Introduction

This article presents a newly discovered rock art site, Dhambalin, in Somaliland. It results from perhaps the first ever survey initiated and led by a Somali archaeologist in a Somali country and it was carried out from September to December 2007. This survey resulted in the findings of many different types of archaeological sites, including rock art sites in the provinces of Hargeysa, Togdheer and Sanaag. However, this article is mainly concerned with the unique site of Dhambalin rock shelter. It is a sand stone rock with polychrome paintings including the first sheep depictions in Somali archaeology as well as extraordinary bovine paintings. The site is located about 60 km from Berbera towards the east and 30 km from the coast of the Red Sea into the hinterland of Togdheer region. The nearest town is called Beenyo Dhaadheer (see Fig. 1, map).

Fig. 1
figure 1

Map of Somaliland showing the location of the site of Dhambalin

Full size image

In this article I describe preliminarily the paintings of Dhambalin rock shelter (see Fig. 2). The polychrome paintings of sheep have characteristic rump and muzzle. There is an extraordinary originality in the paintings of the sheep and goat compared to the sites with the Horn of Africa’s sheep depictions in Ethiopia. There is an extraordinary quality of colours and styles of sheep and goat, cattle and hunting scenes with human figures with bow and arrows. Some human figures have arms stretched out and what appear to be masks on their heads. The richness of fauna depicted includes antelopes, dogs, giraffes, snakes and a turtle. I suggest a relative dating is possible perhaps through the superimposition of various styles of paintings that might be of the similar styles of the known Neolithic or protohistoric rock art in the Horn of Africa.

Fig. 2
figure 2

The Dhambalin sandstone rock shelter

Full size image

Hence, the paintings of Dhambalin seem to belong to what has been termed Arabian–Ethiopian Style (Červiček 1971). Furthermore, since there appear to be similarities as well differences in pastoral rock art of North East Africa including Nubia and the Horn, I also call for collaborative projects between scholars active in the Sahara, Nubia the Horn and Arabia in order to generate a holistic picture of particularly the symbolic significance of rock art of this region of the world.

The Discovery of the Site and the Naming

Two inhabitants of Beenyo Dhaadheer, a small rural nomadic settlement 60 km east of Berbera, on the Red Sea coast, who were staying at the time in Hargeysa, the capital of Somaliland, saw my heritage awareness public educational programs on Somaliland TV at the time when I was the Director of the Archaeology Department that I helped to establish within the Ministry of Tourism and Culture. The men from Beenyo came to my office and insisted I go and look for a dhambalin with rock art paintings near their settlement. I went to the site as part of my survey on ritual archaeological landscapes.

The word dhambalin describes the shape of the rock and in Somali it means ‘half mountain cut vertically’ due to the peculiar shape of the mountain with the paintings on the façade of the vertical side of the mountain.

Geomorphology of the Berbera Area and Paleoenvironments of the Horn of Africa

Dhambalin is a rock art site located about 60 km from the coast of the Red Sea within a few kilometres from a small settlement called Beenyo Dhaadheer. The geomorphology of coastal Berbera, has been classified as Quaternary surfaces from quaternary cycles of erosion resulting from the coast into the escarpment and into the hills (Pallister 1963). The coast east of Berbera is thought to be possibly from Pliocene times as it is characterized by succeeding uplifts (Hunt 1951, and Pallister 1963, p. 184–6). Early Holocene environmental conditions worsened and around 7,000 years b.p., in the Sahara, Nile Basin and the Horn of Africa there were warm and humid conditions that seem to have occurred at the same time as farming and pastoral subsistence economy develops in the Sahara and Nile Valley (see Brandt and Carder 1987; Hassan 1985). According to the locals, there are constant sand dunes and it is a vast land which was scarcely populated with many rock shelters of sand stone. Furthermore, there seems to be an ongoing desertification of the area. Hence they keep very little livestock due to continuing droughts.

