Ethnoarchaeological Approaches to the Study of Prehistoric Agriculture in the Ethiopian Highlands
A.C. D'Andrea
D.E. Lyons
Department of Archaeology, Simon Fraser University, Burnaby, British Columbia, Canada.
Mitiku Haile
Mekelle University College, Mekelle, Ethiopia.
E.A. Butler
Institute of Archaeology, University College, London, UK.

This paper can be found in The Exploitation of Plant Resources in Ancient Africa, edited by Van der Veen. Kluwer Academic: Plenum Publishers, New York, 1999.

This paper reports on the preliminary results of an ethnoarchaeological project underway in the highlands of northeastern Ethiopia. The work is based at Adi Ainawalid, a small farming village located about 20 km northwest of Mekelle in south-central Tigrai. Investigations are focussing on selected aspects of traditional farming systems with a view to generating testable models on the nature and development of prehistoric agrarian societies in the region. Field studies are utilising interviews and direct observation to document crop processing of selected cereals and legumes in an effort to examine the effects of these activities on the composition of archaeobotanical assemblages. In addition, observations on domestic architecture, craft production, as well as general refuse disposal patterns are being conducted to aid in the interpretation of site formation processes. Plant husbandry and crop processing activities are placed into a broader cultural context by examining the socio-economic organisation of Adi Ainawalid, based on household studies.





Ethiopia has long been recognised as having an immense but untapped potential for research into the origins of plant domestication as well as related ethnoarchaeological studies (e.g., Vavilov, 1926, 1951; Porteres, 1962; Harlan, 1969; Brandt, 1984; Clark, 1988; Fattovich, 1988; Phillipson, 1993a). Harlan (1969) in particular, draws attention to the wide range of agricultural practices of great antiquity that have survived to the present-day, representing a vast reservoir of information for ethnoarchaeologists and archaeobotanists. The lack of investigation into these issues has been the result of regional political insecurity rather than scholarly neglect. Over the past five years however, concerted efforts have been made to rectify this situation. Two major archaeological projects are now underway in northeastern Ethiopia, both of which have made the study of subsistence and the recovery of ancient plant remains significant research priorities (Phillipson, 1993b, 1994, 1995, 1996; Bard and Fattovich, 1995; Fattovich, 1995; Bard et. al., 1996, 1997). Ethnoarchaeology has developed rapidly in sub-Saharan Africa since the late 1980s, but has only slowly made inroads into Ethiopia (MacEachern, 1996). While several studies are available on hide-working and associated lithic tool production (e.g., Clark and Kurashina, 1981; Gallagher, 1977; Brandt, 1996), as well as pottery (Messing, 1957a; Perlingieri, pers. comm.), there is now some interest in the examination of traditional practices associated with domesticated plants. Ethnoarchaeological studies of enset (Enset ventricosum (Welw.) Cheesman) cultivation are in progress in southern Ethiopia (Brandt, pers. comm., Hildebrand, pers. comm.), and the research described herein represents a first contribution to this area of study in the northern highlands.

This paper presents preliminary results of a joint Simon Fraser University-Mekelle University College ethnoarchaeological project underway at the highland farming community of Adi Ainawalid in south-central Tigrai (Figure 1). The ethnoarchaeological work falls into two main categories, each of which includes a variety of research objectives (D'Andrea et. al., 1997). For the purposes of this report, the focus will be on the implications of results for archaeobotanical studies. The first component involves actualistic studies of crop processing along with an analysis of residues at each stage of processing. Studies are concentrating on cereals including tef (Eragrostis tef (Zucc.) Trotter), finger millet (Eleusine coracana (L.) Gaertner), sorghum (Sorghum bicolor (L.) Moench), emmer (Triticum dicoccum (Schrank) Schübl.), intercropped wheat (Triticum aestivum L., T. durum Desf., and T. dicoccum) and barley (Hordeum vulgare L.) and legumes such as grass pea (Lathyrus sativus L.), chick pea (Cicer arietinum L.), lentil (Lens culinaris Medik.), pea (Pisum sativum L.), and fenugreek (Trigonella foenum-graecum L.). Similar investigations of harvest and processing of West Asian cereals have shed considerable light on processes of domestication (e.g., Wilke et. al., 1972; Hillman and Davies, 1990), and have proven useful in the interpretation of archaeological samples (Dennell, 1972, 1974, 1976; Hillman, 1973, 1984; Jones, 1984). The work of Hillman (1984) and Jones (1984) has documented the harvesting and processing of wheats and barley, however, apart from a few studies on Far Eastern millets (e.g., Yabuno, 1987), very few data exist on the processing of millet and other small-seeded cereals and legumes (cf. Butler, 1992; D'Andrea, 1992, 1995; Reddy, 1991). The data that are available suggest that West Asian wheat and barley processing models are not applicable to small-seeded cereals (Reddy, 1991). Given Harris' assumption (1984:64) that traditional harvest and processing techniques used by farmers are probably quite similar to those employed by non-agrarian peoples harvesting wild grasses, ethnographic models developed for millets and other small-seeded cereals such as tef may represent closer analogues for prehistoric wild grass processing. In addition, with the exception of Jones and Halstead (1995), few studies have considered the effects of intercropping on archaeological assemblages of plant remains. This issue is being addressed in this project by examining the processing of intercropped wheat and barley. With regard to legumes, Butler (1992, in press) has argued that traditional methods of cultivating Southwest Asian founder crops in Ethiopia, which include intercropping, may involve elements of technologies introduced with those crops in antiquity, and thus provide the basis for models applicable to ancient Near Eastern cultivation.

  Jan Feb Mar Apr May June July Aug Sept Oct Nov Dec
finger millet
grass pea+*
Sowing Jan Feb Mar Apr May June July Aug Sept Oct Nov Dec
finger millet  
grass pea+*            
Weeding Jan Feb Mar Apr May June July Aug Sept Oct Nov Dec
finger millet          
grass pea+*                        
Gamsa Jan Feb Mar Apr May June July Aug Sept Oct Nov Dec
finger millet          
grass pea+*            
Harvesting Jan Feb Mar Apr May June July Aug Sept Oct Nov Dec
finger millet
grass pea+*                  
Threshing Jan Feb Mar Apr May June July Aug Sept Oct Nov Dec
finger millet
grass pea+*                  

Figure 2. Seasonal schedule of plant cultivation activities, Adi Ainawalid. All responses are included in this figure except in cases where only one informant indicated that a particular activity took place during a given month.
*Grass pea + includes grass pea, chick pea, and fenugreek.

The second component of this study takes a wider perspective, looking at the social and economic organisation of non-mechanised highland farming communities. This will enable the placement of plant husbandry practices within a framework of domestic and community relations, as well as contribute to an understanding of archaeological site formation processes. A main goal of these investigations is to examine the relationship between cultural perceptions of gender, access to resources, and the configuration of domestic and community space. This builds upon a rapidly growing area of ethnoarchaeological and related research being conducted in Africa and elsewhere (cf., Moore, 1986; Blier, 1987; Lyons, 1989, 1996, in prep.; Donley-Reid, 1990; Hastorf, 1991; Roberts, 1991; Tringham, 1991; Hodder, 1994; Wall, 1994). While recording house dimensions and observations of domestic activities are relatively straight-forward procedures, acquiring an understanding of the symbolic relationships between people and material culture, in particular cultural perceptions of gender, requires a greater time investment in the field. Data presented here will be confined to preliminary observations on domestic activities, and their implications for the formation of the archaeobotanical record.

Beyond archaeology, the documentation of traditional agronomic systems is of value to scientists working on biodiversity preservation in developing countries (Amanor et. al., 1993). In recent years, environmental degradation, famine, resettlement, and other political disturbances have disrupted the continuity of farming practices in northern Ethiopia (cf. McCann, 1989, 1995; Marcus, 1994). Furthermore, as is the case in many other parts of the world, crop germplasm as well as knowledge associated with traditional cultivars is being lost at an alarming rate (Engels et. al., 1991). Recent trends in landrace preservation studies have highlighted the urgent need for investigation of the cultural context in which crop varieties are developed and maintained (e.g., de Boef et. al., 1993). Although several varieties of wheat and barley have been reported lost in Tigrai, many isolated communities continue to practice small-scale subsistence farming of indigenous crops, and remain relatively sheltered from the direct effects of modern agrotechnology. In addition, farmers have an intimate knowledge of crop varieties and play an important role in landrace conservation (Melaku Worede and Hailu Mekbib, 1993). As such, this region is an ideal location to undertake this programme of ethnoarchaeological research, which in turn may have wider implications for rural development studies.



