Tanambelo V.R. RASOLONDRAINY,
(Centre de Documentation et de Recherche sur l’Art et la Tradition Orale de Madagascar)
Abstract:
This article presents the result of an archaeological study on the first prehistoric rock paintings site found in Madagascar and the rest of islands of Western Indian Ocean south of the equator.
Based on systematic methods and techniques, the study provides groundbreaking archaeological evidences that contribute largely to the understanding of the prehistory of Madagascar which was not archaeologically known thitherto to have prehistoric rock paintings, nor human settlement prior to 8th century AD. A Libyco-Berber inscription was discovered on the rock paintings of Ampasimaiky Rock Shelter, in Southwestern Madagascar. Based on this writing -supported by ancient historical records, circumstantial archaeological and paleontological evidences, linguistic data, and to some extent DNA analysis results -, the present study provides minimum age falling in the bracket time around 500 BC to roughly 8th century AD for the rock paintings of Ampasimaiky rock shelter. Furthermore, the study traces back early contacts between Madagascar and Northern Africa during prehistoric times.
Introduction
With the kind assistance of Dr Chantal Radimilahy, National Coordinator of the Network of African Archaeologists in Madagascar, I researched the ethnographic exhibition on the website of the Museum of Ethnography in Geneva (MEG). The theme was the peoples and cultures of the southwestern region of Madagascar, and was based on an exhibition of photographs taken by the French ethnographer Jacques Faublée in 1939. The pictures included rock paintings found in caves and rock shelters in the Isalo region. After discussions with my teachers at the Archaeology Unit of the University of Dar es Salaam, Professors Felix Chami, Fidelis Masao, and Audax Mabulla, I further investigated the significance of these paintings for the prehistoric period of Madagascar. This study is taken from a Master of Arts in Archaeology entitled
“Archaeologial study of the prehistoric rock paintings of Ampasimaiky rock shelter, in the upper Onilahy, Isalo region, Southwestern Madagascar”.
Maps of southern and southwestern Madagascar locating the study site and other archaeological and paleontological sites
Africa has many archaeological sites with prehistoric rock paintings (see Coulson and Campbell 2001; Willcox 1984). However, no such sites have yet been scientifically
scientifically reported from Madagascar, which is only 400 kilometres from southern Africa where they are abundant. All rock art previously found in Madagascar has been judged to have been produced in the recent past (Faublée 1947: 164; Hébert 1971: 218). This judgement is based on the theory that Africans were incapable of having navigated the high seas during prehistoric times (Donque 1965: 43; Grandidier et al. 1902; Southall 1975: 194; Vérin 1975, 1981). It was therefore out of the question to think that they had the possibility of transferring the practice of rock painting to the adjacent islands of Africa such as Madagascar. However, recent archaeological findings have demonstrated that the east coast of Africa and adjacent islands were already in contact since the prehistoric period (see Chami and Kwekason 2003; Chami et al. 2009; Juma 2004; Kessy 2009a,
2009b). These prompt a problem that led me to revise the antiquity of the rock paintings in Madagascar.
On the other hand, the name of the Arabic-Malagasy script – Sorabe meaning “great script” – indicates that there was an ancient script in Madagascar that was distinguished from Arabic (Hébert,
1967:XLVII; Huygues-Belrose 1997: 8). Although Hébert (1967:XLVIII) presumed that such a hypothetical script would have to be of Indian origin and would be discovered from
Although Hébert (1967:XLVIII) assumed that such a hypothetical script would have to be of Indian origin and would be discovered from stele and parietal inscriptions, no such inscription has so far been found in Madagascar.
My problem was therefore to find out whether this would appear on the rock paintings reported by the website. What provoked this idea the most was that Chami (2006:84-85; 2008) argued that the African Geometric and Amorphous Schematic (GAS) rock paintings and engravings are ancient writing.
Another problem is that radiocarbon 14 dating placed the earliest human settlement of Madagascar around the 8th century AD (Allibert 2008: 7; Dewar 1996: 474).
Although earlier dates are still being considered, many researchers have assumed that settlement of Madagascar would not go beyond the year W AD (for more information see Dewar 1997). This assumption has made research on Malagasy material culture going back to the Stone Age more than difficult. Any findings of material cultures dating from prehistoric times have been discarded, overlooked or misinterpreted. However, subfossil animal bone remains bearing traces of butchery found in paleontological sites in the southwestern region indicate the presence of human activity
(For more information, see Blench 2007; Perez et al. 2005). Despite this evidence, an archaeological site has not yet been
archaeological site with similar dates has not yet been found on the entire island.
With the discovery of a rock shelter with rock paintings, discovering new information about the prehistory of Madagascar before the 8th century proved to be feasible. All the more so as some of the paintings and engravings are
Especially since some of the SGA-type rock paintings and engravings found on the African continent – similar to those found in the photos on the MEG website – have been radiocarbon dated from the 1st millennium BC to the early 1st millennium AD (Chami, 2008: 54; Hall and Smith 2000; Lynche and Donahue 1980; Willcox 1984: 26, 83.
Fieldwork and results
Despite various challenges and constraints – mainly caused by the lack of archaeological reconnaissance prior to the fieldwork, insecurity in the study area, and the sacredness of the site to be studied – we had only somehow managed to conduct survey, archaeological excavation, and documentation of the paintings in the Ampasimaiky rock shelter. It is only the result of the latter that will be presented in this article.
The rock painting site is a rock shelter located at the foot of a hill immediately next to a small stream called Ampasimaiky, from which the name of the site derives. The presence of this water source may have influenced the painter’s choice of location.
The shelter itself is 26.8m long (south-north axis) and 5m wide (inwards).
Its ceiling begins to slope inwards from 4m high at its entrance to approximately 1.5m high at its irregular wall. The deepest part of the rock shelter – a chamber – is currently used by the royal family “Bara Zafimanely” as a burial place. Dung and ox hoof prints found on the ground indicate that cattle frequent the place. This occurrence should have disturbed the soil.
However, the areas around the site are dominated by rocky hills covered with grass. Only in the deep canyon near the confluence of Ampasimaiky and the Sakamarekely River is a patch of thick vegetation up to 10 m high observed along the stream. This situation would make agricultural activity almost impossible around the site, unless one had to move 1.5 km to the bank, where rice cultivation takes place today.
An archaeological survey of the southern Isalo region (Figure 1) was carried out and a rock painting site was discovered. Professor Chami participated in the fieldwork, advising on the approach to be followed in the treatment of these previously unstudied cave cultures in Madagascar.
Madagascar. History students from the University of Toliara, Mex Tafitasoa, Nadya
Ravololonirina, and Norbert Antilahy, and some villagers participated in the work. The Network of African Archaeologists in Madagascar, under the direction of Dr Chantal Radimilahy and Prof Barthélemy Manjakahery, provided the logistical equipment used. The various signatures The various signatures of national and local administrative approval, the security provided by the Gendarmerie Brigade of Benenitsy, and above all the kindness and openness of King Zafimanely who allowed us access to the (burial) site – contributed enormously to the accomplishment of the expedition.
Madagascar. History students from the University of Toliara, Mex Tafitasoa, Nadya
Ravololonirina, and Norbert Antilahy, and some villagers participated in the work. The Network of African Archaeologists in Madagascar, under the direction of Dr Chantal Radimilahy and Prof Barthélemy Manjakahery, provided the logistical equipment used. The various signatures The various signatures of national and local administrative approval, the security provided by the Gendarmerie Brigade of Benenitsy, and above all the kindness and openness of King Zafimanely who allowed us access to the (burial) site – contributed enormously to the accomplishment of the expedition.