Description of Dhambalin

Dhambalin is a rock shelter with rock paintings. As a result of the great collection of sand heaps that cover most of the mountain, the lower part of the rock shelter is buried under the sand. The sandstone is so fragile that the paintings are deteriorating. The paintings suffer great corrosion. Furthermore, on the ground there is debris, broken off from the rock shelter perhaps due to wind erosion and porosity of the stone. Some painting residues still remain visible on the uncountable fragments of rock on the ground and in the sand.

However, there are small patches without sand; these are the places on the edges where the curving rock gives some shelter from the strong winds of sand dunes, that otherwise fill the shelter with sand. Other than that most of the site is open to the wind, since there is no overhang in this vertical wall. It is astonishing in fact that what remains of the delicate paintings on the wall actually still survive.

The Paintings of Dhambalin

General Remarks

The wall of Dhambalin is richly painted and because of the concentration of paintings on this wall, the image appears strikingly with a mosaic of colours (see Fig. 3). There are paintings only on one side of the shelter. The paintings cover about 4 m depth and 12 m width. Dhambalin is a site of successive layers of paintings. It was not possible to undertake measurements. The paintings depict among other things, human figures in hunting scenes with wild animals and herds of domestic animals. Some of the paintings have fallen off or faded away, however, it is still possible to deduce some of the shapes from the worn-out figure that still remain on parts of the rock.

Fig. 3
figure 3

The great panel of Dhambalin

Full size image

Regarding the origin of the colours used at Dhambalin, it was possible during the survey to pick up from the ground natural stones that were of the colours of yellow-brown, terracotta red, brown-red, and red. Such stones seem to have been employed in the paintings as well as white, black and a combination (grey) for the different paintings. Also at places, an earlier painting in white is visible under the new red and some times produces a pinkish colour. In terms of style and juxtaposition, it seems that sets of images that belong together can be identified. This might prove useful in the process of relative dating.

Human Figures

There are about eight–ten human figures visible. In addition, places where the actual paint has come off or faded, there are the worn away figures still visible on the rock. Scenes with human figures are dominated by herds of wild animals and hunters in action with bows and arrows. There are at least four hunting scenes with four hunters each holding what appear to be bow and arrows (see Fig. 4). One depicts a male hunter with bow and arrow, who seems to be wearing his hair down with headgear on top of the head. He is surrounded by two dogs and at least one antelope (see Fig. 5). Another human figure is surrounded by animals and holds what looks like a club in one hand and a long object in the other hand. One of the hunting scenes depicts two hunters together, one standing and another hunter sitting on an animal, perhaps a horse, and holding a bow and arrow in position to hit antelopes surrounding him Horses are still found/kept in Somaliland. I have seen and photographed some in Sanaag region, the north eastern part of Somaliland. There is also a human figure standing behind the rump of a bull which is also might have a symbolic significance.

Fig. 4
figure 4

Human figure in hunting scenes with wild animals such as antelopes and giraffes

Full size image
Fig. 5
figure 5

Human figure wearing headgear and carrying bow and arrow, hunting with dogs

Full size image

Two human representations have the male gender marked, and arms stretched out on the sides with a headgear or a mask. These resemble paintings of human representations from Saka Sharifa, Harar (Joussaume 1995: 51) with arms like wings in the air. Another human representation with what looks like some kind of a headgear, also arms stretched like the wings of a flying bird, but this figure lacks gender specification. Is this a dancer with a mask? There is also one depiction painted in white, which resembles a human (a child?) with a big head in relation to the body.

Wild Animals

Dhambalin paintings offer a unique variety of wild animals. There are at least eight giraffes, of different styles, sizes and colours. A couple of them painted in brown-red and one in red colour (see Fig. 4). There are also about four giraffes painted in dark brown colours accompanied by what seems to be a turtle painted in the same dark brown colour. The turtle has typical distinctive shell with pattern. A few of the giraffes are in conjunction with a hunting scene with other wild animals such as antelopes.

