Urban precursors in the Horn: early 1st-millennium BC communities in Eritrea. (Special section).

Schmidt, Peter R. ; Curtis, Matthew C.

Eritrea fought a war of liberation for three decades between the early 1960s and 1991. While professional research stagnated because of the war, amateur archaeologists provided the sole source of information for ancient material culture in the country during this era. With the coming of independence in 1993, awareness of the potential value of Eritrea's heritage resources began to grow, leading to an initiative in 1997 to teach archaeology and heritage management at the University of Asmara.

Out of the combined training and research programmes conducted by the University of Asmara have come several major discoveries that change the way that the rise of urbanism is seen in the Horn of Africa. We highlight research showing that between 800 BC and 400 BC the greater Asmara area of Eritrea supported the earliest settled agropastoralist communities known in the highlands of the Horn. These communities pre-date and are contemporaneous with Pre-Aksumite settlements in the highlands of southern Eritrea and northern Ethiopia. (1)

The agropastoralist settlements around Asmara were vital precursors to later 1st-millennium BC and early 1st-millennium AD urban developments in the southern highlands of Eritrea at Keskese, Matara and Qohaito. Matara, 90 km to the south of Asmara, was an urban centre of between 20 and 40 ha, possibly even larger. It was likely an Aksumite administrative centre that also had a significant Pre-Axumite settlement that has been dated to approximately 500 BC by the French archaeologist Francis Anfray (1967; 1974), suggesting that the communities around today's Asmara were the first in the region to show an organic growth toward demographic complexity. Another urban center, Qohaito, located approximately 70 km south of Asmara, was an ancient garden city (Schmidt & Wright 1995) surrounded by hundreds of satellite towns, villages and homesteads located on the 13x3 km Qohaito plateau (Wenig 1997) and connected to a larger urban hinterland (Curtis & Libsekal 1999). Qohaito remains unexcavated, but survey evidence indicates that its urban character derives from a tradition that goes back to Matara and the communities of the Greater Asmara area.

We also discuss evidence that suggests the possible presence of humped cattle (Bos indicus) in the greater Asmara area about 500 BC, revising previous ideas about the arrival of this species in the Horn and assessing what importance it has for the development of a settled agropastoral way of life.

Setting and background

Asmara, the capital of Eritrea, is located at 2350 m a.s.l on a portion of the Eritrean highlands called the Asmara Plateau, a peneplain that ranges from approximately 2200 to 2500 m (Abul-Haggag 1961) (FIGURE 1). In comparison to other parts of the highlands, the greater Asmara area is blessed with relatively fertile soils, a more moderate climate, a relatively flat plain for agriculture and a reliable supply of water.

[FIGURE 1 OMITTED]

Only Italian amateur archaeologists, V. Franchini and G. Tringali, focused on this region; their most significant identifications centred on what are called `Ona' sites. (2) The basic characteristics of the `Ona' sites and their material culture, particularly ceramics and ground and chipped stone figurines called `Bulls' heads', were described by Tringali in a number of Italian-language publications (Tringali 1965; 1967; 1969; 1973-77; 1980-81; 1987). These sites were mentioned and identified as Pre-Aksumite by Anfray (1970). The potential importance of the Ona finds went mostly unnoticed in the archaeological world until Rodolfo Fattovich drew attention to their significance for understanding early complex societies in the Horn. Calling these sites both the `Ona Culture' and `Ona Group-A', he argues that the Ona ceramics bear affinities to the black-topped ware of the Sudanese Nile Valley dating to approximately 1500 BC (Fattovich 1978; 1980; 1988; 1990). (3)

Fattovich also subscribes to the presence of south Arabian cultural influence among Ona peoples. He links undated petroglyphs at sites around Asmara to figures in Arabian rock art that date between the 3rd and 1st millennia BC (Fattovich 1983; 1988). (4) He also argues that the Tihama coastal culture along the south Arabian littoral of the mid 2nd millennium BC has ceramic affinities to Kerma and the C-Group in Nubia, stating that `A possibly Arabian influence is noticeable in the Ona Group-A culture', where the `ceramics reflect a local tradition, partly comparable to the Tihama' (Fattovich 1997a: 481). Fattovich's positions have a larger design: that there was an interregional interaction zone that linked the Nile Valley and the lowlands of eastern Sudan and western Eritrea with the highlands of Eritrea and northern Ethiopia as well as with the Tihama culture area. (5) The Ona culture is central to his diffusionist construct because of its geographical location at the confluence of these cultural areas. As our archaeological investigations were launched in Eritrea, these were the interpretations that dominated thinking about the ancient history of the Eritrean highlands.

