BOOK. THE ORIGINS OF THE BORANA GADA SYSTEM

[48]
THE ORIGINS OF THE BORANA GADA SYSTEM
C. R. HALLPIKE

A Discussion of Gada: Three Approaches to the Study of African Society.
By ASMAROM LEGESSE. New York: Free Press, 1973

GADA, or generation-grading, systems are found only in a few Cushitic-speaking
societies of Ethiopia. Unlike a true age-grading system a gada system is distinguished
by the rule that all sons follow their fathers in the sequence of grades at a single
fixed interval, irrespective of their actual age. In order to maintain the constancy of this
interval, the length of time which a man spends in each grade is fixed at a specific number
of years, as is the number of grades separating him from his father. It is evident,
therefore, that in such systems the various grades will include at the more senior levels
young boys and adolescents together with mature men, a confusion that will not arise
in true age-grading systems.
These systems, moreover, have an innate tendency to progressive distortion by
which either men are born too high up the system, and may be retired before they reach
maturity (under-ageing), or are born too low down in the system, and never reach the
senior grades at all (over-ageing). To prevent over-ageing it is necessary to restrict
legitimate procreation to those men who have reached a sufficient level of seniority in
the system: and to prevent under-ageing it is necessary that men should not go on
begetting children after they have reached a specific grade. It may, however, be possible
to dispense with the second rule, since the higher a man is born into the system, the
sooner he may marry. Thus, other things being equal, under-ageing will be reduced
by'the capacity of men born into the higher grades to marry sooner than men born into
lower grades, and if 'under-age' men take advantage of this, their sons will be born
into the lower grades appropriate to their real age.
Generation-grading systems are therefore cumbersome and difficult to operate,
requiring a sophisticated calendar and a number of irksome restraints of a sexual or
marital nature, and are clearly less effective than true age-grading systems in stratifying
society into such basic categories as boys, warriors, and elders. It is perhaps for
this reason that they exist in very few societies-notably the various branches of the
Galla, the Sidamo, the Darasa, and the Burji, Konso, and Gidole-although the
generational principle is applied to age-sets in a less rigorous form by a few societies
outside Ethiopia e.g. the Kikuyu, Karimojong, and Masai. The Ethiopiangada systems
differ widely from one another both in their internal working, and in their relation to
other social institutions, but in view of the close proximity of these societies to one
another, and the basic homology of the systems, it is obvious that they have a common
origin-just where need not concern us, at present.
The high degree of coordination required by gada systems and the fact that the
Konso, for example, have indisputably modified their gada systems in the past shows
that such systems have not just evolved without some clear purpose; their originators
knew what they were about and in the case of the Konso were able to design systems
that show no tendency either to over- or to under-ageing. Our problem therefore is to