The rock shelter at Ampasimaiky faces east. This arrangement makes the painting panels exposed to the sun in the morning and in the shade beyond noon. This arrangement suggests that painting activities are conducted either early in the morning while the heat of the sun is still moderate, or beyond noon when shade is available below the shelter.
This is based on the fact that the height of the panels does not exceed 130 cm, suggesting that the paintings were intended for people sitting on the floor of the undercroft shelter.
Minimal, therefore, is the likelihood of the painting activity – involving seated people – having been conducted during the time when the sun is heating the shelter floor.
However, the view outside the site is limited. The opposing hills – on the other bank of the Ampasimaiky stream – prevent a person standing under the shelter from having a good view of the surroundings . It is much more likely that the view outwards – which implies a defensive strategy (Smith 1995) – was not among the painter’s concerns when he/she chose the site. This demonstrates that the layout of the Ampasimaiky rock shelter favours invaders from unexpectedly attacking its residents. The place would therefore not be suitable for permanent habitation.
Especially since a person standing on the hills of the other bank would be able to look directly at the other one positioned under the rock shelter, unless he hides in the deeper part of the shelter. Since the wall of this “chamber”, according to the local population, is not painted, it is plausible that the activities related to the paintings in the Ampasimaiky rock shelter are not clandestine. However, the paintings were not intended to be seen by people passing close to the site. They are not seen from a distance. Not only are the panels placed below 130 cm in height, but the slope of the ground before the access
to the shelter also prevents anyone from seeing them until they get close to the entrance. Nevertheless, the use of darker coloured paints (red and black) suggests the painter’s intention to contrast the paints he/she was using against the pale colour of the sandstone wall. This implies his concern for the visibility of his paintings.
The walls of the rock shelter have red, reddish orange, burgundy red, black, and white paintings in monochrome, bichrome and polychrome patterns. The images
are painted with the fingertip. They depict zoomorphic (mainly zebu) and anthropomorphic (semi-naturalist and stylised) figures, as well as schematic-geometric-amorphic symbols (Figure 3). The latter dominate and are represented by quadrangular, circular and elliptical signs, dot/dash lines, and alphabet-like signs. In total, the Ampasimaiky rock paintings have 166 recognisable images.
Figure 3 : Typologie de forme des images dans les peintures rupestres d’Ampasimaiky.
Analysis and Interpretation
Based on comparative studies of the images, it has been found that the images of the Ampasimaiky rock paintings are not different from the images depicted on the prehistoric SGA rock art of the African continent (see Chami, 2008; Chaplin, 1974; Coulson and Campbell 2001;
Eastwood and Smith, 2005; Lynch and Donahue, 1980; Mabulla, 2005; Smith, 1995; Teka 2008; Willcox 1984). The similarities apparently suggest links between the two regions (Madagascar and the mainland) in the
(Madagascar and the continent) in prehistoric times. This suggests that the Ampasimaiky rock paintings are of African origin and could be relatively dated to the same period as those on the continent (Thackeray, 1983: 22). Comparative studies of the images have also shown that more than 30 types of images in the Ampasimaiky rock paintings are closely related to different ancient scripts such as Libyco-Berber, Protosinaitic, Meroitic (cursive), Punic Iberian, Ancient Phoenician, Tifinagh, and Aegean, Cretan and Proto-Egyptian linear signs respectively (Figure 4). These similarities would also suggest that the SGA-type rock paintings of Ampasimaiky are ancient writing.
Analysis and Interpretation
Based on comparative studies of the images, it has been found that the images of the Ampasimaiky rock paintings are not different from the images depicted on the prehistoric SGA rock art of the African continent (see Chami, 2008; Chaplin, 1974; Coulson and Campbell 2001;
Eastwood and Smith, 2005; Lynch and Donahue, 1980; Mabulla, 2005; Smith, 1995; Teka 2008; Willcox 1984). The similarities apparently suggest links between the two regions (Madagascar and the mainland) in the
(Madagascar and the continent) in prehistoric times. This suggests that the Ampasimaiky rock paintings are of African origin and could be relatively dated to the same period as those on the continent (Thackeray, 1983: 22). Comparative studies of the images have also shown that more than 30 types of images in the Ampasimaiky rock paintings are closely related to different ancient scripts such as Libyco-Berber, Protosinaitic, Meroitic (cursive), Punic Iberian, Ancient Phoenician, Tifinagh, and Aegean, Cretan and Proto-Egyptian linear signs respectively (Figure 4). These similarities would also suggest that the SGA-type rock paintings of Ampasimaiky are ancient writing. them (Willcox, 1984: 210); the same has been applied to that of Professor Barry Fell, a specialist in Libyan-Berber writing. The same has been applied to that of Professor Barry Fell, a specialist in Libyan-Berber writing, who translated the glyphs of Driekopseiland in southern Africa as Libyan-Berber writing. But it has also been refuted on the grounds that the Bushmen could not write (Willcox 1984: 210); it is important to recall
It is important to recall in this connection that Libyco-Berber inscriptions inscribed on petroglyphs in the Canary Islands were not accepted as such, under the pretext that the Guanches (indigenous people of the Canary Islands) could not have acquired or adopted any writing system in prehistoric times (Cline 1953; Farrujia de la Rosa et al. 2010: 16-17; Willcox 1984: 48). This erroneous model persisted until reputable scholars succeeded in translating an inscription reading “save me from javelin annihilation” (Cline 1953: 273). Today, it is commonly accepted that the groups of aligned glyphs from the Canary Islands are Libyco-Berber inscriptions. Recently, Chami (2006: 84-5) insisted that African SGA rock paintings and engravings are ancient scripts. He has shown that some, such as those at Namoratunga in Kenya, are arranged. For the accounts of those that
are not arranged, he argued that at the time when the ancient scripts of Ethiopia, India, China, Egypt and Mesopotamia were still using ideograms, the characters too were not aligned (see also Casson 1966: 141). This can be amplified by the case of some Sahelian writing systems such as the ‘Nsibidi. This script has no ‘order in the arrangement’, but it is practically recognised as a writing system (Dayrell
1910; Macgregor 1909). Zahan (1950) identified the Dogon and Bambara pictographs as scripts based on the argument that
“writing consists essentially in the communication of thought by means of conventional signs, and the pictographs of both Dogon and Bambara serve to convey thoughts from one writer to another or from writer to lay public” (pp.137) « Une écriture consiste essentiellement de la communication de pensée par le moyen de signes conventionnels, et les pictographes de Dogon et de Bambara servent à transporter des pensées d’un auteur à l’autre ou d’un auteur au public”.
Therefore, it is also possible that without the “order in the arrangement”, the Ampasimaiky SGA-type rock paintings could have served to communicate thoughts and information. This would qualify them as a WRITING SYSTEM! Especially since some groups of symbols are horizontally and vertically aligned (Photo 4). Since “order in arrangement” has been established by some scholars as the sine qua non for a group of symbols/signs to qualify as an inscription (Willcox 1984: 210), one can reasonably assert, based on the above evidence, that the
AMPASIMAIKY RUPESTRATED ROCK SHELTER PAINTINGS ARE INSCRIPTIONS! This assertion is clearly supported by the vertically arranged group of symbols in panel 8 (see Photo 4b). Indeed, the repetition of the first sign in the third character (read downwards) is undoubtedly demonstrating the USE OF THE ALPHABET, and thus the
ALPHABET, so this is WRITING! All the more so as all five characters constituting the inscription are closely related to the Libyan-Berber alphabets (see
www.ancientscripts.com), the proto-Egyptian, Aegean and Cretan linear signs respectively
(Figure 5).