The nature of modern Ethiopian agriculture has been the object of many scientific investigations since the late 1930s (McCann, 1995). Several important contributions have been made by Italian agronomists, especially Ciferri and others (e.g., Chiovenda, 1928; Ciferri and Baldrati, 1939, 1940; Ciferri and Giglioli, 1939a, 1939b), and more recent sources are available (e.g., Simoons, 1960; Huffnagel, 1961; Westphal, 1974, 1975). In addition, rural economic surveys in Tigrai (e.g., Hunting Technical Services, 1976; Holt and Lawrence, 1993; Simon Adebo, 1993) have produced some useful insights into prevailing local conditions. Ethnographic studies undertaken by Bauer (1975, 1977, 1987, 1989) and Messing (1957b) provide a window on the nature of culture change over the past two decades, during which the region has experienced severe environmental and socio-political stresses. These works have proven extremely valuable as background preparation as well as sources of comparison for this study, which differs in its focus on the examination of selected aspects of highland farming societies as they relate to the archaeological record. Ultimately, the main goal of this project is to generate testable hypotheses about the nature and development of prehistoric agrarian societies in the region.

The archaeological questions that have fueled this study involve the nature of prehistoric and early historic agriculture and the origin of plant domestication in Ethiopia. Ancient peoples in this region domesticated several indigenous species, including tef, finger millet, noog (Guizotia abyssinica (L.) Cass.), enset, coffee (Coffea arabica L.), chat (Catha edulis Forsk.), as well as adopted the Near Eastern suite of domesticates such as wheat, barley, chickpea, lentil, linseed or flax (Linum usitatissumum L.), faba bean (Vicia faba L.), pea, and safflower (Carthamus tincorius L.) (Vavilov, 1926, 1951; Porteres, 1962; Harlan, 1971, 1992; Zeven and Zhukovsky, 1975). The status of an indigenous, non-shattering Ethiopian oat (Avena abyssinica Hochst.) as a domesticate remains uncertain (Simoons, 1960:108; Ladizinsky 1975:240; Harlan 1992:68). Although it is often described primarily as a tolerated species in fields of barley and wheat (Ladizinsky, 1975:240; Harlan, 1992:68), recent sources indicate that the Ethiopian oat is cultivated for grain in northern Ethiopia, more frequently than the common oat (Avena sativa L.) (Phillips, 1995:37). These conflicting reports have a long history that extends back to the early 19th century (see discussion in Simoons, 1960:108). The route and timing of the arrival of Near Eastern crops and their impact on local agricultural systems is virtually unknown (Brandt, 1984), but it is thought that some form of indigenous cultivation already was in existence (Harlan, 1992). It is possible that these two agricultural complexes co-existed for thousands of years (Harlan, 1969). In addition, there were dispersals of domesticates to the Ethiopian region from other parts of Africa, including sorghum (Sorghum bicolor (L.) Moench) and pearl millet (Pennisetum glaucum (L.) R. Br.). The processes involved in the integration of these many crops into the present-day diverse farming pattern of the highlands remain to be elucidated.

Ethiopia and Eritrea have been a pivotal region for prehistoric contacts between Africa, Asia, Egypt, and Southwest Asia for the past 4000 years. There is fragmentary evidence that indicates an early dispersal of sorghum, finger millet, and tef out of Africa, suggesting the existence of earlier local domestication events. Although the African origin of these species has been established by botanical and cytogenetic studies (e.g., Mehra, 1962, 1963; Hilu and de Wet, 1976; Hilu et. al., 1979; Harlan, 1982, 1989, 1992), in several cases the earliest evidence for these crops is outside Africa. For example, the earliest known archaeological finger millet is found in India dating to the second millennium BC (Harlan, 1993). Tef appears in the form of pottery impressions at Hajar bin Humeid, a site in southern Yemen dating to the first century BC (Soderstrom, 1969). Although there is a report of sorghum in the Khartoum area dating to the sixth millennium BC (Harlan, 1993), the earliest confirmed archaeological specimens appear in India by at least 2000 BC, probably arriving via trade routes from Yemen or other parts of the Arabian peninsula (Kajale, 1988; Harlan, 1989, 1993; Rowley-Conwy, 1991). Claims have been made for the presence of sorghum in North and South Korea during the Plain Pottery Period (ca. 2000-500 BC) at the sites of Hunam-ni and Songguk-ni (Yim, 1978 and Kang et. al., 1979 in Higuchi, 1986:123; Kim et. al., 1978). Finally, Tsukada (1986) notes the occurrence of pollen derived from an endemic African melon (Cucumis melo L.) in lake sediments from southwestern Japan dating to 3000 BP. The possible presence of African-derived domesticates in regions as far away as Korea and Japan by about three thousand years ago certainly strengthens the case for early domestication in Africa, however, until archaeobotanical data are obtained, arguments such as these remain in the realm of speculation. This lack of information notwithstanding, the role of Ethiopia and Eritrea is pivotal, since they are located in close proximity to Yemen and Saudi Arabia, and may have been on the dispersal route out of Africa to South and East Asia via the Sabean Lane. Ethiopian crops such as tef and finger millet, and other African species, such as sorghum and pearl millet could have moved along this route either overland or by sea traffic.

Since the publication of several comprehensive reviews on Ethiopian agricultural origins (e.g., Brandt, 1984; Clark, 1988; Phillipson, 1993a), there are few new data to add. The remains of domesticated cattle at the sites of Besaka (Clark and Williams, 1978; Brandt, 1980) and Laga Oda (Clark and Williams, 1978) dating to the second millennium BC constitute the earliest evidence for food production in the Ethiopian region. Domesticated camel may be represented at Gobedra Rockshelter, but the estimated date of 7000-3000 bp awaits confirmation by radiocarbon dating (Phillipson, 1993a). Based on ecological and related data, Bulliet (1975) suggests that camels were introduced to sub-Saharan Africa sometime between 2500-1500 BC, however, this hypothesis remains untested by zooarchaeological evidence. At Laga Oda, cattle remains are associated with microblades bearing edge-gloss, possibly the result of grass harvesting activities involving tef or sorghum (Clark and Prince, 1978; Clark and Williams, 1978), but whether the gloss was produced by domesticated grasses is moot. The earliest unequivocal archaeobotanical evidence for domesticated plants in the region dates to at least 500 BC, and consists of hulled and naked barley, chickpea, and several legumes recovered at Lalibela cave in central Ethiopia (Dombrowski, 1970, 1971), as well as wheat and barley identified by J. Hansen at Ona Nagast (Aksum) dated by association with charcoal to 2335 +/- 220 b.p. (390 cal. BC) (Bard and Fattovich, 1995; Bard et. al., 1997). Recent flotation sampling by Boardman (this volume) at Aksum and continuing work at Ona Nagast (D'Andrea, 1997) is promising to deliver new data on the role of agriculture prior to and during the height of the Aksumite Kingdom during the first millennium BC/AD. Excavation and flotation sampling of earlier periods is critical and already underway (Phillipson, 1996; Agazi Negash, 1997, pers. comm.; Finneran, pers. comm.)

Despite the dearth of archaeobotanical data for the Ethiopian region, several workers have attempted to provide explanations for the origins of plant domestication and agriculture based on other data (e.g., Arkell, 1954; Murdock, 1959; Clark, 1976, 1980, 1988; Fattovich, 1975, 1977; Brandt and Carder, 1987). These theories generally attribute the origin of farming as the result of cultural contact, which Clark (1988) associates with increasing desiccation in East Africa beginning by 7000 bp. Several workers have noted similarities between Nubian C group ceramics from sites in the eastern Sudan (Fattovich, 1975, 1977; Clark, 1988). Clark (1988) suggests that Sudanese Nile pastoralists moved from lowland regions to adjacent Ethiopian uplands. These peoples then either began to intensify their use of local wild plants, leading to domestication or they influenced local inhabitants to cultivate highland species such as tef and finger millet. It is further suggested that pastoralists or mixed farmers followed a pattern of transhumance between cultivation in fertile highlands and livestock grazing in lowland areas. This interpretation is based on ceramic similarities between highland and lowland sites (Clark, 1988), and on the nature of rock art (Brandt and Carder, 1987). Although this sequence of events is quite plausible, the proposed timing of plant domestication remains untested by archaeobotanical data.