THE ORIGINS OF THE BORANA GADA SYSTEM
discover what advantages such systems could have over the ordinary age-grading
systems which clearly preceded them.
The interest of Legesse's book is that he proposes a solution to this problem for the
gada system of the Borana Galla, and it is upon this aspect of his book that I intend
concentrating here. A brief preliminary description of the Borana system will be
necessary first, however.
The system is a cycle, or rather spiral, of ten grades each of eight years' duration, so
arranged that boys in the first grade are the sons of men in the 6th grade; in the 2nd
grade, of men in the 7th grade, and so on-the generations being divided by a five
grade or forty year interval, all brothers being in the same grade. Not all grades are in
fact of 8 years' duration: the 5th grade lasts 3 years, the 6th 8 years, and the 7th 3
years, while the 8th grade is of normal duration, but the basic symmetry of the system
is preserved nonetheless. There is an eleventh grade, gada mojji, whose members assume
ritual responsibilities of purity and special sanctity, in many ways resembling that of
the youngest boys in the first grade, who are the grandsons of gada mojji. After gada
mojji men become iarsa, retired men.
The system applies to men only, and the basic categories through which they pass
are those of sexless boys attached to their mothers; adolescent boys whose masculinity
is socially acknowledged; young warriors; senior warriors; political and ritual leaders;
semi-retired men with ritual authority; and finally men of special sanctity who live a life
of ritual purity.
All the members of each grade belong to a named gada class, luba, in which they
remain all their lives, and these classes retain their corporate identity even when they
become iarsa.
Men are not allowed to marry before the 3 2nd year of the cycle, though they may have
mistresses as soon as they are capable of attracting them, and may only beget sons in the
4oth.year of the cycle, and daughters in the 48th year, but there is no upper limit on
procreation.
It appears that at present men are being born on average about 5 grades too high in
the cycle, and that the membership of the lowest grades has therefore been partially
shifted to the upper grades. As a result, a series of true age-sets, harjyya,h as been established
to compensate for this. These sets have a. span of 8 years, and the youngest set
includes boys of about 12-19 years of age, but the age sets, like the gada classes, are
only formally incorporated at the transition from the 3rd to the 4th grade. The agesets
are distinct from the gada-classes, since they include men from different gadaclasses;
the age-sets have the function of organising cattle raids and war-parties.
If, in the I6th and i7th centuries, the distortion of the age structure of the gadaclasses
or luba had resembled that of today, too many warriors would have been in
those luba nominally above the warrior grades, and some other institutional means of
recruitment and organisation for warfare would have been necessary, and the agesets
clearly provide this. Legesse argues (p. I 3 8) that the presence of infants, young
children, and adolescents in the warrior luba would have prevented them from carrying
out their tasks, but this is fallacious-they could simply have left the toddlers at home.
The real problem was the untimely promotion of too many warriors into the upper
grades.
From this description of the gada system it will have become clear that each of the
49

50 THE ORIGINS OF THE BORANA GADA SYSTEM
5 gada classes or luba in the first half of the cycle has a fixed relationship with one of the
5 luba in the second half, and this relationship will also extend to the luba of the retired
men. The Borana have a term, gogessa, patri-class, for men linked in this way, and since
there are 5 luba in each half of the cycle, it follows that there must be 5 gogessa.
In one of the most interesting chapters of the book (VII), the author describes
how a cycle of 7 names, the mak.abasac ycle, is superimposed on the gogessac ycle for the
purpose of ordering Borana chronology, whereby the makabasa name returns to the
same gogessa after seven generations, or 36 gada periods, and the Borana believe that
events repeat themselves in an intelligible manner that reflects the gogessa or makabasa
cycles:
Dacci is the mystical influence of history on the present course of events. It may be transmitted
from genealogical ancestors to their descendants, or it may operate in accordance with the
makabasa cycles. (Legesse: 194) . .
The returno f historicali nfluencei n accordancew ith the makabasacy cles is a source of great
concern for the gada leaders. (Legesse: 195).
By the use of the makabasa and gogessa cycles the Borana have an historical map of
their society reaching back to the mid-I 7th century. The gada system is also based on
an accurate lunar calendar which is too complex to discuss here.
The groups of the gada system and the age-sets cut across the descent groups of
moiety and clan, and Legesse has some most illuminating observations on what he
calls the 'functional redundancy' or Borana institutions, especially the gada system in
this respect:
The individual is free to take political action as a member of an age-set or a gada class or a
moiety (clan). It is this phenomenon of cross-cutting categories that introduces considerable
freedom to individual choice into an otherwise rigid pattern of social differentiation.
(Legesse: 224).
It is particularly encouraging that another anthropologist should have proposed,
quite independently, exactly the same explanation as the reviewer's own for the multiplicity
of cross-cutting ties in Konso society:
It is a means of escaping from the tyranny of exclusive allegiance. Aid and assistance can be
obtained from quite different sources, with the result that neither ward nor lineage can
exercise the same control as it would be able to if it were the sole object of allegiance. (Hallpike
1972: 127).
But while one can accept that for a society like the Borana, which places high value
on equality, individual freedom, and also on social order, the functional redundancy
of institutions is a means of obtaining the maximum of choice within a society of
ascribed statuses, it does not of itself explain the existence of the gada system, and we
must now consider Legesse's specific explanations of this institution.
His first approach is an appeal to universal human needs (in some ways very similar
to the theory of age-grading systems proposed by Eisenstadt ( 95 4, 1956), who is not
mentioned, however. These needs are the provision of a conceptual framework within
which the individual can learn new social roles; the support of the community as he

