social anthropology and development planning - a case study in ethiopian land reform policy

Social Anthropology and Development Planning - A Case Study in Ethiopian Land ReformPolicyAuthor(s): Allan HobenSource: The Journal of Modern African Studies, Vol. 10, No. 4 (Dec., 1972), pp. 561-582Published by: Cambridge University PressStable URL: http://www.jstor.org/stable/160014 .

Accessed: 09/05/2014 07:08

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp

.JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact [email protected].

.

Cambridge University Press is collaborating with JSTOR to digitize, preserve and extend access to TheJournal of Modern African Studies.

http://www.jstor.org

This content downloaded from 195.78.109.51 on Fri, 9 May 2014 07:08:51 AMAll use subject to JSTOR Terms and Conditions

http://www.jstor.org/action/showPublisher?publisherCode=cup

http://www.jstor.org/stable/160014?origin=JSTOR-pdf

http://www.jstor.org/page/info/about/policies/terms.jsp


The Journal of Modern African Studies, I0, 4 (I972), pp. 561-82

Social Anthropology and

Development Planning- a Case Study in Ethiopian Land

Reform Policy

by ALLAN HOBEN*

SINCE World War II anthropological research in the new nations has come under increasing criticism on both ideological and pragmatic grounds. To those who regard the creation of a modern life-style and cultural integration as urgent national goals, the concern with traditional institutions seems conservative, if not reactionary, and any interest in ethnicity seems divisive. To those responsible for planning rural development, the anthropologist's detailed report on a single community all too often seems unnecessarily obscured by professional quibbles, devoid of concrete policy recommendations, and of unknown

representativeness. A general assessment of the validity of these criticisms is beyond the

scope of this article. It is not my purpose to defend anthropological research or to apologise for it. I do, however, want to address myself to the more modest question of whether, and if so in what ways, social anthropology can contribute to development planning. Specifically I want to ask what, if anything, of relevance can be learned from an intensive social anthropological micro-study? I will try to answer this question with an illustration drawn from my research on land tenure among the Amhara of Ethiopia.

During the past three decades the Ethiopian Government, urged on by foreign advisers, has tried to introduce a series of apparently minor changes in land taxation1 - changes intended as a first step towards a major land reform programme designed to enable rural Ethiopians to increase their productivity and achieve a higher standard of living.2

* Assistant Professor of Anthropology and Research Associate of the African Studies Center, Boston University. Research for this article was supported by the Ford Foundation, the Wenner-Gren Foundation, and the National Science Foundation.

1 For details of these changes, see Gizachew Adamu, 'A Historical Survey of Taxation in Gojjam, 19o0-69'; Department of History, Haile Sellassie I University, Addis Ababa, 1971.

2 For an excellent review of this programme, see Harrison C. Dunning, 'Land Reform in Ethiopia: a case study in non-development', in U.C.L.A. Law Review (Los Angeles), xvIII, 2, 1970, PP. 271-307.



In Gojjam Province, where I conducted much of my field-work, each of these changes in land taxation was opposed with armed resistance by the peasants they were intended to help. The most recent example of this was the revolt of 1967-8, which is reported to have taken the lives of several hundreds of persons.

Many factors, including administrative errors and peasant mistrust of the Government, contributed to these outbreaks of violence. The

underlying source of conflict, however, is that official policy has been

predicated on an inaccurate set of assumptions about the nature of land

rights, the processes through which people acquire land, and the value

peasants place on their land-tenure system in its present form. After

commenting briefly on the character of Amhara society, I will outline these assumptions, and re-examine them in the light of my own research. Finally, I will return to the more general problem of how such an intensive local study can be used in development planning.

LAND RIGHTS AND SOCIAL STRATIFICATION IN AMHARA

The Amhara people, numbering perhaps five million, occupy a central position on the great plateau that rises between the Somali desert to the east and the lowlands of the Sudan to the west. Both

politically and culturally they are the dominant ethnic group in

Ethiopia as a whole.l As heirs to the Axumite civilisation which once flourished on the

northern reaches of the Ethiopian plateau, the Amhara differ from the other peoples of Eastern Africa in several important respects. Their

language is Semitic, their religion is a variant of Coptic Christianity, they have a written tradition, and their mixed-farming agriculture is based on the ox-drawn plough. Internally, they are characterised

by a distinctive pattern of socio-economic and cultural differentiation.2

Traditionally their society was composed of a number of distinct segments based on occupation, power, and honour. The most important of these included slaves and low-caste artisans, probably making up less than five per cent of the population; independent tax-paying farmers,

1 Perhaps the best general description of Amhara society and culture is to be found in Donald N. Levine, Wax and Gold: tradition and innovation in Ethiopian culture (Chicago and London, 1965).

2 For further discussion of the Amhara cultivator as a peasant, see Lloyd A. Fallers, 'Are African Cultivators to be called Peasants?', in Current Anthropology (Chicago), 2, I961,

pp. Io8-Io; Frederick Gamst, 'Peasantries and Elites without Urbanism: the civilization of

Ethiopia', in Comparative Studies in Society and History (Cambridge), xII, 4, 1970, pp. 373-92; and Allan Hoben, 'Social Stratification in Traditional Amhara Society', in Arthur Tuden and Leonard Plotnicov (eds.), Social Stratification in Africa (New York, I970).

