Amhara Marriage: The Stability of Divorce

Amhara Marriage: The Stability of Divorce*

W . W E I S S L E D E R / University of Toronto

Le mariage, le divorce et le remariage Ctablissent des distinctions significatives de statut entre les laics et les clercs dam la miCt6 rurale Amhara. Les avantages soci&conomiques ne peuvent pas rendre compte des difE6rences dans le t a m des divorces entre les deux group &ant donnC le fort degrC d’homog6nbit6 qui existe parmi ceux-ci. Les satisfactions psychologiques, les contraintes de mEme que les pressions semblent Etre comparables chez les laics et les clercs. Dans cet article, l’incidence du divorce n’est pas assmi6 B la dCt6rioration des liens maritaux individuels, mais 2 la facon dont on extkriorise les justifications sociales sur lesquelles les manages se fondent. Le divorce n’est pas vu ici comme une manifestation d’une guerre des hommes contre les femmes, mais comme un concomitant de l’intentionalitk du mariage en tant qu’institution.

Marriage, divorce, and remarriage delineate significant status distinctions between the clergy and laity of rural Amhara society. Socio-economic advantages d one group over the other cannot be held accountable for differential divorce rates between the two groups in view of the far-reaching homogeneity that encompasses both. Psychological gratifications, cortstraints, and pressures appear to be the same in clergy and in laity. In this paper, the incidence of divorce is related not to deterioration of individual marital bonds, but to the playing-out of the societal rationale upon which marriages are based. Divorce is here treated not as a manifestation of the war of men against women, but as a concomitant of the purposive institution of marriage itself.

I N T R O D U C T I O N

Social anthropologists have tended to deal with marriage in its institutional and societal aspects, while at the same time preferring to explain divorce in psychological terms. The underlying assumption seems to be that society establishes the rationale of why and when individuals are to get married, while individuals determine why and when the societal rationale is to be thwarted through divorce.

Ronald Cohen, for instance, accepts the dichotomy for situations such as the Kanuri where “descent and bridewealth become less important and interpersonal conflict and personal desires increase as explanatory factors” (Cohen, 1971 :7). Here, descent and bridewealth may be assumed to repre-

* The fieldwork of which this paper is a by-product was carried out in 1963 and 1964 under the auspices of the Foreign Area Trainiig Fellowships. I am grateful to Torn McFeat who contributed valuable criticism and to those colleagues in the department who appealed to my unenlightened self-interest in urging me to publish this paper. 67

Rev. canad. Soc. & AnthXanad. Rev. Soc. & Anth. 11 (1) 1974

sent that large array of vested societal practices which promote and maintain the institution of marriage; personal incompatibilities, unrecognized or disregarded at the time of marriage, supply the explanatory psychological motivations of divorce. Duality of rationales is even more pointedly ex- pressed in a summary of the Kanuri case: “The durability of the union is a matter outside the control, more or less, of the alliance that created it” (Cohen 1971:181).

The discussion of Amhara marriage will lay ethnographic groundwork for a unifying hypothesis to permit direct explication of the marriage and divorce dyad from social-anthropological data and theory without a detour through psychology. The data suggest that both the formation of marriages and their eventual dissolution are traceable to identical stimuli, to one corpus of societal determinants. Imperatives that lead to divorce are not to be distinguished from imperatives that underwrite the institution of marriage. Assuming this, the rationale that accounts for marriage can also account for divorce, and divorce may be taken to be as “normal” as marriage. In fact, not divorce, but stability of marriage then constitutes the special case that needs to be explained in each instance.

This hypothesis, supported even by the single instance of the Amhara, raises theoretical implications and considerations which I shall pursue in a subsequent paper.

T H E AMHARA PARADIGM

Critical distinctions between the clergy and the laity of the Ethiopian Ortho- dox church are tied intricately to marriage, divorce, and remarriage. I hope to show that a notable differential between the divorce rates of the clergy and the laity cannot be attributed to substantially better economic, social, or political opportunities available to the clergy alone. I shall then discuss critical factors that differentiate clergy from laity and, lastly, quantify and compare dBerential marriage, divorce, and remarriage rates.

Field data were gathered at two rural communities in Ankober District, Shewa Province, Ethiopia; one predominantly ecclesiastical (Gadam), the other predominantly secular (Laye Madar) .l Details describing these communities will be given later.

THE SHARED BASE

I Students of agrarian monarchical systems, especially of absolute ones, habitually overemphasize stratification. In Ethiopia, European travellers, from Cosmas in AD 525 to the most recent arrivals, have tended to stay close

1 Gadam and Laye Madar are cover names. However Ankober, the name of the district (woredu) to which Gadam and Laye Madar belong, neither can be nor

68 need be disguised.

to the courts and capitals, and have based their impressions and generali- zations upon the illustrious personages with whom they dealt there, The same may be said of Abyssinian chroniclers through the centuries. All reported on courtiers and rural greats who put their best and most impressively shod foot forward in the presence of princes of the realm and church. Scholarly investigators of today show similar preference for dealings with educated Ethiopians who, “even if their origins are rural ... have been oriented all their school days toward modern.education and tend, wittingly or not, to ignore what they can of their past” (Levine 1965: 5 5 ) .

The present data were not collected at so high a social plateau, but in rural communities where differentiation operates to interrelate status-distinct persons more than it keeps them apart. In minimal rural communities, status groups are vertical divisions of society whose specialized integrative activities outweigh the importance of their distinguishing titles. Lords, clergy, and peasantry are accessible to each other. Influence and wealth are not cate gorically reserved to one status group or the other. In the broad rural setting, the economic advantages of the clergy over the laity are minimal, if at all discernible.

