Society and Death

14-1. Social Classes

By Stuart Munro-Hay

We have remarkably little information about the stratification of Aksumite society, but some suggestions can be made using indications from archaeological and other evidence. Mobility between classes, inheritance, marriage status or other family arrangements is all at present quite outside our knowledge. Polygamy can perhaps be assumed by analogy with later custom, but there is no actual evidence. Later Ethiopian law followed the Fetha Nagast, `The Law of the Kings’ written in Arabic by a Copt in the mid-thirteenth century, and translated into Ge`ez perhaps in the middle of the fifteenth century (Tzadua 1968), but inscriptions like that of Safra show that there were earlier legal codes in use (Drewes 1962).

We do not know if there was any prestige derived from being an `Aksumite’ (as in the case of the extra privileges bestowed on a Roman citizen), rather than a member of one of the other communities which made up the kingdom. A distinction between Ethiopia/Habashat and Aksum itself is implied when the kings are referred to by South Arabian inscriptions as `nagashi of Habashat (Abyssinia) and Aksum’. It has been noted elsewhere that the tribes such as the Agwezat, presumably part of Habashat but not Aksumites, retained their identity for a long while as a distinct people; but after a while, any such Aksum/Habashat dichotomy may have blurred.

Social class may well have been based on the ownership of land, perhaps entraining more or less feudal commitments down the scale, but there is little reliable evidence to affirm this from the Aksumite period. Copies of land-grants to individuals and institutions are preserved (Huntingford 1965), but no originals survive from Aksumite times. Huntingford notes, however, that there is a good possibility that the early charters might be genuine transmissions; they all include Christianised sanction clauses which resemble those on the Aksumite inscriptions (Ch. 11: 5, Geza `Agmai, DAE 10, DAE 11, Kaleb inscr.). If this is true, although all the examples given are grants to Maryam Tseyon cathedral, we can imagine that individuals might have been similarly rewarded by the kings with estates and villages to support their rank and that land-registers of some sort were maintained. The only actual `land-grant’ we know of from Aksumite times is that of King Ezana to the six Beja chieftains (Ch. 11: 5, DAE 4, 6 & 7, Geza `Agmai); and this is exceptional being a forcible removal of a population. However, it does illustrate that the king possessed land to bestow, as we might expect from the Monumentum Adulitanum’s statement about conquered peoples; “I reserved for myself half of their lands and their peoples. . . .”

Slaves, perhaps largely prisoners of war or criminals, are alluded to occasionally. Kosmas seems to imply that the majority of those at home and in the hands of foreign merchants came from Sasu and Barbaria, roughly the western Sudan and south-eastern Ethiopia or Somalia (Wolska-Conus 1968: 378). Such unskilled basic tasks as field work and rough quarry work, hauling, and domestic work could be expected for them. Exceptions would be prisoners of some special quality, like Frumentius and Aedesius, destined for tasks of greater responsibility, who probably were not actually considered as slaves. Procopius speaks of `slaves’ (d?????) in the Aksumite army in Arabia, but these seem to have been allowed to remain in Arabia, and were included among those who later rebelled against Sumyafa` Ashwa` (Esimiphaios), which leaves their actual status unclear (Procopius, ed. Dewing 1914: 189).

Ultimately, life in Aksumite times, as today, was based on the work of the peasant toiling in the fields. Plowing with oxen, sowing, clearing, reaping, and threshing would have occupied his day, and very likely the land he worked as part of another’s estate from which he could take only basic subsistence products for himself. Shepherding the flocks and herds, and tending vegetable and fruit gardens, would have been other countryside occupations. We have no information about land-tenure systems in Aksumite times, through gifts of land by the king to the gods or to the church are mentioned, the former in inscriptions, and the latter in both inscriptions and land-charters. Those of the latter which claim to be of Aksumite times are all in reality much later but may preserve some genuine information (Huntingford 1965). Possibly the prisoners offered to the gods were destined, if not as human sacrifices, to work on such lands?

It is also not known whether the peasants were free, or tied to the land. Probably the houses of such people, as today, were constructed of perishable materials and contained little besides essential tools, skins for clothing and bedding, a few storage vessels, (including wooden or basketwork ones?) and perhaps one or two extras for the richer peasant. Such houses may have been round, like a clay house-model from Hawelti, or perhaps, in more prosperous circumstances, of the type found by de Contenson at Mazaber in the Hawelti-Melazo region (de Contenson 1963ii, pl. XXXVIIb-c; 1961iii: 44). The latter was a stone dry-walled house with the typical Aksumite steps or rebates in the wall, consisting of two rooms only, altogether about 9 m in length by about 4 m wide. Its only remaining contents were sherds from a few pottery vessels and fragments of household objects in bronze (a pin-head and a hook).

