A paper presented at the 17th Annual Meeting of the Amerian Anthropological Association,
December 4, 1975, San Francisco

The social processes and cultural forms of East African pastoralists can be understood in large measure as being patterned by, or adapted to, constraints inherent in the ecology at cattle. Maintaining herds from day to day and from generation to generation involves problems of monitoring, of flexibility, and of continuity quite different from these faced by agriculturalists.

By their nature, cattle are controlled and defended by men. Each animal is distinguished individually by a system of descriptive terms. With few exceptions ownership resides with individual men or sets of full brothers. Inheritance is overwhelmingly from fathers to sons. Because cattle beget cattle in a very direct and limited way (as opposed to gardens or craft skills} continuity of ownership, i.e. the presence of male heirs, must be achieved in every possible case.

Corporate groups are defined in relation to family herds; the size and composition domestic groups and local communities must reflect viable herding strategies. The principal ties between different family units are by marriage, the transfer of cattle to the bride's family legitimating the relationship and establishing the husband's paternity over all children born to her. In sum, definitions of rights and obligations in the idiiom of kinship are, to large extent, a transform of the rights and obligations involved in the ownership and use of cattle.

Much of the social dynamics of pastoral societies can be traced to the problem of keeping to these basic principles in the face of the vicissitudes of life. In doing so there is value in keeping all internal boundaries porous and all local imbalances temporary.

Ownership is never absolute, agnates share responsibilities to various degrees, and the limits of a cattle-owning or cattle-controlling unit are subject to some negotiations. Similarly, the collection and distribution of bridewealth cattle frequently involve cooperating units larger than that of direct ownership. High rates of polygyny serve to distribute dependents among families according to resources, while inequalities are also tempered in part by large payments of marriage cattle. The "house-property complex" (Gluckman 1950:195) prevents the accumulation of wealth over generations in specific families. Shorter term discrepancies in the distribution of cattle to land and people to cattle are reduced through systems of individual loans or cattle partnerships. Such loans also permit individual units to underwrite or reinsure each other's risks.

While a family exists as a unit in relation to a set of cattle, pastoralism does not require the physical presence of the owner at the herd, the geographical aggregation of all the people sharing rights to the herd, or even the geographical aggregation (or herding together) of the cattle they share. Indeed, in East Africa fluctuations of rainfall, herd size, and many other variables require some or all of the human and bovine populations he be involved in transhumant grazing, dispersing and coalescing in ever-changing arrangements. Geographical subdivisions are conceived, not as coherent, we--bounded land blocks, but as areas of ongoing interaction. Generally more people have rights to use a given locale than can be sustained there at any one time. Each cattle operating unit seeks to establish a variety of relationships with many others to determine its own strategy according to its options. Individual autonomy is jealously guarded, group action requires consensus, and influence arises from being well connected rather than from any system of constituted authority. Typically some system of age-sets defines etiquette and facilitates cooperation among men with regard to differences in community of origin, wealth, family, or genealogical position, while at the same time rewarding the experience of elders and instituting the duty of young men to defend cattle aggressively.

There are many more resonances between the constraints of pastoralism and the various sub-systems of pastoral societies that have been noted. Suffice it to say that the model is well established. The social and psychological particulars of the "cattle=complex" are now appreciated not as peculiar curiosities but as parts of a highly redundant and finely tuned ecological adaptation. One does not have to appeal to any irrational fixations with traditionalism to understand why people are reluctant to tamper with or neglect, with obviously irreversible consequences, a system of proven efficiency in order to embrace uncertainties.

Among East African pastoralists the Kipsigis (and the closely related Nandi) pose a notable exception. At the start of the twentieth century the Kipsigis probably numbered less than 100,000, with an economy based on pastoralism and on hoe cultivation of finger millet by women. Unlike most East African pastoralists, the Kipsigis live in a highland environment, the western slopes of the Mau Narok, ranging from grasslands at 5,000 feet with 40 inches of rain a year up to forest zones beyond 7,000 feet with up to 70 inches of rain a year. in such a plush environment little transhumance was required. The majority of the population lived in settled communities of dispersed homesteads with gardens, milk cows, sheep and goats, while the men and elder boys grazed cattle in the grasslands in the lower areas to the west. In most cases these were not many miles from the home communities. With an economy designed to exploit the juxtaposition of two zones, the Kipsigis were able to displace gr incorporate elements of the more numerous agricultural Gusii and the formidable but less numerous pastoral Maasai, securing all but the last corner of a coherent land mass when British administration was imposed. The Kipsigis thus started the colonial era with a vast land base; Pilgrim (1961233) estimates that only one-fifth of it was being used at that paint for permanent settlement. This situation also drew the attention of the British who alienated two large sectors. One, prime grazing land to the west and southwest, was uses for dairy farming and justified in terms of separating the Kipsigis from the Gusii and aiding in their "depastoralization". The other alienated area, in the north, was predominantly fertile highlands. Tea was first planted here in the early nineteen-twenties and in the next few decades vast plantations were established.