562 ALLAN HOBEN



SOCIAL ANTHROPOLOGY AND DEVELOPMENT PLANNING 563

including laymen and clerics,l comprising the great majority of the

Amhara; and a small non-farming titled elite, composed of secular, military lords, and the high officials of the Ethiopic Church.2

The difference between these major segments of Amhara society corresponded quite closely to the type of land rights potentially available to their respective members, and a person's standing within his

respective group also tended to correspond to the extent of his control over the appropriate type of land rights. Low-caste artisans (including smiths, weavers, and tanners in the area I studied) have until recently had access to land only through tenancy. Ordinary tax-paying farmers have held land by hereditary usufructuary rights known as rist. Members of the elite have held large estates through a class of administrative and

judicial rights known as gult, different types of which are distinguished according to the terms of the grant and the identity of the recipient. Finally, the imperial throne has, at least in theory, claimed rights of eminent domain over all the land in the Empire.

Land rights were not merely attributes of membership in the various segments of Amhara society; they also served to define political and economic relations between individual members, and hence, in an

important sense, they provided the vertical axis of Amhara social

organisation. The most significant rights in this respect were gult and rist. Gult rights over land were given to members of the ruling elite as a

reward for loyal service to their lord, and to religious institutions as endowment. The individual or institution that held such land had the right to collect taxes from those who farmed it, and also judicial and administrative authority over those who lived on it. Such rights were thus far more than just a type of land tenure. They were an integral part of the Amhara polity: they represented the granting away by a regional ruler of an important part of his taxational, judicial, and administrative

authority. Virtually all arable and inhabited land was held by someone or some

institution as gult. A particular great lord or monastry might hold many such estates, and some of these might be contiguous. However, from an administrative standpoint, and in the eyes of the peasants, each estate of gult was a distinct unit with its own internal organisation. Indeed, they were the minimal and most enduring units of secular administrative organisation in traditional Amhara society.

Rist rights, by way of contrast, were - and still are - land-use rights.

1 I would estimate that one out of ten Amhara farmers is a member of the clergy. 2 Social stratification in Amhara society is discussed by Levine, op. cit. passim; and Hoben,

loc. cit.



564 ALLAN HOBEN

It is of fundamental importance to bear in mind that rist and gult are not different types of land, but distinct and complementary types of land

rights, both of which normally extend over the same estate. There has been considerable confusion on this point in the literature, often creating the impression that gult holders had rights of usufruct over all the land

they held. But a single estate of gult land, comprising perhaps one or two

square miles, often included within its boundaries strip-fields held as rist by scores or even hundreds of farmers. Moreover, a locally born gult holder usually had rist fields within his estate, and peasants living there were obligated to give several days of labour each year to help the gult holder cultivate these fields.

A person who held land by rist right could cultivate it as he wished, subject only to the limitations imposed by the fallowing pattern of his

neighbours. By virtue of possessing the land, however, he assumed all the obligations and privileges with which it was entailed, notably: the

obligations to the gult-holding local lord, already mentioned; the neces-

sity to support the local church, or some part of its yearly cycle of masses; and the right to periodic eligibility for a number of minor offices of secular and ecclesiastical administration.

TWO VIEWS OF LAND REFORM

In the light of the close association between land tenure, status

honour, authority, and administration in traditional Amhara society, it is not surprising that many foreigners, and an increasing number of educated Ethiopians, have come to call the system 'feudal'. They have cast it somewhat uncritically in an image believed appropriate to other feudal societies: a corollary of this assumption is that economic development and social 'modernisation' among the rural Amhara can occur

only after an extensive overhaul of the traditional system of land tenure.1 Gult rights, and the administrative structure they entailed, have already been drastically curtailed by the introduction of a governmental bureaucracy down to the district level.2 The demand for further reform is a demand for reform of rist rights, and it is axiomatic with most

observers, from 'moderates' to 'radicals' that this is thought to be a

necessary first step towards rural development.3

1 For a good description and re-examination of this assumption, see Dunning, loc. cit.

pp. 293 ff. 2 See Allan Hoben, Land Tenure among the Amhara of Ethiopia (Chicago, I973), ch. Io. 3 It should be remembered that my discussion is concerned with only the Amhara rist

areas. There are many other peoples and land-tenure systems in Ethiopia and it would be interesting to know more about the effects of land reform policy on each of them.




Despite the widespread conviction that the rist system of land tenure must be changed, there has been very little concern with the way it functions at present; that is, with the processes through which rist lands are acquired and lost, and with their relationship to natural, social, and economic pressures. For the most part this has been viewed in terms of familiar western concepts of freehold tenure, and conse- quently has been found to be confused, inefficient, and an impediment to development. Actions which do not seem rational when judged in terms of these concepts have been attributed to the residual category of 'customary behaviour'. Thus, for example, the peasants' resistance to land measurement and registration, and to tax reform, is seen as evidence of an irrational, tradition-bound attitude towards land.

Two rather different images of the rist system have currency among educated Ethiopians and foreign advisers: that it is 'static' and/or 'communal'. The 'static' image, which is the older and more widely held of the two, enjoys great popularity among students and younger bureaucrats who generally favour drastic reform. It rests on four assumptions derived from the 'feudal' view of Amhara society. First, it is sup- posed that authority and status honour are based on hereditary control of land. Secondly, it is thought to follow from this that there is little social mobility among the peasants, or between them and the gentry. Since land is not bought or sold in the old Amhara provinces, a man's social standing is, in this view, more or less determined by chance of birth. The peasant is seen as toiling year after year on the lands held by his father before him, with little incentive to improve his techniques or hope of bettering his social standing. Thirdly, disputes over land rights, which are indeed frequent, are held to be senseless and wasteful squabbles resulting from insecure title, and poorly-marked boundaries.1 Finally, it is generally assumed that rural Amhara society consists of numerous land-poor peasants dominated by a much smaller, landholding gentry.