Some workers have dealt with the importation of the Near Eastern complex from the Arabian peninsula based on ceramic and other criteria (Seligman, 1957; Fattovich, 1990), but these studies also lack corroborating bioarchaeological data. Ehret (1974, 1979, 1982, 1984) has discussed the role of linguistic evidence in establishing the origins of plant and animal domestication in the Ethiopian region. Most languages spoken in Ethiopia today belong to the Afroasiatic language family, including Cushitic, Omotic, and Semitic groups. Cushitic and Omotic language groups are thought to have a much longer time depth in the area, and may be descendants of original inhabitants of the Ethiopian plateau. Although many people today speak Semitic languages (e.g., Amharic and Tigrinya), the earliest Semitic language speakers are thought to have arrived from Arabia during the first millennium BC. Ehret (1979) has argued that Afroasiatic languages developed in northeastern Africa in the vicinity of the Red Sea some 15,000 years ago, and the speakers of these languages were involved in the intensive exploitation of wild grasses. He points out that several Semitic terms relating to the plough and cereal cultivation were borrowed from Cushitic sources, suggesting the presence of agrarian practices in Ethiopia as early as the seventh millennium, BC. Ehret (1979) further concludes that sheep and goats were introduced to East Africa as early as 5000 BC, several thousands of years before the arrival of Semitic language speakers. This hypothesis, although intriguing, awaits testing by a systematic programme of palaeoethnobotanical research, as do the other models discussed.



Ethnoarchaeological investigations on their own will not be able to directly solve the problems outlined above; achieving this will require excavation and sampling of Late Stone Age and earliest agrarian sites in Tigrai for preserved biological remains. However, ethnographic analogy can provide valuable insights into cultural interactions with the surrounding environment, as well as generate testable hypotheses about the relationships between the ethnographic and archaeological records (cf. Wylie, 1985). Michels (1990:187) has noted that continuity in material culture, domestic technology and land-use patterns in northeastern Ethiopia until at least 1974, presents a useful starting point in producing archaeological interpretations. Although this apparent continuity affords certain advantages in this study, it is not assumed that there will be a direct relationship between Adi Ainawalid and prehistoric Ethiopian societies. The focus instead is to use ethnographic analogies to generate testable models of prehistoric lifeways.

It must be stressed that the results presented below are preliminary. As such, some of the reporting is anecdotal at this stage, and will almost certainly be revised as more time is spent in the field.

 The Study AreaThis study is based at Adi Ainawalid, a small village of approximately 180 households in south-central Tigrai. It is situated about 20 km northeast of Mekelle in Inderta administrative region (Figure 1). Adi Ainawalid is part of a collective involving four other villages: Adi Bakal (180 households), Chin Feras (270 households), Mesahil (250 households), and Tselwa (300 households). It is being used as a basis for both household and crop-processing components of the research, but other villages are visited to record activities not taking place in Adi Ainawalid, such as metal working, pottery-making and the cultivation of some crops.

Three environmental zones have been outlined for Ethiopia based on temperature and latitude: cooler high altitude (>2400 m asl), warmer medium elevation (1700-2400 m asl), and hot lowlands (<1700 m asl). In addition there are two rainy seasons experienced by the region: the long kremt rains fall from mid-June to early October and lead to the meher harvest between November and January while short belg rains are experienced from February to April (Westphal, 1975). Adi Ainawalid is classified as medium altitude, which is the most amenable for agriculture, and where the greatest variety of crops is grown. For the past several years, belg crops have not been grown in the area because of poor short rains (Hunting Technical Services, 1976; Holt and Lawrence, 1993), consequently farmers have concentrated their efforts on the meher harvest (Simon Adebo, 1993).

 Preliminary ResultsThe results to date of ethnoarchaeological research at Adi Ainawalid are presented in three sections: seasonal schedule, crop processing studies, and households. In the initial survey undertaken in November-December of 1996, 64 households were randomly selected, and members interviewed on the timing and nature of various cropping activities. Of the 64 interviews, 41 were sufficiently detailed to form the basis of seasonal data presented in Figures 2 and 3. Specific crop processing activities that were underway during the field season also were observed and recorded. Harvesting and threshing were the main occupations of farmers at the time, and it was possible to observe threshing of several crops in six threshing areas scattered throughout the village. Observations on crop processing will be refined during an upcoming field season, during which samples of residues from all stages of processing will be conducted. Detailed household interviews were undertaken in May-June 1997.



Seasonal Schedule

The interviews on seasonal activities presented in this section were completed in 1996, and concentrated mainly on plant husbandry. These data will be enlarged in upcoming field seasons to include other farming tasks related to terrace field maintenance and livestock management. The results of 41 detailed interviews on seasonality of crop-related activities are summarised in Figures 2 and 3. Figure 2 illustrates the seasonal scheduling of major plant cultivation activities over the course of one calendar year, while Figure 3 illustrates the timing of activities for each crop. The overlap in the timing of activities exists because farmers were asked when activities normally happen, rather than what took place in any specific year. As a result, the answers should incorporate variation in the timing of activities over several years, and this kind of information may be of more relevance to the archaeological record. The annual cycle can be roughly characterised by three different periods. From January to June the main activities are land preparation and sowing, with residual threshing and harvesting early in the year. In addition, much effort is expended on the repair and building of terraces in communal work groups. The region around Adi Ainawalid is intensively terraced in an effort to reduce soil erosion. The nature of activities changes from June to September during the kremt rains when farmers first concentrate on sowing, and later on tending of crops. During September the early harvest of maize begins, but from October to December, farmers are occupied exclusively with harvest and threshing.

  Jan Feb Mar Apr May June July Aug Sept Oct Nov Dec
Finger Millet Jan Feb Mar Apr May June July Aug Sept Oct Nov Dec
Sorghum Jan Feb Mar Apr May June July Aug Sept Oct Nov Dec
Wheat/ barley Jan Feb Mar Apr May June July Aug Sept Oct Nov Dec
Maize Jan Feb Mar Apr May June July Aug Sept Oct Nov Dec
Grass Pea +* Jan Feb Mar Apr May June July Aug Sept Oct Nov Dec
Lentil/Linseed Jan Feb Mar Apr May June July Aug Sept Oct Nov Dec

Figure 3. Seasonal schedule of agricultural tasks for main crops, Adi Ainawalid. Thses data are based on 41 interviews. All responses provided by respondents are included in these diagrams except in two cases. In the first instance, an answer given by one informant regarding plowing was no included because of translation inconsistencies. Secondly, answers given were excluded if only one informant from the entire group indicated that any particular activity took place during a given month.
*Grass pea + includes grass pea, chick pea, and fenugreek.

In Figures 2 and 3, data for the following crops were combined by farmers in their responses because of similarities in the timing of activities: grass pea, fenugreek and chick pea; wheat and barley; lentils and linseed. For the most part, this presents an accurate picture, however there are some minor differences, for example, the harvesting of chick pea sometimes takes place later than grass pea and fenugreek. In general, less attention is given to grass pea, fenugreek, chick pea, lentil and linseed, in terms of ploughing and weeding. The vast majority of respondents do not plough for lentils and linseed, and if so, only for a relatively short time. Several informants stated that they often sowed lentil and linseed on unploughed surfaces (see also Butler et. al., this volume).

In general, ploughing for most crops takes place during the dry season (January to June or July), with the exception of grass pea which can continue until August. This is accomplished using a maresha or scratch plough as described by Goe (1990:9). Because this instrument does not penetrate the soil deeply, several passes must be made before sowing begins. According to Adi Ainawalid farmers, the number of times a field is ploughed is directly related to access to oxen (cf. Bauer, 1975; McCann, 1995:78-79). Although farmers plough as many as five or six times for tef and usually fewer for other cereals, most emphasised that they plough their fields as many times as possible. The activities of sowing, weeding, and gamsa overlap for most crops over the course of a year. Gamsa is defined as a shallow plough, usually for the purpose of removing weeds or covering a newly sown crop. Weeding for sorghum, maize, and finger millet takes place before gamsa. The term 'weeding' normally refers to uprooting weeds by hand, however, in the case of maize, sorghum, and finger millet a hoe-like implement is used in weeding before gamsa. In tef and grass pea cultivation, gamsa can take place before sowing. Harvest and threshing for most crops occurs from October to December, with the exception of maize, lentil/linseed, and finger millet. Maize and lentil/linseed are harvested earlier than most crops, while finger millet is normally left in the field after harvesting for up to three months before threshing begins. Harvest time (November-December) is identified by most respondents as the busiest time of the year followed by sowing time (April to June).