Figure 5: The vertically arranged group of signs from Ampasimaiky compared with Libyco-Berber alphabetic characters and Proto-Egyptian, Aegean, and Cretan linear signs (Table: Evans, 1897:386).
Although Proto-Egyptian, Cretan, and Aegean linear signs were also vertically written, I would favor the Libyco-Berber origin of the Ampasimaiky inscription under the influence of Walter Cline’s description of the Libyco-Berber inscriptions as follows.
“They [Libyco-Berber inscriptions] are usually very brief, consisting of personal names and a few phrases … The characters are simple and predominantly rectilinear. Variant alphabet may have existed. Most of inscriptions are to be read upward beginning with the left-hand column, though horizontal rows reading from right to left are not uncommon and some columns read downwards. The words are not separated. Though these signs sometimes occur in groups of four or five, they often stand singly…” (Cline 1953: 273)
All the underlined sentences qualify the vertically arranged inscription of Ampasimaiky as being of Libyco-berberber origin. Although this ancient script was written either horizontally or vertically, it should be remembered that the Libyco- Berber peoples used two different alphabets for each row (see http://www.ancientscripts.com). All five characters of the Ampasimaiky inscription are exclusively related to the vertical Libyco- Berber alphabet (Figure 6). Transliterating them to Latin alphabets – based on the transliteration of www.ancientscripts.com – one could read “ğ r ğ t f” (downwards) or “f t ğ r ğ” (upwards). The phonetic value ‘r’ for the presumed circle/square sign can be read as ‘3’ (= a), if the represented sign is perceived as a triangular symbol (see Fell 1975: 271). In the present state of the study, decoding the meaning of the inscription is not yet within the scope of the author. This task would rather be reserved for palaeographers and epigraphers.
Figure 7: The group of vertically arranged signs of Ampasimaiky transliterated in the Libyco-Berber alphabet
It is important to point out here that certain Libyco-Berber inscriptions brought back from the Canary Islands are not different from that of Ampasimaiky (see Figure 8). You can clearly see their resemblance.
Figure 8: The group of vertically arranged signs from Ampasimaiky compared with a Libyco-Berber inscription from the Canary Islands (Drawings by Farrujia de la Rosa et al. 2010: 29)
Discussion and Conclusion
Some of the images found at Ampasimaiky are similar to those of the rock art of continental Africa.
However, the painting tradition of southwestern Madagascar does not seem to originate from the continental region closest to Madagascar, i.e. Mozambique and the rest of the central Limpopo basin. In fact, Professor Benjamin Smith, an internationally recognised authority on the study of SGA rock paintings in sub-Saharan Africa, has taken the view that they are not typically affiliated with the Khoekhoen SGA rock paintings, nor with the geometric art of central southern Africa (pers. comm.). The exclusive representation of humped cattle associated with SGA-type symbols is particular to Ampasimaiky. Since no ancient representation of zebu (Bos indicus) has yet been found in the region between Sudan in the north and the Zambezi River in the south (Magnavita 2006: 62), it can be inferred that the tradition of rock paintings in Ampasimaiky originated in North Africa. Evidence of a bronze zebu figurine (of the type Bos indicus) inscribed with SGA signs dating back to 200 CE at Zeban Kutur, Eritrea (Photo 4) would confirm the early use of SGA signs among pastoral zebu-breeding peoples such as those depicted in the Ampasimaiky rock paintings
Photo 4: Bronze figure of zebu inscribed with SGA type signs in Zeban Kutur, Eritrea (Photo: Magnavita 2006:63)
The Libyan-Berber inscription discovered at Ampasimaiky also reinforces this proposal. The Libyan-Berber script emerged in North Africa around 500 BC and ceased to be used around 800 AD (Cline 1953).
around 800 AD (Cline 1953). This implies that the Libyan-Berber inscription of Ampasimaiky should have been written during the period between this bracket of dates (500 BC – AD 800). It is worth mentioning that Iamboulos, a Greek merchant who lived in the 3rd
century BC, reported about a vertical order of writing that was practised by the inhabitants of the ‘island of the sun’ (Allibert 2010: 111; Kobishschanow, 1965: 138; Oldfather, 1961 vol.2: 65-67; Winston 1976: 221). This island is believed to be geographically located adjacent to the east coast of Africa, south of the Equator. Although some of the geographical and ecological information in Iamboulos’ account is very much in line with that of the Malagasy west coast (for more details see the
Although some of the geographical and ecological information in Iamboulos’ account closely matches that of the Malagasy west coast (for more information see Allibert 2010), some scholars have consistently judged this account to be fictional, imaginary, and fanciful (Cary and Warmington 1963: 233-245; Winston 1976). One of the criteria used to censor this narrative is the use of vertical order writing by the islanders. Certainly, it was difficult for some scholars to accept the existence of a pre-Islamic script used in the adjacent islands of the east coast of Africa such as Madagascar (Janvier 1975: 19). However, with the discovery of the Libyan-Berber inscription at Ampasimaiky, this denial no longer holds. New evidence is uncovered to confirm that the “island of the sun” of Iamboulos could have been Madagascar. In any case, man was already present in the vicinity of Ampasimaiky when the Greek merchant passed through the island. Human activity is attested by subfossil lemur bones bearing cutting marks and radiocarbon-dated from the 5th to the 4th century BC at Taolambiby, in the upper Onilahy valley (see Perez et al., 2005). All the more so since ancient written documents (see Jones 1960: 235, 269, 315; Strabo 17, 1, 13 in Lacroix 1998: 95), archaeology (see Cary and Warmington 1963: 130; Chami 2007: 6; 2009; Wainwright 1940: 165), genetic study (see Ricaut et al. 2009: 5-6), and zooarchaeology (Chami 2006: 137; Grigson 1991: 124) all confirm the ancient contacts that existed during this period between the Onilahy Valley and the Onilahy Valley,
During this period, the Nile Valley and the Mediterranean world on the one hand, and the eastern coast of Africa and the adjacent islands on the other.
