Originally published in Human Ecology, 1989, 17:299-319. Reprinted in Case Studies in Human Ecology, 1996, edited by D. Bates and S. Lees, pp. 327-348. Plenum Press, NY.
| Note: We are still working on this file. Pardon the slightly confusing footnotes (in small font in the text) and endnotes (at the end of the text); a revised version will be posted soon. |
INTRODUCTION
According to all accounts Nigeria in the 1970s went through an agricultural change of alarming proportions with dizzying speed. Exports of palm products, cocoa, groundnuts, rubber, and cotton declined both relatively and absolutely, while domestic food supplies failed to keep pace with demand, resulting in spiraling imports and inflation (Watts, 1984; Berry, 1984; Iyegha, 1988). The oil boom fueled explosive urban growth and rural exodus without a sign of Green Revolution crops or technology changes on the farm, and with a marked deterioration in rural terms of trade and even nutrition (Matlon, 1981; Rimmer, 1981). It was conclusively demonstrated that major agricultural development projects had disastrous cost/benefit ratios (Andrae and Beckman, 1985; Mortimore, 1987). Total production of food crops and total land under cultivation in Nigeria was said to be going down (Iyegha, 1988, pp. 42-50). Cheap food supplies that the national government regarded as necessary for industrialization (Bates, 1983) and the indirect support of a burgeoning governmental and commercial urban elite (Lipton, 1976) were not forthcoming. There appeared to be little surplus to be captured from the peasantry (Hyden, 1980). Vast arable areas remained underutilized while the population ballooned.
Nigeria, it would seem, was ripe for the opening of agrarian frontiers, the adoption of new crops and more productive techniques-and for a more inclusive movement from subsistence to market-oriented agriculture. Nationally, there was no evidence of this happening, and indeed no one seems to have imagined that increased food production could possibly take place without capital-intensive mechanization and applied science, without decreasing labor and increasing farm size, and without the planned provision of seeds, fertilizers, extension services, and rural credit. The dominance of a Western model of agriculture was unquestioned, and "indigenous agricultural revolution" (Richards, 1985) was a contradiction in terms.1
For these reasons, the experience of the Kofyar (Netting, 1968) in Plateau State is instructive. In a little over 30 years, thousands of farm households have migrated voluntarily from the southern edge of the Jos Plateau to largely vacant bushlands of the Benue valley, and they have begun producing yams on a vast scale for the Nigerian internal market. They have continued to provide their own subsistence crops with traditional hoe techniques while implementing new systems of land use. In addition, they have mobilized substantially increased labor supplies by existing social means, and they have used their cash earnings enthusiastically to purchase a variety of manufactured goods and modern educational, medical, and transportation services.2 Though a new all-weather road and artificially low gasoline prices facilitated this process, there was almost no direct governmental intervention to establish extension services, to provide agricultural inputs or subsides, or to influence prices, wages, or market arrangements. Because agricultural expansion is taking place- without sophisticated technology or bureaucratic planning and control, it is officially unobserved and its existence may even be denied. The Kofyar story may be unrepresentative (though there must certainly be other unheralded farmers increasing their food supplies to a national population estimated at 88.5 million in 1991) and their present trajectory of growing prosperity may yet encounter declining economic returns and ecological deterioration, but a case study of uncoerced, self-generated development may still illuminate significant variables in a particular historic process of genuine change.
SPONTANEOUS MIGRATION AND PIONEER SETTLEMENT
Kofyar movement from their traditional village areas of Kwalla, Doemak, and Merniang (Kwa, Kwang) at the base of the Jos Plateau southern escarpment and the hills above did not become important until the 1950s. Following British conquest in 1909 and the establishment of colonial rule, there had been some occupation of Guinea savanna lands adjacent to Kwalla and Doemak for shifting cultivation of outfield plots, but this reflected little more than a minor expansion of subsistence by local populations with densities often exceeding 300 persons per square mile. Several walled, nucleated communities to the south, like Namu (linked ethnically to the Kofyar), Kurgwi with its Hausa traders, and Kwande, a Goemai settlement, maintained a desultory trade in salt and farmed adjacent bits of the forested plain by slash-and-burn techniques. Seasonal rainfall of 1143 - 1270 mm is generally adequate to sustain both tubers and grains. A few Kofyar had begun to head-load food crops from this area to sell in the tin mining areas of the high Plateau. A 1946 government report mentions farms of Kofyar, Tiv, and other tribes east of the Shemankar and in the southern Goemai bush (Rowling, 1946). Permanent colonization began in 1951 when around 30 Doemak households established farms in Mangkogam south of Kwande. In 1953, the Chief of Kwa opened a farm south of Namu and encouraged his people to follow him, growing extra food for subsistence and perhaps for trade and tax payments. Kofyar from the most crowded home areas with the lowest per capita crop and livestock production (Stone et al., 1984) were the first to begin seasonal treks to the area, killing the trees with fire in the dry season and planting sesame, followed by yams, and then staple grains, sorghum and millet. They continued to maintain their permanent, intensively-tilled and fertilized homestead fields in the old villages, moving household members back and forth to the bush farms as their labor was needed.