THE ORIGINS OF THE BORANA GADA SYSTEM
enters a new phase of life; and the removal of excessive family influence during
adolescence:
It is . . . in response to such universal human needs that Borana have institutionalized the
generationaolr derg overning their central institution. (Legesse: I o. My emphasis).
But even if these were universal needs, they would be sufficiently satisfied by an
age grading system, and would not require a gada system. He attempts to supplement
these arguments by introducing another 'human universal'-the fear of incest; since
power over the family herds lies with the father or eldest brother, and since cattle for
bridewealth are essential for marriage, the despotic use of these resources creates
friction between fathers and sons, and between brothers:
Just as the father is a potential threat to the maturation of the sons, so too are the sons a threat
to the polygynous dominions of their father. The polygynous family contains wives of such
disparate ages that the sons of the first wife are often the same age or older than their
youngest co-mothers. The Borana male continues to marry younger and younger wives
even after he has ceased to be a viable husband. Under these circumstances, keeping the
generations straight and the incest rules in force becomes more difficult than it is in monogamous
families. The danger of incestuous liaisons between the sons and the wives of the same
man is therefore quite real. The universal human fear of incest becomes especially acute
because it is reinforced by the inequities of Borana family structure. (Legesse: III).
Since a man may not have sexual relations with the women of his father's or his son's
gada class (women take their husband's gada class), Legesse concludes that 'the generational
structure of the gada system is probably a response to these basic fears'.
The reader might suppose that polygynous families with wives of widely differing
ages were a significant proportion of the total, but in fact this is not so. The census in
Appendix I (an invaluable contribution to Ethiopian enthnography) shows that of
131 married women only I6 were co-wives, and i an established mistress, while there
are no instances of a man with more than 2 wives. These figures make it obvious that the
fear of incest is likely to trouble only a small minority of the Borana population, and
to have had little significance in the evolution of the system.
These explanations for the existence of the gada system have little connection with
the demographic arguments, which are as follows. Legesse sugests that for most of its
history the gada system was a simple age-set system, and that in the I6th century, when
the great northward expansion of the Galla was taking place, stimulated by a rapidly
expanding population, the generational rules were introduced to limit the population:
The rules restricting the position of the generations, marriage, and childbearing were
introduced some time during the i6th century to set limits on the rapid expansion of the
population that occurred in that century. Once these rules were introduced, the gada system
became an extremely unstable institution and began to undergo a process of structural
transformation [under-ageing]. (Legesse: 154).
Legesse reaches these conclusions as the result of a computer simulation of the
system, and the simulation further suggests that the system was introduced in I623.
Legesse concedes that this date is too recent, since the chronicler Bahrey, writing in the
latter part of the i6th century describes the gada system as fully operational before then,
5
THE ORIGINS OF THE BORANA GADA SYSTEM
and Legesse suggests that this partial failure of the simulation may be attributable to
the lack of the necessary ecological data incorporated into the model.
Let us first examine the possibility that the gada system was a simple age-grading
system at some time in the i6th century. One of the deficiencies in Legesse's argument
is that he fails to consider the gada systems in societies other than the Borana; in fact we
must assume either that the Borana invented the basic principles of the system, which
were then diffused, or that they borrowed them from some other society. Let us first
assume that the Borana invented them.
I have shown (Hallpike 1972:192n) that the most recent form of one regional gada
system among the Konso was adopted in 604, replacing an earlier defective one; both
the present and earlier systems differ considerably from that of the Borana, however,
and according to Konso tradition their ancestors borrowed the original system from
the Burji, not the Borana. The Konso have three different systems in different regions,
and in these the equivalents of the Kallu (Borana moiety leaders) are responsible for
the gada system and have no special relation with descent groups; there are no gada
councils; the Konso systems have, or had, a principle of sibling-seniority grading
apparently unknown to the Borana; there are no true age-sets; there are no gada
classes; the oldest men remain in the top grade and do not pass beyond the system; the
period of each grade is I8, 9, or 5 years; and the generational interval is 36, 27, or 30
years.
While one readily concedes that many of these differences could have been produced
in the process of adapting the basic gada system to different societies, it still seems very
unlikely that it would have been possible for the Borana to have invented the gada
system even at the beginning of the 16th century, and for it then to have been transferred
to the Burji, then to the Konso, and there operated in two successive forms,
all in a period of less than Ioo years, or less than four generations. Indeed, since it is
improbable that such a system would be borrowed until it had become well-established
in its parent society, this leaves very much less than oo years for the differences already
described to have evolved. But if the Borana invented the system earlier than this, in
the I5th century, or before, the simulation, which gives an adoption date of I623,
ceases to be an even vaguely accurate model of the system's evolution.
But it is possible that the Borana borrowed the system from some other people,
such as the Burji or the Konso; in this case Legesse can indeed argue that this could
have occurred in the 6th century for the purpose of restricting their birth rate, but this
line of argument leaves the origin of the system as obscure as ever.
Whether the Borana borrowed or invented the system, however, it is extremely
unlikely that they adopted it as a means of restricting their population. The increase of
the Borana population in the 6th and I 7th centuries is central to Legesse's explanation
of their expansion and conquests at this time (though for some reason he ignores the
effects of Mohammed Graf's devastation of the Ethiopian state in accounting for the
Galla successes). He very reasonably claims that expansion and more intensive use of
land are responses to population increase:
they adopted techniques of intensive cultivation in most of the newly occupied territories:
ensete( false banana)g ardeningi n the southwest and grain cultivation in the north and east.
(Legesse: I55).