Within two years of independence, large housing projects and light industrial complexes began to spring up around Asmara. This spurt of development intersected with the initiation of an archaeological teaching programme during which we conducted several survey courses outside of Asmara in areas in which `Ona' sites had been generally but not precisely documented. We also embarked upon a programme of regional documentation and analysis of settlement in concert with strategic test excavations to establish baseline culture histories. We now turn to the results of these inquiries over the last three years.

Initial archaeological tests

Sembel site

The Sembel site in 1998 was under threat from rock quarrying, ploughing, rock removal for grave markers, terracing for erosion control, and new road development (See FIGURE 2 for location). On this 12-13 ha site the area strewn with dense amounts of building stone and wall features associated with mounds is limited to approximately 4 ha. Recent terracing for tree planting on the northwest side of Sembel had exposed a large number of small bowls and cups (5-7 cm. diameter at mouth) (Schmidt 1999). (6)

[FIGURE 2 OMITTED]

The presence of a cup and small bowl in situ on the surface of one of the central mounds provided an opportunity to establish a stratigraphic context for this aspect of `Ona' material culture. A 2x2-m test was placed where the vessels were located, on a mound with stone walls exposed at its base. This test upon the mound, among others on the site, provided the most useful information for building a more refined idea of `Ona' culture. From the surface to the deepest deposits at -220 cm, this test unit was characterized by a mass of stone rubble from collapsed walls, filling spaces between ancient stone walls.

The stone walls exposed in the initial test were made of closely fitted fieldstones held together with a mud mortar, It appears that the test was placed in such a manner to intersect a passageway or niche between two houses that were joined by a common wall. Ash features within this niche cannot be definitively tied to cooking. At approximately 60 cm depth, an extensive ash layer was first observed and eventually documented to as deep as 140 cm, having been piled against the northern stone wall exposure.

`Bulls' heads' figurines were found throughout these deposits, as were small red ware cups, burnished and slipped red ware, and coarse brown ware vessels. Indirect evidence for grain agriculture was found throughout the deposits in the form of many upper and lower grindstones. Faunal materials were also plentiful and were found in discrete clusters often associated with ash, suggesting a discard context (Schmidt 1999).

In 1999 an advanced archaeological field school was conducted at Sembel with the express goal of expanding knowledge about the architectural characteristics. These additional excavations yielded significant information on agricultural food processing, with the recovery of many upper and lower grindstones. As well, the exposure and definition of two in situ hearths provided excellent evidence for additional paraphernalia used in food preparation.

Sembel Kushet

Approximately 1 km to the northwest of Sembel is another `Ona' culture site of equally large size, about 12 ha. A central portion of the site has mounds and a very dense concentration of architectural stone. One of the distinctive characteristics of this site is a large zone of ashy soil, approximately 60x80 m, that has been partially disturbed by ploughing. One test in this zone showed plentiful remains of burnished and slipped red ware. Another test located in an area of dense ceramic scatters to the south of the mounds and ashy zone showed that many of the ceramics may be redeposited from the higher mounds to the north.

Sembel II

On the western margins of the Sembel site we noted another discrete `Ona' period site that was adjacent to another component marked by black ware ceramics. Regional survey shows that black ware sites are relatively few in number when compared to Ona sites. A large portion of the Sembel II site had been destroyed by rock quarrying, exposing human burials, hearths and charcoal concentrations. Also present was a hole drilled and ground into the surrounding bedrock, descending into what appears to be an underground chamber. Two exposed hearths and an exposed wall were tested. The presence of human burials, including a fetus burial in one test, suggest that a cluster of hearths may be associated with the underground chamber. Black-ware ceramics were recognized by Tringali in other parts of the Asmara plateau, mostly on the far eastern side of Asmara (Tringali 1965; Munro-Hay & Tringali 1993). One radiocarbon date from a hearth in association with black ware indicates that this component dates to approximately the 11th century AD, or the post Aksumite era,

Regional survey

Results from the Greater Asmara Regional Archaeological Survey Project (carried out in 1999-2000) amplify insights concerning the development of ancient complex society around Asmara. The survey universe was designed to capture a variety of physiographic and environmental zones and to build upon the survey projects carried out during earlier University of Asmara field schools. An area of mostly nonurban landscape of 145 sq. km surrounding Asmara was demarcated, with a simple random sample of 15 1-sq. km survey units resulting in a 10.3 % sample of the survey universe (FIGURE 2).