Migration increased rapidly in the 1960s, going from 22.4% of 655 households surveyed in 1961 to 45.5% of 677 households in 1966 (Stone et al., 1984, pp. 95-96). With limited swidden lands available in Kwa where population densities exceeded 200 per km2, 76% of the households had established migrant bush farms in the Namu area by 1966. The hill village of Bong that had depended solely on its homesteads and local shifting cultivation in 1961 had 26.5% of its households going to distant plains farms by 1966 (Stone et al., 1984, p. 96). Over the next 18 years, the movement increased, and growing numbers of people settled permanently on the bush cash-crop farms. A 1984 survey of 979 households showed that 74% had migrated and 518, over half the sample, had transferred their sole residence to the bushlands. The original homesteads of these migrants were abandoned and rapidly disintegrated. Their homestead fields were untilled or taken over by neighbors for growing groundnuts or grazing. A hill hamlet of Bong village was almost deserted, with only 2 of the original 17 homesteads still occupied. At the same time, the cash-crop farms changed from small collections of makeshift sleeping huts and crop-drying racks to permanent residences with a number of round huts or rectangular, metalroofed houses plus kitchens, brewing huts, yam barns, pigsties, goat shelters, and granaries. Remaining homesteads in the traditional villages often housed elderly or retired people as well as children who had temporarily left the cash-crop farms in order to attend primary and secondary schools.
The magnitude of population migration to the frontier and demographic growth in place can only be estimated because of the absence since 1963 of an officially accepted national census in Nigeria. Within a mapped 36.22 km2 sector of bush farms south of Namu, the 138 homesteads visible in 1963 aerial photograph increased to 546 in 1984. Using an average household size of 6.58 based on our 1984 survey, population in this area rose from 902 to 3593. This represents an almost 300% increase in density from 24.9 to 99.2 per km2 (Stone, 1988). A biennial registration of population compiled for tax purposes in Namu District shows the numbers of rural Kofyar rising from 7050 in 1972 to 12,602 in 1984 with increments of about 1000 per year until 1978. A declining rate of growth, 500 a year (1978-1980) and 150 a year (1980-1984), may represent the filling-in of frontier lands and the beginnings of significant out migration to more distant bushlands. The size of Kofyar population in Namu town has increased rapidly, up 24% to 1661 between 1980-1984 alone. Adjoining Namu District is Kwande District where 11,364 Kofyar, mostly originally from Kwalla, were settled by 1984. It appears probable that a minimum of 25,000-30,000 Kofyar now occupy the Benue Valley lowlands south of their original homeland, and that their migration has taken place within the last 40 years. The entire Kofyar population was calculated in 1952 at 55,174 people (Netting, 1968, p. 111).
The timing and extent of involvement in migratory bush farming was always a matter of individual decision-making. Some families never participated. People from the same village who began making the thirty-mile trip to Namu at the same time might live temporarily with the Chief of Kwa or form a large temporary compound for a few years. But households quickly dispersed along the paths, building huts on their fields that straddled a ridge and often reached down to a shallow stream valley (Stone, 1991). There was little sign of patrilineal lineages forming residential units, and neighborhoods could incorporate members of several Kofyar home villages and even an occasional family of the Angas or Mwahavul (Sura) ethnic groups .3 Younger men might work for kin or friends until they had accumulated some seed yams, and then select a vacant spot with good soil and easy access to drinking water. As yields from shifting cultivation declined and grass fallows became harder to work, households would seek new field sites in the bush.
The gradual extension of settlement and commuting between home and bush farms contrasted with an earlier attempt at forced relocation of Kofyar. Following the killing of a British district officer in Latok in 1930, seven villages and some 3000 hill Kofyar had been rounded up and evacuated to the plain (Netting, 1987). Only after annual military patrols had removed people and livestock and repeatedly burned huts for some 9 years were the Kofyar allowed to return to the hills. Despite the availability of unoccupied plains land, the attempt at permanent resettlement failed, and all the Kofyar returned to their ancestral homesteads.
The more recent movement of Kofyar into migratory, and then permanent, cultivation has converted the frontier from a patchwork of fields with dead trees and partially burned trunks, bush and grass fallows, and undisturbed forest into a relatively continuously cultivated, open farmland with perhaps 20% short fallow (1-3 years) land. Hardwoods such as mahogany and iroko in the better watered valleys were cut in the late 1960s, and the less valuable species (silk cotton, copaiba) are now being harvested by timber contractors who use chain saws to reduce the trees to 4 x 6 inch, 12-foot beams for further processing into planks by sawmills in the towns. Only locust bean (Parkia clappertoniana), sheanut (Butyrospemum paradoxum), and other economically useful trees remain in the fields. Our earliest aerial photographs from 1963 show about 20% of the frontier land in use, but by 1978, 65% was under cultivation with another 11% in fallow. The original plots, typically 4-6 ha in size were laid out along watersheds on well-drained, nonlateritic soils. As the landscape was filled in, more recent arrivals or new households fissioning from those of original settlers took up less desirable plots that were farther from water (Stone, 1986). A number of farmers tried low-lying areas in Danka (the name itself means "heavy clay") but found after a few years that the fields became waterlogged and there were few house sites not subject to swampy conditions. Often they returned to lands that had been farmed and vacated. By 1984, farms in the areas of early settlement were being fragmented, with smaller portions being inherited, given to kin and friends, or in some cases sold.
A rather typical case is that of Datoegoem of Ungwa Kofyar who moved to the frontier in 1963. Datoegoem's original 14 ha farm was more than his household needed, and he gave a 5 ha section to his friend Doewat in the early 1970s. When Datoegoem's eldest son Christopher began to acquire wives, Datoegoem moved to a new compound across the road, leaving almost 4 ha of land to the son. The son's household had grown (he took his fourth wife in 1984), and he has received another parcel from his father as well as gaining control over part of the farm given to Doewat, whose own household has shrunk. Another friend has farmed the parcel that had been Doewat's, but has not been allowed to build a compound there.