The Dhambalin site includes hunting scenes with small and large antelopes (see Fig. 4). Some are as big as the goat paintings. Others are bigger than the goat paintings and have large bodies. Some of the large ones are close to the size of some of the cattle paintings. One particular image depicts what looks like an oryx, a male, except the tail is perhaps too short and thin and also, unusually, the lower body is bigger than upper body. However, it has straight and long distinctive horns and there is another one similar to this antelope with also straight horns. One other large antelope, associated with a hunting scene, has horns similar to those of an ibex.

Other wild animals include one particularly striking image of what looks like a wild cat with multicoloured paintings such as red and white patches. However, the head is unique with what looks like rays coming out on all sides. Or it could be that these rays represent perhaps blood coming out from the neck of the headless body.

Also there seems to be lions painted in grey colours but also in red colours [though not very visible on the shot]. The dogs are painted in red or reddish brown colour with typical upright ears and tail (see Fig. 5). They are in connection with some of the hunting scenes described above.

There are also paintings of animals that remind of baboon(s) in white and red colour polychrome or either red or white with distinctive form of big belly which sloped-down upper body and with a characteristic head and up-pointing back-body with distinctive rump. Another painting shows similar figures next to each other in a row, one white in colour and two red. There are also two more a group of four at least with the same style but are bigger in the size. These possible baboons are clearly not part of a hunting scene.

There are serpents (free standing) with distinctive body, head and spiral and curling ends painted in red and white colours (see Fig. 6). Also here they come both in white and red.

Fig. 6
figure 6

Snakes, sheep and goat polychrome with symbolic elements

Full size image

Bovine

The representations of cattle, like most of the other animals, are in profile (see Fig. 7). They are painted in many different colours including red-brown, blood red, red-yellow, terracotta-red, also some pinkish colour resulting from superimposition the red with previous white paint. The size of the cattle varies and there are both big and small cattle. Most paintings depict female fully grown cattle with udder that are filled and some times with their calves close by. Stylistically, no distinguishable difference between depiction of a cow and a bull are demonstrable, except when the udders or other gender characteristics are clearly depicted. There is at least one painting of a bull with clear gender. There are two more possible bulls. All bovine depictions seem to possess tails. Some of the figures of the cattle are painted over one another; the old ones appearing behind the new ones. Curiously, the cattle have different styles of heads. There is only one cow with a big neck, while most of the cattle have a small neck or no neck at all. Some of the cattle are painted headless (see Fig. 7).

Fig. 7
figure 7

Polychrome paintings of bovines including headless ones

Full size image

I will discuss this further below in comparison with other archaeological and anthropological data of the Horn of Africa. However, first I introduce the foremost sheep paintings in Somali archaeology at Dhambalin, Somaliland.

First Evidence of Sheep in Somali Rock Art

Dhambalin depicts the sheep paintings (see Figs. 6 and 8) with polychrome painting of at least 14 clear sheep. The precision of the paintings is extraordinary. All the sheep possess the characteristic rumps, distinctive head and muzzle, and the relatively big bodies and slender legs. In contemporary Somali country there are the black-headed sheep exclusively. At Dhambalin, we see no black head on the sheep. Nevertheless, colours of white and red with decorative elements are present. Three sheep are painted in red colour with a white belt across the belly. Eleven others are painted in white colour and possess decorative elements of red. The forelimbs like the hind-limbs are painted separately. The limbs are slim nevertheless the fatness of these sheep is accentuated by their big rumps. Since the sheep from Dhambalin are depicted from the side, the ears are not marked out.

Fig. 8
figure 8

First sheep depictions in Somali archaeology at Dhambalin

Full size image

There are at least five goats of polychrome paintings in white or red. Three painted in white with the decorative elements in red colour. One of them has three (dotted) red lines painted on the neck (see Fig. 6) resembling necklaces. The fact one of the goats is depicted with ornaments of what appear to be necklaces makes the argument for a domestic goat rather than an antelope more convincing. It is also clear from the juxtaposition with sheep as well as the fact that some of the sheep and goat are painted in the same style. For example there are also two goats, which are painted in red with a white belt across the belly (see Fig. 6).