Field survey incorporated an intensive surface survey of these sample units. Walking linear transects, surveyors recorded detailed archaeological, physiographic and environmental observations. A total of 80 archaeological sites were documented in the 15 sq. km sampled (FIGURE 2). Average site size of the 80 documented sites is 1.64 ha. Of the 80 sites documented, 47 sites are less than I ha, 22 sites are 1-3 ha, 6 sites range between 3 and 6 ha, 3 sites are between 6 and 12 ha and 2 sites are larger than 12 ha. Most Ona sites are located in upland areas above or adjacent to ploughed and/or fallow agricultural fields and situated within a few hundred metres of seasonal and/ or perennial streams.

Of the 80 Ona sites documented, half have substantial intact mound features composed of architectural rubble and/or architectural features within anthropogenic soil deposits, often exhibiting exposed stonewalls, platforms, terraces, and sometimes cisterns. Mounds range in height from fewer than 30 cm to over 4 m in elevation. A diverse range of local igneous and metamorphic rocks were used for building.

Mai Hu tsa

A site of more than 12 ha located 4 km north of Asmara, Mai Hutsa stretches from the edge of Mai Hutsa stream to the top of a 30-m high ridge approximately 400 m east of the Asmara-Keren highway. This site, at the time of excavation, was threatened by construction of irrigation channels and other major agricultural earthmoving. A mound in the site's centre once rose to 4 m in height. Stone walls were visible on the surface, with 14 linear patterns clearly visible. Visible architectural features included a small cistern, terraces, and platforms. A test was placed at the top of the main mound over an exposed wall feature, an apparent room with adjoining walls that were 1.7 m high (FIGURE 3). Foundations made of andesitic rock underlay regularly shaped fieldstones mortared with mud. Of particular interest was a thick deposit of faunal material, including ovicaprines and cattle, some of which was burned and found in association with lithic tools. Also of note were the presence of three whole and four nearly complete vessels of both Ona red ware and coarse brown ware ceramic traditions, apparently propped against one wall; these were found in mid-way in a 230-cm deposit.

[FIGURE 3 OMITTED]

Ona Gudo

This site was mentioned by Tringali (1965). Oral traditions link it closely with the adjacent site of Ona Hachel. It lies on a ridge above the Mai Bela stream and next to Ona Gudo village. The site is marked by a series of 5 terraces/platforms on which are found much eroding ash and exposed ancient walls. There are large numbers of Bulls' heads and abundant quartz and obsidian lithics on the site surface. One test on the topmost platform showed a series of ashy deposits filled with artefacts such as slag, textile fragments, stone beads, ceramics and Bulls' heads. Two hearths were documented during excavation, one with similarities to a hearth excavated at the Sembel site.

Radiocarbon dates

An initial series of five radiocarbon dates were processed in 1999 from Sembel. These dates range from the 9th to the 4th centuries BC, indicating that Sembel dates to a period that may overlap with the so-called Ethio-Sabean culture in the highlands of northern Ethiopia and southern Eritrea. One Sembel date is an AMS determination with a low standard deviation that fixes the deepest deposit at the juncture of the 9th and 8th centuries BC. The results from Sembel are amplified and affirmed by three radiocarbon dates from Mai Hutsa and four dates from Ona Gudo (TABLE 1). At both of these sites the basal deposits, or deepest 10 cm directly above bedrock, date to as early as the late 9th century BC, dates that are congruent with the Sembel dates. Dates from upper deposits, while showing one reversal at Ona Gudo, fit well with the idea that these Ona sites represent contemporaneous settlements.

Faunal assemblage

The most notable characteristic of the fauna excavated in Ona sites is the dominant presence of domestic animals, mostly cattle and ovidcaprines. There is only incidental evidence for wild animals -- an unidentified shrew species and a bird, suggesting that the Ona people had made the transition from an economy in which hunting played a significant role (Shoshani & Boza 1999; Shoshani & Beyin 2000). The assemblage from Sembel, Mai Hutsa and Ona Gudo shows a very low incidence of butchery cut-marks. Most long bones are broken, sometimes in multiple places -- suggesting extraction of marrow and perhaps meat division. There are no pigs, horses or donkeys -- results that mirror excavations elsewhere in the Horn where Semitic speaking peoples may have resided.

There is a decided preference for cattle over sheep and goats, with a ratio of nearly 2:1, favouring cattle (TABLE 2). Although the sample is too small to comment definitively on herd management, at Sembel it tentatively appears that sub-adult cattle were preferred for meat consumption. The emerging picture of the pastoral part of the subsistence economy is one in which domesticated cattle played a dominant role.