Any schema of rigid stratification, even if it were to describe satisfactorily a reality near the complex centre of traditional or modernizing Ethiopia, is largely meaningless at the level of the rural community. Of course, the aristocracy (rnukwunentenat) is normally present in one form or another and so is the church. To specify a peasantry would appear unnecessary in most eyes. Not, however, because we should deduce that everyone who is neither aristocrat nor clergy must be a peasant, but rather because clergymen (and even rnukwonnen) are likely to be more “peasant” than anything else in the traditional rural setting. Minimal rural communities neither invite nor permit absolute occupational specialization. Therefore, the governmental, military, and seigniorial activities of the great, as well as the spiritual and ritual services of the clergy, are adscititious and conjoined to mundane and pedestrian activities of householder and farmer. Men and women of seigniorial and clerical status are deeply involved in their respective daily routines of farming, animal husbandry, domestic management, and housekeeping chores.

The clergy usually has an income in kind or specie. In most rural parishes this does not add up to significant concentrations of wealth. It turns out to be what it was originally intended to be - compensation for time a clergyman must take from his own agricultural pursuits to devote to the church, No “salary” or “gift” is in itself large or regular enough to support a priest or an acolyte permanently as a buyer of essential goods and services. In traditional rural settings, the path to wealth and distinction (in form

of land, money, or position) adheres to the same pattern for all estates.2 2 Rules by which rights to land use are allocated according to an individual’s place

in ambilineal descent apply to clergy and laity alike. (Reported in Hoben, 1963 and Weissleder, 1965). Access to land “owned” by the church is not in any rigid way correlated with membership in the clergy or service to the church (Weissleder: n.d.). 69

Landed wealth is acquired by all either through marriage, descent and in- heritance (rest), “grants” (gulf) based on merit or service, or the judicious use of influence in the right quarters. These processes are essentially status- maintaining and do not produce much social mobility, though mobility exists. In any event, rich peasants, lords in modest circumstances, and dirt- poor priests are by no means exceptions; traders or lawyers may be wealthier (or poorer) than any of them. These points must be stressed if we are to guard against the trap of using class terms as if they denoted rigidly super and subordinated strata in every case. Clergymen are first of all Amhara peasants. In all but specialized ecclesiastical activities, the life of the Amhara clergy resembles that of t hehha ra cultivator. AU are members of the greater Amhara cultural community within which they must prove themselves able in every respect. The clergy’s place in society is not so isolated and specialized that ecclesiastical duties alone would describe and define it. Rarely is an Ethiopian priest a full-time specialist who devotes all of his time and energies to his ecclesiastical office. Most of his days are taken up with pre- cisely the same duties, worries, or satisfactions as those of his parishioners. “He plows his field and harvests his crops. He may do business in the markets, or add to his income by weaving” (Levine 1965: 169). He is almost always self-supporting, a householder who maintains a life style much like that of any other Amhara peasant, a married man with family who eats the same food and drinks the same drinks, keeps the same fasts and does the same work. To the unalerted observer, Amhara clergy and laity could be practically indistinguishable.

As I see him, the Ethiopian priest is not a religious specialist who aZso engages in agrarian activities. He is an ordained farmer.

I1 Marriage is an institution shared by Amhara clergy and laity. Its rationale and regime applies to both estates. In the rationale of Amhara marriage, we are primarily dealing with factors which influence its time boundaries. We are therefore not concerned with structural universals, such as alliance and descent, but with specific expedients and tactics, many overtly taken by families or individuals who are purposefully making decisions according to what they conceive to be their own best interest. It is not so much a question of “how the system really works,” but of “how the system is really manipulated.”

The concept of descent presents few problems to the Amhara. Ambilined reckoning permits every individual to trace kinship through multiple lines on father’s as well as mother’s side. In practice, most of these connecting links are activated only when specific rights to position or land are sought and must be justified.

Alliances between families can constitute reasons for marriage only in the least technical sense. Levine’s statement (1965: 101) that Amhara marriage should be “regarded as a bond between families, not individuals” can 70

refer at best to the way in which a marriage, especially a first marriage, is brought about. The process does indeed involve families rather than the prospective spouses in most cases. But permanent co-operation and coalition between families is hardly an expected result. The high incidence of divorce alone would have demonstrated long ago that permanent bonds between families would be an unrealistic and illusory hope. It is probably more pro- ductive in the long run to view such situations as Cohen ( 197 1 : 18 1 ) regards presumed alliance and foreseeable divorce: “whether or not the union lasts, the noteworthy point for all concerned is the fact that the event has taken place.”

At the heart of an Amhara marriage and wedding is neither spiritual nor social union, but the negotiation and validating d contracts, hallowed by tradition and recently sanctioned by the modern civil code (1960). It has customarily taken written form and the code now demands this to be binding (article 629). An Amhara marriage contract is a very businesslike document. As a time-

honoured practice, recast in the up-to-date legalisms of article 627, it regu- lates “the pecuniary effects of the union and may specify the spouses’ reci- procal rights and obligations in matters concerning their personal relations.” Marriage contracts are little more than inventories drawn up in calculated expectation of marital disaster, so that the separation of possessions can, when necessary, be accomplished with the same facility as the separation of persons. The customary formula at Ankober stipulates that each spouse is to take out of the marriage what he or she brought into it, plus half of any gain due to common effort, according to the equitable principle of kupte ba Raptesh, the gain of one is the gain of the

If alliances or descent entered the thoughts of the Amhara at all, it would be in a very secondary sense at best. Marriage, in the hard-headed and realistic world of the peasant, is thought about in the pragmatic terms of house hold formation and organization of a livelihood. The rationale of marriage is coterminous with the allocation of opportunities, marshalling of resources, providing for self and offspring. In agrarian societies everywhere, the household seems to have proven itself to be the most efficient and convenient mechanism and general model to attain and maintain access to land, non- landed resources, wealth, position and status, the services of household- connected kin and non-kin, security and protection.

