^{Page 362 note 1} See D. Biebuyck, ‘Land Holding and Social Organization’, and Bohannan, P. J., ‘Land Use, Land Tenure and Land Reform’, in Herskovits, M. J. and Harwitz, M. (eds.), Economic Transition in Africa (Evanston, 1964), pp. 77–88 and 133–50.Google Scholar

^{Page 363 note 1} This pattem is ably described by Levine, Donald in Wax and Gold: tradition and innovation in Ethiopian culture (Chicago, 1965),Google Scholar and by Hoben, Allan, ‘Social Stratification in Traditional Amhara Society’, in Plotnicov, L. and Tuden, A. (eds.), Social Stratification in Africa (New York, 1970), pp. 187–224.Google Scholar Both scholars eschew the use of class typology, drawing attention to the existence of sturdy vertical lines of integration, an uninterrupted Continuum in the structure of the social hierarchy, the absence of cultural differences among the social strata, and other traditional features that minimise social distance between groups.

^{Page 364 note 1} Imperial Ethiopian Government, Ministry of Land Reform and Administration, Report on Land Tenure Survey of Tigre Province (Addis Ababa, 1969),Google Scholar and Report on Land Tenure Survey of Begemdir and Semien Province (Addis Ababa, 1970).Google Scholar A holding comprises all the land being worked by the members of one household.

^{Page 364 note 2} While rist – the traditional land tenure – is recognised by the Ethiopian civil code, its abolition has been insistently advocated by the Ministry of Land Reform and Administration, the Ministry of Finance, and by U.N., F.A.O., and I.B.R.D. reports, as well as by numerous other agencies concerned with taxation reform and rural development in Ethiopia. The imposition of agricultural income taxation in 1967, requiring individual assessment and registration of holdings, is a major step towards this goal.

^{Page 365 note 1} Source: compiled from the 1968–70 land tenure surveys of the Ministry of Land Reform and Administration. These report that the number of absentee landlords is likely to have been understated by the Chika Shums who provided the information. The share of unmeasured land is not shown in some surveys.

^{Page 365 note 2} Source: Central Statistical Office, Report on a Survey for the Province(s) … as listed in the Table, from July 1966 to Novembers 1968. The share of cultivated area is not shown in some surveys.

^{Page 365 note 3} Ministry of Land Reform and Administration, ‘Justification for Agricultural Tenancy Regulation’ – preface to draft legislation (Addis Ababa, 1968), p. 30.Google Scholar

^{Page 366 note 1} The partial exceptions are the small-holders of northern Ethiopia, and the Christian clergy who also continue to hold rights over land according to traditional arrangements.

^{Page 366 note 2} The size and distribution of this group during 1968–9 was estimated by the former Ministry of Planning and Development as follows: building and construction, 100,000; manufacturing, 60,000; transport and communication, 30,000; electricity, 2,500. Labour organisation was allowed in 1962, and has made slow progress since then; by 1970, the Confederation of Ethiopian Labour Unions claimed 55,000 members.

^{Page 367 note 1} For an elaboration of this viewpoint, see Lloyd, P. C., The New Elites of Tropical Africa (London, 1966), introduction.Google Scholar

^{Page 367 note 2} For example, see Skiar, Richard, ‘Political Science and National Integration’, in The Journal of Modern African Studies (Cambridge), v, I, 05 1967, pp. I–IIGoogle Scholar; Plotnicov, L., ‘The Modern Elite of Jos, Nigeria’, in Plotnicov, and Tuden, (eds.), Social Stratflcation in Africa, pp. 269–302Google Scholar; and Southall, A. W., ‘Stratification in Africa’, in Plotnicov, L. and Tuden, A. (eds.), Essays in Comparative Social Stratification (Pittsburgh, 1970), pp. 231–72.Google Scholar Many scholars use class as one analytic concept; see, for example, Cohen, Richard, ‘Social Stratification in Bomu’, in Social Stratification in Africa, pp. 225–68.Google Scholar Indeed, various connotations of class appear parenthetically or supplementarily in most studies of social structure in contemporary Africa. Lloyd, P. C. himself takes note of ‘incipient class conflict’ in his Africa in Social Change (Baltimore and Harmondsworth, 1967).Google Scholar A contrasting doctrinaire approach is followed by Diop, Majhemout, Histoire des classes sociales dans l'Afrique de l'ouest, Vol. I, Le Mali (Paris, 1971).Google Scholar ‘Committed’ studies normally take for granted the appearance of class formation in Africa. See Nkrumah, Kwame, Class Struggle in Africa (New York, 1970)Google Scholar; Arrighi, Giovanni and Saul, J. S., ‘Socialism and Economic Development in Tropical Africa’, in The Journal of Modern African Studies, VI, 2, 08 1968, pp. 14169CrossRefGoogle Scholar; and Amin, Samir, ‘The Class Struggle in Africa’, in Revolution (Paris), I, 9, 1964, pp. 23–45.Google Scholar

^{Page 367 note 3} Holders of higher qualifications, including university degrees, numbered 7,000–8,000 at most in 1970. The number of secondary school ‘graduates’ is impossible to calculate precisely; but though this is a much larger group, its inclusion here does not significantly alter the minute proportion of the educated elite in comparison to other social groups. There are many more who have had primary and some secondary education; however, they cannot be considered as belonging to the incipient middle class. Nevertheless, this ‘semi-educated’ group is also an urban element that aspires to modem middle-class status, and its members are oriented towards the westemised elite as a reference group rather than the traditional nobility.

^{Page 368 note 1} The starting salary for a university graduate – regardless of professional specialisation, rank, or branch of service – is Eth. $500 per month (= U.S. $217.4). Those who leave secondary school begin with a little less than half that amount. The estimated income per capita in Ethiopia in 1970 was Eth. $150 per annum. Relatively few educated Ethiopians work for private enterprise whose management is largely in the hands of expatriates – either recently arrived custodians of foreign investment, or members of the resident Italian and Greek communities.

^{Page 369 note 1} Skiar, loc. cit.; see also his Nigerian Political Parties (Princeton, 1963), passim.Google Scholar

^{Page 372 note 1} A minor exception, forced by political consideration, allows the printing of Tigrinya and Arabic in a newspaper published by the Ministry of Information in Asmara.

^{Page 377 note 1} ‘Class, Status, Party’, From Max Weber: essays in sociology, edited and translated by H. H. Gerth and C. W. Mills (New York, 1946), pp. 181–3.Google Scholar

^{Page 377 note 2} In a subsistence economy the tenant is seldom evicted, since uncultivated land has little value. In this setting, the tenant, as Karl Marx pointed out, ‘is the landowner's representative – the landowner's secret; it is only through him that the landowner has his economic existence’; Struik, D.J. (ed.), Economic and Philosophic Manuscripts of 1844 (New York, 1964), p. 123.Google Scholar The provincial land surveys reported by the Central Statistical Office in Addis Ababa show that the average tenant has been on the same land for a considerable length of time.

^{Page 379 note 1} In a survey of the 1966 freshman class in which ethnic affiliation was identified with first language learnt, 55 per Cent proved to be Amhara, and 25 per cent Tigre; C. R. Langmuir and J. E. Bowers, ‘Language Learning Patterns in Ethiopia’, Haile Selassie I University Testing Centre, October 1967, mimeo. A similar survey by Langmuir and Getachew W. Selassie of those who sat for the Ethiopian School-Leaving Certificate in 1968, showed 60 per cent to be Amhara and 22 per cent Tigre. There may be some exaggeration in these figures, stemming from the superior status of Antharinya which induces members of other ethnic groups to claim this as their mother tongue.