Bulls' heads

Bulls' heads figurines were excavated from all tests at Sembel, Sembel Kushet, Mai Hutsa and Ona Gudo. The figurines from excavated contexts tend to be predominantly ground stone artefacts, made of metavolcanic rocks, quartz and various igneous andesitic rocks, chloritic schist and, occasionally, ground ceramic sherds. Found throughout the excavations, the Bulls' heads seem to be manufactured from locally available material, with laterized metamorphic rock sources the easiest to work and the most readily available in the Sembel and Sembel Kushet areas. As the name suggests, these objects are likely symbolic representations of cattle. Several radiocarbon dates are associated with Bulls' heads, which continue deep into the deposits to at least 700 BC at Sembel and Mai Hutsa, and in the deepest level at Ona Gudo.

One of our more significant finds at Sembel is a stone Bull's head that shows a hump between the horns -- dated to approximately 500 BC (see FIGURE 4). This appears to depict a humped cattle species, perhaps Bos indicus, which has been indirectly documented in Eritrea from a bronze figurine from the Aksumite site of Zeban Kutur in the Akele Guzai region, dated to the 2nd century AD (Ricci 1955-58; Drewes 1962; Clark 1976; Drewes & Schneider 1976; Fattovich 1977) and by clay figurines of humped cattle found in excavations at Matara and ascribed to the Aksumite period (Anfray 1967). The presence of Bos indicus, with its ability to tolerate drier climates, to resist disease and parasites (Du Toit 1936; Epstein 1971:198), and its capacity for high milk productivity in arid conditions may be linked to the development of this early agropastoralist culture in the Horn. The genetic material of Bos indicus may have tipped the scales towards permanent settlement in these early Eritrean communities, making it possible to tolerate a mid-Holocene period of aridity that seems to have affected the Horn about 3000 years ago. Bos indicus would have provided a more stable subsistence base, and in combination with grain agriculture may have been a primary catalyst for the rise of large permanent villages and small towns in the greater Asmara area in the early 1st millennium BC. The timing of these highland developments corresponds to what some scholars believe to be rapid changes in the size and numbers of pastoralist communities in eastern Africa about 3000 BP (Bower 1991; Marshall 1994). We must also keep in mind known corridors of interaction between the highlands and the lowland zones of the northern Horn. The coming of new breeds of cattle and the transhumant interaction of cattle keepers and agriculturalists of the highlands are ideas that have been discussed for some time in the archaeology of the Horn (Clark 1967; 1976; 1980; Brandt 1984; and Brandt & Carder 1987). The greater Asmara area may provide the substantive archaeological evidence to test the appropriateness of these models of interaction.

[FIGURE 4 OMITTED]

Ceramics

One surprise of these excavations was the relatively high proportion of a coarse brown ware, largely unrecorded by earlier investigators. These thick-walled, heavy-duty vessels range in height from 40 to 70 cm, with a mouth diameter of 19-30 cm (see FIGURE 5). A notable vessel attribute is an extruded coarse temper, often exaggerated on the interior by significant exfoliation of the clay. Decoration is limited to several forms of occasional vertical incising located usually above small horizontal lugs. Burning on the bottom and lower sides of vessels and the extreme exfoliation of clay from the interior of vessel walls both fit with recent ethnoarchaeological evidence on vessels used in beer production, suggesting their use for beer brewing (Arthur 2000; 2001).

[FIGURE 5 OMITTED]

Red-ware ceramics have commonly characterized the Ona culture. They predominate in number of vessels represented as well as numbers of sherds. Red wares are dominated by red slipping with a burnish applied. Sometimes the top of the pot is black and burnished. These vessels are adorned with finely executed hatched incising (often in triangular fields), punctates and wavy-line comb incising, the latter applied to the excavated pot from Sembel (see FIGURE 5). The geometric motifs are seen by some to resemble redware decoration common to the mid 2nd-millennium BC Nile Valley, Kassala in the eastern Sudan, and Pre-Aksumite redwares of southern Eritrea and northern Ethiopia (Fattovich 1988; 1997a; 1997b). Our radiometric dating of Ona red ware to the early 1st millennium BC now throws such attributions into question. These vessels have a very fine paste, are finished with a high burnish and sometimes show signs of cooking. There are a few affinities between the red wares of Yeha and other Pre-Aksumite sites and the Ona ceramics. Each tradition includes vessels that have very fine pastes, similar colours including black tops, and slipped and burnished exteriors. Yet the shapes, sizes and decorative applications on more southerly `Pre-Aksumite' vessels are distinct in important ways from the Ona wares. (7) This suggests that there are regional commonalities in finish and paste that cross cut time and space in northeastern Africa, but these commonalities are so broadly distributed that they do not constitute sufficient criteria on which to base ideas about culture contact or interaction.