The division of labour in Amhara households follows predictable lines, though possibly with somewhat more rigidity in some respects than in other cultures. Women work in the house and kitchen, while men toil in the fields. Among the Amhara, these activities are firmly sex-allocated by rules of conduct that have the compelling force of taboos. It is considered im-

3 Alan Hoben names buluqul (equal partners) as the most common form of marriage at Dega Damot. “It is a civil form of marriage which requires. as the name implies, the c&tribution of equal movable propertiesio the new &upIe-by each set of p&en& (1963: 113) .” 71

possible that (a ) a man cook for himself, or that ( 6 ) a woman work on the land. Hence, neither men nor women can carry on households independently of each other, since the basic agricultural resources either could not be produced or could not be converted into edible form. As a set of imperatives, this stringent enforcement of the division of labour governs much of the Amhara’s householding r6gime. In turn, of course, this protects and safe- guards the central importance of marriage to all aspects of Amhara life. In one way or another, practically all essentials are organized, procured, and guaranteed through the institution of marriage.

It is instructive to compare motivations for divorce with motivations for marriage in this context - not the causes in an abstract sense, nor the grounds in a legal sense, but rather those motivations which are offered as reasons for divorces. The announced reasons are rarely the operant ones. Operant reasons for divorce are bad housekeeping, laziness of the provider, advancing age, and debilitating illness. These are urgent considerations that call for immediate action and redress. When they apply, they are much less likely to be overlooked than such often heard grounds for divorce as adultery, maltreatment, bad temper, or insulting behaviour, which reveal little of the real underlying rationale. The latter are transgressions to which the offended partner might shut his eyes for some time, if they are not too public and flagrant. They are weighed in the balance against the more compelling reasons of householding imperatives. Similar value scales are applied to situations where hitherto unrealized “opportunities” arise during an on- going marriage: another marital union, for instance, which offers more land to be ploughed, or even one which gives a farmer access to land at different locations and altitudes where specific, desirable crops can be grown.

Several categories or levels of motivations for initiating divorce can be identified. I should like to think of them as threshold factors: ( a ) imperative factors, such as chronic illness, debilities of age, insanity, incorrigible laziness, occupational incapability, etc. ; ( b ) oprtunistic factors, such as more valuable or profitable social ties, access to more and better land in different localities, other economic reasons; and ( c ) optional factors, such as adultery, maltreatment, shrewishness, “insulting behaviour,” or “bad character.”

The last category, though the one most often specified as grounds for divorce in casual gossip or in the litigation that usually accompanies the separation of persons and property, rarely emerges as the sole or main agent when the situation is probed. It usually turns out that the first or second category, and often a combination of both, needs to be added for fuller understanding. Incompatibility, lack of respect, or lack of affection are hardly the effective causes - not surprisingly, since compatibility, respect, and af€ection are not the primary agencies which generate marriages in the first place. The resultant congeries of announced causes is naturally almost impossible to dissect and sort out in detail, but my data support the con- tention that imperative and opportunistic reasons are invariably active in the break-up of an Amhara marriage, while optional reasons can be and 72

often are set aside even though evidence is abundant. At times, the often dramatic intra-family or courtroom recitations of grievances seem phrased in terms of the optional category in order to obscure more pragmatic, but emotionally less palatable reasons. Disappointment, character defects, be- trayal, wrongdoing, and even hatred are culturally more acceptable ex- planations for divorce than cynically rational or even rationally rational reasons.

The first and second category of threshold cause5 stand in positive r e lationship to the reasons and purposes upon which marriages are predicated.

Before proceeding to the comparison of marriage and divorce among clergy and laity, it will be necessary to clarify the cultural idiom of Amhara marriage.

111 Hardly a traveller in Ethiopia from the fifteenth century to the present has failed to take note of three basic (statutory) forms of Amhara marriage: sacramental (qerbun), contractual (samanya), and wage (damoz) marriage. The three forms exist side by side throughout the Amhara cultural precinct. Only the first is characterized by a religious ceremony which takes place in the inner ambulatory of the church edifice and includes the cele bration of the Eucharist. It is indissoluble even after the death of one of the spouses. Bu qerbun is the only form of wedding ceremony which is, in the strict sense, recognized and sanctified by the Ethiopian Orthodox church. The other forms are condoned and sanctioned through long permissive practice. Sumanyu, which includes the greatest number of marriages, is essentially secular, and devoid of solemnizing ritual, although the priesthood is usually represented among the invited guests. It is ended as easily as it is begun - or with as much difficulty, when one considers the earnest and often edgy negotiations which precede the drawing up and signing of the marriage contract, paralleled in turn by equally elaborate, if even less amiable, negotiations when the marriage must be di~solved.~ The third form of marriage, damoz, is the one most regularly misunderstood and even disparaged by commentators. Hostile reporters have called it time-marriage and even a sort of concubinage since it involves stipulated payments from husband to wife. The husband obligates himself by contract to provide a specified sum of money, a fixed number of garments, and other designated items during a specified period of time. Levine’s statement that in this form of marriage husband and wife simply separate when the husband has to leave for another part of the country or the couple quarrel (1965:259) is not fully cor- roborated by the Ankober evidence. The marriage may of course be ter- minated by agreement, as all other marriages. The legally binding sense of the marriage is manifested by the fact that unilateral termination on the