Other Kofyar had sought virgin bushlands up to 80 km from the homeland at Kuka not far from Ibi on the Benue River and in Lafia division south of Assaikio. Everywhere the phase of frontier shifting cultivation appears to be transitory, with a tendency to bring fields into permanent production by crop rotation, frequent weeding, intercropping, manuring, and chemical fertilization (Netting et al., 1993). Unlike the Tiv who migrated from the south into the same area and maintained large areas for bush fallow rotations around their compounds, Kofyar intensified their agriculture as bush population density increased and farm size declined.
MOBILIZING LABOR
Pioneer occupation of a large forested frontier zone, followed by permanent settlement and intensifying farming, required increased application of labor. The opening of these extensive tracts of land for shifting cultivation meant that increases in production came directly from more workers and additional hours in the fields. Neither Kofyar migration nor greater output was contingent on the adoption of a new technology. The broadbladed Sudanic hoe remained the principal tool, and the traditional farming system based on interplanted millet, sorghum, and cowpeas was not discarded, but rather modified to include a yam rotation or relay cropping (Stone et al., 1990). Kofyar households usually function as independent economic units, and they quickly reproduced their traditional dispersed settlement pattern, with each of the new homesteads occupying its own cultivated land and fallow.
The bulk of labor in both the home and migrant communities was, in the 1960s, and continues to be drawn from, the household labor force. In a sample of 15 households for which daily labor records of all adults were kept in 1984, household labor applied to household fields constitutes 62.8% of all hours devoted to farming and remains particularly important for the cultivation of the traditional cereal crops of millet and sorghum (Stone et al., 1990). Of all the hours devoted to cereals, 75.8% is contributed by the household to the household fields. Relatively more of the labor time on the cash crops of rice and yams comes from interhousehold cooperative work or from individuals cultivating their independent plots, but 49.9% of all hours spent on rice farming and 59.7% on yams are performed by the household itself.
To meet the increasing needs for labor, households grew in size. Large households could both concentrate labor power to meet the seasonal demands of cash-cropping and also divide their forces to maintain the old homestead farm that provided a secure subsistence. Whereas non-migrant Kofyar households had averaged 4.54 (1961) and 4.17 (1966) members, those with migrant bush farms averaged 6.44 in the 1960s (Stone et al., 1984, p. 99). By 1984, the average household size in the Namu area was 8.38 people. Growth in household size has continued progressively with number of years spent in bush farming, peaking at a mean of 9.26 for those households 16-20 years into their developmental cycles. Larger size reflects, in part, increasing frequency of polygyny and multiple family formation (Netting, 1965, 1968). Cash income from the sale of yams, millet, and rice could be used for bride wealth and additional wives, increasing the number of adult workers immediately and providing children as a future source of labor. Keeping married sons on their fathers' farms also enlarged the household work force.
Although polygynous households and the extension of a residential family to include two or more married couples (multiple family households) had always existed as culturally acceptable options, they were less frequently realized in the home communities where farm sizes were constrained by population pressure and larger labor groups were unnecessary for most tasks. The proportion of polygynous family households among non-migrant Kofyar was 32.3% in 1966 (Stone et al., 1984, p. 99) and had climbed to 51.5% among 1984 bush farmers. The corresponding proportions of multiple family households had risen from 7.3-17.0%.
With the filling of the frontier and the increasing abandonment of hill farms, the utility of large extended and multiple households may be reduced. Cash-cropping households formed in the last 5 years have tended to be less complex and smaller, averaging only 5.70 members. Of these predominantly younger families who have moved permanently to the cashcrop farms, 41 % are polygynous and 10% are multiple. Households with a longer history of migration and a continued maintenance of both home and bush farm residences tend to be complex, with 69% having at least one polygynous marriage and 26% being multiple. Three hundred households that are either extended polygynous or multiple polygynous families have an average membership of 12.09. There are also growing numbers of children per household. The proportion of the population in the 0-15 age range has grown from 37% in 1966 to 47% 18 years later. But for the present, increasing household size and labor potential is regularly associated with higher grain and yam production, more use of consumer goods, and higher household incomes.
While the resident family household remains the fundamental labor and consumption unit, households also depend on agricultural work performed by cooperative groups of friends and neighbors. Though Kofyar have always exchanged some assistance in such tasks as field ridging, harvesting, and thatching, we had predicted that the major emphasis on cashcropping and the end of the free land for frontier expansion would lead to more wage labor and less interhousehold collaboration. Indeed, agricultural monetization in Africa is often seen to involve the replacement of allegedly expensive, inefficient traditional exchange labor with the hiring of local or seasonally migrating workers (Saul, 1983). This wage labor is seen to be drawn from the impoverishment of rural smallholders and the landless (Hill, 1980; Watts, 1983a) with increasing land consolidation and economic inequality (Reyna, 1987). Hiring of labor does indeed exist among the Kofyar, but its contribution to total work inputs has remained relatively minor.
Cooperative work groups, in contrast, may have increased in their incidence and variety. One form is a 30-80 person voluntary group composed of neighbors from the same-named neighborhood (ungwa) of 20-50 households. These mar muos (farming for beer) parties assemble to accomplish one farming task and are entertained by their host with millet beer at the end of the day. Such festive labor (Erasmus, 1956; Moore, 1975) is scheduled well in advance and may last a few hours or half a day, with some groups absorbing a full day's work. Men and women work side by side at ridging, making yam heaps, and storing millet after the early August harvest, but men do somewhat more of the heavy hoeing while women grind and cook the millet beer. It is not unusual for 400 pounds of millet to be prepared, so that each participant will have about two gallons of the nutritious brew to drink (Netting, 1964). Some beer may be brought to the field, and the work is spurred by young men who set a fast pace and drummers who accompany the hoeing.