Lithics

Flaked stone lithics are characterized by quartz material, which comprises more than 75% of the surface and excavated assemblages. Lesser amounts of obsidian, chert and various other igneous and metamorphic rocks were also utilized. Scrapers, including end-, side-, convergent, circular and thumbnail forms make up the most frequent artefact class. Perforators, burins and unifacial points also are abundant in assemblages. In addition, a diverse range of quartz and obsidian microliths, including crescents, triangles, outil ecaille and backed microblades are important components of the tradition. (8)

Conclusions

Research in the greater Asmara area shows that we cannot substantiate earlier ideas that the highlands of Eritrea owed their cultural genesis and their urban development to interactions with the South Arabian Peninsula. Comparison of the latter ceramic traditions with the ceramics of the Ona culture suggests that the communities around Asmara were endogenous. There is currently no evidence that the Ona communities were influenced by Sabean incursion(s) to the south or that they were an integral part of the Ethio-Sabean cultural complex.

In the area around Aksum in northern Ethiopia it appears that the earliest Pre-Aksumite settlements and ceremonial sites date from the mid 1st millennium BC (Fattovich 1988; 1990; Michels 1994; Bard et al. 2000). The ancient Ona communities of Greater Asmara show signs of growth toward urbanism in the very early 1st millennium BC. Tentatively, we see a corridor of intensifying urbanism beginning around Asmara in the early 1st millennium BC, extending to southern Eritrea at Matara and Keskese possibly in the mid 1st millennium BC.

The One settlements are sedentary communities practising a mixed economy of grain agriculture and pastoralism, a significant development that marked a major transition from pastoral economies that seem to have prevailed in the highlands of the Horn up to Ona times. These prosperous people, living in villages and small towns made of solid stone walls, also made images of cattle that included depictions of Bus indicus. These representations provide us with the first evidence that this hardy species was in the Horn by the mid 1st millennium BC, some 700 years before previously thought. The presence of Bus indicus, with its higher milk production under arid conditions and its natural resistance to disease, may well have been a key element of predictability and stability that allowed settled life to take root in the Asmara area and that possibly opened the way to urbanism in the Horn. Asmara's ideal location in the highlands with accessibility to good soils and water as well as its location at a crossroads in regional trade were also critical factors that favoured the growth of a complex culture here in the early 1st millennium BC.

TABLE 1. Calibrated radiocarbon dates for Ona sites.

site laboratory excavation conventional
test number level and radiocarbon
excavation depth age

Sembel A Beta 130119 Level 4D, 2440 [+ or -] 70 BP
 -77 to -86 cm

Sembel A Beta 130120 Level 5, Lens A, 2460 [+ or -] 60 BP
 -102 to -107 cm

Sembel A Beta 130121 Level 7, 2370 [+ or -] 70 BP
 -124 cm

Sembel A Beta 130122 Level 8, 2550 [+ or -] 60 BP
 -135 to -142 cm

Sembel A Beta 130123 Level 11, 2600 [+ or -] 40 BP
 AMS -167 cm

Mai Hutsa Beta 152960 Level 5, 2390 [+ or -] 70 BP
MH07-A Feature 2,
 -40 to -50 cm

Mai Hutsa Beta 152961 Level 10, 2480 [+ or -] 80 BP
MH07-A Feature 4,
 -90 to -100 cm

Mai Hutsa Beta 152963 Level 23, 2560 [+ or -] 70 BP
MH07-A Feature 7,
 -220 to -225 cm

Ona Gudo Beta 152964 Levels 6-8, 2400 [+ or -] 80 BP
OG01-A -50 cm to -80 cm

Ona Gudo Beta 152965 Level 13-15, 2200 [+ or -] 80 BP
OG01-A -129 to -140 cm

Ona Gudo Beta 152966 Level 17, 2360 [+ or -] 80 BP
OGO1-A Feature 3,
 -161 to -170 cm

Ona Gudo Beta 152967 Level 22, 2480 [+ or -] 60 BP
OG01-A Feature 4
 -211 to -220 cm

Sembel II Beta 130124 hearth: associated 1030 [+ or -] 50 BP
 with Black Ware

site laboratory calibrated calibrated
test number 1[sigma] 2[sigma]
excavation (68% prob.) (95% prob.)