These observations at Adi Ainawalid are generally comparable to those reported by at least one rural economic survey undertaken in Inderta administrative region (Simon Adebo, 1993). This survey employed "rapid rural appraisal" techniques which involved five days of fieldwork at Debri village collective located approximately eight kilometres southeast of Mekelle. The seasonal schedule reported by Simon Adebo (1993) is roughly the same as that observed at Adi Ainawalid, except for a more tightly defined sowing season (June to August). He also reports that the most labour-intensive times occur during the October-December harvest and June-July sowing periods, but unlike Adi Ainawalid farmers, his informants believed the latter was the busier time. More detailed comparisons with the findings of Simon Adebo (1993) and Holt and Lawrence (1993) will be attempted as fieldwork progresses.

An interesting cultivation technique observed at Adi Ainawalid and other villages is the intercropping of wheat and barley (hanfetse), in addition to growing each crop separately. A similar practice is described in detail by Jones and Halstead (1995) in their study of traditional farming practices on the Greek island of Amorgos. At Adi Ainawalid, farmers generally sow two varieties of wheat with two varieties of barley in ratios of 50:50, and always in the same combination. However, because the proportion of wheat gradually increases after two or three plantings, harvested mixtures vary from 50:50 to 70:30. This situation is similar to that observed by Jones and Halstead (1995:104) who found that the composition of wheat:barley mixtures varies from 50:50 to 80:20. Additional comparisons with the Amorgos study await further sampling of crop processing products and by-products at Adi Ainawalid. According to Adi Ainawalid farmers, growing hanfetse affords several advantages, including the production of generally higher yields, protection against lodging, and some reported that the presence of barley reduces the occurrence of fungal attack on wheat. The mixed crop is sown, harvested, threshed, and winnowed together, followed by hand separation of barley which is then pounded to remove hulls. Women were observed to spend long hours hand-sorting grains of threshed hanfetse. Several farmers emphasised that the practice of intercropping free-threshing wheats and barley is relatively recent, whereas the "traditional" method involves mixtures of emmer and barley. These two cereals are sown, harvested, threshed, winnowed, pounded, and ground together, and then used in bread making. When asked why emmer was no longer grown in Adi Ainawalid, a wide variety of answers was given. However, several informants did emphasise that this crop was discontinued because of a reduction in land availability and the large amount of work involved in removing hulls during processing.

The antiquity of hanfetse in Tigrai is unknown. Although there are historical references to mixed cropping of several cereals (especially wheat and rye) in northwestern Europe dating to at least the 14th century AD (Slicher van Bath, 1963:263; Jones and Halstead, 1995:104), surveys of 19th century agronomic literature on Ethiopia have not produced any accounts of cereal intercropping, other than references to mixtures of tef varieties (McCann, 1995, pers. comm.). Field mixtures of wheat and barley are briefly mentioned in some rural surveys (Holt and Lawrence 1993:68; Simon Adebo 1993:38), but in general this practice seems to have been overlooked by many observers. One exception is Simoons (1960:70-71) who, in his economic survey of northwestern Ethiopia, states that the most common intercropping technique involves the cultivation of wheat and barley. Other combinations include tef-linseed or barley, sorghum-finger millet (mentioned by one farmer in Adi Ainawalid), and several mixtures involving pulses, such as green pea-horse bean and sorghum-chick pea (see also Butler, in press). Based on observations made in Gondar, Simoons (1960:71) further suggests that mixing wheat and barley is an effective strategy against crop failure.

Several lines of evidence suggest that this practice has been in existence for some time, however definite conclusions cannot be reached at this point. As stated above, emmer-barley intercropping is viewed as an older method by several Adi Ainawalid informants. Furthermore, although historical records are silent on this practice, they do indicate that the spectrum of crops and conditions under which they were cultivated have remained more or less constant in many regions of Ethiopia since at least the 19th century AD (McCann, 1995:102). Finally, related evidence points to a natural association between barley and emmer that may have extended into the more distant past. In a discussion of the rationale for selecting potential grain cultigens in the prehistoric Jordan Valley, Bar-Yosef and Kilsev (1989:640) state that close morphological and ecological relationships between barley and wild emmer in that region enabled the latter to survive in fields of domesticated or "pre-domesticated" barley for several generations. Following parallel mutations for a tough rachis, the selection of wheat eventually took place, and it became an independent crop. Interviews at Adi Ainawalid and observations at near-by emmer-growing villages indicate that emmer-barley intercropping is a deliberate practice, and not the result of farmer tolerance of emmer in barley fields. Farmers who are still growing emmer as pure stands and intercropped with barley or free-threshing wheat have been located, and investigations on these practices are in progress (D'Andrea and Mitiku, in prep.). This work may contribute to the growing ethnoarchaeological database on processing of glume wheats (e.g., Hillman, 1984; Nesbitt and Samuel, 1996).


Crop Processing Studies

Several crop processing activities were observed in the autumn of 1996. The main activity during this time was threshing, although some food preparation was observed (see Butler et. al., this volume). Processing stages were described and photographed for tef, hanfetse, sorghum, and grass pea, and a simplified flow chart for tef processing outside the household is illustrated in Figure 4. The following are some preliminary observations on tef processing, while grass pea is detailed in Butler et. al., (this volume).

     Fig 4.   Preliminary Tef Processing Sequence, Adi Ainawalid.

The processing sequence for tef is a compilation of observations made in three threshing areas. Farmers stated that although there is no general rule regarding the location of threshing yards, the same positions are often used over several years. Threshing tef requires special preparation of the ground surface. This is accomplished the day before when a thin coat of a dung and water mixture is placed on the ground surface within a circular area surrounded by stones. This creates a hardened surface to reduce seed loss, because tef grains are exceedingly small.

Tef is harvested by a combination of uprooting using a sickle. Plants are not left to dry in the field, but are transported immediately to the threshing area. Entire tef plants are piled in the centre of the floor during First Threshing, and while oxen are trampling, straw is turned over and gradually moved to the edge of the threshing area using a two-pronged fork. During Second Threshing, the harvest is returned to the centre for further trampling. Straw is then slowly removed using hands and a threshing fork. At the end of threshing, the harvest is swept into two piles using a hand-held brush, and the surface of each pile is raked to remove larger fragments of straw. First Winnowing involves the use of a tray where material is gently poured into the wind. At the foot of the farmer who is winnowing, a pile of semi-clean grain and chaff accumulates, and after each basketfull the assistant gently rakes away fine straw, chaff, and gurdie3. This creates a second accumulation which is swept aside and sieved later. First Sieving is accomplished using a coarse basketry riddle whereby materials are sieved into the wind, and after each basketfull, the pile is raked to remove gurdie. This results in the production of three fractions: 1) cleanings which remain on top of the sieve and are discarded; 2) semi-clean grain that falls through the sieve and is placed into sacs for transport home via donkey; and 3) gurdie which accumulates immediately downwind of the semi-clean grain. Second Sieving of materials produced during First Winnowing takes place next, and this produces two fractions: 1) cleanings remaining on top of the sieve which are used as fodder; and 2) gurdie which is combined with gurdie produced during First Sieving. All gurdie is transported home but is stored separately from semi-clean grain. Cleanings from the Second Sieving are combined with the straw, and bundled up onto a donkey for transport home where it is stored in a fenced yard adjacent to the main residential compound. In the household, both semi-clean grain and gurdie are processed in small amounts as required. Detailed investigations of household processing methods are still in progress.

Microscopic analyses of residues produced in the processing of tef and other cereals are not complete, however, a brief examination of samples indicates that the nature of weed seeds and chaff fractions may be useful criteria in defining stages of processing. The archaeological preservation potential of tef processing stages has not yet been assessed in detail, however these early observations suggest the probability of preservation will be low, given the fragility of tef chaff and the lack of exposure to fire during processing.



There is little comparative ethnographic research available on Tigrayan domestic compounds and related material culture. Simoons (1960:57-63) provides a brief description of house structures typical of Tigrai, but offers little detail on the internal arrangement of living spaces. In his ethnographic study of Adi Hareyna, a village located about 30 km northeast of Mekelle, Bauer (1977:26-29) emphasises how components of house design can reflect social status. The results presented below are restricted to observations regarding food processing, serving, and storage within domestic settings, all of which can affect the archaeological preservation of plant remains.

To date the spatial organization of domestic buildings, food processing areas, and crop storage facilities have been mapped and photographed in 36 compounds in Adi Ainawalid. Figure 5 illustrates one such compound mapped in 1997. Compounds generally consist of a main house, a separate kitchen building, and depending on family wealth and stage of domestic life cycle, buildings for newly married children.