Smaller groups (called wuk) bring together 8-20 people that regularly exchange work, and women and adult sons have organized groups of their own to work on their individual fields (Stone, 1988a). No cash or beer accompanies the labor, although close track is kept of reciprocity. New forms of collective labor have also emerged in the plains, including church and school groups and social clubs who farm collectively for their members and to generate group funds.
In our labor records, 21.4% of all agricultural work is some form of extrahousehold exchange. The festive labor groups account for 8.2%. Of the remaining work groups, the largest part is accounted for by cooperative beer brewing (7.6%).
Almost as significant, in terms of hours, is the time spent on cultivating personal fields. Individuals may grow yams, rice, peanuts, and minor crops for sale. Women derive a significant independent cash income from such activities (Stone, 1988b). Overall, 19% of agricultural hours are devoted to personal fields within the household. When combined with festive labor groups and exchange labor applied to personal fields in other households, work devoted to individual production reaches 22.4% of all agricultural labor.
The use of beer party labor has not been supplanted by wage labor. In our larger survey of 799 cash-cropping households, two-thirds of them had used beer party labor on at least one occasion in the last year. The average number of these events was 0.9, with a median size of 50 workers. In comparison, 37% employed hired labor, and these households use a median of 20 person days annually at a median cost of N90. In general, household size and use of paid labor went up together, suggesting that these two modes of labor mobilization were not alternatives but were complementary.
While many studies have shown that rural wage labor in Africa is actually cheaper than cooperative labor (Latour Dejean, 1975; Charsley, 1976; Saul, 1983), the stated preference of Kofyar household heads is for beer party or exchange labor as opposed to wage workers. They contend that (1) costs of agricultural labor including food are high in comparison to the cost of buying millet and making beer for a cooperative work party, (2) hired workers are demanding, undependable, and less careful in agricultural operations than neighbors, and (3) there are not enough hands available at bottleneck periods when 40-100 workers can complete a crucial task in 1 day. Most hired workers are young men coming down in groups of two or three from the Angas and Mwahavul (Sura) ethnic areas on the high plateau. They may hill up yam heaps for N5 per 100 (about 1 day's work), and they are also entitled to board and lodging. Kofyar claim that some workers ask for oil and meat, threatening to leave before the task is completed if these are not forthcoming. Mean costs of a workday are approximately N5.41 plus food, while comparable effort from a Kofyar would be recompensed with N1-1.30 for brewing grain. Two bags or 400 lbs of millet costing about N100 are considered adequate for approximately 80 workers. The malted millet is ground and cooked by perhaps 20 women who exchange labor and receive a portion of the beer. Firewood and water must also be fetched. Even when the time of these women, as well as the loose obligation of reciprocal work on the fields of others, is considered, the Kofyar feel it is cheaper (and certainly less requiring of immediate cash) to sponsor beer parties than to pay workers. In fact, the time women spend brewing for work parties is only about 15% of the total labor hours invested in this type of group labor.
As intensification increases, the quality of work and the lack of a need for close supervision also become important factors.4 The prevalence of interplanting means that there may be millet, sorghum, cowpeas, and groundnuts at different stages of growth in the same field. Weeding and ridging operations must be discriminating as well as fast. Heaps for next year's yams are made among stalks of maturing sorghum in the same field where millet has just been harvested. Kofyar rely on the skill and experience of their neighbors, who will expect similar attention when they in turn receive help.
In savanna cultivation, a long dry season (November-March) and often inconsistent rains set universal planting periods and create successive bottlenecks for cultivation, weeding, and harvesting. Kofyar work parties both mass labor in order to finish a task near the optimum time and supply incentives of social competition, emulation, and rewards (beer and the relaxation of drinking in a group are highly valued) for increasing the vigor and amount of labor applied. Any slacking or failure of a household to send a worker is publicly noted and ultimately penalized. Though households and individuals continue to work on their own fields during other parts of the day, Kofyar regard their cooperative work as both effective and pleasurable. If hoeing a single field is stretched out over weeks, it cannot be planted at one time, or planting is seriously delayed. The crop may not ripen all at once. Growing weeks may also drastically lower yields in part of a big field not reached promptly by a small household labor group. During the agricultural season, Kofyar labor input is so steady that there are practically no breaks in the tight schedule, and a longer time spent on one crop will interfere with necessary tasks for other cultigens.
Only those who have not been able to schedule a festive labor party, who are attempting to increase their cultivated area, or who pursue other occupations in town are likely to hire labor. Though new farms are becoming smaller and harder to establish, most Kofyar can still acquire some land. They are not pressed by lack of resources to do wage work, and there is not yet a landless proletariat with no choice but to accept depressed pay scales. At crucial times in the agricultural year the demand for labor fai outstrips supply. For a few specialized dry season tasks such as well digging and cereal threshing, crews of Hausa long-distance seasonal migrants from the Kano area are now beginning to appear. Just how soon Kofyar economic inequality will force some individuals into the labor market on a permanent basis is unclear. As long as population density on the frontier remains below the level at which small holdings are economically viable, intensification maintains adequate returns on labor, and out-migration reduces pressure on land, the role of wage labor may not appreciably increase. Intensive, careful techniques also put a premium on the skills and unsupervised responsible efforts of household members and neighbors with common understandings and shared economic interests.