Références Allibert, C. (2008), “Austronesian Migration and the Establishment of the Malagasy Civilization: Contrasted Readings in Linguistics, Archaeology, Genetics and Cultural Anthropology”, Diogenes, SAGE, Volume 218, pp.7-16 Allibert, C. (2010), “L’île de Iamboulos est-elle Madagascar? », Pount, no.4, pp.103-119 Blench, R. (2007), “New Palaeozoogeographical Evidence for the Settlement of Madagascar”, Azania, Volume XLII, pp.69-82 Cary M. and Warmington, E. (1963), The Ancient Explorers, Middlesex, Penguin Books Casson, L. (1966), Ancient Egypt, Nederland, Time-life International Chami, F. A (2006), The Unity of African Ancient History, 3000 BC to AD 500, Dar es Salaam, E&D Limited Chami, F. A (2008), “The Great Lakes: a complexity of cultural wellsprings”, in Arnold, M. (eds.) Art in Eastern Africa, pp.47-64, Dar-es-Salaam, Mkuki na Nyota Chami F. A and Kwekason A. (2003), “Neolithic Pottery Traditions from the Islands, the Coast, and the Interior of East Africa”, African Archaeological Review, Volume 20, No. 2, pp.65-80 Chami, F. A, Tabibou, I. A and Abderhamane, B. (2009), “Preliminary Report of Archaeological Work Conducted on the Southern Ngazidja Island”, in Chami, F. (eds.), Zanzibar and the Swahili Coast from c30.000 years ago, pp.115-128, Dar es Salaam, E&D Vision Publishers Chaplin, T. (1974), “The Prehistoric Rock Art of the Lake Victoria Region”, Azania Volume 9, pp.1-50 Cline, W. (1953), “Berber Dialects and Berber Script”, Southwestern Journal of Anthropology, Vol.9, No.3, pp.268-276, New Mexico, New Mexico University Press Coulson D. and Campbell A. (2001), African Rock Art, Paintings and Engravings on Stone, New York, Harry N. Abrams Incorporated Dayrell, E. (1910), “Some “Nsibidi” Signs”, Man, Vol.10, pp.113-114. Dewar, R. E (1996), “The Archaeology of the Early Settlement of Madagascar”, in Reade J. (eds.) The Indian Ocean in Antiquity, pp.471-486 Dewar, R. E (1997), “Does It Matter That Madagascar Is an Island?” Human Ecology, Volume 25, No. 3, pp. 481-489 Donque, G. (1965), « Le contexte océanique des anciennes migrations : vents et courants dans l’océan indien », Taloha, Volume 1, pp.43-59. Eastwood, E. B and Smith, B. W (2005), “Fingerprints of the Khoekhoen: Geometric and handprinted Rock Art in the Central Limpopo Basin, Southern Africa”, Goodwin Series, RSA, South African Archaeological Society, Volume 9, pp.63-76 Evans, A. J (1897), “Further Discoveries of Cretan and Aegean Script: With Libyan and ProtoEgyptian Comparisons”, The Journal of Hellenic Studies, The Society for the Promotion of Hellenic Studies, Volume 17, pp.327-395 Farrujia de la Rosa, A., Pichler W., Rodrigue A., Marín S.G. 2010, The Lybico-Berber and Latino-Canarian Scripts and the Colonization of the Canary Islands, African Archaeological Review, vol.27, no.1, pp.13-41 Faublée, J. (1947), Récits Bara, Travaux et mémoires de l’Institut d’ethnologie XLXII, Paris, Université de Paris Fell, B. (1975), America BC: Ancient Settlers in the New World, New York, A Demeter Press Book Grandidier, A., Charles-Roux J., Delhorbe C., Froidevaux H. and Grandidier G. (1902), Collection des ouvrages anciens concernant Madagascar, Tome II, Paris, Comité de Madagascar Grigson, C. (1991), “An African Origin of African Cattle? Some Archaeological Evidence”, The African Archaeological Review, Volume 9, pp.119-144 Hall, S. and Smith, B. (2000), “Empowering Places: Rock Shelters and Ritual Control in FarmerForager Interactions in the Northern Province”, Goodwin Series, RSA, South African Archaeological Society, Volume 8, pp.30-46 Hébert, J. C (1967), « Essai d’interprétation de la stèle indéchiffrée d’Ambilobe », Taloha, Volume 2, pp.XLVII-LVII Hébert, J. C (1971), « Les tatouages de la côte est de Madagascar, d’après Chapelier (1794- 1806) », Taloha, Volume 4, pp.211-223 Huygues-Belrose, V. (1997). Les supports anciens de l’écriture à Madagascar, Etudes Océan Indien, vol.22, pp.7-19, INALCO Janvier, Y. (1975), « La géographie gréco-romaine a-t-elle connu Madagascar? Le point de la question », Omaly sy Anio, No.1-2, pp.11-41 Jones, H. (1960), The Geography of Strabo, Volume 1-2, London, William Heinemann Ltd. Juma, A. (2004), Unguja Ukuu on Zanzibar – An Archaeological Study of Early Urbanism, Sweden, Uppsala University Kessy, E. (2009a), “Analysis of Lithic Artefacts From Mwanampambe and Kuumbi Caves, Zanzibar”, in Chami F. (eds.), Zanzibar and the Swahili Coast from c30,000 years ago, Dar es Salaam, E&D Vision Publishers, pp.131-143 Kessy, E. (2009b), “Stone Artefacts from Male”, in Chami F. (eds.), Zanzibar and the Swahili Coast from c30,000 years ago, Dar es Salaam, E&D Vision Publishers, pp.184-88 Kobishchanov, Y.M. (1965), “On the Problem of Sea Voyages of Ancient Africans in the Indian Ocean”, The Journal of African History, Vol.6, No.2, pp.137-141, Cambridge, Cambridge University Press. Lacroix, W. F (1998), Africa in Antiquity, Nijmegen, Catholic University Lynch, B. M and Donahue, R. (1980), “A Statistical Analysis of Two Rock-Art Sites in
Northwest Kenya”, Journal of Field Archaeology, Volume 7, pp.75-85
Mabulla, A. Z (2005), “The Rock Art of Mara Region, Tanzania”, Azania, Volume 40, 19-42
Macgregor, J. K (1909), “Some Notes on Nsibidi”, The Journal of the Royal Institute of Great
Britain and Ireland, Volume 39, pp.209-219
MacPhee R. D and Burney D. (1991), “Dating of modified femora of extinct dwarf
Hippopotamus from southern Madagascar: implications for constraining human
colonization and vertebrate extinction events”, Journal of Archaeological Science, No.
18, pp. 695-706
Magnavita, C. (2006), “Ancient Humped Cattle in Africa: A View from the Chad Basin”, The
African Archaeological Review, Volume 23, No. 3/4, pp.55-84
Oldfather, C. H (1961), Diodorus of Sicily, 12 Volumes, London, William Heinemann
Perez, V. R Godfrey L. R, Nowak-Kemp M., Burney D. A, Ratsimbazafy J. and Vasey N.
(2005), “Evidence of Early Butchery of Giant Lemurs in Madagascar”, Journal of
Human Evolution, No. 49, pp.722-742
Radimilahy, C. (2011), « Contribution á l’Archéologie du Sud-Ouest de Madagascar »,
Civilisations des mondes insulaires (Madagascar, îles du canal de Mozambique,
Mascareignes, Polynésie, Guyanes), Paris, Éditions Karthala, pp.825-853
Rasolondrainy, T. (2011), Archaeological Study of the Prehistoric Rock Paintings of
Ampasimaiky Rock Shelter, in the Upper Onilahy, Isalo Region, Southwestern
Madagascar. M.A. Dissertation unpublished, University of Dar es Salaam.
