dysfunction; them can be triggered by a particular, possibly unique, (c) that African rural populations only increase, and historical conjuncture of ecological factors. From this do so fast. Population increase is as such environ- viewpoint, there is no basis for identifying a region’s mentally and socially damaging; fundamental, archetypal vegetation. Vegetation is in (d) that African society is essentially sedentary and continual transition, and its trajectory is determined by subsistence oriented with an anti-commercial senti- the legacy of past vegetation paths and present eco- ment (e.g., in the popular imagery of “anti-com- logical conditions. mercial” traditional hunters). Money, mobility and Ideas of environmental optima dovetailed neatly trade are modem and lead to socioenvironmental with ideas of static social maxima - of tradition and dysfunction. African history consisted of the con- structure-typical of, but persisting beyond, colonial tinuous reproduction of tradition until it began to anthropology. But notions of society with a given become “modern,” whether with markets and social structure and order, maintained by functional mobility, colonial intervention, or (in some work) adaptation and/or by rules and regulation, are chal- the arrival of Islam. lenged by more recent social theories giving weight to The romantic links forged between these assump- social action, processes and their capacity to shape and tions mean that vegetation change carries very pro- determine rules. Such continual structuration, over found moral messages. “Original climax vegetation” time and through social change, challenges the notion and “traditional functional society” provide funda- of a baseline “traditional” societal state. That African mental baselines, so that whether the concern is about social forms have been in constant transition dovetails society or the environment, it is possible to judge that with the view of vegetation as m continual transition. something is wrong and assess the extent of damage. There is no baseline in terms of how society values From such a vantage point, the imperative is to vegetation (and therefore no basis for the moral argu- intervene. ment that indigenous values once preserved a more These assumptions, stabilized and sometimes hid- “natural” ideal). Vegetation values are shifting in den within social science analysis, are destructive and accordance with social, economic and polltical ultimately have no policy relevance. “The hard fact” changes, often of quite a conjunctural nature. The val- as Sayer (1992) puts it, “is that most aid projects, and ues placed on different vegetation types, conditioned especially those in foresty fail.” As the Kissidougou by prevailing social conditions, are also socially dif- and Ziama cases exemplify, misleading narratives are ferentiated: the high forest and wildhfe priorities ol fundamental to this failure. Moreover, just as for these today’s global conservation planners are very difler- cases there are counternarratives which better fit the ent from the agricultural bush fallow prioritich ot FALSE FOREST HISTORY 1033

today’s Toma inhabitants. population is removed at that point, and given the In the West African context, these social and eco- legacy of people’s previous land use practices, the nomic transitions have taken place within a long area may develop into high forest, as happened early historical context of movement and migration, agri- this century in Ziama. culture and commerce, and political and religious This regional counternarrative provides different, turbulence. The relationship between social and and more appropriate, guidelines for policy. In pre- environmental change does not turn on the dramatic senting socioenvironmental change in a way which increase in any of these, but rather on people’s better fits local experience, it provides a more effec- responses to changing signals within this broader, tive basis for dialogue and participatory development dynamic continuity. Thus Kissidougou villagers have work with local populations. In removing the baseline adapted forest island form to meet changing needs for link between social and vegetation form, it removes fortification and different cash crops. Demographic the justification for external intervention in the orga- change, rather than consisting always of unilinear nization of resource management to reestablish a lost population increase, involves periods of stability social order, whether by replacement with external and decline, of shocks as well as secular trends. control or by the externally promoted “community Depending on prevailing ecological and economic reorganization” of recent more decentralized conditions, the effects of population growth periods approaches. It suggests that more important priorities can be positive as well as negative. are to create the enabling policy and economic con- In the West African forest margin zones, climatic ditions in which local resource management constel- transition appears to have involved rehumidification lations can act effectively, to support the diverse since the mid- 19th century, following a long relatively existing local institutional forms, and to build on the dry phase (Nicholson, 1979). Where the combination beneficial environmental implications of broader rural of ecological factors makes conditions marginal for development and pricing policies - an approach forest, creating a precarious balance between forest which now finds support in some regional policy insti- vegetation and fire-maintained savanna, people’s tutions (e.g., ENDA, 1992). Finally, as McNeely activities can make the difference, allowing forest argues, “because chance factors, human influence and vegetation to develop in grassland. Where people have small climatic variation can cause very substantial socioeconomic or political reasons to create forest changes in vegetation, [the biodiversity for] any given they do so. in small patches, triggering transitions in landscape will vary substantially over any significant small parts of the landscape, as has happened, for time period - and no one variant is necessarily more example, in Kissidougou. In open savanna and with ‘natural’ than the others” (1993). From this perspec- low population densities, fire is harder to control, but tive, environmental policy can call on no moral high as populations increase and transitions to forest are ground in recreating the natural (or the social that went provoked in more places, fire is reduced and may with it). It becomes very clearly a question of social or eventually be eliminated. Agricultural priorities may political choice about what vegetation forms are desir- mean large areas are maintained as bush fallow rather able at any given time in social history, and about than allowed to develop into high forest. As popula- ensuring that conflicting perspectives on this - such tions increase further, fallow periods may need to be as between local, global and intergenerational inter- shortened and some resavannization can occur. But if ests - are adequately articulated and addressed.