CASH-CROPPING AND DEPENDENCE
Selling agricultural products in Africa is often seen as being synonymous with incorporation into a market economy, capitalist penetration, and the capture of a peasantry. by urban, commercial, and bureaucratic sectors of the economy (Hyden, 1980; Shenton, 1986). A postulated earlier "Golden Age" of economic self-sufficiency, political independence, and communal equality inevitably falls victim to forces from the city, the nation, and the world system that systematically exploit the countryside.5 • Though the Kofyar are subject to the same local government, taxation, terms of trade, and mercantile institutions as other Nigerian farmers, they appear to have voluntarily embraced a growing degree of market involvement and to behave in a manner that furthers their economic goals. In so doing, they have not sacrificed their own ability to provision themselves, maintain their nutritional status, and preserve their household autonomy. They have not become dependent on wages, indebted, or subject to administrative control. They appear to have chosen to migrate, to work longer and harder than they did in the past, to sell crops on the market, and to buy new goods and services. The economic situation is assuredly still in flux and its positive aspects may be transitory, but the Kofyar do not react to either the state or the market in terms of loss of control, insecurity, or powerlessness. Their conditions arise from an environment and a recent economic and political history in which constraints, choice, and chance all play their parts.
The characteristics of the Benue valley are undeniably relevant. A wellwatered area of reasonably fertile soils was (1) almost devoid of settlement, and (2) near enough to the Kofyar Jos Plateau homeland to be reached and cultivated without pulling up stakes and moving from still-productive traditional village homesteads. The Guinea savanna climatic and edaphic conditions would support both the staple crops familiar to the Kofyar, and the yam, a valued food for the increasingly urbanized Ibo and Yoruba peoples from southern Nigeria. There was also a market for millet and sorghum grain that could absorb any surplus the Kofyar produced (Little and Horowitz, 1987). The geographic centrality of the Nigerian Middle Belt that includes the Benue and Niger river valleys, and the dry season that made possible the evacuation of produce along rudimentary bush roads meant that transport costs for bulky food crops were economically feasible. The Kofyar had comparatively easy access to land and markets.6
The fact that the Kofyar area lay between the core regions that supported Nigeria's great colonial export crops-cocoa, palm oil, and groundnuts-meant that they came later to cash production, but possibly more propitiously. The Kofyar were never encouraged to concentrate on a crop monopolized by state marketing boards, so they were not subject to fluctuating world prices or the irresistible temptation for national governments to depress producer prices and skim profits for public projects or private speculation (Hart, 1982, pp. 92-93; Bates, 1983, p. 131). Kofyar yams and cereals are bought for cash by entrepreneurs who bring lorries to the farm gate, or the products are carried by bike, motorcycle, pick-up, or taxi to local periodic markets where middlemen bulk and ship to cities. There is competition among buyers and prices that appear to respond to seasonal, regional, and local changes in supply and demand.7 This indigenous system of food supply was certainly affected by national events including: The Biafran war effort of the late 1960s that impeded transport and drained wealth, vaulting urban demand powered by the oil boom of the 1970s (Watts, 1984), and the cheap imports of flour (Andrae and Beckman, 1985) and rice that the government encouraged to feed (and quiet) the cities. Raising the price of bread by restricting wheat imports, as the present Nigerian military government has proposed, may increase demand for internally produced yams and millet, for which there is no direct equivalent in international trade, but we cannot estimate what this will mean to the Plateau State farmer. We lack the daily market reports or the price series that would measure changes in supply and demand for Kofyar crops. Nor can we follow the distribution links, though the completion in 1982 of an all-weather eastwest road connecting Namu with Shendam and Lafia, along with an explosive increase of motor vehicles powered by plentiful, inexpensive petrol, enabled the routine shipping of crops and livestock as far as Port Harcourt or Kano. We can say that thousands of Kofyar farmers are finding prices in active national markets sufficiently attractive that they now devote up to half of their labor time to producing food for sale.
Neither migration to the plains nor entry into the market was hasty or particularly risky for the Kofyar. As we have mentioned, they gradually increased their time and effort on Namu bush farms while maintaining the homestead as a secure source of food. Burning trees and advancing the farm boundary took place over a number of years. Any produce that was not sold could potentially be eaten, and a bad harvest meant less money income rather than severe local food shortages.8 Moreover, individual farmers had the leeway to experiment with crops, learn the characteristics of the soil catena on their farms, and try out new variants on intercropping and rotations (Richards, 1986). Because farmers were self-selected and self supported, they could make individual decisions on how to use resources, when and what to sell, and how to organize their labor. Payment of bridewealth, keeping a son in the household by helping him to buy a motorcycle, giving a sister's son a field, and retiring to the hill homestead were all personal choices that directly affected the agricultural enterprise.
Because no one told the Kofyar what to do or when to do it, they have made mistakes, but both successes and failures have been their own and they have had the opportunity, with time and experience, to learn from them. The original yam variety grown (akuki) has been replaced with one (danganitcha) from lboland that may have smaller tubers but that is of better quality and stores without shrinkage into the wet season. Almost every Kofyar farmer now raises pigs, which were entirely unknown as recently as 1960. A few innovators are now growing cassava or bananas for the market. Commercial fertilizer has been widely adopted since 1979, but, lacking extension advice, dosages may be too little or too much. Kofyar sorghum yields are being threatened by parasitic striga, for which they have no effective remedy. Purchased seed dressings are frequently used. The Kofyar are, we believe, willing to accept further cost-effective inputs and scientific advice, but these must be geared to their particular needs and aimed at improving their existing farming system rather than replacing it with a "package" of expensive inputs and technology.