Sembel A Beta 130119 2720-2355 BP 2740-2335 BP
 770-405 BC 790-385 BC

Sembel A Beta 130120 2720-2360 BP 2740-2345 BP
 770-410 BC 790-395 BC

Sembel A Beta 130121 2465 to 2335 BP 2720-2310 BP
 515-385 BC 770-360 BC

Sembel A Beta 130122 2750-2710 BP 2770-2450 BP
 800-760 BC 820-500 BC

Sembel A Beta 130123 2760-2735 BP 2770-2720 BP
 AMS 810-785 BC 820-770 BC

Mai Hutsa Beta 152960 2480-2340 BP 2730-2320 BP
MH07-A 530-390 BC 780-370 BC

Mai Hutsa Beta 152961 2740-2360 BP 2760-2340 BP
MH07-A 790-410 BC 810-390 BC

Mai Hutsa Beta 152963 2760-2710 BP 2780-2370 BP
MH07-A 800-760 BC 830-420 BC

Ona Gudo Beta 152964 2500-2340 BP 2740-2320 BP
OG01-A 550-390 BC 790-360 BC

Ona Gudo Beta 152965 2330-2120 BP 2350-1990 BP
OG01-A 380-160 SC 400-40 BC

Ona Gudo Beta 152966 2460-2330 BP 2720-2300 BP
OGO1-A 520-380 BC 770-350 BC

Ona Gudo Beta 152967 2730-2370 BP 2750-2350 BP
OG01-A 780-420 BC 800-400 BC

Sembel II Beta 130124 970-925 BP 855-810 BP
 AD 980-1025 AD 1095-1140
TABLE 2. Minimum Number of Individuals in
faunal assemblages.

sites Bos, Bos, ovicaprine, ovicaprine,
 adult sub-adult adult sub-adult
 to juvenile to juvenile

Sembel 4 9 4 3
Sembel 2 1 1 1
 Kushet
Sembel II 2 nil nil nil
Mai Hutsa 2 1 2 nil
Ona Gudo 1 nil 1 nil

Acknowledgements. We want to thank the University of Asmara for the opportunity to conduct archaeological research in Eritrea. Special thanks go to President Wolde-Ab Yisak, Dean Asmerom Kidane and Dr Yosief Libseqal for their steadfast support. Our appreciation, too, goes to Jeheskel `Hezy' Shoshani and Ana Boza for their help with the faunal analyses, to Lalemba Tsehaie for artefact illustrations and to Alvaro Higueras and Henry Wright for their insights and assistance. We also thank Amanuel Beyin, Asmeret Ghebrezgabiher, Daniel Habtemichael, Dawit Okubatsion, Michael Halle, Werede Okubay, Yoseph Mobae, Zelalem Teka and all of the other students who assisted us. In addition, thanks go to the National Geographic Society for critical support for the Sembel investigations and the US Fulbright Program for underwriting our stays in Eritrea in 1998, 1999 and 2000.

(1) In recent publications, Rodolfo Fattovich contends that the `Pre-Aksumite Culture' period dates from approximately 800/700 BC to 400/300 BC (e.g. Fattovich 1997b; 2000).

(2) `Ona' is a term in the Tigrinya language that describes ruins or old homesteads where either ancient populations once resided or where now departed families once resided within a village. We use this term to refer to ancient sites only.

(3) Fattovich (1980: 62; 1990: 10) assigns one Ona site from the greater Asmara area, Ona Hachel, to the Pre-Aksumite culture period based on his study of affinities with Yeha and Matara ceramics of that era. However, in his 1990 article and in other publications (e.g. Fattovich 1997a and 1997b), Fattovich argues that the Ona sites date to the 2nd millennium BC.

(4) While the Arabian rock art dates between 3000 BC to 500 BC, Fattovich settles on dates of 2nd and 1st millennium BC for Eritrean examples (1997a).

(5) Fattovich (1997a) calls this the Afro-Arabian cultural complex (Tihama cultural complex).

(6) These cups are smaller than most Pre-Aksumite and Proto Aksumite cups known from southern Eritrea and northern Ethiopia, which range from 7 to 15 cm in diameter, have slightly curving rims, sometimes have vertical handles, and are mostly undecorated: Bard et al. 1997: 395, 397; Fattovich & Bard 1998: 25. A few `miniature' cups at Yeha are within the size range of these finds: Fattovich 1980: 24, plate 27.

(7) The Mai Temenai site in northern Asmara has been dated to approximately 400 BC. It appears to be a discrete funerary context in which bronze knives, bracelets and tweezers as well as finely finished vases were documented. We cannot assess possible affinities between Mai Temenai ceramic vessels and the ceramics excavated from the tombs of Yeha in Ethiopia until these salvage excavations are published,

(8) General elements of these lithics are similar to Pre-Aksumite and Aksumite lithic materials described from the highlands of the northern Horn (Puglisi 1946; Franchini 1953; Tringali 1969; Fattovich 1972; Phillipson 1977; Phillipson 2000).