In the opinion of those who hold these views, the best way to help peasant farmers to improve their lot is to move towards a system of freehold tenure. This could be accomplished, it is maintained, by means of a cadastral survey, and individual registration of title, neither of which presently exists. The result would be an increased security of tenure against the claims of kinsmen, the introduction of land sale, and the possibility of establishing agricultural credit using land titles as

1 In the district where I carried out my field-work, nearly half of all the disputes which reached the government court concerned land. Well over half the farmers interviewed had been involved in a land dispute of some kind during the previous I2-month period.

38



collateral. Under these conditions, it is argued, peasants would for the first time have the opportunities and the incentives to work harder, to save money, to purchase additional land, and to substantially improve their lives. As a corollary, it is assumed that most peasant farmers will welcome a cadastral survey and individual registration of title, once their objectives are properly understood.

The 'communal' image of the rist system has become increasingly popular with officials of the Ministry of Land Reform and Administra- tion since its formation in 1966, to some extent based on a rather selective reading of studies by S. F. Nadel and myself, and an influential F.A.O. report.' This image has three major tenents. First, it is assumed that in some obscure sense the land is held communally, and hence is

subject to re-allocation by local elders or officials. Secondly, title to land is regarded as insecure, not only because of the possibility of re-alloca-

tion, but also because boundaries are poorly marked and land records are not kept. Thirdly, it is asserted that because of this insecurity of

title, farmers have no incentive to improve their land - in fact, in this

view, improving land only increases the probability that it will be claimed by others with 'communal' rights in it.

Most who hold these views favour precisely the same reforms as those who have a static image of the rist system, namely the establishment of freehold tenure through a cadastral survey and the registration of individual titles to land. This paradox is, however, more apparent than

real, for both images have been distorted by viewing rist through the lens of freehold concepts and categories. Both are models of how the

present system should work, rather than how it does work. Both can be

supported by 'the facts', because these are themselves the product of

concepts implicit in the questions from which they are derived. An Amhara farmer will readily state in response to a query, for example, that he 'owns' a certain field, or that he 'inherited ' it: but this does not

clarify what is meant by 'ownership' or 'inheritance'. Similarly he will, if asked, affirm that he 'shares' an interest in a large tract of land with his 'kinsmen'; but this tells us little about the meaning to him of either 'share' or 'kinsmen'. The rist system can be termed individualistic, as is

implied by the static image, for it has such features; and it can be termed

communal, for it has these characteristics. It could also be termed a

1 See S. F. Nadel, 'Land Tenure on the Eritrean Plateau', in Africa (London), XVL, I,

1946, pp. I-21 ; Allan Hoben, 'Land Tenure and Social Mobility among the Damot Amhara', in The Journal of Ethiopian Studies (Addis Ababa), 197I, first circulated in mimeographed form as part of the Proceedings of the Third International Conference of Ethiopian Studies (Addis Ababa, I966), vol. 2; and Lawrence and Mann, 'F.A.O. Land Policy Project', in The Ethiopia Observer

(Addis Ababa), ix, 286, I966, pp. 314-I5.

566 ALLAN HOBEN



SOCIAL ANTHROPOLOGY AND DEVELOPMENT PLANNING 567 mixed system, having aspects of both; but to do so would be as meaning- ful as to say that a squirrel is a mixture of a rat and a rabbit.

If a programme of land reform is to be effective it must be based on a model of what the present system of tenure is, instead of what it is not; a model which indicates what the system actually accomplishes, instead of what it does not. Such a model illuminates the rational processes through which people make decisions about land, instead of simply attributing these to the dead hand of tradition. The relevance of social anthropology to land reform policy, and more generally to development planning, is that it attempts through an intensive study of limited territorial scope to construct such a model.

THE AMHARA LAND TENURE SYSTEM

The communities I studied most intensively are located I50 air- miles to the north-west of Addis Ababa, in the Dega Damot district of Gojjam Province.1 They are perched atop a narrow windswept plateau II,ooo feet above sea-level, a day's walk from the nearest road and telephone line in the lowlands to the south-west. The rolling landscape on the high plateau is unfenced, except for an occasional temporary barrier of thorny brush, hastily erected to discourage livestock from grazing on a stand of ripe barley, or in an irrigated onion patch. Most of the uplands are in field or fallow, while the lower land bordering the plateau's many small rivers is in lush permanent pasture. These rivulets, together with major trails and the cliffs that bound the plateau on two sides, mark the boundaries of the territorial units I shall discuss presently.

The settlement pattern in Dega Damot, as in all Amhara regions, is one of scattered sprawling hamlets rather than nucleated villages - the density is of the order of 130 persons per square mile. Each homestead of from one to three huts is generally inhabited by a married couple, their unmarried children, and, if the household head is wealthy, perhaps a few dependent kinsmen. Clusters of neighbouring homesteads, often strung out along a minor pathway, make up loosely structured hamlets; and these in turn are grouped into territorially bounded parishes and estates of gult. The parish, with several hundred souls served by a single church, is the basic unit of ecclesiastical administration, while the estate still serves as the basis of secular administration, despite the fact that the powers of the gult-holding local lord have been

1 I was able to conduct field-work in this area during two periods: from April I96I for I8 months, and from April 1966 for four more months.