Ricaut, F., Razafindrazaka H., Cox, M., Dugoujon J., Guitard E., Sambo C., Mormina M.,
Mirazon-Larh M., Ludes B., and Crubézy E. 2009, A New Deep Branch of Eurasian
mtDNA Macrohaplogroup M Reveals Additional Complexity Regarding the Settlement
of Madagascar. BMC Genomics, vol.10, pp.1-9
Smith, B. W (1995), Rock Art in South-Central Africa. A Study Based on the Pictographs of
Dedza District, Malawi and Kasama District Zambia, Unpublished Phd Thesis,
University of Cambridge
Southall, A. (1975), “The Problem of Malagasy Origins”, in Chittick H. N and Rotberg R. I
(eds.), East Africa and the Orient, Cultural Syntheses in Pre-Colonial Times, New
York, Africana Publishing Company, pp. 192-215
Teka, Z. (2008). “Distribution and Significance of Ancient Rock Art Sites in Eritrea”, in
Schmidt P. et al (eds.), The Archaeology of Ancient Eritrea. Eritrea: The Red Sea Press,
Inc., pp.49-61
Thackeray, A. (1983), “Dating the Rock Art of Southern Africa”, Goodwin Series, Volume 4,
pp.21-26.
Vérin P., (1975), “Austronesian Contributions to the Culture of Madagascar: Some
Archaeological Problems”, in Chittick H. N and Rotberg R. I (eds.), East Africa and the
Orient, Cultural Syntheses in Pre-Colonial Times, New York, Africana Publishing
Company, pp. 164-191Vérin, P. (1981), “Madagascar”, in Mokhtar G. (eds.) General History of Africa, Volume 2, 693-
717
Wainwright, G. A (1940), “The Egyptian Origin of the New Year’s Sacrifice at Zanzibar”, Man,
Volume 40, pp.164-167
Willcox, A. R (1984), The Rock Art of Africa, New York, Holmes and Meier Publishers
Winston, D. (1976), “Iambulus’ “Island of the Sun” and Hellenistic Literary Utopias”, Science
Fiction Studies, SF-TH Inc., Volume 3, No. 3, pp.219-227
Zahan, D. (1950), “Pictographic Writing in the Western Sudan”, Man, Volume 50, pp.136-138
The research has been removed from the academia.edu site, that’s why I put the screenshot that shows its origin
Northwest Kenya”, Journal of Field Archaeology, Volume 7, pp.75-85
Mabulla, A. Z (2005), “The Rock Art of Mara Region, Tanzania”, Azania, Volume 40, 19-42
Macgregor, J. K (1909), “Some Notes on Nsibidi”, The Journal of the Royal Institute of Great
Britain and Ireland, Volume 39, pp.209-219
MacPhee R. D and Burney D. (1991), “Dating of modified femora of extinct dwarf
Hippopotamus from southern Madagascar: implications for constraining human
colonization and vertebrate extinction events”, Journal of Archaeological Science, No.
18, pp. 695-706
Magnavita, C. (2006), “Ancient Humped Cattle in Africa: A View from the Chad Basin”, The
African Archaeological Review, Volume 23, No. 3/4, pp.55-84
Oldfather, C. H (1961), Diodorus of Sicily, 12 Volumes, London, William Heinemann
Perez, V. R Godfrey L. R, Nowak-Kemp M., Burney D. A, Ratsimbazafy J. and Vasey N.
(2005), “Evidence of Early Butchery of Giant Lemurs in Madagascar”, Journal of
Human Evolution, No. 49, pp.722-742
Radimilahy, C. (2011), « Contribution á l’Archéologie du Sud-Ouest de Madagascar »,
Civilisations des mondes insulaires (Madagascar, îles du canal de Mozambique,
Mascareignes, Polynésie, Guyanes), Paris, Éditions Karthala, pp.825-853
Rasolondrainy, T. (2011), Archaeological Study of the Prehistoric Rock Paintings of
Ampasimaiky Rock Shelter, in the Upper Onilahy, Isalo Region, Southwestern
Madagascar. M.A. Dissertation unpublished, University of Dar es Salaam.
Ricaut, F., Razafindrazaka H., Cox, M., Dugoujon J., Guitard E., Sambo C., Mormina M.,
Mirazon-Larh M., Ludes B., and Crubézy E. 2009, A New Deep Branch of Eurasian
mtDNA Macrohaplogroup M Reveals Additional Complexity Regarding the Settlement
of Madagascar. BMC Genomics, vol.10, pp.1-9
Smith, B. W (1995), Rock Art in South-Central Africa. A Study Based on the Pictographs of
Dedza District, Malawi and Kasama District Zambia, Unpublished Phd Thesis,
University of Cambridge
Southall, A. (1975), “The Problem of Malagasy Origins”, in Chittick H. N and Rotberg R. I
(eds.), East Africa and the Orient, Cultural Syntheses in Pre-Colonial Times, New
York, Africana Publishing Company, pp. 192-215
Teka, Z. (2008). “Distribution and Significance of Ancient Rock Art Sites in Eritrea”, in
Schmidt P. et al (eds.), The Archaeology of Ancient Eritrea. Eritrea: The Red Sea Press,
Inc., pp.49-61
Thackeray, A. (1983), “Dating the Rock Art of Southern Africa”, Goodwin Series, Volume 4,
pp.21-26.
Vérin P., (1975), “Austronesian Contributions to the Culture of Madagascar: Some
Archaeological Problems”, in Chittick H. N and Rotberg R. I (eds.), East Africa and the
Orient, Cultural Syntheses in Pre-Colonial Times, New York, Africana Publishing
Company, pp. 164-191Vérin, P. (1981), “Madagascar”, in Mokhtar G. (eds.) General History of Africa, Volume 2, 693-
717
Wainwright, G. A (1940), “The Egyptian Origin of the New Year’s Sacrifice at Zanzibar”, Man,
Volume 40, pp.164-167
Willcox, A. R (1984), The Rock Art of Africa, New York, Holmes and Meier Publishers
Winston, D. (1976), “Iambulus’ “Island of the Sun” and Hellenistic Literary Utopias”, Science
Fiction Studies, SF-TH Inc., Volume 3, No. 3, pp.219-227
Zahan, D. (1950), “Pictographic Writing in the Western Sudan”, Man, Volume 50, pp.136-138
The research has been removed from the academia.edu site, that’s why I put the screenshot that shows its origin
Tanambelo V.R. RASOLONDRAINY,
(Centre de Documentation et de Recherche sur l’Art et la Tradition Orale de Madagascar)
Abstract:
This article presents the result of an archaeological study on the first prehistoric rock paintings site found in Madagascar and the rest of islands of Western Indian Ocean south of the equator.
Based on systematic methods and techniques, the study provides groundbreaking archaeological evidences that contribute largely to the understanding of the prehistory of Madagascar which was not archaeologically known thitherto to have prehistoric rock paintings, nor human settlement prior to 8th century AD. A Libyco-Berber inscription was discovered on the rock paintings of Ampasimaiky Rock Shelter, in Southwestern Madagascar. Based on this writing -supported by ancient historical records, circumstantial archaeological and paleontological evidences, linguistic data, and to some extent DNA analysis results -, the present study provides minimum age falling in the bracket time around 500 BC to roughly 8th century AD for the rock paintings of Ampasimaiky rock shelter. Furthermore, the study traces back early contacts between Madagascar and Northern Africa during prehistoric times.
Introduction
With the kind assistance of Dr Chantal Radimilahy, National Coordinator of the Network of African Archaeologists in Madagascar, I researched the ethnographic exhibition on the website of the Museum of Ethnography in Geneva (MEG). The theme was the peoples and cultures of the southwestern region of Madagascar, and was based on an exhibition of photographs taken by the French ethnographer Jacques Faublée in 1939. The pictures included rock paintings found in caves and rock shelters in the Isalo region. After discussions with my teachers at the Archaeology Unit of the University of Dar es Salaam, Professors Felix Chami, Fidelis Masao, and Audax Mabulla, I further investigated the significance of these paintings for the prehistoric period of Madagascar. This study is taken from a Master of Arts in Archaeology entitled
“Archaeologial study of the prehistoric rock paintings of Ampasimaiky rock shelter, in the upper Onilahy, Isalo region, Southwestern Madagascar”.