References

ABUL-HAGGAG, Y. 1961. A contribution to the physiography of Northern Ethiopia. London: University of London/Athlone Press.

ANFRAY, F. 1963. La premiere campagne de fouilles a Matara, pres de Senafe (Nov. 1959-Jan. 1960), Annales d'Ethiopie 5: 87-112.

1967. Matara, Annales d'Ethiopie 7: 33-88.

1970. Notes archeologiques, Annales d'Ethiopie 8: 31-42.

1974. Deux rilles axoumites: Adulis et Matara, in Atti IV Congresso Internationale di Studi Etiopici -- Roma 1972: 745-65. Rome: Accademia Nazionale del Lincei.

ANFRAY, F. & G. ANNEQUIN. 1965. Matara, deuxieme, troisieme et quatrieme campagnes de fouilles, Annales d'Ethiopie 6: 49-86.

ARTHUR, J. 2000. Ceramic ethnoarchaeology among the Game of southwestern Ethiopia. Ph.D dissertation, University of Florida.

2001. Reading pots: an ethnoarchaeological use-alteration analysis of pottery. Poster presented at the 66th annual meeting of the Society for American Archaeology, New Orleans, April 18-22.

BARD, K., M. COLTORTI, M. DIBLASI, F. DRAMIS & R. FATTOVICH. 2000. The environmental history of Tigray [northern Ethiopia) in the Middle and Late Holocene: a preliminary outline, African Archaeological Review 17(2): 65-86.

BARD, K., R. FATTOVICH, A. MANZO & C. PERLINGIER1. 1997. Archaeological investigations at Bieta Giyorgis (Aksum), Ethiopia: 1993-1995 Field Seasons, Journal of Field Archaeology 24: 387-403.

BOWER, J. 1991. The pastoral Neolithic of East Africa Journal of World Prehistory 5: 49-82.

BRANDT, S. 1984. New perspectives on the origins of food production in Ethiopia, in J.D. Clark & S. Brandt (ed.), From hunters to farmers: the causes and consequences of food production in Africa: 173-90. Berkeley (CA): University of California Press.

BRANDT, S. & N. CARDER. 1987. Pastoral rock art in the Horn of Africa: making sense of udder chaos, World Archaeology 19(2): 194-213.

CLARK, J.D. 1967. The problem of Neolithic culture in sub-Saharan Africa, in W.W. Bishop & J.D. Clark (ed.), Background to evolution in Africa: 601-27. Chicago (IL): University of Chicago Press.

1976. The domestication process in sub-Saharan Africa with special reference to Ethiopia, in Origine de l'elevuge et de la domestication: 56-115. Nice: Union International des Sciences Prehistoriques et Protohistoriques.

1980. The origins of domestication in Ethiopia, in R.E. Leakey & B.A. Ogot (ed.), Proceedings of the Eighth Panafrican Congress of Prehistory and Quaternary Studies: 268-70. Nairobi: The International Louis Leakey Memorial Institute for African Prehistory.

CURTIS, M. & Y. LIBSEKAL. 1999. Archaeological survey in the Adi Qeyeh area, Eritrea, Nyame Akuma 51: 25-35.

DREWES, A.J. 1962. Inscriptions de l'Ethiopie Antique. Leiden: E.J. Brill.

DREWES, A.J. & R. SCHNEIDER. 1976. Origine et development de l'ecriture ethiopienne jusqua' a l'epoque des inscriptions royales d'Axoum, Annales d'Ethiopie 10: 95-107.

DU TOIT, P.J. 1936. Bovine tuberculosis, Quarterly Bulletin of the Health Organisation, League of Nations, V. 1.

EPSTEIN, H. 1971. The origin of the domestic animals of Africa 1. New York (NY): Africana Publishing Corporation.

FATTOVICH, R. 1972. Sondaggi stratigrafici, Yeha 1971, Annales d'Ethiopie 9: 65-84.

1977. Some data for the study of the cultural history in ancient northern Ethiopia, Nyume Akuma 10: 6-18.

1978 Traces of a possible African component in the pre-Aksumite Culture of Northern Ethiopia, Abbey 9: 25-30.

1980. Materiali per lo studio della ceramica pre-Aksumita Etiopica, Annali dell'Istituto Universitario Orientale di Nepali 40(4). Supplemento 25.