38-2



greatly diminished.l Approximately two-thirds of all the parishes in

Dega Damot are co-terminous with estates of gult. Most of the remain-

ing parishes are composed of more than one such estate, while in a few instances they include more than one parish.

The nature of the research

At the beginning of my field-work I had only an unfocused interest in rural Amhara social organisation, despite a working knowledge of the

many other societies I had read about in anthropological monographs. I was not particularly interested in land tenure, much less in the specific institutions through which the rist system was organised- for these had not even been hinted at in the previous literature on the Amhara.

Nevertheless, I spent more and more of my time investigating these institutions as I came to understand their central role in Amhara social, economic, and political life.

During the early part of my research in Dega Damot I relied heavily on the participant-observer techniques ordinarily used by social anthro-

pologists in intensive micro-studies of this kind. I spent most of my time in the same communities, and among the same people, and tried to establish personal ties with as many of them as was practical. I talked with farmers and priests, with low-caste artisans and titled gult holders, with government officials and learned monks, with women and children. I asked them about themselves and about each other, about what they did and thought and felt and hoped and feared.

Perhaps most important of all, I tried to relate what people told me about themselves to what they did. Sometimes I could do this by

participating in their activities, as when I attended a wedding, a land

dispute, or a church festival. Sometimes the best I could do was to record case-histories of events from what various villagers told me. In this

way, over time, I was able to gather a large fund of information about the lives and actions of a limited number of people. As I did so, I began to see how they were constrained by customary rules, and how they were able to manipulate them. I was able to see how the total complex of rights and duties, of interests and sentiments a person had in virtue of all his many roles could affect his decisions in specific situations -

including those involving land. Indeed one of the major goals of any analyst making such a micro-study is to understand the point of view of the individual actors in the society being studied - to understand the

1 See Allan Hoben, 'From Feudalism to Bureaucracy: a case study of social change in

Ethiopia'; Occasional Paper 2, Committee of Ethiopian Studies, Michigan State University, East Lansing, 1972.

568 ALLAN HOBEN




way they classify and evaluate their experience, and make decisions

concerning what to do about it. It was only gradually that I became aware of something puzzling

about rist rights, particularly about the sense in which they are here-

ditary. Finding the solution took much longer, and led me in directions I had not anticipated.

Initially, I found little to contradict the 'static' image of Amhara

society and the land tenure system, which I had accepted as more or less accurate. Men spoke frequently and with fervour of their ancestral rist - the land of their forefathers. Their sentiments about land were also evident as a recurrent theme in song and verse. Above all, the strength of their attachment to land was brought home to me by their willingness to fight about it and, if need be, to die for it. People also insisted that all rist was hereditary - that it could not be bought or sold. My questions about inheritance were met with the apparently simple statement that when a person died his/her land was divided equally among all his/her children, regardless of sex or birth-order. When I asked a man about a specific field, I was invariably told that it had come to him as his father's or his mother's or his wife's rist.

As the months passed, however, I found myself confronted with

increasingly anomalous data. Most peasants, despite their strong attachment to ancestral land, did not seem to know just where their ancestors had lived - not only great-grandparents, but often even some

grandparents. They were even less certain as to the exact location of their ancestor's fields. It also became apparent that a number of the

largest landholders in the area, including benefice-holding members of the elite, were born of humble parents; yet they claimed, and others

agreed, that all of their land, even that acquired through litigation, was their ancestral rist. Even more anomalous was the fact that almost half of all the fields in the community I lived in longest, had been

acquired by their present owners from men or women who were not their

parents, or persons whom they considered to be significant kinsmen! It thus became evident that land might legitimately be considered

hereditary rist, whether or not it was inherited. More generally, I found that the status and legitimacy of a person's right to a field were not determined by the way in which he acquired it, or from whom he acquired it. Paradoxically, then, a man might inherit a field from his father only to later have it taken from him on the grounds that it was not his hereditary rist; or, conversely, he might obtain it through litigation, or by clearing forest land, only to later defend it against the claims of others by arguing that it was his ancestor's rist!



The more I came to appreciate the central importance of rist in the lives of the people I was studying, the more I began to understand that the associated basic ideas, rights, interests, and processes differed fundamentally from those associated with freehold tenure. At the same time I found that the rist system does not resemble very closely systems of kin-group or village ownership of the types commonly found in other parts of Africa.

The rist system Most of these distinctive characteristics of rist are related to the fact

that a field so held is not thought of as an enduring unit with fixed boundaries and a permanent location. It is not an object, like a cow or a gun, which a person can dispose of as he pleases. Instead, it represents a share of a much larger tract of land, held corporately by men and women who recognise one another to be the descendants, through any combination of male and female ancestors, of a legendary figure who is believed to have first held the tract of land as his rist. Those who hold fields in an ancestral land tract because they recognise one another to be descendants of its first holder, constitute what is known to anthropologists as a

cognatic descent group - or, as I prefer to call it, a corporation. The individual's rights in a particular field are subject to the control of this

corporation. The peculiar sense in which rist is hereditary can now be better under-

stood. All land so held is considered to be held by hereditary right, because the holder is ipsofacto a descendant of the ancestral first holder. It does not matter how the present holder actually obtained possession of the field - from a parent, by litigation, by clearing, or even through tenancy - so long as his pedigree of descent from the first holder is

accepted by the appropriate corporation. The land-holding descent corporation is the central institution of the

rist system. The nature of these rights, and the processes through which men and women compete for land, can be understood only in relation to the organisation and regulatory functions of these Amhara institutions. They differ from most African descent groups or lineages in a crucially important respect: they are cognatic rather than unilineal.1 This means that the shareholders in such a corporation may validate their rights to land with pedigrees traced through any combination of male and female ancestors, instead of through lines of ancestors of one sex only as in unilineal systems. The significance of this is that people may claim

1 Land-holding cognatic descent groups have also been described in West Africa by P. C.

Lloyd, roruba Land Law (London, I962).