     Fig 5.   Adi Ainawalid Compound Map.

The main building is where crops, fodder, and personal effects are stored, food is consumed, and where the family sleeps. These buildings are well-constructed using shaped and mortared stone, and have a life-span of 60-100 years. All interior furnishings are constructed of stone slabs covered in a mixture of clay, straw and dung. At the east or west end of the house there is a sleeping platform divided from the main room by short partitions which probably function as buttresses for the north-south walls. On the opposite side of the house there are two compartments divided by a wall running north-south which supports the upstairs loft. The downstairs compartments are used mainly for pottery storage, while the upstairs loft is enclosed and contains grain storage bins. The central living area is furnished with low benches built along the base of walls. Sleeping platforms and sometimes benches are covered in leather which is protected from abrasion by a layer of leaves or straw. Interior walls of the main building and the entry are covered in clay and painted in two colours: a red-brown or yellow clay along the lower third of the wall and the upper section in white clay paint. The base of the white section is sometimes decorated with generic plant designs, crosses, geometric patterns, and other motifs with ink or plant-based paints yet to be identified. Walls are repainted every three or four months.

Main buildings usually have hidmo-style roofs of stone and clay supported by wooden beams. The preferred wood is juniper (Juniperus spp.) which is no longer available locally because of deforestation. Consequently, juniper beams from old buildings are either reused or more commonly replaced with eucalyptus (Eucalyptus spp.) or olive (Olea europaea) wood. Main buildings usually have foundations approximately 0.6 m wide and 1.0-1.5 m deep. The floor is made up of the original ground surface covered with a clay and dung mixture which is re-surfaced every three to four months coincident with wall repainting.

Separate kitchen buildings are typical of most compounds but are not as well constructed as main buildings. Kitchen walls are made using unshapen stone with no mortar, and interior surfaces are rarely painted. Often kitchens have only a partial and light daubing of clay to prevent wind from extinguishing the stove fire. Roofs are made from sorghum stalks, and floors usually have no preparation over the original ground surface. Subterranean foundations for walls and floors are normally absent in the kitchen, but when present, they are about half the depth of those in the main building. The poorer quality of construction for kitchens is in keeping with the fact that these are impermanent structures that are frequently moved.

The kitchen is much smaller than the main house with little space for anyone except the cook. It is furnished only with a stove made of clay covered with stone slabs measuring approximately 1.0-1.5 x 1.0 x 0.5 metres (l x w x h). Stoves are fueled with dung and sorghum stalks, and usually have three elements: a large ceramic griddle for cooking breads; an open element for boiling sauces in pots; and an iron griddle for roasting grains. Chimneys are a recent introduction by the Home Economics agent for the Ministry of Agriculture in order to improve women's health. Stoves without chimneys produce considerable smoke which can cause lung and eye diseases as well as blacken the roof and walls of the kitchen with soot. Occasionally a box-like structure of clay covered stone is constructed in the kitchen to burn leaves of O. europaea or Euclea schimperi to produce an aromatic smoke used in women's cleansing practices for the month after giving birth. In homes of widows or elderly couples, kitchens may be altogether absent. In this case, a single griddle stove is built inside the main house in one of the compartments under the loft or in a corner of the main room.

The architecture and spatial impermanence of kitchens vis-a-vis other domestic buildings has important implications regarding the potential recovery of archaeobotanical remains and for understanding site formation processes in general. In this regard, it is important to note that food preparation and processing is mobile, and even fixed equipment such as grinding tables can have a variety of spatial locations. It has been observed that grinding tables occur in kitchens, in the yard along compound walls, and in main buildings. Furthermore, although the main kitchen stove is usually fixed, a small mobile stove is used anywhere in the compound for a variety of cooking tasks. Women and men actively participate in entertaining guests in the central area of the main house, and considerable ceremony goes into the preparation of coffee and popped grains. Roasting and popping occurs in the central area, usually near the door and natural light source, and in view of the guests.

Grain is typically stored in the main house, but never in the kitchen. Three types of granaries are used to store cereals. Several farmers observed that tef and finger millet are more resistant to insect infestations (see also McCann 1995:67) and mold damage during storage than wheat, barley, and sorghum. This may suggest a longer presence of tef and finger millet in the region, but this is pure speculation at this point. Tef and finger millet can be stored for up to ten years in a tall, unfired clay cylinder-shaped granary or less frequently in a similarly shaped granary basket, both of which can be located on the ground floor. If other cereals are stored on the ground in these types of granaries, they will quickly become infested by insects. Tef and finger millet granaries are located on either side of the sleeping platform and sometimes in the compartments under the loft. Other cereals, such as wheat, barley, and sorghum are stored in the loft in open bins made of upright stone slabs, wood, and clay. The elevated floor of the loft offers more in the way of protection from insect pests. Although this spatial separation of grain storage is typical, it is not absolute. In houses without lofts, bins for all cereals are commonly found on the ground floor.

Women sweep house floors, courtyards, and kitchens on a regular basis. This continuous sweeping action on kitchen floors that are not treated with dung tends to erode surfaces, resulting in their displacement by several centimetres below the surrounding courtyard. Sweepings are discarded outside compounds in clearly defined areas. Domestic compounds in the community tend to cluster around open areas where footpaths from nearby residential clusters converge. In these areas, it is common to dispose of house sweepings and dung separately from hearth ash cleanings. The ash is then used to mix mortar for wall construction, or it can be used as fertilizer for house gardens.

Investigations into the symbolic and functional use of pottery was initiated in 1997. In general, Adi Ainawalid households have very few western goods. When such goods are found they are most commonly plastic containers and tin cans used to substitute for the most frequently broken pottery types (water carrying pots and beverage cups). What is remarkable about the pottery is that although there are approximately a dozen different functional types, only three are used directly over the fire, and require frequent replacement during the year. Flat breads are prepared on a ceramic griddle reducing the need for cooking pots. Other ceramic vessels are used to store and brew beverages, and if damaged, they are reused for grain or flour storage. Such pots have relatively long life-cycles of 7 to 40 years. The result of this emphasis on griddle cuisine is a low concentration of pot sherds around compounds and surrounding fields as well as in middens. Close (1995:32) has suggested that the volume of pottery sherds increases with sedentism, however, in the case presented here, additional factors may intervene.

One avenue that will be explored is the correlation between type of cuisine, local ecology, and division of labour in Adi Ainawalid. There are four flatbreads produced in this region, the most common of which is taita. It is made of a fermented batter of various combinations of tef, barley, sorghum, and finger millet flours. Taita cooks on the griddle in a few minutes, and a stack is stored in baskets for consumption as needed. Brumfiel's (1991) study of pottery in Aztec Mexico revealed a correlation between griddle (tortilla) and boiled cuisines with both environmental conditions, and the economic role of men and women within the Aztec state. The role of griddle cooking in the highlands is of particular interest. It may be possible that griddle cooking developed in this area partly to increase fuel efficiency. High altitudes require greater temperatures and quantities of fuel for boiling, but at the same time, deforestation limits fuel choices to dung, woody shrubs or straw. Since women are responsible for cooking and collecting fuel, griddle cooking may be a technology developed to reduce fuel needs at a cost of greater time investment in food preparation. If so, this adjustment may have influenced other aspects of food production and processing, such as the division of labour for household and farming tasks. There is some archaeological evidence to support this suggestion, as oven remains, griddle rims, and lid fragments are present in deposits at Lalibela and Natchabiet Caves, which date to the early second millennium AD. At Lalibela, similar remains are present but much rarer in lower levels dating to the middle of the 1st millennium BC (Dombrowski, 1971:123, 130). Historical accounts from the 16th century AD indicate that highland environments were already lacking in forest cover, and animal dung constituted the main fuel source (McCann, 1995:36-38). In fact, deforestation of the highlands may date at least to the late 1st millennium BC based on preliminary palynological studies undertaken on Pre-Aksumite contexts at Bieta Giyorgis, Aksum (DiBlasi, 1996). Further investigation of modern households as well as the archaeological record of domestic activities are needed to clarify the nature and development of this culinary practice.



The preliminary nature of this investigation precludes the formulation of definite conclusions at this point, however some general observations can be made. Now that a general seasonal schedule for Adi Ainawalid has been established, it is possible to begin a detailed examination of various farming activities, as well as an investigation of land-use patterns, and the influence of culture and ecology on the lifeways of small-scale highland farmers. Although detailed analyses of cereal processing residues have not yet been completed, a gross examination of samples collected to date indicates that it may be possible to differentiate stages of processing (see also Butler et. al., this volume). However, the preservation potential of these components in the archaeobotanical record has yet to be determined. With regard to cropping systems, although the antiquity of intercropping cereals as well as legumes is unknown, archaeobotanists may have to consider this practice as a possibility in the interpretation of archaeobotanical assemblages in the Ethiopian highlands.