Hence, the goats are not to be mistaken for antelopes due to the context in juxtaposition with the sheep and the simple fact that they are associated with each other further through the painting styles with white belt or red patches. There is also one curious distinction relating to the painting of the legs of the goat; the forelimbs and the hind-limbs are fused. The paintings of sheep and goat seems to be earlier than at least some of the cattle painting since at places the paintings of the bovines are on top of the sheep and goat ones (see Fig. 6). This might indicate a long-term use of the site with succeeding various traditions and styles of paintings and perhaps cultures. Dhambalin style of sheep seems unique in the Horn.

Burial Grounds and Surface Material

A couple of burials stand less than 100 m from the paintings, which seem to be facing the burial ground at the site. The burial grounds might well be connected to the site. The types of burials are locally known as Arawelo and locals credit these burials with great antiquity, suggesting pre-Islamic times. Also more such burials are found a few hundred meters from the shelter.

Due to the sand dunes that cover the ground, there were no objects collected from the surface of the shelter. However, there were small pieces of stones that would also suggest that some of the red paintings are executed utilizing local coloured rocks. A future thorough survey of the area of the shelter might include surface collections of the areas possibly leading to more specific identification of the burials and settlement.

Dating

The paintings of bovines at Dhambalin seems to fit well in the descriptions of what is named the Ethiopian–Arabian style (Červiček 1971), a style that has been lengthily discussed in other contexts of the Horn (Červiček 1971, 1979; Graziosi 1964; Joussaume 1981; Brandt and Carder 1987). The rock art of the Ethiopian–Arabian style is dated cautiously to 5,000–3,000 years b.p. through comparison with Nubian C-Group pottery (Červiček 1979). Some of the cattle paintings at Dhambalin perhaps relate to the early phase of what is known as Dathami style Proper due to lack of neck and/or heads. Furthermore, similarly many of the cattle also relate to the “long-horned” humpless cattle in the Sorre–Hamakiya style. In places, the paintings are superposed one above the other. The successive strata of paintings might make it difficult to pin down the earliest time of inhabitation of the site. However, the styles of painting might be of use since these can be compared with other sites in the country and beyond, in the Horn of Africa.

Rock art sites are spread across a belt stretching from Somalia’s east highlands to the Harar region in Ethiopia and further to the west in Ethiopia (Brandt and Carder 1987; Gutherz et al. 2003). Karin Heegan has polychrome paintings of bovine, and was first mentioned by Clark (1954, 1972) and visited by Jönsson (1983). However, it was studied first by Brandt and Brook, who excavated it (Brandt et al. 1984; Brandt and Carder 1987). The dates from charcoal samples of the middle levels date to c. 1,600 and 2,100 b.p. (Brandt et al. ibid.). It is however possible as also suggested by others (see Gutherz et al. 2003) that these dates might not correspond to the earliest human settlement of Karin Heegan and the rock art. The well known difficulty in dating rock art remains a problem also for the Horn of Africa.

The paintings of the Horn of Africa as that of Somali region such as the extraordinary paintings of Karin Heegan (Brandt and Carder, ibid. p. 199) and the recently visited site Laas Geel (Gutherz et al. 2003: 235) are, too, of the style that is known as Arabian–Ethiopian style. However, as also pointed out by Brandt (1986), there are chronological problems relating to the lithics which need updating and also to the fact that we are lacking ceramics at rock shelter sites in the Horn of Africa. According to local informants, when it comes to Dhambalin, the area around the coast is known for good clay sources (cf. Pallister 1963: 186–187). I have found potsherds of ceramics including imported material at various settlement sites in Hargeysa and Togdheer regions and it seems possible to find also on rock shelter sites if we employ thorough survey and investigations beyond current single test pit excavations employed at rock art sites in the Horn. While pottery is used by farmers in the riverine region of Somalia, northern Somalia and Somaliland, however, containers of metal or of organic material including calabash is used.