57o ALLAN HOBEN




Jkl

2

i ! <>A3 ! ll

Key Shaded ancestors have had land divided in their name by 'father'... A 0*

Marriage... 4_) and A 0 divorced Numbered ancestors are wife first settlers of descent ego corporations

Figure I. Pedigree through which an Amhara Farmer Validates his Rights to Rist Land

rist in any descent corporation in which any of their ancestors (their two

parents, their four grandparents, their eight great-grandparents, and so

on) have ever held such land. Since the pedigrees through which rist claims are validated (see

Figure I) are usually from 8 to 12 generations in depth, there is virtually no theoretical limit to the number of descent corporations in which



each Amhara peasant can trace such claims with the help of relatives and genealogical experts, should he think it in his interest to do so. Yet

empirically, an ordinary farmer usually holds rist land in only two or three corporations; an influential elder may have holdings in five to

ten, while a leading regional political figure may hold land in 20 to

30 corporations. There is thus a fundamental difference between a claim to rist,

established by tracing a pedigree to an ancestral landholder, and the actual possession of land by rist right, although the Amhara make no such distinction. A rist claim - that is, of potential entitlement to rist land validated by a pedigree - is considered to be inalienable and in-

extinguishable; it cannot be sold, and it cannot die through disuse. It does not, however, in itself guarantee that the claimant will be able to obtain rist land; for what is critical is not merely to have a legitimate claim (that is to say, a pedigree), but the willingness of the corporation of landholders to recognise this. A rist claim is a necessary, but not a

sufficient, condition for obtaining rist land. To grasp the dynamics of this system, then, it is necessary to under-

stand the organisation of the cognatic descent corporation, and the

processes through which individuals are able, with varying degrees of

success, to obtain rist land in many corporations through their widely ramifying rist claims. This can only be done by realising that it is solely with respect to their shared interest in this land that the individual holders constitute a group - in contrast with most unilineal descent

groups described in the literature on Africa. A descent corporation's land tract is usually from one-half to one-

and-a-half square miles in area. It bears the name of its legendary first

holder, a titled noble, a famed warrior or a priest, who is thought to have received it from his superior when the Amhara first settled there. Each

parish and estate of gult is made up of an integral number of these

tracts, and the origins and inter-relations of all three types of territorial unit are accounted for in a single cycle of founding legends.

Typically, the number of men and women who are shareholders in a

corporation - that is, who hold fields as rist land within the tract - varies from 50 to I50, or more. All consider themselves to be descendants of the first holder, and they refer to him/her as the main ancestor of the tract. One of their members is selected to act as his representative - that is, of the corporation - in situations that involve the entire piece of land. It should be remembered that the active shareholders in the

corporation represent only a small fraction of those who believe they have potential rist claims in its land tract through their pedigrees.

572 ALLAN HOBEN




Figure 2. Genealogical Charter of Corporation X

vdivision 0 A 4 C Di n T of

Figure 3. Division in Land Tract of Corporation X

The shareholders are not a solidary group of kinsmen bound together by strong common interests, sentiments of loyalty, and collective rituals. On the contrary, most of them do not consider one another to be

significant kin, they do not even necessarily all know each other, and their communal concern in the estate of one corporation is balanced by the similar interests most of them have as landholders in others. Nor are

they united by common obligations, other than paying the land tax,' and defending their tract against the encroachments of neighbouring

1 Since there is no individual registration of title in Gojjam, the land tax is assessed on the descent corporation's land tract, and it is the responsibility of the shareholders to make sure that this is collected and paid at the district tax office by one of their number.



574 ALLAN HOBEN

corporations. With regard to the allocation of fields, they do not constitute an undifferentiated group - as the term 'communal' suggests - but divide and subdivide amongst themselves. The pattern of this segmentation within a corporation must be understood in relation to the way the land tract is divided, and the genealogical 'charter' it is thought to represent.

There are two modes of arranging this. Division by 'father' is a way of partitioning land bearing the name of a remote ancestor (including the first holder himself) into equal shares in the names of his remembered children, each of whom thereby becomes the focal point of one of the corporation's segments. Division by 'allotment' is a way of assigning individual fields in a segment's land to living descendants without further subdivision by 'father' - that is, without regard for the exact structure of intermediate genealogical ties. These modes of division are illustrated schematically in Figures 2 and 3.

Division by 'father' is used to split the first holder's tract of land into smaller sections in accordance with the per stirpes rule. This means that it is divided into equal shares in the names of his putative chilrden, or - more precisely - in the names of all those of his children through whom living landholders trace their descent. The land tract of the first holder X, for example, has been divided into equal shares for each of his three children A, B, and C. Note that the entire piece of land has, as is usual, first been divided according to soil type, drainage, and exposure, and that each of these divisions (T1-5, in Figure 3) has been further split into

equal parts for A, B, and C. The shares of the first holder's children, in turn, are usually sub-

divided into equal shares in the name of those of their respective children through whom living landholders trace their descent. Thus each of A's five fields have been subdivided into equal shares for his two children 1 and 2; and each of B's five fields have been equally subdivided for her three children 3, 4, and 5. Note that C's fields have been subdivided into three equal shares for his children 6, 7, and 8, since there are at present no individuals holding rist land as descendants of his fourth child, 9. Should a descendant of g make a successful claim, however, he would be theoretically entitled to one fourth of each of C's fields.