4 Divorce negotiations normally involve the same sets of individuals that initiated the marriage in the first place; a minor point, but hinting at a common rationale for marriage and divorce. 73

part of the husband is at times actually more diflicult than in the other types of marriage. A husband does not simply ‘‘walk away” from such a marriage before his contractual obligations are fulfilled. Though defaulting on the dumoz payment would not, of itself, dissolve the marriage, it would provide the wife with prima facie evidence for a legal clairn to be pursued in court. Only if the husband were not to comply with a verdict in her favour would she divorce him. Rather than being the loosest form of marriage, it tends to give maximum protection to the economic interests of the woman; more so than do the other forms. That it should be so is not surprising, for the dumoz union is established with this in mind. It is not a prestigious form of marriage and is usually entered into by poor women who bring no land or wealth to the union. If it is remembered that a divorce customarily involves a division of all gains and profits jointly derived from individually owned property (kupte bu kuptesh), the man’s and the woman’s, obviously no possible profit can arise to her. She would have contributed her work and services for nothing, a state of &airs not readily acceptable to the very reciprocity- minded Amhara peasant.

While the Ethiopian Orthodox church may have dogmatic reservations regarding all but the sacramental type of marriage, custom and the state recognize and sanction all.

By definition, the clergy is to live in ba qerban marriages only. The extent that this is borne out by observable fact, or that deviations from the abso- luteness of the statement occur, offers a test which provides information on the rationale of marriage and divorce itself.

T H E S T R A T E G Y O F CLERICAL M A R R I A G E

I Levine distinguishes between ecclesiastics “who are significant at the na- tional level and those who are primarily local dignitaries” (1965:167). Only local dignitaries of the church live at Gadam and Laye Madar, and only four grades or roles of the clergy need to be considered and discussed in specific detail.6 Three of the four imply a certain state of religiosity and unction: diaqon (acolyte), qes (priest), and munakse (monk). The fourth category consists of the dubtmu (precentors, succentors, scribes, and cantors all in one), who perform necessary and essential adjunct services in the ecclesiastical as well as administrative life of their parish. Their status is completely secular, though quite extraordinary because of their dual involvement in world and church. Marriage and wedding ceremonials are critical to the definition of all four categories.

Often the diuqun begins his career informally, by memorizing ritual and

5 Titles of the regular and secular clergy in the Western church coincide only partially with similarly designated categories in the Ethiopian church. Neither is there complete congruency between the translated terms or activities of acolyte, priest, and precentor, etc., and their Amharic counterparts. 74

text through practice and imitation; some begin as pupils in the traditional church schools which are attached to the major parishes. Boys enter when they are seven or eight years old, about the time that children not destined for this training take up their apprentice duties in field or home. More often than not, the young scholars are held to their household tasks in addition to studies and exercises. After two or three years they are considered suffi- ciently advanced and may take an active part in the liturgical services. Only the ritually pure may enter into the inner precincts of the church, and therefore the very young are chosen for the acolyte’s office, since their innocence may be assumed without burden of proof. In early adolescence and through puberty their active participation usually ceases for the converse reason and many young men abandon the clergy altogether. Normally a diaqon will marry after he has reached a qualifying level in his ecclesiastical training, is of suitable age, and can be absorbed into the priesthood of the community, (that is, means for his support are available). Unless he wishes to forgo marriage forever, it must take place prior to his ordination as a priest. Two conditions must be fulfilled if the diuqon is to become a member of the secular priesthood and is to be eligible for advancement to higher ecclesiastical dignities: the marriage must be of the sacramental form, that is it must entail the “big vow’’ of perpetual indissolubility, even beyond death; and it is further required that the bride be a virgin. Virginity is always prized as an attribute of the young Amhara bride but only in this instance is it substantively important. It gives incontrovertible assurance of the chastity whose inattestibility forces boys into a hiatus of their diuqon services. Marriage of the qerbun form is followed by ordination and consecration by a bishop, thus admitting the diuqun into the full privileges and benefits of the priesthood. A priest (qes) is exempted from the duties of the warrior, though he may follow the army into battle when the occasion arises. He will be in charge of the spiritual welfare of specific parishioners whose confession he will hear and whose “soul father” he will be, giving counsel and receiving gifts. Above all, however, he will be privileged as celebrant of the Mass to enter the mukdm, the inner sanctum of the church, where he may approach and hold the tubot, the very sacred representation of the Ark of the Covenant. It is this distinction and the valum sincerely associated with it which more than anything else sets the qes (priest) apart from others in the community.

Monks (manakse) are received directly into the monastic estate and may not marry at any time. They are therefore dependent upon a religious establishment prepared to minister to the household needs of a number of unmarried males in accordance with cultural and ritual requirements.

Dabfara are indispensable to the ecclesiastical establishment although there seem to be no canonical requirements for the existence of this role, Either ordained clergy or laymen could perform their duties in the liturgy; they are supportive and require no consecration or state of grace. But the dabtmu are a formidable group both in number and in influence. They have 75

commonly passed through the training of the diuqon and usually substantially beyond it. In fact, dubtaru, not priests, are considered the learned men of the church, and are respected and feared as such. Considering their background and preparation, one wonders why such men have not chosen to enter the priesthood. I found a number of reasons. For one, flagrant dis- regard of the minimum standards of the chaste life (or for circumspection) demanded of a diuqun prior to ordination as a priest; then, in some instances, marriage to a girl who, despite all parental assurances to the contrary, turned out not to be a virgin on her wedding night. The most common reason, however, is compelling economic necessity which can force a diuqun to accept an unsuitable marriage simply for the sake of a livelihood. I record two instances at Gadam in which a diuqun entered into marital unions destructive to clerical ambitions - one with a widow, the other with a divorced woman - for the sake of land, Both women possessed land, while the men had not been bequeathed or could not be given any by their respective parents. The men’s families could not hold out for more appropriate matches and the exigencies of the moment forced their hands.