1983. I `rilievi' rupestri di Daaro Caulos presso Asmara (Etiopia), Annali dell'Istituto Universitario Orientale di Napoli 43: 241-7.

1988. Remarks on the late prehistory and early history of northern Ethiopia, in T. Beyene (ed.), Proceedings of the Eighth International Conference of Ethiopian Studies -- Addis Ababa 1984: 85-104. Addis Ababa: Institute of Ethiopian Studies.

1990. Remarks on the pre-Aksumite period in northern Ethiopia, Journal of Ethiopian Studies 23: 1-33.

1997a. The Near East and Eastern Africa: their interaction, in Vogel (ed.): 479-84.

1997b. Northeastern African States, in Vogel (ed.): 484-9.

2000. Aksum and the Habashat: state and ethnicity in ancient northern Ethiopia and Eritrea. Boston (MA): Boston University African Studies Center. Working Papers in African Studies 228.

FATTOVICH, R. & K. BARD. 1998. The archaeological expedition in Ethiopia (Aksum) of the Istituto Universitario Orientale and Boston University (Documentary Exhibition). Naples: Istituto Universitario Oriental&

FRANCHINI, V. 1953. Stazioni litiche in Eritrea, Bollettino (Asmara) 1: 25-30.

MARSHALL, F. 1994. Archaeological perspectives on East African pastoralism, in E. Fratkin, K. Galvin & E. Roth (ed.), African pastoralist systems: an integrated approach: 17-43. Boulder (CO): Lynne Rienner Publishers, Inc.

MICHELS, J.W. 1994. Regional political organization in the Axum-Yeha area during the Pre-Axumite and Axumite eras, Etudes ethiopiennes 1: 61-80. Paris: Societe francaise pour les etudes ethiopiennes.

MUNRO-HAY, S. & G. TRINGALI. 1993. The Ona sites of Asmara and Hamasien, Rassegna di Studi Etiopici 35: 135-70.

PHILLIPSON, D.W. 1977. The excavation of Gobedra rockshelter, Axum, Azania 12: 53-82.

PHILLIPSON, L. 2000. Aksumite lithic industries, African Archaeological Review 17(2): 49-63.

PUGLISI, S.M. 1946. Industria litica di Aksum nel Tigrai Occidentale, Bivista di Scienze Preistoriche 1: 284-90.

RICCI, L. 1955-58. Ritrovamenti archeologici in Eritrea, II, Rassegna di Studi Etiopici 14: 48-68.

SCHMIDT, P. 1999. Preliminary report on the results from the 1998 training excavations at the Sembel site. Report submitted to the University of Asmara and the National Museum of Eritrea.

SCHMIDT, P. & H. WRIGHT. 1995. The potential of heritage resources in Eritrea. Report submitted to the National Museum of Eritrea.

SHOSHANI, J. & A. BEYIN. 2000. Report on faunal remains from 1999-2000 excavations at Mai Amber, Mai Hutsa, and Ona Gudo archaeological sites, Eritrea. Report on file in the National Museum of Eritrea.

SHOSHANI, J. & M. BOZA. 1999. Final Report on the faunal analysis of zoo-archaeological material from the Sembel site, and Sembel Qushet site, Asmara, Eritrea. Report on file in the National Museum of Eritrea.

TRINGALI, G. 1965. Cenni sulle `Ona' di Asmara e dintorni, Annales d'Ethiopie 6: 143-52.

1967. Necropoli di Curbacaiehat (Asmara), Journal of Ethiopian Studies 5(1): 109.

1969. Varieth di asce litiche in `ouna' dell'altopiano eritreo, Journal of Ethiopian Studies 7: 119.

1973-77. Necropoli di Cascasse e oggetii sudarabici (?) dalla regione di Asmara (Eritrea), Rassegna di Studi Etiopici 26: 47-98.

1980-81. Note su ritrovamenti archeologici in Eritrea, Rassegna di Studi Etiopici 28: 99-111.

1987. Reperti antichi di scultura minore e di omamenti dall'Eritrea e da Aksum, Rassegna di Studi Etiopici 31: 213-18.

VOGEL, J.O. (ed.). 1997. Encyclopedia of Precolonial Africa: archaeology, history, languages, cultures, and environments. Walnut Creek (CA): AltaMira Press.

WENIG, S. 1997. German fieldwork in Eritrea, Nyame Akuma 48: 20-21.

MATTHEW C. CURTIS, Department of Anthropology, University of Florida, 1112 Turlington Hall, Gainesville FL 32611, USA. [email protected] [email protected] (Schmidt also Department of Archaeology, University of Asmara, PO Box 1220, Asmara, Eritrea.)