The process of division by 'father', and hence the segmentation of the

corporation, is not usually carried beyond the third or fourth descend-

ing generation from the first holder in Dega Damot- none of the 12 great-grandchildren of X (represented by a to 1, in Figure 2) have had land divided in their names. Whatever has been divided in the name of an ancestor becomes the focus of interest to those of his descendants who




hold it as rist land, and the ancestor himself becomes the symbol of their

corporate identity. He is thus a point of structural segmentation in the

genealogical charter, and those who hold his land constitute a segment of the larger descent corporation, with a representative to look after their collection of interests. Ancestors in whose name land has been so divided are termed 'divisional fathers', while all others are simply 'fathers'.

It is essential to realise that division by 'father' is an on-going process, and not an account of historical events. It is a principle of Amhara land law used by living men to divide, subdivide, and re-divide descent

corporation land tracts in response to changing political and demo-

graphic pressures. The observable pattern of division among the segments results from the way this principle has been applied, through time and under such pressures, to decisions concerning land disputes - not from an actual division of land carried out long ago between heirs of remote ancestors at the time of their deaths. In Figures 2 and 3, for

example, it may be that C had no land until recently; and it is possible that the present pattern will be altered in the future by claims through g, or by division by 'father' in the names of some of X's great-grandchildren.

The smallest fields produced by division by 'father', or parts of them, are assigned directly to living people by 'allotment' - these are all the fields labelled 1-6. In division by 'allotment', the exact nature of the

intervening genealogical links is ignored (though they must be lineal), and the fields are allotted to accepted descendants of the segment's ancestor by its representative, in accordance with a number of pragmatic considerations: such as how much of the segment's land the

recipient already holds as rist land, the location of his homestead in relation to the land in question, his standing in the local community, and his regional political influence.

It should be evident, then, that the representative of a descent corporation segment is of pivotal importance, with both external and internal duties. As representative, he must defend his segment's land against encroachments, and make certain that it receives its fair share of land in relation to the segments of its ancestors' siblings. As adminis- trator, it is his responsibility to receive all new claims for rist land in his segment's share of the corporation's land tract, whether these claims involve further division by 'father' (and hence much land), or merely the reallocation of a few fields through 'allotment'. If he rejects the validity of the claimant's pedigree, he can be sued in court. If he accepts the validity of the rist claim, he must either supervise the re- division of his segment's land (if the claim requires division by 'father'),



or he must reallocate fields already held by someone else (if the claim involves only 'allotment').

Descent corporation representatives do not have arbitrary power; they consult with other landholders on important matters, and they can be removed from office by their constituents. Nevertheless, they play the

major role in making decisions about the allocation of their segment's land. It is not surprising that they also tend to hold a disproportionately large share of that land. It would be incorrect, however, to view their

power as deriving primarily from their control over land and descent

corporation office. In fact it is the other way around. Men are asked to become representatives, precisely because they already are influential and respected elders, gult holders, or government officials. In part, this is because in the descent corporation, as in any Amhara collectivity, only a 'big man' can successfully arbitrate the quarrels of others; and it is also because only a powerful representative can successfully defend the land of the corporation, or one of its segments, in a major dispute.

The position of the representative is thus crucially important, not

only in the organisation of the descent corporation, but also in the dynamics of the relationship between land, social status, and social mobility. For it is through this position that an ambitious,

politically astute, individual can translate his increasing influence and

power into a commensurate holding of'hereditary' rist land.

The acquisition of rist land

I would like now to shift the perspective of my discussion from the

descent corporation, and the way its land tract is divided among many rist landholders, to the individual and the way he obtains fields in

descent corporations. My purpose is to illuminate more fully the

strategies through which Amhara farmers seek to acquire rist land, the

value they place on this system of tenure in its present form, and their

attitudes towards any proposed changes. The land a man depends upon to support his household consists of a

number of dispersed strip-fields scattered over the countryside, usually within two or three miles of his homestead. Miost farming households

cultivate at least four fields averaging about three acres each, or a total

of I2 acres.1 One household in four cultivates more than 18 acres, and

one out of seven more than 30 acres, while I would estimate that not

more than one out of 50 households cultivates more than 45 acres.

Tenancy is important to some individuals, especially younger men,

1 Here I am excluding the households of widows, divorcees, and pensioners; they are not

ully involved in agricultural activities and are not self-supporting.

576 ALLAN HOBEN



SOCIAL ANTHROPOLOGY AND DEVELOPMENT PLANNING 577 but with the exception of artisans, who make up less than 5 per cent of the population, there is no class of landless people.

The scattered fields on which a household depends for its livelihood do not constitute a clearly delimited estate which passes intact from the household head to a principal heir, as in systems of impartible inheritance.1 They rather represent a collection of fields brought together under the management of the head through diverse processes and strategies, of which the most important are inheritance, gift in anticipation of inheritance, clearing woodland, pursuing rist claims against the descent corporation, and tenancy.

The relative importance to an individual of these various ways of acquiring rist land, and the institutional context involved, varies with his position in the domestic cycle, in the local community, and in the wider sphere of regional politics. A young married man usually obtains most of his land through either inheritance, or gifts in anticipation of inheritance. The institutional setting in which such land transfers occur is largely defined by the norms, interests, and sentiments of household and kinship relations.

