For a long time historical writings on agrarian differentiation in Zimbabwe focused on the racial division of land and on inequalities between white commercial farmers and an increasingly marginalised subsistence sector; the rural population was usually conceptualised as a `peasantry' or `proletariat' (Arrighi, 1970; Palmer, 1977; Moyana, 1984). In posing two mutually exclusive categories, scholars tended to assign each Zimbabwean farmer to one or the other idealised, homogenous category.
Recent writings on Zimbabwe and elsewhere in Southern Africa (see Neocosmos, 1987) have enhanced our understanding of the process of agrarian change and differentiation in the rural areas. They have argued that most rural areas experienced considerable commercialisation of agriculture during and after the colonial period, and that commercialisation gave rise to various forms of socio-economic differentiation among rural households and communities. Scholars in this group have advanced various arguments about how differentiation occurred in rural Zimbabwe. They have pointed to differences in access to land; differences in households' access to off-farm income; the privileged position and political manouevring of chiefs; and differential access to labour, which again is related to various factors, including access to off-farm income, polygamy and co-operative labour (Phimister, 1988; Masst, 1996; Ranger, 1985).
However, even studies that recognise patterns of differential land holding have not shown clearly how such differences occurred over time. Research into land distribution among the Zimbabwean peasantry has largely been limited to surveys documenting the amount of land held by each household at a particular time. There has been little investigation of the process by which land is actually accessed, thus into the means by which some households have acquired significantly larger land holdings. The lack of such a study is attributable to the tendency among scholars to argue that the distribution of land in the `communal' areas is less skewed than other variables and acts as a brake on individual accumulation and on the process of differentiation. Metre Masst's study of peasant differentiation in north-eastern Zimbabwe epitomises this line of argument. She argues that `the differentiation process within the peasantry is ... constrained by the communal land tenure system, which governs land distribution in Zimbabwe's communal lands. This system prevents commoditized exchange of land, and preserves the existing pattern of small landholdings' (1996: 382, emphasis added). She further argues that `land distribution within the communal areas is quite egalitarian', and she attributes this egalitarianism to what she calls `the predominance of a peasant ideology of egalitarianism within the communal areas' (p. 383). Other observers, including the present government, have attributed the egalitarian structure of land holdings to egalitarian values they claim are inherent in the communal land tenure system (Zimbabwe, 1985; Yudelman, 1964).
This article challenges the generalisation of the land tenure system and landholding practices in the communal areas of Zimbabwe. It argues that such generalisation arises from the misconceptualisation of `communal' tenure, a term used in the literature to reflect an African attitude to land rather than a European one. The idea that `customary' land law in general was created under colonial rule has gained wider currency among scholars. Several scholars have demonstrated how the colonial authorities sought to enforce long-established custom rather than current opinion, thereby freezing what had previously been dynamic systems (Colson, 1971; Ranger, 1983; Chanock, 1985). Other writers have extended this argument to the post-colonial period, suggesting that African governments have kept alive the colonial myth of communalism in order to justify increased state control over land. The new governments asserted total control over the lands in the name of protection of African customs. …