Ritual and Symbolic Significance of Dhambalin Paintings

Some Curious Elements at Dhambalin and Stylistic Comparison with the Horn and Beyond

There seem to be symbolic meanings of the rock art of Dhambalin in general. For example, what does the white belt on the paintings of sheep and goats in red signify? There is also some sort of headless wild cat with rays surrounding what seems to be the neck. The Dhambalin human representations give the impression that only the action is significant, namely the hunting mostly. The human heads are not clear, and seem some times to wear what looks like headgear. It is curious that certain figures are painted perfectly, such as the sheep, while the human figures are painted with the least accuracy.

As noted earlier, the Dhambalin sheep forelimbs and hind-limbs are painted separately, while the forelimbs and the hind-limbs of the sheep painted at Laga Oda and Ourso in Harar region, Ethiopia, are each fused into one thick line like the goat limbs at Dhambalin and the depictions of cattle limbs in most of the Horn of Africa.

There are more similarities in the rock art of the Horn and that of the Sahara. As noted earlier there are headless cattle at Dhambalin (see Fig. 7). This element seems to be also found in other parts of Africa. Fekri Hassan also pointed out to me that there are also headless cattle at Gilf el-Kebir (Hassan, pers. com.) At the tenth International Congress of Egyptologists Luc Watrin, (with Khaled Saad and Emmanuelle Honoré) talked about new data on mythologies from the Egyptian Sahara including headless cattle (Watrin et al. 2008).

The white belt of the sheep and goat at Dhambalin might be compared to what looks like a white belt on the human figure and cattle painted at Jilib Rihin (see Fig. 9) and cattle at Haadh (see Fig. 10), two new unpublished rock art sites from my survey at Sanaag region in Somaliland in November 2007. At Jibil Rihin we have the human figure which bears similarities to, as Fekri Hassan pointed out to me later, the Predynastic female figurines from Nagada II with arms stretched above the head (for image of the ancient Egyptian figurine see Hassan 1998, 107).

Fig. 9
figure 9

A new rock art site of Jilib Rihin in Sanaag region showing headless bovine

Full size image
Fig. 10
figure 10

A new rock art site of Haadh in Sanaag showing bovines

Full size image

Furthermore, in the context of Somali archaeology itself, the bovine style of scene from the lower panel, as pictured by Brandt and Carder (1987: 203), shows polychrome painted cow and human figure with similar head and body shapes to those of the head of a decorated cow and human figure on the central shelter of the upper level as seen in Laas Geel (Gutherz et al. 2003: 231). Also the distinctive head of the Karin Heegan cow is evident also of the Las Geel cow. Also both cows have the large ‘neck’ or “plastron” and the human with the arms stretched out on the sides. However, Bouakaze-Khan (2002) suggested the neck lines (referred to by others as “plastron”) of the cow paintings in the Horn of Africa such as at Karin Heegan, are lines that mark fat folds in the neck.

Karin Heegan and Las Geel share great similarities in the rock art of Haadh [Haadh] and Jilib Rihin [Jilib Rihin1] in Sanaag region, Somaliland.

A Call for a Multi-disciplinary Approach to Religions and Ritual Landscapes in Africa

The Horn of Africa needs investigations into the symbolism of rock art to determine the nature of past belief systems and changes within these over time and space. Although African archaeology, like many other archaeologies, benefited from Hodder’s (1982) Symbols in Action, the archaeology of the Horn of Africa still lacks studies of symbolism of material culture and I would argue the symbolism of the rock art is almost untouched.

