The Civil Administration

By Stuart Munro-Hay

1. The Rulers

The government of Aksum, as far as can be discerned, was administered through a pyramid of authority expanding as it passed from the king to the lower echelons. There is some slight evidence that at times there may have been two kings reigning contemporaneously (Ch. 7: 3), but in such a case one of them would have presumably been recognized as pre-eminent. The structure of power appears to have been that of an absolute monarchy, with a form of kingship implying a semi-divine ruler, and with the king’s immediate family retaining important supportive military and administrative posts. At the next level were provincial governors or chiefs and sub-kings.

Various stratagems were employed to keep power vested in the royal family. This appears to have also included female members of the ruling dynasty since Rufinus (ed. Migne 1849) tells us that the dowager queen acted as regent during Ezana’s minority. If the theory suggested by de Blois (1984) is valid, the royal wives may have been chosen from the chief clans of the country, thus encouraging their support for the regime. Those royal wives who became the mothers of future kings would pass on to their sons the clan-names used by the kings in their `

What information we have concerning the theory of government and the working of the administration and other government functions is episodic, and we may assume that over the centuries the machinery of government did not remain static, but was subject to gradual changes as situations altered. One of the more important changes must have occurred with the acceptance of Christianity. The pre-Christian kings whose inscriptions have come down to us called themselves `son of the invincible god Mahrem’, the royal tutelary deity, and thus asserted their own claim to divine honors; they may also have been high-priests of the state cult. This semi-divine status, though implicitly abandoned by the Christian kings — or at least transformed by the anointing and coronation rituals — seems to have lost them little if any authority, since they remained the de facto heads of their own church; one of the advantages of having the titular head of the church in faraway Alexandria (Ch. 10: 2). Ezana, the first Christian king, in order to keep the customary protocol intact in an inscription (Ch. 11: 5, DAE 11) simply replaced the divine filiation with his own real filiation `son of Ella Amida, never defeated by the enemy’.Bisi’-titles (see below).

Sons of the nagashis, like Baygat and Garmat (Ch. 4: 4), led military campaigns in South Arabia, and Ezana later employed two of his brothers on similar tasks. Sometimes things may not have worked out so well; the story of the rebellion of Abreha, who is said by Tabari (ed. Zotenberg 1919: 184) to have conquered Yemen after a previous general, Aryat, had failed, illustrates this. Tabari thought that Abreha belonged to Kaleb’s family, and was accused of trying to take power in Yemen; the najashi sent Aryat back to Yemen to take control, but Abreha killed him and then later managed to make an accommodation with the najashi.

However, Tabari was writing some centuries after the event, and is not necessarily reliable; his source, Ibn Ishaq, does not mention any relationship between the najashi and Abreha (Guillaume 1955). Procopius, a contemporary of these events, says rather that after Abreha’s coup, another army was sent to punish him, under a relative of Kaleb Ella Atsbeha, and that this army defected to Abreha’s camp, killing their royal general (Procopius: ed. Dewing 1914). Abreha’s assumption of command of the army, or at least of those parts of it which had remained in Arabia, far from the supervision of Aksum, was almost inevitable if the local leader (at that time Sumyafa` Ashwa`, apparently a Himyarite by birth) was incompetent or for some reason disliked by the army, and the same thing often happened among generals in charge of the Roman legions. It is interesting to note, however, that Tabari gives various reasons for Abreha’s eventual submission to his overlord, including the fact that the army would not have supported him against the najashi himself (Zotenberg 1919: 187). Whilst military power in the hands of any subject must have been a possible threat, the reserving of it in the royal family, though impossible to guarantee, must usually have seemed the safest course.

The delegation of authority was an evident necessity as the state expanded, and royal responsibilities for defense and administration became strained. The custom of confirming local rulers was one way of responding to this. Kirwan saw the `archons’ known from a sixth-century account of Aksum as indicating an organized civilian administration on the Byzantine model, in contrast to the regime of petty kings under a supreme monarch (Kirwan 1972: 171). They may, in fact, be those same kings under a new title, and it is, after all, only the Byzantine subject Kosmas who gives them the Byzantine title `archon’. Malalas’ use of the title, as a comparative term, when describing the chariot of King Kaleb, seems to refer to the Byzantine archons, rather than to Ethiopian ones (Kobishchanov 1979: 220; Munro-Hay 1980: 151). It is not known whether Aksumite soldiers formed garrisons in the provincial towns, but in cases where there was some uncertainty, it would seem a likely form of check on the local rulers’ loyalty.