52
THE ORIGINS OF THE BORANA GADA SYSTEM
but suggests that the restrictions on procreation embodied in the gada system were a
third response. Yet if the Borana dealt with over-population by territorial expansion
and more intensive land use, it is difficult to see why they should also have needed to
restrict their birth rate. To Legesse's amazement-and to mine-his computer simulation
showed however that:
the population declines by about 40% during the first eighty years after the rules of the
Gada system are imposed on the normal age-graded population. (Legesse: 155).
This decline continues for a total of 240 years after which the population stabilises. The
simulation was repeated 13 times, with increasing birth rates, but even at the highest
levels ever recorded in any society, the population decline was very substantial. Such a
conclusion makes nonsense, however, of the admitted facts of Galla expansion during
this period, and the simulation must therefore be fallacious. (I believe that a basic error
in the simulation is the assumption that birth rates are independent of the father's age,
but lack of space prevents me developing this argument here. It should be noted however
that one of the disadvantages of computer simulations, as I discovered for myself
in The Konso, is that they themselves become 'black boxes' whose processes are as
obscure as the institutions they model). Not only is the conclusion of the simulation
contrary to the facts, but there is no reason to suppose that, provided the food supply
could have been expanded by migration and more intensive land use, there would have
been any advantage in population control for a militarily aggressive pastoral people
engaged in the largest expansion in their history.
The simulation is based on the assumption that the gada system has an inherent
tendency to under-age, and that it could never have been in equilibrium. But since the
Konso systems are in equilibrium it is conceivable that the Borana system is not
intrinsically unstable either, and that its under-ageing is produced by some other factors
peculiar to Borana society that were not always present. If this is so, there is no reason
why the system might not have been invented by the Borana well before the 16th
century, and only later become subject to distortion.
The most obvious means by which under-ageing can be produced is by women
bearing children to men in the higher grades, and secondly, when men who have been
born into high grades do not take advantage of the lower age of marriage automatically
permitted them by the system. Why should a high proportion of women marry men
in the higher grades? The Borana, like the Konso, think that warriors should not have
children. As Legesse says,
the social system is so organised that the man must first demonstrate his capacity to wage war
before he is allowed to become a father. Furthermore, it is obvious that if the warriors were to
have many children their mobility would have been curtailed. This was probably the function
of infanticide in past centuries when warfare was a dominant aspect of Borana life. (Legesse:
73-4).
While one must distinguish between marrying and begetting sons, it seems likely that
the exigencies of warfare would have made marriage itself more difficult for warriors
in their twenties. This being so, it is likely that in the 6th and I 7th centuries, during
an epoch of large scale warfare, many more women than normal would have been
available as wives for older men, and as mistresses for the warriors, and this would