The dramatic transformation of Kipsigisland which ensued has been discussed in detail by Robert Manners (1967). Quickly pacified, the Kipsigis adopted maize with some reluctance were drawn into wage labor and the edges or the cash economy, learned plow cultivation from European settlers, and turned with official encouragement and some enthusiasm he cash cropping maize to supply the plantation work force. When the first person successfully fenced a complete farm in 1931, land enclosure spread throughout the Reserve with rapidity. By the mid nineteen-fifties virtually all land was privately owned and fenced.

Visually, most of the countryside is new a rich agricultural area covered with a quilt of rectangularly fenced fields of several different crops and pasturage, thick with homesteads, dotted with primary schools, dissected by reads supplying over a thousand shops. There has been a four-fold increase in population, and in the most fertile areas individual tea holdings, European dairy cattle and tractors are well established.

Whet happened to all that deep-coded pastoralism? In describing the situation in 1962, Manners states

Concerning the role of cattle he says

Granted Manners 'excellent discussion of the factors which made the agricultural transformation overdetermined or "ordained"; the phenomenon of pastoralists actively adapting traits supposedly anathema to their basic values remains somewhat paradoxical, particularly if viewed in terms of a one dimensional model with modern economics replacing native cattle. The paradox is heightened by the continuing pastoral "feel" to much of Kipsigis life -- about which more in a few minutes.

I contend that we are faced with a more complex situation, one in which pastoralism is alive and not all that unwell 1n the midst of tractors and super phosphate.

My own fieldwork¹ focused on the community of "Kapsuswek" in a grassland area just at the edge of the more fertile hills. Although primarily a grazing area, and not noted for progress, all homesteads in the area grow maize and most market some. Grade cows and tracters have now been tentatively introduced. While not wishing to diminish the evolving differences between the two major zones, I feel that the differences between Manners' research and my own are not simply where we looked but how we looked; the resulting images are, I hope, complementary,

To put the case most simply, the Kipsigis continue to keep large numbers of native cattle. Data from 26 homesteads in Kapsuswek (collected in 1966 and 1967) show 219 people and 575 head of cattle. The average number of people per homestead was roughly 9; the median 8. Homestead herds ranged from 13 to 32 head. Thus, on the average, there were 2.6 cattle per person.

In comparison, Evans-Pr1tchard"s estimate for the Nuer, whose herds were still depleted by rinderpest, was less than one and a half cattle per person (1940:2O). Thomas gives figures suggesting about three and a half cattle per person for a Dodoth neighborhood (l965:22). 1952 government figures for the Pokot, who are an example of "pastoral resistance" (Schneider 1959) work out to "an average or 10 to 20 bead of cattle per adult man in the main pastoral areas and an average of from 2 to 5 head in the more heavily agricultural areas" (Schneider 1957:279). The numbers tor Kapsuswek, comparable to the first figure, are 17.4 cattle per married man (including dependent sons) or 22 per homestead. I do not have figures for the Kipsigis highland areas but they are certainly higher than those cited for the Pokot.

Furthermore, the two zones continue to be coordinated in a larger system of seasonal grazing. Although the combination of pastoral privacy in such matters and the presence of a government quarantine against moving cattle made quantified data collection impossible, I encountered many herds, 20 to 50 in number, being moved at night along back roads between the two zones as wet and dry seasons alternated. Some of these cattle were moved to plots belonging to relatives, others to cattle partners, and still others, it was explained to me, to people leasing pasturage for cash.