The specialist potter, metalworker, leather-worker or other artisan, in the urban setting, may similarly have lived in a relatively humble house and exchanged his work for food or money at one or other of the markets, or he may have traveled, doing work where needed. The only excavated urban areas which could give an idea about the dwellings of such people are at Matara, but they have not yet been fully published. However, a certain idea can be gained from published plans (Anfray 1974: 756 and fig. 7), which show a sharp contrast to the neighboring mansions. The symmetrical arrangement of the former is replaced by an irregular series of square or rectangular rooms, entered by twisting streets and through courts. The impression given is of an organic process, the residence building, rebuilding, adding, or removing rooms and walls as their needs required. Hearths, ovens, and abandoned pottery indicate living floors in these simple two- or three-room dwellings. Complete publication of these quarters of Matara may eventually give us an idea as to the sort of people, and the way of life, to be found in the humbler echelons of an Aksumite urban population. Whether those peasants or artisans who lived and worked in defined areas were obliged to join the armies when required is not known, but seems very likely.

Certain specialists, smiths and so on, must have been necessary to minister to the armies on campaign, and staff such as cooks, porters, and grooms or herdsmen to tend the animals would also have been taken along. The local trader in the market towns was probably not much better off. But the merchant in the larger centers, the larger independent farmer (if such existed), and the various city officials may have constituted something of a middle class, dwelling in rather better houses, perhaps like those illustrated by the clay models found at Aksum. These were apparently equipped with wooden doorways and window surrounds, and layered thatch roofs (de Contenson 1959i: pl. XIX, fig. 8; Chittick 1974: fig. 21a). A greater quantity of tools and fittings, with some occasional luxuries, can be imagined among their possessions. Good quality pottery, some glassware, and decorative metalwork, jewelry, perhaps an Indian or Egyptian cloth robe or cloak, and meat and wine on the table are the sort of extras to be expected. They may have employed artisans and servants or been able to afford a few slaves. Possibly the burial goods found in a tomb in the Gudit Stele Field (see below) belonged to someone from this level of society.

In the central area of the towns, and in country mansions, the landowners and rulers of the dominant class would have led a rather more pleasant way of life, surrounded by households comprising slaves and servants living in the outer wings of their houses where the domestic offices probably were. The great distinction among the élite residences appears to have been one of size, and, as one might expect, the largest was the metropolitan palaces. We can approximately divide the buildings into two groups, the very large `palaces’ and the lesser `villas’ or `mansions’, and these may reflect two echelons of the Aksumite élite; the rulers themselves, and the nobility and great officials.

Those we may term palaces were at Aksum, with the length of the four sides of their central pavilions ranging from 24-35 m; the smallest of these, Ta`akha Maryam, was surrounded by outbuildings measuring 80 × 120. After an intermediate structure, the 21 m sq pavilion at Dungur (Anfray 1972: pl. I), where the outbuildings measured c. 64 m sq, measurements of the pavilions of the next size of building down (`villa’ or `mansion’) varied as follows; 17.50 m, Tertre B at Matara (Anfray and Annequin 1965); the outbuildings measured 59.50 × 49 m; 15.20 m, Tertre C at Matara (Anfray and Annequin 1965); 15 m, `Addi Kilte villa (Puglisi 1941); 12.60 × 11.20 m, Tertre A at Matara (Anfray 1963); the outbuildings, if symmetrically arranged, were about 17 × 15.50 m.

In the central pavilions of these structures, we might expect to find the reception rooms, and, upstairs perhaps (Buxton and Matthews 1974), the main living quarters. The quality of the fittings would have varied with the rank of the owners, from the monarchs to perhaps different grades of noble or official. From tomb finds we can furnish these with gilded and decorated furniture, with vessels and other equipment of gold, silver, bronze, glass, and stoneware. To this, we can probably add certain more costly furs and fabrics, perfumes and incense, carved wood and ivory work, and luxuries of the table both local and imported. Such establishments may have employed a number of specially-skilled retainers, such as musicians and singers, artisans of various sorts, clerks, accountants, bailiffs or stewards. We can imagine a fairly considerable population for the larger dwellings and dependencies; for example, Matara Tertre B had over thirty rooms in its outbuildings, the Dungur mansion, with its several courtyards, contained about fifty rooms, while Ta`akha Maryam had probably around eighty.