NOTES

I This. like 311subsequent quotations. have been translated Leach, forthcoming; Leach and Fairhead, 1995). from the origmal French by the authors. 4. For more on the management-use continuum (i.e. the 7 hlore details of the following historical vegetation way people use a resource in the way they manage it), see klys~q arc given In Leach and Fnirhead (199-L). Roe and Fortmnnn (1982).

1 Such alternntlw sow11 science analysis and its consid- 5. Further details concerning this case are given in Fairhead cmblc evidence IS documented fully elsewhere (Fairhead and and Leach (1994). WORLD DEVELOPMENT 1034

REFERENCES

Adam, J. G., “Les reliques boisCes et les essences des savnnes versity,” Mimeo (Gland, Switzerland: IUCN, 1993). dans le zone ptiforestitre en Guinee fancaise,“Ecrllerin de Nicholson, S. E., “The methodology of historical climate la Sucit% Botaniyue Francaise, Vol. 98 (1948). pp. reconstmction and its application to Africa,” Journal of 22-26. African History, Vol. 20, No. I (1979). pp. 3 l-49. Anderson, B., Narrative of a Journey 10 Musardu: capital of Ponsart-Dureau, M.-C., “Le pays Kissi de GuinCe forestiere: the western Mandigoes (New York: S. W. Green, 1870). contribution a la connaissance du milieu; problematique Baum, G. A. and H-J. Weimer, “Participation et dCvtlop- de dCvCloppement,” Memoire (Montpellier: Ecole pement socio-Cconomique comme conditions ptialables Superieure d’Agronomie Tropicale, 1986). indispensables d’une implication active des poulations Port&es, R., “Les noms des riz en GuinCe: VI - Les noms riveraines dans la conservation de la for& classke de des vtiCtCs de riz chez les Toma,” Journa! d’Agricu!rure Ziama,” Report (Conakry: RCpublique de Guinte: Tropicale er de Boranique AppliquCe, Vol. I2 (1965), pp. Deutsche Forst-Consult/Neu-Isenburg. RFA/KfW, 1992). 687-728. Chevalier, A., “Les bois sac& des Noirs de I’Afrique tropi- Programme d’Am&agement des Bassins Versants Haute tale comme sanctuaries de la nature,” Revrre de /a Sock% Guinb (Projet Kan II), “Plan d’Op&ation,” Report de Bio&ographie (1933), pp. 3742. (Conakry, RCpublique de GuinCe: Programme Chevalier, “Rapport sur les nouvelles recherches sur les les d’Amtnagement des Hauts Bassins du Fleuve Niger, plantes a caouchouc de la Guinee francaise,” 16276 1992). (Dakar: Archives du Senegal, 1909). Programme d’Am&agement des Hauts Bassins du Fleuve Cleaver, K., “Deforestation in the western and central Niger, “Etude sociologique,” Report (Conakry: African forest: the agricultural and demographic causes, RCpublique de GuinCe: Programme d’AmCnagement des and some solutions,” in K. Cleaver, M. Munashighe, M. Hauts Bassins du Fleuve Niger, n.d.). Dyson, N. Egli, A. Peuker and F. WencClius (Eds.), RCpublique de GuinCe (Jean, B.), “La gestion des ressources Conservation of West and Central Africa’s raitzforests, naturelles,” Contribution to the lettre. politique de Environment Paper No. I (Washington, DC: World Bank, dCvCloppement agricole (Conakry: RCpublique de GuinCe: 1992). pp. 65-78. Minsttre d’Agriculture et des Ressources Animales, Dupri G., “Les a&es, le fourrk et le jardin: les plantes dans November, 1989). la socittC de Aribinda, Burkina Faso,” in G. Duprk (Ed.), RCpublique de GuinCe and IUCN (Bourque, J. and R. Savoirs Paysans et Dhkloppement (Paris: Karthala- Wilson), “Guinea forestry biodiversity Study - Ziama ORSTOM, 1991). pp. 181-194. and Diecke Reserves,” Report (Gland, Switzerland: IUCN ENDA, Avenir des Terroirs: la ressource humuine (Dakar: for RCpublique de GuinCe. September, 1990). ENDA/GRAF, 