The power of economic necessity, in the domestically compelling sense, does not affect the diuqun alone. In fact, the majority of dubtaru come from the ranks of the ordained priests, that is of men who had been living in sup- posedly indissoluble matrimony, Though there are such potential grounds for official relegation as a fall from Grace because of doctrinal transgressions or generally reprehensible conduct, in verifiable practice divorce and remarriage in contravention of marital vows are the main and almost the only causes of this downward adjustment of status. It is definitely a downgrading. The prestige and privileges of the priesthood are highly prized. They are irretrievably lost by divorce and remarriage and these steps are therefore not taken lightly. If they are taken at all it is because of pressures for which the spiritual values of ecclesiastical unction and the pride which attaches to it cannot compensate.

Significantly, in all casa that I was able to investigate, the pragmatic reasons that led to divorce were in the area which I have called “imperative,” that is, chronic illness, debilities of age, insanity, incorrigible laziness of the female spouse. It was invariably the inability to carry on the domestic rationale of marriage which forced the farmer-priest to vacate his ecclesiastical distinction. Opportunistic motivations are, of course, not easy to separate from the imperatives. Remarriage can readily provide access to more or better land, stronger social ties, and other gains of this nature. After the fact, it is easy to assume that such benefits have been the original motivations. In one case, for instance, a woman was persuaded by her brother to leave and divorce her husband because a better match presented itself. Unfor- tunately, I could never establish conclusively whether this husband was qes or dubturu at the time. In any case, the price is paid in irreplaceable status.

As to the third category of divorce motivations, the “optional,” I was assured by my informants that clerical marriages are by no means more 76

tranquil than those of the lay population, Unfaithfulness of one of the spouses is neither more nor less venial a sin for the clergy than for the laity. Adultery is a frequently cited ground for divorce; it is endemic. But far more husbands and wives are unfaithful far more regularly than get divorced in a given time span. Many men (both of clergy and laity) do not discard wives who have been suspected of or repeatedly caught in acts of adultery. They will scorn them, even mistreat them, but will not divorce them as long as their pride is not too patently damaged. Yet, as we shall see, the remarriage rate of priests is lower than that of the laity, which might indicate that what- ever else the advantages of the priesthood may be, they tend to make the clergy even more tolerant, if not necessarily easier to live with than laymen.

A widowed priest who has neither daughter, sister, nor any other female relative to run his household is practically forced to remarry. Were he to take a female servant into his house, the arrangement would be interpreted to his moral disadvantage or as a dmoz arrangement and hence “remarriage.” An old dubturu told me that many years before, when he was qes, he had to send away his wife who had been in failing health ever since the birth of a child, even though “she had been a good worker and I liked her very much.” He remarried, of course, and became dubturu.

After divorce divests a priest of his consecrated state, he usually re- marries. The status of the divorcC does not exist among this group for any length of time, for freedom from marital bonds confers no advantages. This is hardly surprising; after all, deficiencies and disadvantages which formed the compelling reasons for the dissolution of the marriage can only be com- pensated for by another state of marriedness. The socio-economic efficiency of matrimony has no substitute in celibacy. Thus another union is sought by the divorced priest almost immediately, following either the sumunyu or d m o z wedding pattern. In this, his actions again correspond to those of divorced laymen.

I was never able to confirm Levine’s observation that the status of dubtara is sought by priests because they find the “strict conditions of the priesthood ... too much to live with” (1965: 171). Why would they, in view of the no- toriously lax and dissolute mode of life of which the majority of priests is accused of living, even by Levine in an earlier passage (1965: 169). Quite to the contrary, it is not strictures governing the life of the priesthood but mundane economic or householding problems which beset priest-farmers, no less than lay-farmers, that tend to become “too much.”

I doubt, in fact, that the status of dubtmu is sought with deliberation by anyone. Can we speak of a vocation of dubturu? Admittedly, they hold a unique and fascinating place in Amhara culture. Writers have traditionally created the impression that dubtarenat constitutes a “professional” option in its own right. However, not a single one of about forty cases known to me bears this out. Where there was a direct transition from diuqon to dubturu, it could always be established that something had gone awry. The progres- sion from diaqon to qes to dubtmu is the most common route for the recruit- 77

ment of dubmu. In every instance we can find some element of unrealized hopes or of the “unforseeable.” There is more than a small measure of disappointment, disillusionment, and resignation in the biography of almost every dubtmu. Though their position in church and lay society may be influential, even powerful and feared, none seem to start out with the ambi- tion of becoming dabtaru as long as the priesthood beckons - as we do not set our sights on becoming an assistant cashier or second violinist.

I I The Qerbun initiates a marriage like any other. The married state that follows is not characterized by intrinsically distinct or specifically religious attri- butes. In fact, the particularity of qerbun weddings is not an attribute of the ceremony itself but of the social classification of persons who engage in it. It differs in practical significance and meaning for the clergy and the laity.

Through qerbun marriage af€ects the position and status of the clergy, it is not the clergy’s sole “property.” Not only diuqon on their way to priesthood marry in this manner. Laymen with neither training nor inclination for the priesthood will sometimes choose to be married ba qerban. This is relatively rare, however. The usual explanation, namely that the average lay person does not credit himself with the moral strength or calibre to under- take so serious a commitment, covers the case only partially. One major reason for its rarity among lay persons is its appreciably greater cost; another, that it offers no significant, concrete advantages for the laity. Spiritual gratification, a treasure laid up in heaven, is its main and only reward. (The realistic Amhara considers this a highly speculative investment). After all, divorced persons no less than other sinners remain members of the congre- gation, fast, confess and are forgiven, make their contributions, and are buried near their church when they die, though they are slightly dimini6hed in spiritual stature by their transgression. Generally only wealthier people and those of superior social standing choose to be married bu qerban.