Several observations can be made on the nature of households and their influence on the formation of the archaeological record, in particular, the preservation of archaeobotanical remains. In Adi Ainawalid, kitchens are easily distinguishable from main residence buildings. However, the impermanence of kitchen buildings, where charring accidents are more likely to occur, as well as the existence of mobile stoves, may result in fewer discrete concentrations of charred plant remains in residential compounds. Furthermore, charred materials that do remain in the living area may be more indicative of sweeping and disposal activities rather than specific food processing tasks. As a result, other classes of archaeological remains may be required to help isolate specific food-producing activities. The spatial separation of cereal storage in residential compounds with lofts, as well as the storage of cereals in different containers at Adi Ainawalid, for the most part seem to reflect varying degrees of resistance to pest infestations. Finally, the possible association of highland griddle cooking with a reduction in fuel availability, and adjustments in social relations are tentative observations awaiting further study.

More intensive interviews are planned on crop processing activities, household food preparation and associated social dynamics. The mapping of Adi Ainawalid as well as adjacent fields is in progress, and this will allow the spatial analysis of subsistence activities, disposal patterns, land-use, and family relationships.



This project was made possible by the cooperation and hospitality of the people of Adi Ainawalid, to whom we are indeed grateful. We thank the Ministry of Information and Culture (Addis Ababa) and the Bureau of Culture, Tourism, and Information of Tigrai (Mekelle) for granting permission to undertake this study. This research is funded by a Social Sciences and Humanities Research Council of Canada Research Grant (No. 410-96-1520). We would like to thank our principal interpreters Mr. Shewaye Belay and Mr. Zelealem Tesfay (Mekelle University College) for their assistance in the field, as well as Ms. Shannon Wood (Simon Fraser University) for producing Figures 1 and 5. We gratefully acknowledge the assistance of the following individuals: Ms. Fetein Abay, Mr. Yemane Asgedom (both of Mekelle University College), Dr. Jean Hanson (International Livestock Research Institute, Addis Ababa), Dr. Sue Edwards (National Herbarium, University of Addis Ababa), and Prof. Rodolfo Fattovich (Istituto Universitario Orientale, Naples, Italy).



Agazi Negash. (1997). Preliminary results of an archaeological reconnaissance of Tigrai, northern Ethiopia. Nyame Akuma, 47:27-32.

Amanor, K., K. Wellard, W. de Boef and A. Bebbington. (1993). Introduction. In (W. de Boef, K. Amanor and K. Wellard, eds.) Cultivating Knowledge. London: Intermediate Technology Publications, pp. 1-16.

Arkell, A.J. (1954). Four occupation sites at Agordat. Kush 2, 33-62.

Bard, K.A. and R. Fattovich. (1995). The I.U.O./B.U. Excavatioin at Bieta Giyorgis (Aksum): an interim report. Nyame Akuma 44, 25-27.

Bard, K.A., M. DiBlasi, R. Fattovich, A. Manzo, C. Perlingieri, L. Crescenzi. (1996). The B.U./I.U.O. Archaeological Excavations at Bieta Giyorgis (Aksum, Ethiopia): a preliminary report on the 1996 field season. Nyame Akuma 46, 21-23.

Bard, K.A., R. Fattovich, A. Manzo and C. Perlingieri. (1997). Archaeological investigations at Bieta Giyorgis (Aksum), Ethiopia: 1993-1995 Field Seasons. Journal of Field Archaeology 24, 387-403.

Bar-Yosef, O. and M.E. Kilsev. (1989). Early farming communities in the Jordan Valley. In (D.R. Harris and G.C. Hillman, eds.) Foraging and Farming. London: Unwin Hyman, pp. 632-642.

Bauer, D.F. (1975). For want of an ox...: land, capital and social stratification in Tigre. Proceedings of the First United States Conference on Ethiopian Studies. African Studies Centre, Michigan State University, East Lansing, pp. 235-248.

Bauer, D.F. (1977). Household and Society in Ethiopia. Occasional Papers No. 6, African Studies Centre, Michigan State University, East Lansing.

Bauer, D.F. (1987). The dynamics of communal and hereditary land tenure among the Tigray of Ethiopia. In (B. McCay and J. Acheson, eds.) The Question of the Commons: The Culture and Ecology of Communal Resources. Tucson: University of Arizona Press, pp. 217-230.

Bauer, D.F. (1989). The sacred and the secret: order and chaos in Tigray medicine and polities. In (W. Arens and I. Karp, eds.) Creativity of Power. Washington, D.C.: Smithsonian Institution Press, pp. 225-244.

Blier, S.P. (1987). The Anatomy of Architecture: Ontology and Metaphor in Batammaliba Architectural Expression. Cambridge: Cambridge University Press.

Brandt, S.A. (1980). Archaeological investigations at Lake Besaka, Ethiopia. In (R.E. Leakey and B.A. Ogot, eds.) Proceedings of the 8th Panafrican Congress of Prehistory and Quaternary Studies. Nairobi: International Louis Leakey Memorial Institute for African Prehistory, pp. 239-243.

Brandt, S.A. (1984). New perspectives on the origins of food production in Ethiopia. In (J.D. Clark and S.A. Brandt, eds.) From Hunters to Farmers. Berkeley: University of California Press, pp. 173-190.

Brandt, S.A. (1996). The ethnoarchaeology of flaked stone use in southern Ethiopia. In (G. Pwiti and R. Soper, eds.) Aspects of African Prehistory. Harare: University of Zimbabwe Press, pp.733-738.

Brandt, S.A. and N. Carder. (1987). Pastoral rock art in the horn of Africa: making sense of udder chaos. World Archaeology 19, 195-213.

Brumfiel, E.M. (1991). Weaving and cooking: women's production in Aztec Mexico. In (J.M. Gero and M. Conkey, eds.) Engendering Archaeology: Women in Prehistory. Oxford: Blackwell, pp. 224-251.

Bulliet, R.W. (1975). The Camel and the Wheel. Cambridge: Harvard University Press.

Butler, E.A. (1992). Pulse agronomy: traditional systems and implications for early agriculture. In (P. Anderson-Gerfaud, ed.) Préhistoire de l'Agriculture; Nouvelles Approches, Expérimentales et Éthographiques. Paris: Éditions du CNRS, pp. 67-78.

Butler, E.A. (1995). The role of small legumes as a resource in prehistory with particular reference to temperate Africa. Acta Palaeobotanica 35(1), 105-115.

Butler, E.A. (in press). Traditional seed cropping systems in the temperate Old World: Models for Antiquity? In (C. Gosden and J.G. Hather, eds). Change in Subsistence Systems: Social Theory and Biological Processes. London: Routledge.

Chiovenda, E. (1928). Le piante alimentari nelle nostre colonie. Atti della Societa Italiana per il Progresso delle Scienze 17, 543-558.

Ciferri, R. and G.R. Giglioli (1939a). I frumenti dell'Africa orientale studiati su materiali originali. Vol. I, I Cereali dell'Africa Italiana. Florence: Regio Istituto Agronomico per l'Africa Italiana.

Ciferri, R. and G.R. Giglioli. (1939b). La cerealicoltura in Africa orientale, I-IV. L'Italia agricola 76, 247-257; 379-387; 765-774; 837-844.

Ciferri, R. and I. Baldrati. (1939). Il 'Teff' (Eragrostis Teff). Cereale da panificazione dell'Africa Orientale Italiana montana. Vol. II, I Cereali dell'Africa Italiana. Florence: Regio Istituto Agronomico per l'Africa Italiana.

Ciferri, R. and I. Baldrati. (1940). La cerealicoltura in Africa orientale. VI. Il 'Teff' (Eragrostis Teff). L'Italia agricola 77, 170-176.

Clark, J.D. (1976). The domestication process in sub-Saharan Africa with special reference to Ethiopia. Proceedings of the 9th Congress of the U.I.S.P.P., Nice 1976, pp. 56-115.

Clark, J.D. (1980). The origins of domestication in Ethiopia. Proceedings of the 8th Panafrican Congress of Prehistory and Quaternary Studies. Nairobi: International Louis Leakey Memorial Institution for African Prehistory, pp. 268-270.

