¹ I was engaged in fieldwork in Lesotho from August 1972 until November 1974, based mainly in a village in Leribe district which I call Ha Molapo. I should like to acknowledge financial support under the Anglo-American Advanced Research Fellowship scheme; and to thank Maurice Bloch and Simon Roberts for their helpful comments on an earlier draft of this article.

² For a tribute to this aspect of Schapera's work see Gluckman, M., “Anthropology and Apartheid: The Work of South African Anthropologists”, in Fortes, M. and Patterson, S., eds., Studies in African Social Anthropology, London, 1975.Google Scholar

³ For the period up to 1870 see Thompson, L., Survival in Two Worlds, Oxford, 1975, 190–201;Google ScholarGermond, R. C., Chronicles of Basutoland, Morija, 1967, 319–328, 431ff.;Google Scholar and J. Kimble, “Aspects of the Economic History of Lesotho, 1830–85”, History Workshop, National University of Lesotho, July 1976. I have attempted a synthesis of material on the period since 1870 in my unpublished Ph.D. thesis, Keeping House in Lesotho, University of Cambridge, 1976. But more intensive work remains to be done.Google Scholar

^{page 80 note 1} For sources see C. Murray, “Marital strategy in Lesotho: The redistribution of migrant earnings”, (1976) 35 African Studies 104.

^{page 81 note 1} Thompson, op. cit., p. 62.

^{page 81 note 2} Report and Evidence of Commission on Native Laws and Customs of the Basutos, Cape Town, 1873; reprinted Morija, 1966, 24–37.Google Scholar

^{page 81 note 3} For example, Ellenberger, D. F. and Macgregor, J. C., History of the Basuto, Ancient and Modern, London, 1912; reprinted Negro Universities Press, New York, 1969, pp. 272–4;Google ScholarCasalis, E., The Basutos, London, 1861; reprinted Cape Town, 1965, p. 183; Germond, op. cit., p. 540.Google Scholar

^{page 81 note 4} Ashton, E. H., The Basuto, O.U.P. for the International African Institute, 1952, 72–3.Google Scholar

^{page 81 note 5} Report and Evidence, 39–55.

^{page 82 note 1} Poulter, S., Family Law and Litigation in Basotho Society, Oxford, 1976, 318–322.Google Scholar

^{page 82 note 2} Poulter, ibid. Cf. also Hamnett, I., Chieftainship and legitimacy, London, 1975, chapter 3.Google Scholar

^{page 82 note 3} L. Duvoisin, quoted in Germond, p. 540.

^{page 82 note 4} Op. cit., p. 336.

^{page 82 note 5} Germond, p. 472.

^{page 83 note 1} Sekese, A., Mekhoa ea Basotho, Morija, 1893; reprinted 1970, p. 4. Poulter, op. cit., at p. 91, calls this the “conventional scale”.Google Scholar

^{page 83 note 2} They are commonly cited as

1 bovine = horse = 1 donkey = 5 small stock = R20-R30 cash.

^{page 83 note 3} I have reviewed the evidence on all these points in Keeping House. The livestock ratio is the number of stock units per head of the human population. Stock units were calculated here in terms of the grazing equivalents of

1 unit = 1 bovine/equine (large stock) = 5 sheep/goats (small stock).

^{page 83 note 4} Murray, “Marital strategy”, 106–7. For a detailed analysis of this sample by media of payment, see Keeping House, 232.

^{page 84 note 1} The statistical evidence is scanty. Figures are given in the following sources but it is not always clear whether they represent normative statements or observed or recollected transactions. For the Tswana, see Schapera, I., Married Life in an African Tribe, London, 1940, reprinted Penguin Books, 1971, p. 77;Google Scholar and I. Schapera and S. Roberts, “Rampedi Revisited: Another Look at a Kgatla Ward”, (1975) 45 Africa 3 266. For the Rolong, Tshidi, Comaroff, J., Competition for office and political processes among the Barolong Boo Ratshidi of the South Africa-Botswana Borderland, unpublished Ph.D. thesis, University of London, 1973, 436.Google Scholar For the Kgalagari, A. Kuper, “The Kgalagari and the Jural Consequences of Marriage”, (1970) 5, Man (N.S.), 472, 481. For the Pedi, Sansom, B., “A signal transaction and its currency”, in Kapferer, B., ed., Transaction and Meaning, Philadelphia, 1976.Google Scholar See also Seymour, S. M., Bantu Law in South Africa, Capetown, 1970, 144–152.Google Scholar

^{page 84 note 2} GES I, Evaluation Study No. 5, Thaba Bosiu Rural Development Project, September 1975. The transfers were divided into those involving livestock only (Category 1, 40%), those involving cash only (Category 2, 20%) and those involving both livestock and cash (Category 3, 40%). Using grazing equivalents, the mean transfer in each category was as follows:

Category 1: 5.1 stock units;

Category 2: R134.30;

Category 3: 3.2 stock units and R65.50.