It is clear that the major incentive for innovation as it was for the original migration to the frontier is the cash to be gained by marketing food crops. High value yams were, from the beginning, cultivated for sale, and to this day Kofyar prefer to subsist on grain rather than eating the crop. By 1965, truck-loading records indicate that 514,000 yams were being exported from the Kwande hinterland. In 1983, a dry year with a distinctly depressed production, the average Kofyar farming household had a total of 3189 yam heaps, while rich farmers put in 8000-10,000. Our censused population of 729 reporting households sold an average of 561 yams, 3.57 bags of millet, 0.63 bags of sorghum, 0.4 bags of rice, and smaller quantities of groundnuts, locust beans, cowpeas, and vegetables. Yams alone generated N438,925, an average of 13161$ per household, and total mean annual income as N1160. Almost all Kofyar smallholder households produce for the market, and income is not concentrated in a small group of big farmers with wage labor (Cohen, 1989, p. 23). (Incomplete evidence for the 1984 agricultural year of good rains suggests that farm income more than doubled.) Of total farm income, approximately one-fourth was made up of individual, as opposed to household, income, and most of the individual returns were from individual yam and peanut production by women.
Cash crops in the aggregate not only represent large quantities grown and major money earnings, but they also absorb a significant proportion of labor time. Daily records of hourly labor expenditure by the adult members of 15 households over 50 weeks show that more than 36% of all work in the fields is devoted to yams, with-rice, that is also sold, accounting for an additional 6.9% (Stone et al., 1990). Since neither of these crops was grown traditionally, and considering that the cash-crop farms are at least eight times the size of the original homesteads (Netting, 1968, p. 200), it is apparent that both total labor (estimated at 1600 hours per adult annually) and work specifically directed to marketable surpluses have been greatly increased. Because 40 percent of the millet harvest may also be sold, along with other crops and occasional goats, sheep, pigs, and chickens, it is possible that as much as 50 percent of all Kofyar farming effort is now committed to production for the market.
The money income of the Kofyar goes in part for bridewealth, taxes, and cloth as it did in the past, but expenditures for manufactured goods and services are impressive. Every household now buys soap, kerosene for lamps, flashlight batteries, and Maggi bouillion cubes. Round huts have locked metal doors with carpentered frames, and 16% of all compounds have at least one galvanized metal roof. A bicycle was one of the first big purchases of a migrant, and now each household is likely to have several.
Official records show that between 1980-1983, the Kofyar purchased almost 6000 motor vehicles, predominantly light motorcycles, but including taxis, pick-up trucks, cars, and mini-buses. School fees, especially board and books at distant secondary schools, are a charge that many families meet. There has been a quite extraordinary surge of interest in medical care, requiring payment for clinic visits, medicines, hospital maintenance (even when health care is provided free by the state), and transportation to health facilities. A survey of 136 households indicated that in the preceding month, 81, or 60% of these households had at least one member seeking medical attention. One hundred and twenty-nine individuals were treated at a cost of N2595, with the highest total of N204 for four people in one household. Mean monthly cost was over 13132 for each household seeking health care.
A variety of agricultural inputs must now be purchased. More than half of all farmers buy chemical fertilizer and seed dressings. Some people buy the millet for festive labor beer parties rather than using their own stocks. Instead of selling produce to middlemen at the farm gate or carrying their own goods to market, over 50% of farmers pay transport costs for the hauling of their grain and yams. A smaller but still significant proportion, 28%, are purchasing grain to supplement food supplies, to brew, or for resale.9 It is becoming increasingly difficult to cash-crop without monetary investments and costs.
INDIGENOUS DEVELOPMENT: SIGNIFICANT CONDITIONS
Case studies of change are always open to the charge that they reflect unique historical circumstances and that their complex multiple variables resist generalization. The simpler models from environmental and technological determinism to the capitalist world system seldom seem adequate to capture essential regularities in a complex process. Furthermore, our studies are often post hoc reconstructions of what went wrong in a specific project of planned development, and the incongruous appearance of rapid, seemingly nontraumatic indigenous economic change is regarded as a fluke, a sport of nature.
The Kofyar indeed enjoyed some fortunate conditions for their agricultural transformation in space, time, and the market. Necessary as it was to have a frontier of near-vacant, easily-worked, well-drained lands in the Benue trough at their doorstep, open tracts of land are not in themselves sufficient to draw large numbers of permanent settlers. It is possible to have so much land that shifting cultivators can utilize it in extensive fashion, keeping well apart from neighbors and harvesting a leisurely subsistence, as did many Tiv pioneers. Kofyar fields grew gradually and by accretion rather than being marked out, assigned, and regulated by some external authority.10 The Kofyar came from the Jos Plateau where space was scarce. Though they sometimes migrated from their initial bush farms to others, they had a repertoire of techniques, intensive ridging and weeding, intercropping, rotation, and manuring for establishing permanent farms and increasing production per unit area.11 They could cope with the constraint of constricting land resources. The closeness of the new lands to the homeland meant that they could commute to the bush farms on foot or, more rapidly, by bicycle. Attempts to clear big new fields did not mean leaving the security of their old small homesteads and dependable sources of food.