Clark J.D. (1988). A review of the archaeological evidence for the origins of food production in Ethiopia. Proceedings of the 8th International Conference on Ethiopian Studies, Volume 1. Addis Ababa: Institute of Ethiopian Studies, pp. 55-69.

Clark, J.D. and G.R. Prince. (1978). Use-wear in later Stone Age microliths from Laga Oda, Haraghi, Ethiopia and possible functional interpretations. Azania 13, 101-110.

Clark, J.D. and H. Kurashina. (1981). A study of the work of a modern tanner in Ethiopia and its relevance for archaeological interpretation. In (R.A. Gould and M.B. Schiffer, eds.) Modern Material Culture. New York: Academic Press, pp. 303-321.

Clark, J.D. and M.A.J. Williams. (1978). Recent archaeological research in southeastern Ethiopia (1974-1975): some preliminary results. Annales d'Ethiopie 11, 19-44.

Close, A. (1995). Few and far between: early ceramics in North Africa. In (W.K. Barnett and J.W. Hoopes, eds.) The Emergence of Pottery Technology and Innovation in Ancient Societies. Washington, D.C.: Smithsonian Institution Press, pp. 23-37.

D'Andrea, A.C. (1992). Palaeoethnobotany of later Jomon cultures of northeastern Japan. Ph.D. Dissertation, University of Toronto, Canada. Ann Arbor, MI: University Microfilms International.

D'Andrea, A.C. (1995). Later Jomon subsistence in northeastern Japan: New Evidence from Palaeoethnobotanical Studies. Asian Perspectives 34(2), 195-227.

D'Andrea, A.C. (1997). Archaeobotanical remains from Bieta Giyorgis excavations, Aksum, 1996-1997: preliminary report. Unpublished MS.

D'Andrea, A.C. and Mitiku Haile. (in prep.). Emmer processing: A case study from northeastern Ethiopia.

D'Andrea, A.C., Mitiku Haile, A.E. Butler and D.E. Lyons. (1997). Ethnoarchaeological research in the Ethiopian highlands. Nyame Akuma, 47, 19-26.

de Boef, W., K. Amanor and K. Wellard, eds. (1993). Cultivating Knowledge. London: Intermediate Technology Publications.

Dennel, R.W. (1972). The interpretation of plant remains: Bulgaria. In (E.S. Higgs, ed.) Papers In Economic Prehistory. Cambridge: Cambridge University Press, pp. 149-159.

Dennel, R.W. (1974). Botanical evidence for prehistoric crop processing activities. Journal of Archaeological Science 1, 275-284.

Dennel R.W. (1976). The economic importance of plant resources represented on archaeological sites. Journal of Archaeological Science 3, 229-247.

DiBlasi, M.C. (1996). Results of the Analysis of Archaeological Pollen Samples from Bieta Giyorgis, Aksum. Unpublished report submitted to the Ministry of Information and Culture and Ministry of Mines, Addis Ababa.

Dombrowski, J.C. (1970). Preliminary report on excavations in Lalibela and Natchabiet caves, Begemeder. Annales d'Ethiopie 8, 21-29.

Dombrowski, J.C. (1971). Excavations in Ethiopia: Lalibela and Natchabiet Caves, Begemeder Province. Unpublished Ph.D. Dissertation, Boston University.

Donley-Reid, L.W. (1990). A structuring structure: the Swahili house. In (S. Kent, ed.) Domestic Architecture and the Use of Space. Cambridge: Cambridge University Press, pp. 114-126.

Ehret, C. (1974). Ethiopians and East Africans: The Problem of Contacts. Nairobi: East African Publishing House.

Ehret, C. (1979). On the antiquity of agriculture in Ethiopia. Journal of African History 20, 161-177.

Ehret, C. (1982). The first spread of food production to southern Africa. In (C. Ehret and M. Posnansky, eds.) The Archaeological and Linguistic Reconstruction of African History. Berkeley: University of California Press, pp. 158-181.

Ehret, C. (1984). Historical/linguistic evidence for early African food production. In (J.D. Clark and S.A. Brand, eds.) From Hunters to Farmers. Berkeley: University of California Press, pp. 26-35.

Engels, J.M.M., J.G. Hawkes and M. Worede. (1991). Plant Genetic Resources of Ethiopia. Cambridge: Cambridge University Press.

Fattovich, R. (1975). The contributions of the Nile Valley's cultures to the rising of the Ethiopian civilisation: elements for a hypothesis of work. Meroitic Newsletter 16, 2-8.

Fattovich, R. (1977). Some data for the study of the culture history in ancient Northern Ethiopia. Nyame Akuma 10: 6-18.

Fattovich, R. (1988). Remarks on the late prehistory and early history of northern Ethiopia. Proceedings of the 8th International Conference on Ethiopian Studies, Volume 1. Addis Ababa: Institute of Ethiopian Studies, pp. 85-104.

Fattovich, R. (1990). Remarks on the Pre-Aksumite period in northern Ethiopia. Journal of Ethiopian Studies 23, 1-33.

Fattovich, R. (1995). Archaeological excavations at Bieta Giyorgis (Aksum, Tigray): a preliminary report on the 1994 field season. Nyame Akuma 43, 34-37.

Gallagher, J.P. (1977). Contemporary stone tools in Ethiopia. Journal of Field Archaeology 4, 407-414.

Goe, M. (1990). Tillage with the traditional maresha in the Ethiopian highlands. Tools and Tillage 6, 127-156.

Harlan, J.R. (1969). Ethiopia: a centre of diversity. Economic Botany 23, 309-314.

Harlan, J.R. (1971). Agricultural origins: centres and non-centres. Science 174, 468-474.

Harlan, J.R. (1982). The origins of indigenous African agriculture. In (J.D. Clark, ed.) Cambridge Ancient History of Africa, Vol. 1. Cambridge: Cambridge University Press, pp. 624-657.

Harlan, J.R. (1989). The tropical African cereals. In (D.R. Harris and G.C. Hillman, eds.) Foraging and Farming. London: Unwin Hyman, pp. 335-343.

Harlan, J.R. (1992). Indigenous African agriculture. In (P.J. Watson and C.W. Cowan, eds.) Agricultural Origins in World Perspective. Washington D.C.: Smithsonian Institution Publications in Anthropology, pp. 59-69.

Harlan, J.R. (1993). The tropical African cereals. In (T. Shaw, P. Sinclair, B. Andah and A. Okpoko, eds.) The Archaeology of Africa. London: Routledge, pp. 53-60.

Harris, D.R. (1984). Ethnohistoric evidence for the exploitation of wild grasses and forbs: Its scope and archaeological implications. In (W. van Zeist and W.A. Casparie, eds.) Plants and Ancient Man. Rotterdam: A.A. Balkema, pp. 63-69.

Hastorf, C.A. (1991). Gender, space, and food in prehistory. In (J.M. Gero and M. Conkey, eds.) Engendering Archaeology: Women in Prehistory. Oxford: Blackwell, pp. 132-159.

Higuchi, T. (1986). Relationships between Japan and Asia in ancient times. In (R.J. Pearson, ed.) Windows on the Japanese Past. Ann Arbor: Centre for Japanese Studies, University of Michigan, pp. 121-126.

Hillman, G.C. (1973). Crop husbandry and food production: modern models for the interpretation of plant remains. Anatolian Studies 23, 241-244.

Hillman, G.C. (1984). Interpretation of archaeological plant remains: ethnographic models from Turkey. In (W. van Zeist and W.A. Casparie, eds.) Plants and Ancient Man. Rotterdam: A.A. Balkema, pp. 1-42.

Hillman, G.C. and M.S. Davies. (1990). Measured domestication rates in wild wheats and barley under primitive cultivation and their archaeological implications. Journal of World Prehistory 4(2), 157-222.

Hilu, K.W. and J.M.J. De Wet. (1976). Domestication of Eleusine coracana. Economic Botany 30, 199-208.

Hilu, K.W., J.M.J. De Wet and J.R. Harlan. (1979). Archaeobotanical studies of E. coracana ssp. coracana (finger millet). American Journal of Botany 66(3), 330-343.

Hodder, I. (1994). Architecture and meaning: the example of Neolithic houses and tombs. In (M.P. Pearson and C. Richards, eds.) Architecture and Order: Approaches to Social Space. London: Routledge, pp. 73-86.

Holt, J. and M. Lawrence. (1993). Making Ends Meet: A Survey of the Food Economy of the Ethiopian North-East Highlands. London: Save the Children.

Huffnagel, H.P. (1961). Agriculture in Ethiopia. Rome: FAO.