53
THE ORIGINS OF THE BORANA GADA SYSTEM
have produced a higher than average number of sons to men in the higher grades, and
given thereby a significant distortion to the system.
The other factor possibly responsible for under-ageing, and whose presence would
aggravate the consequences of the first factor, is the propensity of men to marry later
than they are entitled to; this could be induced by the absence of cattle for bridewealth,
since the herds are under the control of fathers and eldest brothers who, we are told
(Legesse: 1 I) frequently deprive their sons or younger brothers of cattle for marriage.
This would retard the age of marriage of younger sons-that class of men which is
already inherently born higher up the system-and thus continue to distort the system
even in the absence of warfare.
Since Legesse has failed in all respects to explain the origin of the Borana gada
system we must ask again why such a difficult alternative to the true age-grading system
should have been adopted by a number of societies, notably the Borana and Konso.
Both societies are organised on the basis of patrilineality, and the inheritance of
property and ritual status in particular is determined by generational status, or 'genealogical
level', as Needham (1966) more precisely expresses it. Both societies are also
distinguished by a fundamental concern with classification and order, and, for these
societies,
There is a strong belief in the association of social order, peace and harmony with that
general well-being of men which derives from the physical world-health, fertility, rain,
and success in war and hunting. (Hallpike 1972: zzo).
Now, age-grading systems by making age rather than genealogical level the basis of
social status, are in conflict with the organisation of the descent groups and the whole
basis of the inheritance of property and ritual status. Men who are on a lower genealogical
level than others can, in such a system, have superior status to those of a higher
genealogical level. Moreover, as the population increases the number of such
'ambiguous' men will also increase, and it is therefore possible that the Borana, Konso,
and others, eventually considered that the only way of removing this anomaly was to
make the age-grading system conform to what they believed was the more important
principle of genealogical level.
Societies with age-grading systems for whom genealogical level is a less important
source of status may not object to men being in a higher age-grade than their
genealogical 'superiors'. Among the Nuer, for example, 'a man may be to another a
classificatory paternal uncle or grandfather genealogically, but if he is of the same
generation [age group] he is his "brother" and would so address him and be addressed
by him.' (Evans-Pritchard 19 51: 17 5).
Generation-grading systems of course produce notable anomalies of their own,
since young boys may be in the same grades as old men, but at least in this case they are
of the same genealogical level. Granted that it is impossible, given the nature of human
demography, fully to reconcile age and genealogical level, it seems that a few societies
prefer to impose a single classificatory principle upon descent groups and upon age
groups. (The explanation of the origin of generation-grading systems which I am
advancing here is, of course, a modification of that originally proposed in The
Konso of Ethiopia, and even now I do not regard it as conclusive; it may no longer be
possible to reconstruct the origin of these systems).

54
THE ORIGINS OF THE BORANA GADA SYSTEM
At the conclusion of what is a review article, some general criticisms of the book
would be appropriate. The methodological contributions are the least valuable, although
this reviewer agrees with many of Legesse's individual contentions. His aim
is to show how the perspectives of Levi-Straussian structuralism, American
'behavioural studies', and British 'case analysis' are all necessary but individually
inadequate for understanding society. These are the 'Three Approaches' of the subtitle,
and they are a distortion of their supposed originals. In particular, Legesse thinks
that his description of the groups, categories, and rules of the gada system is a
'structural' model in the manner of Levi-Strauss, whereas it is simply a conventional
description of social organisation that might be found in the ethnography of any
of the 'Three Approaches'.
It is regrettable that this ostentatious discussion of methodology (62 pages, together
with another 57 pages of the theoretical analysis) has drastically reduced the book's
ethnographic content and value. The multiplicity of Legesse's ambitions has produced
a confusing account of the gada system which is rendered more opaque by the sparse
yet disorderly description of the society, ecology, beliefs and values upon which the
system rests.
But despite my belief that his explanations of the origins of the system are
entirely wrong, he has written a very lively and interesting book, and one can only
hope that in due course he will give us the fuller account of Borana society which they
deserve.