At times, older men and women brave the vow of perpetuity when they feel certain that it will in all likelihood be a last marriage for both. In the case of the laity, a sacramental wedding need neither initiate a first marriage nor be an only one. In such circumstances the church readily overlooks prior mamages which it, officially, never had to recognize. A qerbun wedding should, naturally, lead to a last marriage.

Yet, conditions that work generally against the permanence of marriage frequently force the dissolution even of marriages sincerely intended to be “forever.” The clergy may raise a disapproving eyebrow, but there are no severe negative sanctions. Friends and neighbours are likely to gossip and make snide remarks, but the worst that is likely to be said is that the offenders are not “good Christians.”

I have found no way of reliably establishing how many weddings among the laity are of the sacramental type. Once an informant admits a second,

78

third, or nih marriage, he is usually reluctant to volunteer the information that he broke his word to God by dissolving or concluding any of them. Fortunately, precise knowledge of this is not essential to the present purpose since divorce and remarriage does not involve the same status consequences for the laity that it involves for the clergy.

Clearly, divorce and remarriage should never follow upon a qerbun ceremony. Yet they do, and with some regularity, not only among laymen where it has minimal social consequence but even among the clergy where the consequences are permanently disruptive.

Upon this paradox hinges our argument: some priests remarry, others do not; some laymen remarry, others do not. However, the divorce and remarriage rates for both group d&er sigmficantly. This calls for explication and explanation, since in other respects, in the rural Amhara situation, this truly seems to be an instance where “all other things are equal.”

If all marriages are “alike,” but those of one subsector of the society are less brittle than those of the other, we may assume factors outside the homo- genous ambient of marital existence that restrain normal pressures toward divorce and remarriage. Greater marital stability among the ordained clergy constitutes here the sort of deviation which puts a norm to the test.

THE CLERGY: AN EXCEPTION THAT TESTS THE RULE

The investigated communities, Gadam and Laye Madar, are similar in most respects. Both are totally agricultural; in neither is there a market or even any “business” establishment such as a store; neither exports produce for profit in commercial quantities, and the central government is directly represented in neither. But there are also differences.

Gadam is quite remote and inaccessible. As an ecclesiastical and political entity it is organized about an old and renowned church which is very much alive and active. Its population is divided into 30 households. The largest is that of the mulkunyu who is the political lord, local judge, leader of the militia, and head of the church, a l l in one. Nearly all of the clergy of the parish church - and almost half of the married males in Gadam are members of the clergy - live within Gadam. Every parcel of land is entailed to the church.

In contrast, Laye Madar is accessibly located on a high plateau. It is a rural community of about 60 households whose parish church is about three miles distant. Very little of the land is church-entailed. None of the married males of the active generation belong to the clergy.

Relevant data from both communities are combined in Table I. The information in this table was derived from a complete survey of both Gadam and Laye Madar. Although all households were investigated, results were not so complete on al l scores that all could be included in the tabulation. Some individuals were absent when the data were gathered, while some of

79

TABLE I

NUMBER OF MARRIAGES FOR EACH SPOUSE _.

Locatbn

Gadam Laye Madar

Generation Generation

Ji‘go’s Parental Ego’s Parental - _ _ “Model”

Laity

- 1 1 - - 2

- 1 1 1 - 1

1

- - dabtara ( 1 9 1 ) (193) (1,4) (292) (293) (3,3) (3,4) (4.1) (191) (La)

(2?3 (495) ( 1 9 1 ) 4 5 26 28 (1 92) ( 1 3 (291) (22) ( 2 3

(2,3 (3,U ( 3 2 (3.3) (3,4) (491)

(493) (5.3)

(83)

- -

- - - ?

- - - - - 1

- 1 9 10 - 1 - 1

1 2 -

- - -

- - - - - - -

- - 1

2 1 7 5 - - 1 3 - 1 2 13 1 4 11 12 - 2 2 4 - 2 - 2

1 - 1 6 1 - 3 5 2

- 2 - 3 2

- - 1 - 1 - 2 - - - 2

1 1

-

- - - (2.4)

- - - -

(42)

( 7 , ~ - - -

- - - - - - 1

NOTE: Since a plurality of marriages on either the husband’s or wife’s part affects the status of the clergy and is critical for that of the ordained priesthood, the number of marriages is specified for each partner. A pairing formula, a “model” (x,x), expresses this, the f i s t entry indicating the number of the husband’s marriages, the second the number of marriages of his present (or last) wife.

the men would not permit their wives to answer questions, Data for Gadam provide full information for 25 of 30 households, those for Laye Madar represent 55 of 63 households.

The survey gathered data for four generations: that of the main re- spondent (ego)? that of ego’s children, that of ego’s father and mother, and that of ego’s father’s and mother’s parents. Any information regarding the grandparental generation had to be eliminated from consideration as

6 In instances where two generations were simultaneously active, Ego’s generation 80 refers to that of the descendent independent householder.

too vague and spotty. Much the same applied to ego’s children. The very young are not married, of course, while many of the adolescents - especially the offspring of divorced couples - are permanent or temporary residents of some other region.

The survey could not refer to official records; there are none. But the number of marriages reported by informants could often be verified or corrected with the help of friends and neighbours whose zeal for exactitude transcended all consideration for their fellow’s right to privacy. Where deceased members of the parental generation were concerned, I had to rely largely upon my informants’ memory. Here, too, I was usually able to verify or eliminate discrepancies through internal and external evidence. Where members of both generations were available, I followed the rules of best evidence, evaluating the statements of both sources. Naturally, each generation is reported only once. After making allowance for incomplete- ness and inconsistency, information on 55 households of ego’s generation could be reliably matched with information on 94 households of the parental generation at Laye Madar. At Gadam, 25 couples in ego’s generation could give usable information on 37 households of their parents’ generation.