————–
Africa has many archaeological sites with prehistoric rock paintings (see Coulson and Campbell 2001; Willcox 1984). However, no such sites have yet been scientifically
scientifically reported from Madagascar, which is only 400 kilometres from southern Africa where they are abundant. All rock art previously found in Madagascar has been judged to have been produced in the recent past (Faublée 1947: 164; Hébert 1971: 218). This judgement is based on the theory that Africans were incapable of having navigated the high seas during prehistoric times (Donque 1965: 43; Grandidier et al. 1902; Southall 1975: 194; Vérin 1975, 1981). It was therefore out of the question to think that they had the possibility of transferring the practice of rock painting to the adjacent islands of Africa such as Madagascar. However, recent archaeological findings have demonstrated that the east coast of Africa and adjacent islands were already in contact since the prehistoric period (see Chami and Kwekason 2003; Chami et al. 2009; Juma 2004; Kessy 2009a,
2009b). These prompt a problem that led me to revise the antiquity of the rock paintings in Madagascar.
On the other hand, the name of the Arabic-Malagasy script – Sorabe meaning “great script” – indicates that there was an ancient script in Madagascar that was distinguished from Arabic (Hébert,
1967:XLVII; Huygues-Belrose 1997: 8). Although Hébert (1967:XLVIII) presumed that such a hypothetical script would have to be of Indian origin and would be discovered from
Although Hébert (1967:XLVIII) assumed that such a hypothetical script would have to be of Indian origin and would be discovered from stele and parietal inscriptions, no such inscription has so far been found in Madagascar.
My problem was therefore to find out whether this would appear on the rock paintings reported by the website. What provoked this idea the most was that Chami (2006:84-85; 2008) argued that the African Geometric and Amorphous Schematic (GAS) rock paintings and engravings are ancient writing.
Another problem is that radiocarbon 14 dating placed the earliest human settlement of Madagascar around the 8th century AD (Allibert 2008: 7; Dewar 1996: 474).
Although earlier dates are still being considered, many researchers have assumed that settlement of Madagascar would not go beyond the year W AD (for more information see Dewar 1997). This assumption has made research on Malagasy material culture going back to the Stone Age more than difficult. Any findings of material cultures dating from prehistoric times have been discarded, overlooked or misinterpreted. However, subfossil animal bone remains bearing traces of butchery found in paleontological sites in the southwestern region indicate the presence of human activity
(For more information, see Blench 2007; Perez et al. 2005). Despite this evidence, an archaeological site has not yet been
archaeological site with similar dates has not yet been found on the entire island.
With the discovery of a rock shelter with rock paintings, discovering new information about the prehistory of Madagascar before the 8th century proved to be feasible. All the more so as some of the paintings and engravings are
Especially since some of the SGA-type rock paintings and engravings found on the African continent – similar to those found in the photos on the MEG website – have been radiocarbon dated from the 1st millennium BC to the early 1st millennium AD (Chami, 2008: 54; Hall and Smith 2000; Lynche and Donahue 1980; Willcox 1984: 26, 83.
Fieldwork and results
Despite various challenges and constraints – mainly caused by the lack of archaeological reconnaissance prior to the fieldwork, insecurity in the study area, and the sacredness of the site to be studied – we had only somehow managed to conduct survey, archaeological excavation, and documentation of the paintings in the Ampasimaiky rock shelter. It is only the result of the latter that will be presented in this article.
The rock painting site is a rock shelter located at the foot of a hill immediately next to a small stream called Ampasimaiky (Figure 2), from which the name of the site derives. The presence of this water source may have influenced the painter’s choice of location.
The shelter itself is 26.8m long (south-north axis) and 5m wide (inwards).
Its ceiling begins to slope inwards from 4m high at its entrance to approximately 1.5m high at its irregular wall. The deepest part of the rock shelter – a chamber – is currently used by the royal family “Bara Zafimanely” as a burial place. Dung and ox hoof prints found on the ground indicate that cattle frequent the place. This occurrence should have disturbed the soil.
However, the areas around the site are dominated by rocky hills covered with grass. Only in the deep canyon near the confluence of Ampasimaiky and the Sakamarekely River is a patch of thick vegetation up to 10 m high observed along the stream. This situation would make agricultural activity almost impossible around the site, unless one had to move 1.5 km to the bank, where rice cultivation takes place today.
An archaeological survey of the southern Isalo region (Figure 1) was carried out and a rock painting site was discovered. Professor Chami participated in the fieldwork, advising on the approach to be followed in the treatment of these previously unstudied cave cultures in Madagascar.
Madagascar. History students from the University of Toliara, Mex Tafitasoa, Nadya
Ravololonirina, and Norbert Antilahy, and some villagers (Photo 1) participated in the work. The Network of African Archaeologists in Madagascar, under the direction of Dr Chantal Radimilahy and Prof Barthélemy Manjakahery, provided the logistical equipment used. The various signatures The various signatures of national and local administrative approval, the security provided by the Gendarmerie Brigade of Benenitsy, and above all the kindness and openness of King Zafimanely (Photo 1) – who allowed us access to the (burial) site – contributed enormously to the accomplishment of the expedition.
Madagascar. History students from the University of Toliara, Mex Tafitasoa, Nadya
Ravololonirina, and Norbert Antilahy, and some villagers (Photo 1) participated in the work. The Network of African Archaeologists in Madagascar, under the direction of Dr Chantal Radimilahy and Prof Barthélemy Manjakahery, provided the logistical equipment used. The various signatures The various signatures of national and local administrative approval, the security provided by the Gendarmerie Brigade of Benenitsy, and above all the kindness and openness of King Zafimanely (Photo 1) – who allowed us access to the (burial) site – contributed enormously to the accomplishment of the expedition.
The rock shelter at Ampasimaiky faces east. This arrangement makes the painting panels exposed to the sun in the morning and in the shade beyond noon. This arrangement suggests that painting activities are conducted either early in the morning while the heat of the sun is still moderate, or beyond noon when shade is available below the shelter.
This is based on the fact that the height of the panels does not exceed 130 cm, suggesting that the paintings were intended for people sitting on the floor of the undercroft shelter
Photo 2). Minimal, therefore, is the likelihood of the painting activity – involving seated people – having been conducted during the time when the sun is heating the shelter floor.
However, the view outside the site is limited. The opposing hills – on the other bank of the Ampasimaiky stream – prevent a person standing under the shelter from having a good view of the surroundings (Photo 3). It is much more likely that the view outwards – which implies a defensive strategy (Smith 1995) – was not among the painter’s concerns when he/she chose the site. This demonstrates that the layout of the Ampasimaiky rock shelter favours invaders from unexpectedly attacking its residents. The place would therefore not be suitable for permanent habitation.