Somewhat older men whose fathers are dead, or who live in a different community from their fathers, try to obtain additional fields of rist land from descent corporation representatives by pressing claims for a few fields through 'allotment', in virtue of their own or their wife's latent pedigrees. Such claims are formally argued in terms of descent rules, but their success or failure is strongly influenced by interests relating to community organisation and leadership.

Elders and office-holders who attain prominence in the wider regional political community try to claim larger amounts of additional rist land by demanding a further division - or re-division - of descent corporation land by 'father'. Such claims are formally argued in terms of descent corporation ideology and rules, but their success depends heavily on the claimant's ability to mobilise support for his cause through his personal political ties.

To summarise then, every Amhara farmer has a very extensive array of rist claims traced through his upward-branching pedigrees. His chances of obtaining rist land are restricted by the willingness of the leaders and landholders in descent corporations to recognise the validity of his claims.2 This in turn is affected by a number of pragmatic considerations arising from the institutional contexts of kinship,

1 A classic example of such a system of impartible inheritance is described by G. C. Homans, English Villagers of the Thirteenth Century (New York, 1960).

2 This process is considerably more complex than I am suggesting in this brief treatment; see my forthcoming Land Tenure among the Amhara of Ethiopia, ch. 8.



community, and polity in which decisions about land are made, rather than from the formal rules of the rist system as such. The most important of these considerations are: whether his parents or grandparents have held land in the corporation, for close kinsmen have difficulty refusing claims; whether he is a resident elder in the parish where the corporation's land is located, for men would rather recognise the claims of

neighbours, with whom they share collective burdens of church support and statute labour, then the claims of outsiders; and whether he has

political power.

The meaning of rist

The full meaning of rist to Amhara farmers and their attitude towards land reform can now be better understood. As I stated earlier, there is no lexemic distinction in Amharic, between rist claims and rist land: that is, between the potential rights a person believes he has in a corporation's land tract by virtue of a pedigree, and a field he actually possesses in virtue of such a pedigree. This ambiguity is not a mere ethnographic oddity, or a sign of conceptual confusion. It is an accurate reflection of a fluid land-tenure system in which it is a person's hereditary claims, and nothis fields, which are his most enduring possession-his real birth-right.

The dual reference of the term rist also points to a basic condition of the system that simultaneously creates conflict within it and commit- ment to it. This is, as I have noted, that a person's rist claims always far exceed his rist land. In other words, he is always able to trace pedigrees to ancestral first holders in whose land tract he does not yet have fields. To put it yet another way, there are always far more legitimate descendants of an ancestral first holder than there are men who hold

fields in his land tract. Whenever a man tries to obtain rist land through his rist claims he comes into conflict with those already using the land. Yet it is precisely by entering into conflicts of this type successfully that

a man can increase his holding of 'hereditary' land and improve his

social status. The frequency and virulence of conflict over land in rural Amhara

areas tends to obscure the fact that people are deeply committed to the

rist system. In it they see the guarantee of the liberties and opportunities they most cherish. In their eyes it enables them in time of personal

adversity to move to another community, and still be accounted heredi-

tary landowners. It enables their children, no matter how numerous, to

become independent land-owning farmers, and it allows them to increase

their personal holdings and community standing.

578 ALLAN HOBEN




LAND REFORM RECONSIDERED

Proposals to individualise land tenure are, in effect, proposals to abolish the rist system, and the descent corporations through which each person has inalienable potential claims to a share of numerous, widely scattered land tracts. But there can be no doubt that, at present, Gojjam Amhara farmers do not want such a change in the rules under which they compete with one another for land. I would add, somewhat more dispassionately, that any assessment of the direct and indirect effects of radically altering or abolishing the rist system must take into account what it accomplishes in its present form.

From a general demographic and social point of view the rist system is not merely a way of allocating land to people. It is also a way of allocating people to available land in accordance with their social and political prominence. It serves to move people from communities which are densely populated to ones which are not; and, at the same time, it allocates individuals with unusual political skills the lands commensurate with their political attainments. It is thus a method of adjusting the actual ecology of an agrarian society to the political realities of a competitive and fluid polity, and without producing a large class of landless and alienated peasants.

In the absence of another mechanism for allocating land to people and people to land, the transformation of rist to freehold- through a cadastral survey, and the registration of individual title to land as it is currently held - would, in effect, freeze a transitory pattern of landholding and social stratification at one moment in time. It would convert a fluid system of individual inequalities into a permanent pattern of economic and social stratification; and, paradoxically, in the absence of substantial economic development, it might well lead to the creation of many landless peasants.

The relevance of the micro-study for development planning I think there can be no doubt that a study such as mine is very

relevant for assessing the effects of development policy in the communities where the research is carried out. First, it can correct factual errors in policy makers' knowledge of local conditions. For example, despite widely held views to the contrary, large holdings of rist land were found to be most unusual in Dega Damot, and absentee land- lordism is virtually non-existent.l Similarly, it became clear that while

1 This is because the land that a man who leaves Dega Damot gives to relatives or ten- ants is soon claimed by them as rist.



individual landholdings are certainly fragmented, this is not the 'inevitable' consequence of equal inheritance by all children, but rather of the way descent corporation lands are subdivided and allo- cated. Furthermore, at least some degree of fragmentation is highly desirable from the point of view of the Amhara farmer, for by providing him with fields of different qualities it enables him to diversify crops to reduce the risk of a total failure. Nor is the official view valid that fragmentation greatly reduces agricultural output by wasting labour under

present conditions, because most households have more labour than their land requires for cultivation. Finally, the contention that insecure tenure inhibits farmers from improving their land is belied by the fact that fertilised land, and land on which trees have been planted, is not

subject to re-allocation in division by 'allotment'.