Separation of data by generation takes the place of a usable age gradient. While the data for ego’s generation represent Amhara society in a state of normal flux, those for the parental generation show the same society closer to the end of its marriage and remarriage cycle, at least to the extent that some of the older generation is past the possibility or necessity of remarriage. The generational columns should not be assumed to make a statement concerning changing conditions. They permit no inference about whether the Amhara divorced and remarried more or less frequently in the past than they do today.

The groupings under diaqon, dabtara, and qes in the table reflect Amhara usage and convention. It includes in the clergy all those who at one time in their lives have opted for an ecclesiastical vocation, whether or not thiigs worked out as they had hoped. By the strict rule, a priest who married more than once or whose wife had been previously married should not be listed as qes. In practice however, divorce and remarriage do not confront the priest with a simple yes or no, either/or, alternative. The range of options at the critical juncture is far more subtle and varied than the mere question of being or not being a priest. Even when divorced, an ordained priest does not lose all standing within the church hierarchy and by no stretch of the imagination does he revert to layman’s status. It is not a matter of courtesy alone that titles and honorific forms of address (qes, mamre, mamher) are retained. The household of any sizable church makes many functions and offices available to a person trained for the priesthood, though excluded from certain of its arcana: marigeta (teacher), gubez (administrator of parish affairs), liqakahenat (judge in matters affecting the church), parish treasurers, storekeepers, etc. A relegated priest may also, of course, serve in one of the many capacities of the dabtara whose status he partially shares. 81

- Location

Gadam Laye Madar

Generation Generation

“Model” Ego’s Parental EgO’S Parental

Clergy (1J) 10 (67*) 10 (53) 0 1 (20) (n,n) 5 (33) 9 (47) 0 4 (80)

5 (28) 26 (47) 28 (32) 4(40) 13 (72) 29 (53) 60 (68)

Laity (1J) (n.4 6 (60)

* Percentages of clergy and laity.

Clearly, even when divorced, a member of the clergy can continue to par- ticipate in the work of parish and church, and may retain some of the distinction and revenue that accompanies such activities, Some of the offices mentioned above can be and often are filled by qualified laymen who never thought of chosing the clerical path. As far as I know, the literature fails to make the point that divorce and

remarriage actually serve to differentiate the clergy internally, more than they set it off from the lay population. Margery Perham for instance (1969: 116) draws a very strict boundary and goes so far as to consider most of the Amhara laity “technically excommunicate” primarily because they live in marriages not initiated by qerbun ceremonies. Miss Perham sees this con- b e d by the undeniable fact that the great majority of Ethiopian Christians who attend the services remain (physically) outside the church. Exclusion from the church building is weak proof in this respect. No one who is not in a state of ritual purity is to enter a church, an injunction which excludes anyone who had sexual intercourse during the preceding 24 (or sometimes 48) hours. This applies to everyone with equal force, even laymen and priests living in proper sacramental unions. When many who would appear entitled to enter, nevertheless remain outside the church at regular liturgical services, they must have chosen to do so for reasons best known to their own consciences.

There obviously is no simple formula to discover the demarcation line separating clergy from laity. Actually, the line runs through the body of the clergy as commonly defined, not around it. Here too, then, the pervasive homogeneity of Amhara culture obliterates the strict and rigid boundaries between status groups so highly prized by social typologists. The symbolic act of the qerban wedding ceremony establishes a distinct social estate which persists until abrogated. Any meaningful specificity of membership in the clergy must clearly distinguish nominal from actual relationships. We shall therefore first compare clergy and laity at Gadam separately, disregarding generational divisions for the moment. Forty-one per cent (20 82

of 34) of clerical marriages turn into the (n,n) type, despite their initial solemnization under the qerban vow, In the case of the laity, 67 per cent (19 of 28) occupy this category; there are no, or only very few qerbm weddings involved here. Comparative data for the minuscule Laye Madar clergy are so sparse and inconclusive that they will be largely excluded from further consideration. For the Laye Madar laity, the (n,n) category amounts to 62 per cent of all the marriages, that is, 89 of 143. The relatively higher attrition rate at Gadam, which may seem surprising, reflects the fact that, notwithstanding their sense of realism, families are betrayed into ill-advised qerbun marriages by the ever-potent lure of clergical dignity. Evidently, time takes its toll. But, are time and its agents impartial to laymen and clergy? What about relative attrition rates?

Since every marital cycle in the clergy begins with a qerbm rite and lay marriage cycles begin with a first marriage which is more likely than not of the ( 1, 1 ) type, we may consider the respective ratio of ( 1, 1 ) marriages becoming (n,n) marriages as a “short-run” rate. It amounts to 33 per cent for the clergy, 60 per cent for the laity. The rate of attrition reckoned between the generations, which we may consider a “long-run” rate, is less drastic, obviously because there it does not convey the dramatic and absolute change from (1 , 1 ) to not (1 , 1) status. For the (1, 1) clergy it changes from 33 per cent to 47 per cent, that is, by 14 percentage points; for the laity it changes from 60 to 72 per cent respectively, a change of 12 percentage

The transgenerational rates for clergy and laity are not far apart. They would seem to come close to indicating what might be considered a normal attrition rate. This is borne out by equivalent data from Laye Madar where there never was an appreciable concentration of clergy. The short-run rate is 53 per cent, the long-run rate 15 per cent, confirmation that pressures and stresses which favour remarriage operate quite uniformly in both communities and for both “status groups.” Once remarried, the clergy obviously tends to conform to the generalized pattern.