Hunting Technical Services Limited. (1976). Tigrai Rural Development Study. Addis Ababa: Relief and Rehabilitation Commission.

Jones, G.E.M. (1984). Interpretation of archaeological plant remains: ethnographic models from Greece. In (W. van Zeist and W.A. Casparie, eds.) Plants and Ancient Man. Rotterdam: A.A. Balkema, pp. 43-62.

Jones, G.E.M. and P. Halstead. (1995). Maslins, mixtures, and monocrops: on the interpretation of archaeobotanical crop samples of heterogeneous composition. Journal of Archaeological Science 22, 103-114.

Ladizinsky, G. (1975). Oats in Ethiopia. Economic Botany 29, 238-241.

Kajale, M.D. (1988). Ancient plant economy at Chalcolithic Juljapur Garhi, District Amraoti, Maharashtra. Current Science 57, 377-379.

Kang, I.K., M.K. Yi, Y.H. Han and K.S. Yi. (1979). Songguk-ni Site, Volume 1. Report of the Research of Antiquities of the National Museum of Korea, Volume 11. Seoul: Kungnip Chungang Pangmulgwan.

Kim, W.Y., H.J. Im and M.I. Choi. (1978). The Hunamni Site - A Prehistoric Village Site on the Han River: Progress Reports, 1976, 1977. Archaeological and Anthropological Papers of the Seoul National University, Vol. 8. Seoul: Seoul University Museum and Department of Archaeology.

Lyons, D.E. (1989). Men's houses, women's spaces: the spatial ordering of households in Doulo, north Cameroun. In (S. MacEachern, J.W. Archer and R.D. Garvin, eds.) Cultures in Conflict. Calgary: Archaeological Association of the University of Calgary, pp. 28-34.

Lyons, D.E. (1996). The politics of house shape: round vs. rectilinear-shaped domestic structures in Dela households, northern Cameroon. Antiquity 70, 341-367.

Lyons, D.E. (in prep). Men's Houses, Women's Spaces: An Ethnoarchaeological Study of Gender and Household Style in Dela, North Cameroon.

MacEachern, S. (1995). Foreign countries: the development of ethnoarchaeology in sub-Saharan Africa. Journal of World Prehistory 10(3), 243-304.

Marcus, H.G. (1994). A History of Ethiopia. Berkeley: University of California Press.

McCann, J.C. (1989). From Poverty to Famine in Northeast Ethiopia. Philadelphia: University of Pennsylvania Press.

McCann, J.C. (1995). People of the Plow: An Agricultural History of Ethiopia 1800-1995. Madison: University of Wisconsin Press.

Mehra, K.L. (1962). Natural hybridisation between Eleusine coracana and E. africana in Uganda. Journal of the Indian Botanical Society 41, 531-539.

Mehra, K.L. (1963). Differentiation of the cultivated and wild Eleusine species. Phyton 20, 189-198.

Melaku Worede and Hailu Mekbib. (1993). Linking genetic resources to farmers in Ethiopia. In (W. de Boef, K. Amanor and K. Wellard, eds.) Cultivating Knowledge. London: Intermediate Technology Publications, pp. 78-84.

Messing, S.D. (1957a). Further comments on resin-coated pottery: Ethiopia. American Anthropologist 59, 134.

Messing, S.D. (1957b). The Highland Amhara of Ethiopia. Ph.D. Dissertation, University of Pennsylvania, USA.

Michels, J.W. (1990). Excavations at Aksum. The African Archaeological Review 8, 177-188.

Moore, H. (1986). Space, Text and Gender: An Anthropological Study of the Marakwet of Kenya. Cambridge: Cambridge University Press.

Murdock, G.P. (1959). Africa: Its Peoples and Their Culture History. New York: McGraw-Hill.

Nesbitt, M. and D. Samuel. (1996). From staple crop to extinction? The archaeology and history of the hulled wheats. In (S. Padulosi, K. Hammer and J. Heller, eds.) Hulled Wheats. Rome: International Plant Genetic Resources Institute, pp. 41-100.

Phillips, S. (1995). Flora of Ethiopia and Eritrea. Volume 7. Poaceae (Gramineae). Addis Ababa: National Herbarium, University of Addis Ababa.

Phillipson, D.W. (1993a). The antiquity of cultivation and herding in Ethiopia. In (T. Shaw, P. Sinclair, B. Andah and A. Okpoko, eds.) The Archaeology of Africa. London: Routledge, pp. 344-357.

Phillipson, D.W. (1993b). The 1993 excavations at Aksum. In (H.G. Marcus, ed.) Papers of the 12th International Conference of Ethiopian Studies. New Jersey: The Red Sea Press, Inc, pp. 84-96.

Phillipson, D.W. (1994). The B.I.E.A. Aksum Excavations, 1993. Nyame Akuma 41, 15-25.

Phillipson, D.W. (1995). The B.I.E.A. Aksum Excavations, 1994. Nyame Akuma 44, 28-34.

Phillipson, D.W. (1996). The B.I.E.A. Aksum Excavations, 1995. Nyame Akuma 46, 24-33.

Porteres, R. 1962. Berceaux agricoles primaires sur le continent africain. Journal of African History 3, 195-210.

Reddy, S.N. (1991). Complementary approaches to Late Harappan subsistence. In (R.H. Meadow, ed.) Harappa Excavations 1986-1990. Madison: Prehistory Press, pp. 127-135.

Roberts, M. (1991). Living in a Man-Made World. London: Routledge.

Rowley-Conwy, P. (1991). Sorghum from Qasr Ibrim, Egyptian Nubia, c. 1800 B.C.-A.D. 1811: A Preliminary Study. In (J.M. Renfrew, ed.) New Light on Early Farming: Recent Developments in Palaeoethnobotany. Edinburgh: Edinburgh University Press, pp. 191-212.

Seligman, C.G. (1957). Races of Africa. Oxford: Oxford University Press.

Simon Adebo (1993). Report of Diagnostic Survey of Debri Tabia in Enderta Wereda. Internal Report. Addis Ababa: FARM Africa.

Simoons, F.J. (1960). Northwest Ethiopia. Madison: University of Wisconsin Press.

Slicher van Bath, B.H. (1963). The Agrarian History of Western Europe AD 500-1850. London: Edward Arnold.

Soderstrom, T.R. (1969). Impressions of cereals and other plants in the pottery of Hajar bin Humeid. In (G.W. van Beek, ed.) Hajar bin Humeid: Investigations at a Pre-Islamic Site in South Arabia. Baltimore: Johns Hopkins University Press, pp. 398-407.

Tringham, R.E. (1991). Households with faces: the challenge of gender in prehistoric architectural remains. In (J.M. Gero and M. Conkey, eds.) Engendering Archaeology: Women in Prehistory. Oxford: Blackwell, pp. 93-131.

Tsukada, M. (1986). Vegetation in prehistoric Japan. In (R.J. Pearson, ed.) Windows on the Japanese Past. Ann Arbor: Centre for Japanese Studies, University of Michigan, pp. 11-56.

Vavilov, N.I. (1926). Studies on the Origin of Cultivated Plants. Leningrad: Institute of Applied Botanical Plant Breeding.

Vavilov, N.I. (1951). The Origin, Variation, Immunity, and Breeding of Cultivated Plants. New York: Ronald Press.

Wall, D.D. (1994). The Archaeology of Gender Separating the Spheres in Urban America. New York: Plenum Press.

Westphal, E. (1974). Pulses in Ethiopia. Wageningen: Centre for Agricultural Publishing and Documentation.

Westphal, E. (1975). Agricultural Systems in Ethiopia. Wageningen: Centre for Agricultural Publication and Documentation.

Wilke, P.J., R. Bettinger, T.F. King and J.F. O'Connell. (1972). Harvest selection and domestication in seed plants. Antiquity 46, 203-208.

Wylie, A. (1985). The reaction against analogy. In (M.B. Schiffer, ed.) Advances in Archaeological Method and Theory, Volume 8. New York: Academic press, pp. 63-111.

Yabuno, T. (1987). Japanese barnyard millet (Echinochloa utilis, Poaceae) in Japan. Economic Botany 41(4), 484-493.

Yim, H. (1978). The Hunam-ni Dwelling Site, Report 4. Archaeological and Anthropological Papers of the Seoul National University, Vol. 8. Seoul: Seoul University Museum and Department of Archaeology.

Zeven, A.C. and P.M. Zhukovsky. (1975). Dictionary of Cultivated Plants and Their Centres of Diversity. Wageningen: Centre for Agricultural Publishing and Documentation.