1992). Roe, E.. “Development narratives, or making the best of Fairhead, J. and M. Leach, Reversing Landscape History: blueprint development,” World Development, Vol. 19, power, policy and socialised ecolugy in West Africa’s for- No. 4 (1991). est -savanna mosaic. Book manuscript (forthcoming). Roe, E. and L. Fortmann, Season undStrafegy, Special series Fairhead, J. and M. Leach, “Contested forests: modem con- on Resource Management, (Ithaca, NY: Rural servation and historical land use of Guinea’s Ziama Development Committee, Centrc for International reserve,” African Aflairs, Vol. 93 (1994). pp. 481-512. Studies, Come] University, 1982). Fofana. S., Y. Camara, M. Barry and A. Sylla, “Etude rela- Sayer, J. A., “Development assistance strategies to conserve tive au feu aup& des populations des bassins versants Africa’s rainforests,” in K. Cleaver, M. Munashighe, M. types du Haut Niger: Monographies des Bassins Kan I, Dyson, N. Egli, A. Peuker and F. WencClius (Ed%), Kan II, Kiss II,” Report (Conakry: Rtpublique de GuinCe: Canservarion of Wesr and Central Africa’s rainforests, Programme d’ Amtnagement des Bassins Versants Haut- Environment Paper No. 1 (Washington, DC: World Bank, Niger, 1993). 1992). pp. 3-17. Green, W., “Lutte contre les feux de brousse.” Report Scoones, I. (Ed.), Living with Uncertainty: new directiansfor (Conakry RCpublique de GuinCe: Projet DERIK, pastoral development in Africu (London: IT Publications, DCvCloppement Rural Intt?grb de Kissidougou, 1991). 1994). Hair, P. E. H., “An account of the Liberian Hinterland c. Seymour, G. L.. “The journal of the journey of George L. 1780,” Sierra Leone Studies, NS Vol. I6 (1962), pp. Seymour to the interior of Liberia: 1858,” New York 218-226. Colonization Journal, Vol. IX, No. 12, Vol. X, Nos. 6 and Holling, C. S., “Resilience and stability of ecological sys- 8 (1859-l 860). tems,” Annual Review of Ecology and Systematic.% No. 4 Sims, J. L., “The journal of a journey in the interior of Liberia (1973), pp. l-23. by James. L. Sims, of Monrovia. Scenes in the interior of Leach, M. and J. Fairhead, “The forest islands of Liberia: being a tour through the countries of the Dey, Kissidougou: social dynamics of environmental change in Goulah, Pessah Barlain, Kpellay, Suloang, and the King West Africa’s forest-savanra mosaic” (Report to ESCOR Boatswain’s tribes, in 1858,” New York Colonizabon of the Overseas Development Administration, July, 1994). Journal. Vol. IX, No. 12, Vol. X, Nos. 6 and 8 Leach, M. and J. Fairhead, “Ruined settlements and new gar- (1859-l 860). dens: gender and soil-ripening among Kuranko farmers in Sprugel, D. G., “Disturbance, equilibrium, and environmen- the forest-savanna transition zone,” IDS Bulletin, Vol. 26, tal variability: What is ‘natural’ vegetation in a changing No. 1 (1995). pp. 24-32. environment?,” Biologicu[ Conservation. Vol. 58 (199 I), McNeely, J. A., “Lessons from the past: forests and biodi- pp. l-18. FALSE FOREST HISTORY 103.5

Starr, F. (Ed.), Narrative ofthe Expedikm Despatched to Republique de Guinee),” Mimeo (Berlin: 1990). ~~sohd~ by the Liberian ~o~er~rne~r under Benjamin K Zerouki, B., “Etude relative au feu aupms des populations Anderson Esq. in 1874 (Monrovia: College of West Africa des bassins versants types du Haut Niger,” Repott Press, 1912). (Conakry: R~publique de Gun&e: ~ogmmme Stieglitz, F. V., “Exploitation forest&e ruraIe et rehabifita- d’Am&agement des Bassins Versants Types du Haut tion des for&s: Premiers resuhats d’un projet de rechenhe Niger, 1993). interdiscipiinaire en Haute-Guinee (Jaw - Mai 1990,