The range for all three categories in terms of cash equivalents (1 stock unit = R25) was R127.50 to R145.50. These transfers must be assessed against farm household income. Figures from an unpublished survey, show that the mean cash income of these farm households in 1973–74 was R487, while the median income was about R380.

^{page 84 note 3} Fortes, M., Kinship and the social order, London, 1970, 256.Google Scholar

^{page 85 note 1} I estimate in Keeping House, 35–39, that approximately 60% of the average household's cash income in 1974 was derived from migrant earnings.

^{page 85 note 2} The 1970 Census of Agriculture Report, Bureau of Statistics, Maseru, 1972; the Report of the Lesotho Pilot Survey on Population and Food Consumption, May 1973, Maseru, 1976, 20.

^{page 85 note 3} It is possible that mean household size in Lesotho, which is smaller than that in most other African populations in southern Africa, reflects a relatively greater dependence on migrant earnings as against other sources of livelihood within the domestic economy, cf. Keeping House, 162.

^{page 86 note 1} Sheddick, V., Land Tenure in Basutoland, London, 1954. He writes (p. 21): “A man's parents will remind him that he must treat his wife reasonably well because he is after all only her guardian; that she really belongs to them by virtue of the fact that they provided the bohadi payment; and furthermore that her children are in fact their children.” This statement fairly represents Sesotho “custom”, but precisely because parents seldom provide bohali cattle nowadays it does not explain why so many children live with their grandparents.Google Scholar

^{page 86 note 2} See the discussion in Poulter, op. cit., 90–105.

^{page 86 note 3} Murray, “Marital strategy”, 105.

^{page 87 note 1} It is also possible for the protagonists either to invoke alternative principles of customary law or to counterpose alternative evidence as to the facts. Poulter's Family Law and Litigation appears to represent a tradition of legal positivism which endeavours to resolve inconsistencies in the customary law by the application of professional expertise, and thereby to offer unequivocal definitions of status and correct resolutions of disputes. But an emerging consensus of scepticism of the value of such an approach may be discerned in other studies by ethnographers of Sotho-Tswana customary law. Ian Hamnett, op. cit., argues that the emphasis in the application of the law in Lesotho, particularly by the lower courts, lies not in imposing particular substantive resolutions but in teasing out compromise in particular circumstances between principles of customary law that are seen to carry divergent implications. Such inconsistency is not necessarily rationalized by the accumulating weight of precedent but, outside the context of litigation, these principles may revert, like springs no longer in tension, each to its generalized validity. A similar view, although dressed up in different theoretical language, is implied in John Comaroff's analysis of succession disputes amongst the Rolong, Tshidi, in his thesis Competition for office and political processes, op. cit.,Google Scholar and “Chiefship in a South African Homeland”, (1974) 1 Journal of Southern African Studies 36–51. He argues that succession, far from being “determined” by an unequivocal statement of customary law, is a political process in which competing factions attach different weights to various “facts” in order to justify contrary interpretations. Candidates for office present competing genealogical rationalizations of their claims, and the winner revises the official genealogy so as to legitimize his accession. In this way the resolution is always a “correct” one and the process of succession may be represented more realistically in terms of the statement, “So-and-so is the eldest son of the previous chief because he is the present incumbent”, than vice versa.Google Scholar

^{page 87 note 2} The composition of 150 households in five sample villages was investigated in 1974. Detailed results are available in Keeping House, 161–174. The importance of residential attachment through women may be inferred from the following summary analysis of lineal links between alternate generations.

(i) Half (75 out of 150 = 50%) of the households contained one or two generations; the other half contained three or more generations.

(ii) More than half (52 out of 96 = 54%) of lineal links between alternate generations, in households containing three or more generations, were traced through women, and less than half (44 out of 96 = 46%) were traced through men.

(iii) A few (5 out of 52) of the links through women reflected a Sesotho customary expectation that a woman should bear her first child at her natal home; the majority (32 out of 52) of links through women derived from circumstances of pre-marital childbirth or of conjugal dissociation; and the remainder (15 out of 52) reflected some degree of choice on the part of parents whose conjugal association persisted.

(iv) Nevertheless the predominantly agnatic character of inter-household residential alignments (measured in terms of the relationship between household heads) may be expected to persist, for the same high proportion of links through women (42 out of 52 = 81%) and links through men (36 out of 44 = 82%) was traced by reference to virilocally resident parent(s) in the senior generation.

These findings are consistent with those of Schapera and Roberts for the Kgatla in “Rampedi. Revisited”, 267. Although their figures are not comparable with mine, they convey the same general impression, viz. of a high proportion of “unmarried” women, and a high incidence of births to such women.