NOTE
I I am grateful to Dr Martin Brett for some valuable criticisms of an earlier draft of this review.
REFERENCES
EISENSTADT, S. N. 1954. 'African Age Groups'. Africa vol. xxiv: 00oo-3.
1956. From Generation to Generation. Age Groups and Social Structure. New York: Free Press.
EVANS-PRITCHARD, E. E. I95 i. Nuer Kinship and Marriage. Oxford: Clarendon Press.
HALLPIKE, C. R. 1972. The Konso of Ethiopia. A Study of the Values of a Cushitic Society. Oxford: Clarendon
Press.
NEEDHAM, R. 1966. Age, Category, and Descent. Bijdragen vol. cxxii: I-33.
Resume
LES ORIGINES DU SYSTEME GADA DES BORANAS
LES Gada, ou systemes de hierarchisation par descendance se distinguent des veritables
systemes de hierarchisation selon l'hge par la regle qui veut que le rang d'un homme soit
determine non pas par son age mais uniquement par le rang de son pere. De tels systemes sont
plus difficiles a mettre en ceuvre que les systemes de hierarchisation selon l'age puisque des
enfants peuvent naitre avec un rang eleve ou inversement etre nes si bas dans le systeme qu'ils
n'atteignent jamais les rangs superieurs. Pour quelles raisons des systemes aussi incommodes
ont-ils ete adoptis? En se basant sur une representation par ordinateur du systeme Borana,
Legesse soutient que celui-ci fut introduit au I6 eme siecle pour permettre de contr6ler la
55
56 THE ORIGINS OF THE BORANA GADA SYSTEM
demographie et que la promotion excessive des jeunes hommes a des rangs superieurs
represente un aspect integral de cette operation: cette theorie presente les defauts suivants:
T. Elle ne tient pas compte de l'evolution des systemes de hierarchisation par descendance
dans d'autres regions de l'Ethiopie, en particulier parmi les Konsos; le systeme Konso etait
fortement developp6e la fin du 16 eme siecle: ainsi, s'il fut emprunte aux Boranas par rapport
au syst6me desquels il presente de grandes differences, les Boranas ont donc diu inventer le leur
bien avant le 6 6me siecle.
2. Mais une date placee au 1 5 eme siecle ou anterieurement s'ecarte a tel point de celle que
suggere la representation par ordinateur que les conclusions qu'on peut en tirer perdent tout
leur sens. Toutefois, si les Boranas n'ont pas invente ce syst6me, les origines de celui-ci
demeurent aussi obscures qu'auparavant.
3. Quelle qu'en soit l'origine, il est clair que les Boranas n'utilisaient pas ce systeme pour
limiter leur demographie parce que a) la population augmenta au cours du 16 eme et du 17 eme
siecle, b) les Gallas acquirent de nouvelles terres pour y installer les habitants representant cette
augmentation de population et c) cette augmentation de la population aurait represente un
avantage militaire.
4. La deformation que l'on trouve dans le systeme Borana et selon laquelle de jeunes
hommes ont ete promus d'une mani6re excessive a des rangs superieurs est sans doute le
resultat de a) la diminution des possibilites de mariage pour les fils cadets depourvus de capital
matrimonial, b) le peu d'occasions qu'avaient les guerriers de se marier a une epoque de guerre
permanente.

http://intersci.ss.uci.edu/wiki/pdf/GadaKinDatasetReview2.pdf