Perhaps more to the point is the farmers' own observation that they do not produce as much as they could, because 'money is too expensive' - that is, the price they receive for their produce is low. In this respect I would agree with the opinion voiced by the majority that what is most needed to spur development is a road, and not a programme of land reform.' As the inhabitants of Dega Damot become increasingly involved in a cash nexus, and as economic criteria come to replace political authority as the basis of status honour, land reform of some kind will undoubtedly become both desirable and desired. In my view, how-

ever, land reform under present economic and political conditions will have only unintended and deleterious consequences.

Secondly, in a more general sense, a study such as mine contributes to development planning by providing a comprehensive picture of the

way decisions are made - in the present case, decisions concerning land. It can do this by investigating the kinds of interest individuals have in land rights, and the institutional context in which decisions affecting these interests are made. Understanding the point of view of the individual decision-maker is crucially important for understanding attitudes towards development plans; for Amhara farmers, like most

people everywhere, generally evaluate government policy in terms of the

way it will affect them personally, and not in terms of its effects on such abstractions as the 'economy' or 'society'.

Thirdly, an anthropological micro-study can, as I have tried to

demonstrate, provide a rather low-level but empirically verifiable model relating diverse aspects of social, economic, and political

1 One of the grievances voiced during the revolt of I967-8 in Dega Damot was that

'voluntary contributions' for road construction had been solicited three times by government officials, but that there was still no sign of a road.

580 ALLAN HOBEN




organisation. My own work, for example, provides a model relating land tenure, status honour, power, and social mobility with exogenous variables, such as demographic pressures, access to national markets, and tax policy.

The problem of representativeness The crucial remaining question - and one asked frequently by those

interested in development planning in Ethiopia - is, Can the results of an intensive study of a few communities be considered representative of the Amhara in general, or even of other Amhara districts in Gojjam? I would prefer to rephrase this and ask, How can a model based on the intensive study of a few communities be used to understand land tenure in other Amhara areas?

In order to answer this question I spent several months travelling in

Gojjam, and in the old Amhara regions of Shoa, Bejemdir, and Wallo. To anticipate, I found that the model of the rist system was valuable, because it provided appropriate questions and concepts for delimiting the geographic extent of the rist system, and for investigating variations in its form. Throughout the long-settled Amhara heartland I found that it had the same basic features as in the communities I had studied in Gojjam. To be sure I also discovered several important regional variations, including the terminology people use to talk about the rist system. For example, in the Menz district of Shoa Province the

apical ancestor of a descent corporation is referred to as a 'first settler'

(Amhr: aquine) instead of as a 'principal father' (wannaw abbat), and a descent corporation official is called a 'chief' (aleka) rather than a

'representative' (fej). More significantly, there are variations in the landholding pattern

which appear to result from differences in the ecological, demographic, and political parameters of the rist system. For example, in regions which, for one reason or another, still have an abundance of uncleared land, a descent corporation's land may not be divided by 'father'; all comers may be given land to clear, whether or not they are descendants of the first settler - though it is noteworthy that, even in these areas, the names and relationships of the offspring of the first settler are remembered against the day when land becomes more scarce. At the other extreme I found communities in which a majority of the population had no rist rights as the result of a collective punishment for

disloyalty to the Government. Finally, there are regional variations in the formal rules of the rist system. Thus, for example, in one part of Menz district the lands of a descent corporation's segments are sup-

39



posed to be rotated by lot every seven years; and the eldest sibling's segment is given a double share of the land.

A substantive discussion of these variations is beyond the scope of this article. The point I want to make is that underneath these regional variations in form, the rist system has the same structural features and

performs the same basic functions. Everywhere it is grounded in a

system of territorially based cognatic descent corporations. Everywhere, children of both sexes can inherit land from both their mother and father.

Everywhere, most men in each generation must actively strive to put together anew a holding of rist land to support household. Everywhere, the rist system functions to create a comparatively fluid relationship between land, status, and power in Amhara social organisation.

There is no reason to consider the communities I happened to study in Dega Demot particularly representative, or that their form of land tenure is typical of the Amhara rist system as a whole. I would argue that the concept of typicality is not a useful one in this context; for Amhara land tenure and Amhara social organisation, of which it is an

important part, exhibit variations, not only from region to region, but even within a region. What are constant are not the exact patterns of

landholding and of social relationships, but the underlying institutional features and the processes out of which these patterns are generated.

In my research I tried to do two things: to describe the land tenure

system I observed at a particular time and place, and also to analyse the institutional elements out of which, and the processes through which, this system is created and sustained. The detailed description of land tenure in the communities I studied intensively may, or may not, fit Amhara communities in other areas - this is an empirical question. I am certain, however, that the model of the rist system which has been constructed as a result ofintensive, albeit localised, research should be very useful to anyone planning development throughout the Amhara region.

Social anthropological research is relevant to understanding social

change, planned or not, because of the way it seeks to account for

enduring patterns of behaviour. It does this, not by invoking custom or the sheer weight of tradition, but by viewing these patterns as the out- come of reasonable men making reasonable decisions. The problem of

understanding and predicting change is the problem of assessing what kinds of decisions will become reasonable under altering ecological, demographic, economic, and political circumstances.

582 ALLAN HOBEN



social anthropology and development planning - a case study in ethiopian land reform policy

Documents