The operant distinction between clergy and laity is not determined by training for the church and the occupational involvement in ecclesiastical activities alone. The crucial separation is that which isolates the fully qualified priesthood from all other not-so-entitled members of the community. Of this, the inviolate (1, 1) qerban marriage is the main indication. The ratio

points.

( I , 1) clergy: (n,n) clergy + (1, 1) laity + (n,n) laity

expresses this. At Gadam, 32.2 per cent of alI marriages (disgarding generation) are of this type and designate the fully privileged clergy; 67.8 per cent of the married population therefore make up the nominal clergy and the laity.

Clearly, membership in the clergy correlates positively with relatively higher marital stability. At Gadam, 67 per cent of the clergy in ego’s generation live in ( l, l ) unions, while only 40 per cent of the laity are able to do so. The figures for ego’s generation represent, of course, a still unstabilized 8 3

TABLE m PERCENTAGES OF EACH SUBTYPE OF MARRIAGE IN THE MTAL MARRIED POPULATION OF GADAM AND LAYE MADAR, DISREGARDING GENERATION

“Model” Gadam Laye Madar

Clergy (1J) 32.2 0.6 Laity (191) 14.5 36.5 Total 46.1 37.1

condition. But even in the parental generation, after things have “shaken down” and marriages, divorces, and remarriages have largely become a matter of biography, the clergy can look back upon 53 per cent of maintained (1, 1) marriages, while the figure for the laity has declined to 28 per cent.

It now remains to ascertain whether the qerban unions of the clergy do indeed have a higher incidence than all other ( 1, 1) marriages. The comparison in Table III gives an indication. About 37 per cent of all marriages at Laye Madar are of the (1, 1) type. At Gadam, (1, 1) marriage constitutes 46.7 per cent, roughly 10 per cent more than at Laye Madar. But a disproportionate number of (1, 1) marriages at Gadam is of the qerban rite as attested by the fact that they are marriages of officiating clergymen, and even allowing for the possibility that a very small number of laic qerbun marriages may have found their way unrecorded into Laye Madar data.

If the Laye Madar figure of 37.1 per cent represents a normal rate, the excess of 10 per cent at Gadam appears to be meaningful. Six of the 20 maintained qerban marriages lie outside the predicted range, and must be accounted for.

I have assumed and stipulated that there are no unaccounted-for factors within Amhara marriage in general, within qerbun marriage in particular, or within the socio-economic conditions of the clergy, that would explain this additional measure of stability. It would appear that socio-economic pressures and forces urge all marriages toward and past the divorce point, that is, the threshold where the rationale of marriage terminates. The incidence of divorce and remarriage is therefore high throughout the society.

Yet, a significant number of clergymen manage to avoid this, Since all evidence indicates that conditions within their marriages or socio-economic situation do not favour them markedly, we may assume that for them, too, the divorce threshold is reached, passed, but in some instances disregarded. Factors which promote the clergy’s greater incidence of marital permanence are not internal to marriage and are not part of the integrated system of social tasks which form the rationale of marriage in Amhara culture. They must represent considerations outside the common rationale and must be strong and unique enough to counter and even overcome pressures generally operating to enforce the need for remarriage.

Marital stability, as far as the Ethiopian clergy is concerned, is not an 84

automatic result of favourable conditions but a conscious and by-and-large volitional option taken in the presence of other alternatives. Qerbun itself and its indissolubility constitute an instrumentality, a choice, manipulated to attain and sustain goals which have, of themselves, nothing to do with marriage or its operant rationale. The stability and durability of clerical marriage is not the product of any particular social structure or of any personal gift of compatibility. As an option, continuity of a marriage must be measured against its alternatives. Not qerban, in either its ritual or its spiritual sense, but the permanence of marriage itself is the instrumentality.

Clerical marriages are relatively stable because stability itself is a (self- serving) value. At the cost of some freedom of action, stability acts as a price with which to buy and retain a special place in society.

OBSERVATIONS AND DEDUCTIONS

I ( 1 ) Among the Amhara, fundamental purposes and goals can be achieved practically only through the instrumentality of marriage. (2) The remarriage rate is high throughout Amhara society.

IL (1) The remarriage rate among the clergy is lower than that among the laity.

(2) The greater stability of marriage among the clergy constitutes a “special case” to be explained by considerations which frequently override the normal tendency to divorce and remany.

The limited hypothesis which integrates causes of marriage and divorce, obviating a need to postulate multiple causations, can be carried beyond the Amhara paradigm. I hope to pursue its theoretical implications in a follow-up paper in which I shall seek to generalize the analysis of the Amhara case.

R E F E R E N C E S

Civil Code 1960 Civil Code of the Empire of Ethiopia of 1960. Negarit Gazeta: Extra-

ordinary Issue no. 2 of 1960; Proclamation no. 165 of 1960. Addis Ababa. Cohen, Ronald 1971 Dominance and Defiance: A Study of Marital Instability in an Islamic

African Society. Anthropological Studies, no. 6, ed. Paul Bohannan. Washington, D.C. : American Anthropological Assn.

Hoben, Alan 1963 “The role of ambilineal descent groups in Gojjam Amhara social organiza-

tion.” Unpublished dissertation, University of California at Berkeley. Levine, Donald N. 1965 Wax and Gold: Tradition and Innovation in Ethiopian Culture. Chicago:

University of Chicago Press. Perham, Margery 1969 The Government of Ethiopia. Evanston: Northwestern University Press. Weissleder, W. 1965 “The political ecology of Amhara domination.” Unpublished dissertation,

University of Chicago, n.d. Land and Dominion in Highland Ethiopia. (Work in progress.) 85

Documents

Amhara Marriage: The Stability of Divorce