Photo 3
Especially since a person standing on the hills of the other bank would be able to look directly at the other one positioned under the rock shelter, unless he hides in the deeper part of the shelter. Since the wall of this “chamber”, according to the local population, is not painted, it is plausible that the activities related to the paintings in the Ampasimaiky rock shelter are not clandestine. However, the paintings were not intended to be seen by people passing close to the site. They are not seen from a distance. Not only are the panels placed below 130 cm in height, but the slope of the ground before the access
to the shelter also prevents anyone from seeing them until they get close to the entrance. Nevertheless, the use of darker coloured paints (red and black) suggests the painter’s intention to contrast the paints he/she was using against the pale colour of the sandstone wall. This implies his concern for the visibility of his paintings.
The walls of the rock shelter have red, reddish orange, burgundy red, black, and white paintings in monochrome, bichrome and polychrome patterns. The images
are painted with the fingertip. They depict zoomorphic (mainly zebu) and anthropomorphic (semi-naturalist and stylised) figures, as well as schematic-geometric-amorphic symbols (Figure 3). The latter dominate and are represented by quadrangular, circular and elliptical signs, dot/dash lines, and alphabet-like signs. In total, the Ampasimaiky rock paintings have 166 recognisable images.
Analysis and Interpretation
Based on comparative studies of the images, it has been found that the images of the Ampasimaiky rock paintings are not different from the images depicted on the prehistoric SGA rock art of the African continent (see Chami, 2008; Chaplin, 1974; Coulson and Campbell 2001;
Eastwood and Smith, 2005; Lynch and Donahue, 1980; Mabulla, 2005; Smith, 1995; Teka 2008; Willcox 1984). The similarities apparently suggest links between the two regions (Madagascar and the mainland) in the
(Madagascar and the continent) in prehistoric times. This suggests that the Ampasimaiky rock paintings are of African origin and could be relatively dated to the same period as those on the continent (Thackeray, 1983: 22). Comparative studies of the images have also shown that more than 30 types of images in the Ampasimaiky rock paintings are closely related to different ancient scripts such as Libyco-Berber, Protosinaitic, Meroitic (cursive), Punic Iberian, Ancient Phoenician, Tifinagh, and Aegean, Cretan and Proto-Egyptian linear signs respectively (Figure 4). These similarities would also suggest that the SGA-type rock paintings of Ampasimaiky are ancient writing.
figure 4
Analysis and Interpretation
Based on comparative studies of the images, it has been found that the images of the Ampasimaiky rock paintings are not different from the images depicted on the prehistoric SGA rock art of the African continent (see Chami, 2008; Chaplin, 1974; Coulson and Campbell 2001;
Eastwood and Smith, 2005; Lynch and Donahue, 1980; Mabulla, 2005; Smith, 1995; Teka 2008; Willcox 1984). The similarities apparently suggest links between the two regions (Madagascar and the mainland) in the
(Madagascar and the continent) in prehistoric times. This suggests that the Ampasimaiky rock paintings are of African origin and could be relatively dated to the same period as those on the continent (Thackeray, 1983: 22). Comparative studies of the images have also shown that more than 30 types of images in the Ampasimaiky rock paintings are closely related to different ancient scripts such as Libyco-Berber, Protosinaitic, Meroitic (cursive), Punic Iberian, Ancient Phoenician, Tifinagh, and Aegean, Cretan and Proto-Egyptian linear signs respectively (Figure 4). These similarities would also suggest that the SGA-type rock paintings of Ampasimaiky are ancient writing. them (Willcox, 1984: 210); the same has been applied to that of Professor Barry Fell, a specialist in Libyan-Berber writing. The same has been applied to that of Professor Barry Fell, a specialist in Libyan-Berber writing, who translated the glyphs of Driekopseiland in southern Africa as Libyan-Berber writing. But it has also been refuted on the grounds that the Bushmen could not write (Willcox 1984: 210); it is important to recall
It is important to recall in this connection that Libyco-Berber inscriptions inscribed on petroglyphs in the Canary Islands were not accepted as such, under the pretext that the Guanches (indigenous people of the Canary Islands) could not have acquired or adopted any writing system in prehistoric times (Cline 1953; Farrujia de la Rosa et al. 2010: 16-17; Willcox 1984: 48). This erroneous model persisted until reputable scholars succeeded in translating an inscription reading “save me from javelin annihilation” (Cline 1953: 273). Today, it is commonly accepted that the groups of aligned glyphs from the Canary Islands are Libyco-Berber inscriptions. Recently, Chami (2006: 84-5) insisted that African SGA rock paintings and engravings are ancient scripts. He has shown that some, such as those at Namoratunga in Kenya, are arranged. For the accounts of those that
are not arranged, he argued that at the time when the ancient scripts of Ethiopia, India, China, Egypt and Mesopotamia were still using ideograms, the characters too were not aligned (see also Casson 1966: 141). This can be amplified by the case of some Sahelian writing systems such as the ‘Nsibidi. This script has no ‘order in the arrangement’, but it is practically recognised as a writing system (Dayrell
1910; Macgregor 1909). Zahan (1950) identified the Dogon and Bambara pictographs as scripts based on the argument that
“writing consists essentially in the communication of thought by means of conventional signs, and the pictographs of both Dogon and Bambara serve to convey thoughts from one writer to another or from writer to lay public” (pp.137) « Une écriture consiste essentiellement de la communication de pensée par le moyen de signes conventionnels, et les pictographes de Dogon et de Bambara servent à transporter des pensées d’un auteur à l’autre ou d’un auteur au public”.
Therefore, it is also possible that without the “order in the arrangement”, the Ampasimaiky SGA-type rock paintings could have served to communicate thoughts and information. This would qualify them as a WRITING SYSTEM! Especially since some groups of symbols are horizontally and vertically aligned (Photo 4). Since “order in arrangement” has been established by some scholars as the sine qua non for a group of symbols/signs to qualify as an inscription (Willcox 1984: 210), one can reasonably assert, based on the above evidence, that the
AMPASIMAIKY RUPESTRATED ROCK SHELTER PAINTINGS ARE INSCRIPTIONS! This assertion is clearly supported by the vertically arranged group of symbols in panel 8 (see Photo 4b). Indeed, the repetition of the first sign in the third character (read downwards) is undoubtedly demonstrating the USE OF THE ALPHABET, and thus the
ALPHABET, so this is WRITING! All the more so as all five characters constituting the inscription are closely related to the Libyan-Berber alphabets (see
www.ancientscripts.com), the proto-Egyptian, Aegean and Cretan linear signs respectively
(Figure 5).
Figure 5 : Le groupe de signes verticalement arrangées d’Ampasimaiky comparé avec des caractères alphabétiques libyco-berbères et des signes linéaires proto-Egyptiens, égéens, et crétois (Tableau : Evans, 1897 :386).
Bien que des signes linéaires proto-Égyptiens, crétois et égéens aient été également verticalement écrits, je favoriserais l’origine libyco-berbère de l’inscription d’Ampasimaiky sous l’influence de la description de Walter Cline concernant les inscriptions libyco-berbères comme
suit).