Time was also on the side of the Kofyar. In gradual explorations and exploitations of the forest, farmers could choose their own times to begin and, finally, to migrate permanently. They could take their time in experimenting with crops, soils, and water. They could devise new schedules of, for instance, making yam heaps in a field of growing sorghum after the millet harvest, thus combining weeding and shoring up the sorghum with providing heaps for the next year's yam rotation. This used slack time in the grain cultivation calendar, avoided working the hard soils of the dry season, and saved time in the planting bottleneck after the rains. With careful interdigitation of cropping cycles for millet, sorghum, yams, cowpeas, peanuts, and rice, along with sequential harvesting and processing and continuous livestock care, Kofyar adults work through the entire year, averaging less than 4 hours per day of agricultural work in only 16 weeks, with men averaging 1729 hours per year and women averaging 1473 (Stone et al., 1990).12 The outside observer can only conclude that Kofyar willingly use time in farm labor because (1) they know how from previous 'experience in intensive agriculture, (2) they are not coerced to undertake this extra toil, and (3) they receive reasonable prices for their produce. As independent smallholders coping with problems of weather, plant growth, and mobilizing human muscle power on their own, they decide how to allocate labor time. Their efficiency is achieved without time clocks, hourly wages, alarm bells, or any of the temporal trappings of externally regimented time.
The conscious decision of thousands of Kofyar farmers to produce for the market entailed clear and growing costs in agricultural labor, migration, and the sacrifice of leisure time. It is evident that many who came to the frontier could have continued to provide an adequate subsistence for themselves from their homestead farms. Rather than continue at a subsistence level, both men and women opted for new sources of cash, and they produced food, sold it in large and growing quantities, and readily bought manufactured goods, medical services, and education.13 Without the incentive of what they considered an adequate return on their efforts and a market that provided them with motorcycles, house-building materials, iron bedsteads, flashlights, and medicines, it is doubtful if such wholesale changes would have taken place. The rapid expansion of a national market, new roads and motor vehicles, and a burgeoning urban population, all resulting in part from international oil revenues, made cash-cropping profitable, and a flourishing, competitive entrepreneurial sector contributed to a reasonably responsive pricing mechanism. Kofyar individuals, while avoiding full dependency on the market, reached similar positive conclusions on the costs and benefits of market participation in an area where genuine market places had only been in existence since the 1930s.
Regardless of how much their economic ends may be congruent with national goals, the government did nothing directly to foster or to impede Kofyar agricultural development. In fact, the means used by the Kofyarhand tools, traditional cultigens, a farming system that they themselves evolved and adapted, and the organization of labor by households, local exchange, and festive groups-were radically inconsistent with the model of planned, technologically modern, capital-intensive development that government and outside donors have usually advocated. The mistakes made in the course of innovation were made by individuals who had the time, the advice of the community, and their own growing experience to correct them. Government rigidity in providing "scientific" packages of seeds, machines, fertilizers, and schedules without considering local needs and achievements or utilizing local experience and expertise was thereby avoided. Tractor plowing services and irrigation at government installations west, south, and north of the Namu area have not been extended to, or adopted by, the Kofyar, presumably because cost/benefit ratios were unfavorable and because the innovations were not compatible with the growing agricultural intensification and independent flexibility of Kofyar farmers. Contract tractor plowing has been employed sparingly to break lands on old volcanic flows that resist hand tools, but the Kofyar contend that even ox plows are poorly adapted to their methods of dense intercropping and yam mounding. As examples of populist, indigenous development (Richards, 1985; but see Watts, 1983b), the Kofyar may have succeeded in part because outsiders were largely unaware of their existence, and government agencies never tried to "improve" their cultivation or control their economy.
Good, free, accessible land, the time to adapt to new conditions without risking a secure subsistence base, and attractive producer prices in an expanding national economy are all important considerations in Kofyar voluntary, spontaneous change, but there may be further cultural and normative elements that distinguish Kofyar from other ethnic populations (Burnham, 1980, p. 278) that have not taken such dramatic advantage of similar conditions. We may speculate that habits of hard, disciplined work, individual initiative, and household economic autonomy that characterize sedentary intensive cultivators in land-short areas of high population density would dovetail neatly with the requirements of transforming bushlands into high-yielding, permanent smallholdings and producing a surplus for the market.14 Certainly, the traditional background promoting a degree of labor input that other groups might regard as unrelieved drudgery, mobilizing men and women as equal contributors to production, and making the family household a self-reliant and largely self-determining economic entity was in some sense a preadaptation to an individualistic, competitive economy. Kofyar options had not been limited in the past by either hierarchical lineage relationships or an autocratic traditional state, and their own society had not been significantly disrupted by the colonial government. It may be romantic exaggeration to seek some "ethic of intensification" among a group of terrace farmers whose transition from subsistence to cash-cropping on a bush frontier has been relatively rapid and, in the short term, effective and rewarding. But well-socialized habits of work and cultural templates of achievement (Ottenberg, 1959; LeVine, 1966) are neither irrelevant nor epiphenomenal to that distinctive material concatenation of environment, economics, and politics from which agricultural change emerges.
1Though rapid population growth, declining agricultural exports, and higher food prices are all obvious, the indices of sharply declining production in Nigeria are open to question, and charges of "mass starvation and malnutrition" (Iyegha, 1988, p. 51) have not been substantiated. Though there may have been regional African crop declines, subsistence food production estimates are often little more than wild guesses (Berry, 1984). "While the evidence is weak that food production per capita has declined, it is non-existent that food consumption per capita has fallen." (Caldwell and Caldwell, 1987; but see Cohen, 1989, p. 29).