If the Kipsigis are to be regarded as pastoralists "with a capital P", not only cattle but the continued existence of many related cultural practices should be demonstrable. I will cite just a few examples from my research in 1965-1968.

Bridewealth continues to be paid in cattle in every case I could determine, though they are now accompanied by cash and trade goods, ranging from a few hundred shillings to an extreme case in which an automobile was added to insure the blessings of the bride's reluctant father. Many men (I suspect most) who have accepted plots 1n the resettlement schemes, where only grade cattle and native oxen are permitted, continue to maintain native cattle with relatives in the former Reserves.

Compensations for homicide, intentional and accidental, are still being paid, and cattle are still being seized by force in cases of murder and elopement.

Male initiations, although new restricted by law to the six week school vacation, are still virtually universal. Those men of modern outlook who were circumcised by Christian hospital staff have generally bought their way back into traditional ceremonies. The symbolism of initiations is still heavily pastoral, as are the virtues they stress.

The famous seven age-set cycle is shill observed, and although the Kipsigis have not had e turn-over ceremeny since the start of the twentieth century, the transition to the next set was made smoothly in 1966.

Formal military organization has long since vanished but in the 1960s cattle raids still occured occasionally along the K1psigis/Maasai border where Kipsigis expansion continued despite government policy. At least as recently as 1965 an incident provoked mobilization within a few hours of thousands of men armed with spears and swords.

Institutionalized authority is still regarded as foreign by most peoples and the majority of men cannot be described as even passively politicized. Similarly, many cases heard by the government courts are understood and judged differently by local communities, and many cases, including some serious offenses, escape the attention of the authorities.

I was once willing to account for such bits and pieces as examples of cultural lag, but only because it took me some time to perceive the nature at the social organization underlying them, a system still informed in its basics by pastoral requirements.

Polygyny is surely disadvantageous to enclosed peasants with a population problem. Yet the rate of polygyny is high. A sample of 206 men aged 20 to 75 (in 1967) were married to 329 women, 39 per cent of the men being polygynous and the majority, 61 per cent, of the women having co-wives. Older men, of course, are more likely to be polygynous than younger men and such a pattern is found in my data. To my surprise, however, there was no clear indication that by 1967 the younger men had adopted monogamy as a goal. Indeed, for a sample of 58 men it is not possible to show that Iiteracy or schooling influences polygyny if one controls for age. One measure does significantly predict polygyny, however, and that is whether or not there were more girls than boys in the subject's sibling group, i.e. whether more bridewealth payments were received than were necesary foer each brother to marry his first wife (phi =.99 p< .0l,, two=tailed test, N = 58). One must conclude, as least for the lower third of Kericho District [now separated out as Bomet District], that in a majority of cases a sufficient excess of cattle in a "house" will lead to polygyny.

The practice of establishing co-wives in separate communities (Levine 1962) still occurs in a third of the cases in my data. Similarly, the traditional pattern of all but one son of each wife establishing new homesteads elsewhere continues in most cases. Any tie, other than a close agnatic one, can be used to find land. While the agricultural tribes in the Kenya highlands have sought continuities between systems of localized descent groups and land enclosure, the Kipsigis have bought and sold, traded, consolidated and divided plots to accommodate a mobility suitable to their traditional family patterns.

The dispersal of sons and co-wives, combined with a marked tendency toward local endogamy (almost all marriages occur between homesteads less than an hour's walk apart), produces a kinship network that spreads over the whale district. Every homestead has agnatic and affinal ties with others in both nearby and mere distant areas. Communities are not merely groups of adjacent homesteads where members interact on a daily basis but also particularly dense localized clusters of relationships, the majority of which are affinal.

These patterns produce some interesting surprises. Although the Kipsigis have been categorized as patrilocal (Murdock 1967), truly patrilocal residence can occur only in a small proportion of cases, and contrary from what one might expect from the term, a woman is more likely to have a mother, sister, or married daughter in the same community than a men is likely to have a father, brother, or married sons nearby. I wonder about the applicability of the term patrilineal to the Kipsigis, for while sons inherit from fathers, and agnates of the same generation are equated for classificatory uses of kin terms, there is nothing lineal about the system except the presumption of common descent among members of the same descent category (oret).