^{page 88 note 1} Gluckman, M., “Kinship and Marriage among the Lozi of Northern Rhodesia and the Zulu of Natal”, in Brown, A. R. Radcliffe and Forde, C. D., eds., African Systems of Kinship and Marriage, London, 1950;Google Scholar cf. also J. Goody, “Bridewealth and Dowry in Africa and Eurasia”, in Goody, J. and Tambiah, S. J., Bridewealth and Dowry, Cambridge, 1973.Google Scholar

^{page 88 note 2} Barnes, J. A., Marriage in a Changing Society, Rhodes-Livingstone Papers No. 20, O.U.P., 1951, p. 123.Google Scholar There appears to be more evidence to support this proposition than to controvert it. For example, Monica Wilson has recorded a drastic increase in normative and actual bridewealth payments among the Nyakyusa, in “Zig-zag change”, (1976) 46, Africa, 4, 399–409; cf. also Watson, W., Tribal Cohesion in a Money Economy, Manchester, 1958, 40–42; and Kuper, op. cit., 472, 477, who notes an inflationary tendency in rates of bogari amongst the Kgalagari on the fringes of the Sotho-Tswana culture area. The main Tswana tribes appear to be exceptional to the general trend in this respect.Google Scholar

^{page 88 note 3} I have explored the structural implications of delay in the transfer of bohali in “The symbolism and politics of bohali: household recruitment and marriage by instalment in Lesotho”, in Krige, E. J. and Comaroff, J., eds., African Marriage in Southern Africa, South Africa (forthcoming).Google Scholar

^{page 89 note 1} See H. Wolpe, “Capitalism and cheap labour power in South Africa: from segregation to apartheid”, (1972) 1, Economy and Society, 425–456, and “The theory of internal colonialism: the South African case”, in Oxaal, I., Barnett, T. and Booth, D., eds., Beyond the Sociology of Development, London, 1975.Google Scholar Cf. also C. Meillassoux, “From reproduction to production”, (1972) 1, Economy and Society, 93–105; and Arrighi, G., “Labor Supplies in Historical Perspective, A Study of the Proletarianization of the African Peasantry in Rhodesia”, in Arrighi, G. and Saul, J., Essays on the Political Economy of Africa, Monthly Review Press, 1973.Google Scholar For general essays relevant to this theme in Africa see Gutkind, P. and Wallerstein, E., eds., The Political Economy of Contemporary Africa, Sage Publications, 1976.Google Scholar

^{page 89 note 2} Op. cit., p. 435, his emphasis.

^{page 89 note 3} For a critique of “neo-marxist” interpretations see Kantor, B. S., and Kenny, H. F., “The Poverty of Neo-Marxism: the case of South Africa”, (1976) 3, Journal of Southern African Studies, 1, 20–40. One problem they raise is that, in Wolpe's terms, the relationship between wages in the capitalist sector and production in the rural periphery must accommodate apparently contradictory requirements. Agricultural production in the rural areas must be sustained at a level sufficiently high to contribute to the costs of reproduction of labour power; but it must not be so high as to counteract the economic imperative to migrate. As Kantor and Kenny remark, somewhat caustically (p. 23): “Not too high and not too low, a requirement which, it may be thought, should test the ingenuity even of capitalists”. Arrighi's formulation, op. cit., of the second requirement of Wolpe's argument, that the “effort-price” of participation in wage-labour be lower than the “effort-price” of participation in domestic agricultural production, is essentially tautologous.CrossRefGoogle Scholar

^{page 89 note 4} See Murray, Keeping House, 99–129.

^{page 90 note 1} Sansom, B., “Traditional Economic Systems”, in Hammond-Tooke, W. D., ed., The Bantu-Speaking Peoples of Southern Africa, London, 1974;Google ScholarKuper, A., “The Social Structure of the Sotho-speaking Peoples”, (1975) 45, Africa, 1, 67–81, and 2, 139–149.Google Scholar For a historical caveat on the assumption of Sotho-Tswana homogeneity, see Legassick, M., “The Sotho-Tswana Peoples before 1800”, in Thompson, L., ed., African Societies in Southern Africa, London, 1969.Google Scholar

^{page 90 note 2} I. Schapera, Married Life, op. cit., p. 77, and A Handbook of Tswana Law and Custom, London, 1938, p. 143; Schapera and Roberts, op. cit., 266. For a statistical comparison over time of the incidence of types of dispute in Botswana see S. Roberts, “The Survival of the Traditional Tswana Courts in the National Legal System of Botswana”, [1972] J.A.L. 103–129.Google Scholar

^{page 90 note 3} Roberts, S., “The Kgatla marriage: concepts of validity”, in Roberts, S., ed., Law and the Family in Africa, The Hague, 1977.Google Scholar

^{page 91 note 1} For a full account of traditional obligations see Ashton, op. cit., pp. 62–87.

^{page 91 note 2} E. Preston-Whyte, “Kinship and Marriage”, in Hammond-Tooke, op. cit.

^{page 94 note 1} Murray, “Marital strategy”, 112–113.

^{page 95 note 1} Ibid., 115.

^{page 96 note 1} Murray, “The symbolism and politics of bohali”. For details of bohali transfers in the marriages of ’MaNana and two of her brothers, see “Marital strategy”, Table 3, Nos. 6, 12 and 13 respectively.

^{page 96 note 2} L. Duvoisin, 1885, quoted in Germond, p. 536.