2 The re-study of the Kofyar including 15 months of field research in Nigeria by Priscilla and Glenn Stone and 6 months by Robert Netting was supported by grants from the National Science Foundation (BNS-8318569 and 8308323) and the Wenner-Gren Foundation for Anthropological Research. Computer analysis of the data directed by Glenn Stone was aided by additional funding from the University of Arizona Social and Behavioral Sciences Research Institute and the Department of Anthropology. Fellowships from the Ford Foundation Foreign Area Studies Program and the Social Science Research Council underwrote Netting's studies of the Kofyar in the 1960s.
3 Though the neighborhoods integrated households from diverse origins in new cooperative labor groups and religious congregations (Roman Catholic chapels have been built by the residents in most neighborhoods, but there are also Protestants, Moslems, and traditional adherents), economic choices do not appear to be constrained by "communalist relations" or the "economy of affection" (Hyden, 1986). Even in the home communities, where lineage and clan descent groups were significant socially and politically, extra-household kinship played little role in organizing production (cf. Lewis, 1981). and easy access to drinking water. As yields from shifting cultivation declined and grass fallows became harder to work, households would seek new field sites in the bush.
44 Sau1 (1983) argues to the contrary that the big expensive volunteer parties among the Bisa of Burkina Faso are inefficient, careless, cannot be pushed, and try to combine work witt diversion. Only the poor integration of the labor market with few workers available during the bottleneck weeding and harvesting periods prevented more frequent hiring of wage workers. His discussion, however, refers to large, bush-fallow fields at some distance from th( village and with relatively free usufruct land rights. A more extensive system of this kin( with less intercropping and an absence of fine-comb techniques does not require the carefu attention that Kofyar groups regularly provide their hosts.
5 Hill (1986, p. 70) points out that this sentimental, largely unconscious "Golden Age fallacy" is shared by Marxists and non-Marxists alike.
6 The provision of roads has been perhaps the most important contribution of government to the advance of the frontier of settlement onto the Middle Belt plains and the accompanying expansion of cash-cropping (Gleave, 1966).
7A study of marketing margins among Hausa farmers in rural Zaria during 1971-1972 showed that the producer received an average of 68% of the final retail price of millet and almost 70% of that for sorghum (Norman et al., 1982, p. 85). Though up to seven intermediaries such as traders, transporters, and retailers claim part of the profit on grain transactions, it does not appear that their incomes are excessive, considering services provided, and ruralurban marketing functions reasonably efficiently (Norman et al., 1982, pp. 87-90).
8 The generally favorable Kofyar environment meant that they have been spared the serious, recurrent famine conditions of the Sahel to the north (Franke and Chasin, 1980; Mortimore, 1987), and their recent entry into the mercantile economy has rendered them less vulnerable to state capitalism (Watts, 1983a).
9 There is as yet no parallel among the Kofyar for the practice of poorer Hausa farmers who must habitually sell cash crops and buy staple food grains on the market, nor do Kofyar in general have dry season craft, trade, or wage employments (Matlon, 1981). Indeed, Kofyar teachers and low-ranking government employees compare their wages unfavorably with the incomes of full-time farmers.
10 Agricultural development projects too often begin with a set of requirements, including areas to be cultivated, crops and amounts to be grown, inputs and techniques to be used, and types of labor organization. The characteristics, previous agricultural experience, and motivations of the farmers are disregarded. An early, ill-fated colonial project in Nigeria, the Mokwa Scheme, asked local shifting cultivators in an area of abundant land to undertake the hand weeding of machine-plowed 24 acre parcels (Baldwin, 1957, p. 4). The settlers could not accommodate this use of land to their existing labor supplies nor their customary farming calendar, and the need to give two-thirds of their production to the company was a major disincentive to work hard.
11 Unlike the less successful migrants to the Brazilian Amazon, the Kofyar had an excellent working knowledge of the local environment and an effective technology. They also had low capital requirements, were not involved in land speculation, and did not find off-farm employment more lucrative. In these respects, they differed from most frontier populations (Moran, 1988).
12 The high and remarkably sustained labor input surpasses that of Hausa on-farm adult male employment of 609 hours (709 with travel time included), and an additional 453 hours in off-farm occupations (Norman, 1969). Hausa women provide only 9% of total farm labor.
13 It is not far-fetched to see many of the Kofyar hill households of the 1960s as self-sufficient and self-contained units characterized by the production of use values in a generally noncapitalist social formation. These same social groupings and individuals have become petty commodity producers, but unlike historic peasantries, there is no sign that "commoditization is or has been resisted by the reproduction of institutional and social relations limiting the mobility of land and labor" (Watts, 1983a, p. 20).
14 Others have speculated that the best settlers for new lands are not the local extensive cultivators but rather farmers from overpopulated areas where they have only been making a precarious livelihood (Baldwin, 1957, p. 187). The colonial government-sponsored Shendam Resettlement Scheme, not far southeast of the Kofyar, encountered difficulties with establishing ex-soldiers on cleared plots with two years' subsistence, tools, seed, and cattle, but uneducated Hill Yergam settlers from an area of high population pressure on the Jos Plateau made a successful adaptation to the frontier, even though only minimal subsidies were provided for them (Meek, 1957, pp. 283-291). This does not mean, as Hart (1982) contends, that only historically deprived or exploited peasants could meet the labor demands of such intensive production as that of irrigated we rice. "Has the [West African] work force been sufficiently prepared by famine, impoverishment, and proletarianization to constitute a reliable, hard-working, paddy-field peasantry? Our assessment is that it has not. Agricultural intensification means getting people to work harder, and that undertaking usually requires coercion. The compulsion would have to be especially severe in the forest areas, where labor-efficient techniques of production yield up the population's food need with comparatively little effort (hence the popularity of manioc)" (Hart, 1982, p. 48).