The behavior of the social network is not easily captured in our usual terms, and even in the definition of a community given above I have oversimplified. Imagine each of the homesteads as a neuron connected to several of its neighbors and more idiosyncratically to many other neurons scattered more distantly. Add to that that there are several modes in which connections exist; close agnation, common descent, direct and indirect affinal ties, both asymmetrical and symmetrical, special relationships arising out of initiation, certain general rules based on age-grades, a terminological system which allows almost any pair of neurons (households) to replace an attenuated connection with a direct kin relationship if they choose, and secret cattle partnerships connecting the vast majority of household heads. The day to day processing of material and information will cause a constant twinkling with discernable areas of locally increased activity, Various initial occurrences will activate different patterns. Agricultural chores light up small knots of interrelated neighbors. Weddings activate two agnatic nets, and some of their immediate affines quite widely. Annual initiations appear as fairly closely synchronized firings of groups of thirty or so neurons all across the board, located slightly differently each year, lasting longer than most other activities but not reaching every neuron. Over time repeated incidents will appear to indicate the presence of roughly localized clusters of 20 to 50 neurons that act together. Their firings are never exactly the same pattern, they trail off in different directions at different times, and their edges interpenetrate. They would have to be defined probabilistically. Such is the Kipsigis community, the kokwet, literally those people (i.e. household heads) who come together because they are in close proximity to and are implicated in an issue requiring group action.

Warren McCulloch (1965) has demonstrated how idealized neurons can be wired to produce the same logical propositions consistently despite instability in the behavior of individual cells anywhere in the net. I suspect Kipsigis communities maintain a sense of what they are doing and what is to be done, despite erratic behavior by some members, by mobilizing in an analogous way. But rather than attempt a description of the resilience at a net oi 50,000 homesteads, I will return briefly to the date at hand.

Twenty-four of 26 homestead: in Kapsuswek are interconnected by marriage, directly or through people in other communities. Putting this into standard genealogical symbols produced a chart 18 pages wide but only two or occasionally three generations deep for any one family. This 'attention to the horizontal' is part of what I mean by the pastoralist feel of Kipsigis social organization. One or the rules about dividing inheritance and settling bridewealth disputes is that the affairs of one generation cannot be reopened by the next. Three generations become directly involved only when a wife passes menopause without a son. In such cases cattle from her daughters' marriages are used to acquire a bride, the "wife" of the old women - in all practicality a daughter-in-law -- and hopefully a male heir will result. Both the problem and the solution hinge on caettle.e The practice is still very common; of 41 marriages occurring between 1956 and l972 involving people in Kapsuswek, 4 were of this type.

Kipsigis men are acutely involved in keeping track of all living cattle and extent relationships; those or past generations receive little attention. A large proportion of the men I worked with were unable to recall their paternal grandfather's name, except for that part indicated by their father's patronym. Similarly, elderly informants delighted in recalling their early memories for me, but whenever I sought to understand initiations, age-sets and the like in terms of their origins, I met a disinterest one might use for a student who keeps going off in the wrong direction. Rather than being "eager to discard" the past, the Kipsigis. I suspect, have always been concerned with regulating the on-going situation. That situation today still involves, among other things, a fundamental relationship between people and cattle.

If one wishes to take a very long range view perhaps all this is still a case of cultural lag during "convergence", demonstrating merely a time differential in the rate of response of various parts of a traditional system along a path of inevitable change. It is emphatically true that forces over which the Kipsigis have no influence increasingly constrain their actions. And it is also emphatically true that positive feedback processes released within their system, most notably a population explosion, are beginning to have very serious repercussions which are not totally predictable but are surely detrimental to the patterns discussed here. But with the increasing uncertainty about the future of these global forces and ignorance of the thresholds to be crossed, the retention of pastoralism among the Kipsigis, if only for an indeterminant duration, takes on new meaning.

The extensive land base at the start of the colonial era, and the return of much of the alienated land a decade ago, have allowed the Kipsigis to retain much of their domestic organization without drastic increases in population density or drastic decreases in cattle per capita. The fertility of their area, its exploitability for both plantation and small hold cropping, has led to an agricultural response of vast significance. But so far this response seems to have augmented and subsidized the retention of an older birthright more than it has replaced it. The Kipsigis economy is our domesticate new, but the Kipsigis ecological system has a long understanding of the limits of domestication and symbiosis. They might just manage to negotiate the future on a few of their own terms.