Kosmas Indikopleustes mentions two archons or governors in his comments about Aksum during the reign of Kaleb Ella Atsbeha (Wolska-Conus 1968: 360, 368). These rulers both held extremely important posts in the political structure of the state since they controlled vital links in the country’s trade-system. The archon of Adulis, Asbas, was in charge of the port city, and the archon of the Agaw region controlled the gold trade of Sasu (perhaps Fazugli in modern Sudan) and was responsible for forwarding the caravans. These officials, if not members of his own family, or hereditary local `kings’, were certainly highly-trusted administrators in Kaleb’s government. It may even be that Aksumite officials of this rank were appointed to supervise specifically `Aksumite’ interests in the regions alongside the hereditary local rulers themselves.

In general, then, the Aksumites arranged for the administration of the lands under their hegemony by appointing or confirming local rulers, and exacting tribute as a sign of dependence. Failure to pay this was an act of rebellion and a declaration of war against the Negusa Nagast. One inscription bluntly outlines the Aksumites’ political philosophy on the matter; “those who obeyed, he spared; those who resisted, he killed” (Ch. 11: 5). In the titulature of the Aksumite monarchs, the king is called the `king of kings’ (negusa nagast in Ge`ez, and basileus basileon in Greek). He claims authority over many other regions, some of which were not only far distant but evidently under strong governments of their own controlled by their own kings. These can scarcely have been under the authority of the `king of kings’ to any great degree. These kingdoms, such as Saba, Himyar, the Hadhramawt in South Arabia, and African states such as that of the Noba, may have submitted in theory to Aksum, but very little, if any, real control can have been exerted except during actual campaigns, or where garrisons were left. This actually was the case in some South Arabian districts at times (a certain Sabqalum was possibly the resident of the nagashi in Najran; Jamme: 1962, 79, 319) but certainly not in all the three Arabian kingdoms mentioned.

The lesser chiefdoms or kingdoms nearer to the core of the Aksumite empire were controlled by the king’s peripatetic expeditions. These seem to have been designed as tribute-collecting tours combined with a parade of the king’s military might to overawe anyone inclined to withhold their dues. Oddly enough, `rebellions’ seem to have been quite a frequent feature (see Ch. 11: 5 for the texts of the inscriptions which mention these), but the Aksumite cities and towns show no apparent concern with defense.

Possibly these rebellions were extremely localized, and for a good part of the time were easy to deal with. Aksumite military organization seems to have been mobile and efficient, and very likely the occupants of the Aksumite heartland had little to fear from these rebels against the state. Some, like the Agwezat, appear in the fourth century as dutiful subjects under their king SWSWT, bearing gifts to the Aksumite king Ousanas (? DAE 8), then as rebels against Ezana (? DAE 9) under king Abba `Alkeo, and later in the sixth century needed to be `pacified’ by Kaleb as well in a campaign which he undertook against both them and the Hasat people (Schneider 1974). The inscription of Kaleb, later revered as both a Christian king and saint, proudly details the numbers of men, women, and children captured or killed. Ezana is supposed to have boasted — if the inscription is correctly translated (see Ch. 11: 5, DAE 9) — that he seized the Agwezat king and chained him, naked, with his `throne-bearer’. In many cases, these rebellions are recorded as being led by the local kings, who, of course, failed in their bid for independence in all the instances recorded by the pro-Aksumite writers of the inscriptions. The Tsarane tribe of Afan was among those who were the object of one of Ezana’s campaigns, ostensibly as a punishment for interfering with a trade caravan. Their ruler was captured with his children and his people were severely dealt with.

Punishment for rebellion could be death, captivity (resulting possibly in sacrifice — or presentation to the gods as a gift — or slavery), and sometimes transportation to another area. Ezana transported six Beja kings and their tribes, but the numbers are given (a total of 4400) show that these tribes were relatively small units (DAE 4, 6 & 7 and Geza `Agmai). The many lesser chiefdoms or kingdoms mentioned in the inscriptions were not considered to be of sufficient importance to warrant inclusion in the king of kings’ titulature. The Aksumite system of ruling through existing tribal authorities must, in its way, have simply encouraged the spirit of independence among the subject peoples. Since their identity as separate peoples was not lost, weaknesses in the Aksumite state or difficult moments such as the death of an Aksumite king and a disputed succession could always give rise to attempts to shake off the yoke. The inscription DAE 8, with its preamble referring to the king’s `re-establishment’ of his empire, may have resulted from a new king’s need to demonstrate visibly his assumption of power. Possibly the regency of Ezana’s mother explains why Ezana had to spend some time in re-integrating his kingdom after attaining his majority; regencies for child-rulers were often dangerous periods in the life of a kingdom. Nevertheless, eventually, the smaller population groups lost their former separate identities, became absorbed in the larger polity, and with this assimilation disappear from Ethiopian history.