Rock art research (both in Africa and Europe) is in its infancy, and although a lot of important data has been collected, a lot of earlier work, (e.g. Breuil 1934) and later work (e.g. Joussaume 1981, 1995), have more often than not pursued a descriptive approach concerned with establishing a chronology for stylistic change (cf. Brandt and Carder 1987:194; Ucko and Rosenfeld 1967). Beyond this descriptive cultural historical approach other work has focused on the behavioural information that can be gained from pastoral rock art. In this approach Brandt and Carder (1987: 204) argue for a ‘culture ecological framework’ and ‘physical and biotic landscape’ and ‘biological requirements of cattle’ in order to gain insights into the human behavioural aspect of rock art (Brandt and Carder 1987).

African archaeology needs to be studied in its cultural context (Andah 1995). In the Horn of Africa, we need to move on from simply description and recording of styles of paintings, and the ecology and social behaviour when investigating rock art. I suggest we need to study the landscape as a whole including its ritual significance today. I argue this is possible through a multi-disciplinary approach to landscape including rock art of the Horn of Africa. Such a line of enquiry might be useful towards widening our perceptions and contributing to our holistic knowledge of the past. Hence it is crucial to pursue a holistic study covering archaeology, anthropology, historical linguistics, and ethnography to put together the puzzle of past.

Also Červiček’s association between rock art styles in Arabia and Ethiopia and Nubian C-Group pottery (1979) should be taken further in a more multidisciplinary and regional perspective. There is a need to carry out comparative studies of the Nile Valley, Western Sahara and the Horn of Africa as well as the other side of the Red Sea.

Conservation and Further Research

The site of Dhambalin, like many sites in Somaliland and Somalia (see Mire 2007a, b) are in danger. This particular site is mostly threatened because the vulnerable panels, without overhang, are open to the strong winds and sand dunes burying as well as eroding the paintings. There is great erosion caused not only by the wind but also by the friable state of the sandstone. Hence, the paintings are disappearing under the sand. The human threat is not as urgent at this site as in other rock art sites in Somaliland, since the site is located in an environmentally challenging and hence scarcely inhabited area. The Ministry of Tourism and Culture have employed a local man as the guardian of the site. He will visit the site regularly and will prevent any intrusion. However, the site needs urgent documentation in terms of tracing and drawing of the rock art. Dhambalin is a unique site and contributes exceptional data to the archaeology of the Horn of Africa and should be subject to urgent preservation measures in place in the near future.

As evident from the Dhambalin survey, no settlements associated with the rock art groups of this region have been found in the Horn except for the actual shelters and therefore it is difficult to pursue this connection further. There are some arguments for viewing some rock art sites as perhaps having a connection with megalithic burial sites in the Horn. Hassan Abdi in Harar, a dolmen site, dated to c. 1710–1275 BC, is claimed to be associated with the rock art site of Laga Oda on the basis of ceramic types and dating (Joussaume 1995:70). Hence, such a line of enquiry might be relevant. Future excavations of the Dhambalin site with surrounding burials of cairns might be a useful path for adding more to the understanding of the culture(s) and people(s) who painted the various paintings at Dhambalin. Also we need to study anthropologically ideas of what is actually depicted on the rock art from the current cultural context. Archaeological dates of locally found pottery must be employed where possible and we need to investigate potential affinities with the rock art.

Conclusions

The spectacular images of Dhambalin rock art site offer the first unambiguous evidence of paintings of sheep in Somalia and Somali archaeology. The sheep and goat decorations include what looks like belts and necklaces. There are wild animals equally painted with great care such as giraffes and baboons and humans in hunting scenes with antelopes. The representations of humans appear in hunting scenes and in curious positions with arms stretched up.

I suggest we also study the site of Dhambalin holistically, employing as many disciplines as possible, anthropology, ethnography, historical linguistics. Also as we study this site and other Horn of Africa rock art sites, I suggest we look at them in their context with the rest of the landscape. The site of Dhambalin offers great data for the Horn of Africa. However, without the urgent conservation measures that need to take place, the site might not survive the hostile environment that has had a devastating effect on it.