“They [Libyco-Berber inscriptions] are usually very brief, consisting of personal names and a few phrases … The characters are simple and predominantly rectilinear. Variant alphabet may have existed. Most of inscriptions are to be read upward beginning with the left-hand column, though horizontal rows reading from right to left are not uncommon and some columns read downwards. The words are not separated. Though these signs sometimes occur in groups of four or five, they often stand singly…” (Cline 1953: 273)
All the underlined sentences qualify the vertically arranged inscription of Ampasimaiky as being of Libyco-berberber origin. Although this ancient script was written either horizontally or vertically, it should be remembered that the Libyco- Berber peoples used two different alphabets for each row (see http://www.ancientscripts.com). All five characters of the Ampasimaiky inscription are exclusively related to the vertical Libyco- Berber alphabet (Figure 6). Transliterating them to Latin alphabets – based on the transliteration of www.ancientscripts.com – one could read “ğ r ğ t f” (downwards) or “f t ğ r ğ” (upwards). The phonetic value ‘r’ for the presumed circle/square sign can be read as ‘3’ (= a), if the represented sign is perceived as a triangular symbol (see Fell 1975: 271). In the present state of the study, decoding the meaning of the inscription is not yet within the scope of the author. This task would rather be reserved for palaeographers and epigraphers.
Figure 7: The group of vertically arranged signs of Ampasimaiky transliterated in the Libyco-Berber alphabet
It is important to point out here that certain Libyco-Berber inscriptions brought back from the Canary Islands are not different from that of Ampasimaiky (see Figure 8). You can clearly see their resemblance.
Figure 8: The group of vertically arranged signs from Ampasimaiky compared with a Libyco-Berber inscription from the Canary Islands (Drawings by Farrujia de la Rosa et al. 2010: 29)
Discussion and Conclusion
Some of the images found at Ampasimaiky are similar to those of the rock art of continental Africa. However, the painting tradition of southwestern Madagascar is not
However, the painting tradition of southwestern Madagascar does not seem to originate from the continental region closest to Madagascar, i.e. Mozambique and the rest of the central Limpopo basin. In fact, Professor Benjamin Smith, an internationally recognised authority on the study of SGA rock paintings in sub-Saharan Africa, has taken the view that they are not typically affiliated with the Khoekhoen SGA rock paintings, nor with the geometric art of central southern Africa (pers. comm.). The exclusive representation of humped cattle associated with SGA-type symbols is particular to Ampasimaiky. Since no ancient representation of zebu (Bos indicus) has yet been found in the region between Sudan in the north and the Zambezi River in the south (Magnavita 2006: 62), it can be inferred that the tradition of rock paintings in Ampasimaiky originated in North Africa. Evidence of a bronze zebu figurine (of the type Bos indicus) inscribed with SGA signs dating back to 200 CE at Zeban Kutur, Eritrea (Photo 4) would confirm the early use of SGA signs among pastoral zebu-breeding peoples such as those depicted in the Ampasimaiky rock paintings
Photo 4: Bronze figure of zebu inscribed with SGA type signs in Zeban Kutur, Eritrea (Photo: Magnavita 2006:63)
The Libyan-Berber inscription discovered at Ampasimaiky also reinforces this proposal. The Libyan-Berber script emerged in North Africa around 500 BC and ceased to be used around 800 AD (Cline 1953).
around 800 AD (Cline 1953). This implies that the Libyan-Berber inscription of Ampasimaiky should have been written during the period between this bracket of dates (500 BC – AD 800). It is worth mentioning that Iamboulos, a Greek merchant who lived in the 3rd
century BC, reported about a vertical order of writing that was practised by the inhabitants of the ‘island of the sun’ (Allibert 2010: 111; Kobishschanow, 1965: 138; Oldfather, 1961 vol.2: 65-67; Winston 1976: 221). This island is believed to be geographically located adjacent to the east coast of Africa, south of the Equator. Although some of the geographical and ecological information in Iamboulos’ account is very much in line with that of the Malagasy west coast (for more details see the
Although some of the geographical and ecological information in Iamboulos’ account closely matches that of the Malagasy west coast (for more information see Allibert 2010), some scholars have consistently judged this account to be fictional, imaginary, and fanciful (Cary and Warmington 1963: 233-245; Winston 1976). One of the criteria used to censor this narrative is the use of vertical order writing by the islanders. Certainly, it was difficult for some scholars to accept the existence of a pre-Islamic script used in the adjacent islands of the east coast of Africa such as Madagascar (Janvier 1975: 19). However, with the discovery of the Libyan-Berber inscription at Ampasimaiky, this denial no longer holds. New evidence is uncovered to confirm that the “island of the sun” of Iamboulos could have been Madagascar. In any case, man was already present in the vicinity of Ampasimaiky when the Greek merchant passed through the island. Human activity is attested by subfossil lemur bones bearing cutting marks and radiocarbon-dated from the 5th to the 4th century BC at Taolambiby, in the upper Onilahy valley (see Perez et al., 2005). All the more so since ancient written documents (see Jones 1960: 235, 269, 315; Strabo 17, 1, 13 in Lacroix 1998: 95), archaeology (see Cary and Warmington 1963: 130; Chami 2007: 6; 2009; Wainwright 1940: 165), genetic study (see Ricaut et al. 2009: 5-6), and zooarchaeology (Chami 2006: 137; Grigson 1991: 124) all confirm the ancient contacts that existed during this period between the Onilahy Valley and the Onilahy Valley,
During this period, the Nile Valley and the Mediterranean world on the one hand, and the eastern coast of Africa and the adjacent islands on the other.
Références Allibert, C. (2008), “Austronesian Migration and the Establishment of the Malagasy Civilization: Contrasted Readings in Linguistics, Archaeology, Genetics and Cultural Anthropology”, Diogenes, SAGE, Volume 218, pp.7-16 Allibert, C. (2010), “L’île de Iamboulos est-elle Madagascar? », Pount, no.4, pp.103-119 Blench, R. (2007), “New Palaeozoogeographical Evidence for the Settlement of Madagascar”, Azania, Volume XLII, pp.69-82 Cary M. and Warmington, E. (1963), The Ancient Explorers, Middlesex, Penguin Books Casson, L. (1966), Ancient Egypt, Nederland, Time-life International Chami, F. A (2006), The Unity of African Ancient History, 3000 BC to AD 500, Dar es Salaam, E&D Limited Chami, F. A (2008), “The Great Lakes: a complexity of cultural wellsprings”, in Arnold, M. (eds.) Art in Eastern Africa, pp.47-64, Dar-es-Salaam, Mkuki na Nyota Chami F. A and Kwekason A. (2003), “Neolithic Pottery Traditions from the Islands, the Coast, and the Interior of East Africa”, African Archaeological Review, Volume 20, No. 2, pp.65-80 Chami, F. A, Tabibou, I. A and Abderhamane, B. (2009), “Preliminary Report of Archaeological Work Conducted on the Southern Ngazidja Island”, in Chami, F. (eds.), Zanzibar and the Swahili Coast from c30.000 years ago, pp.115-128, Dar es Salaam, E&D Vision Publishers Chaplin, T. (1974), “The Prehistoric Rock Art of the Lake Victoria Region”, Azania Volume 9, pp.1-50 Cline, W. (1953), “Berber Dialects and Berber Script”, Southwestern Journal of Anthropology, Vol.9, No.3, pp.268-276, New Mexico, New Mexico University Press Coulson D. and Campbell A. (2001), African Rock Art, Paintings and Engravings on Stone, New York, Harry N. Abrams Incorporated Dayrell, E. (1910), “Some “Nsibidi” Signs”, Man, Vol.10, pp.113-114. Dewar, R. E (1996), “The Archaeology of the Early Settlement of Madagascar”, in Reade J. (eds.) The Indian Ocean in Antiquity, pp.471-486 Dewar, R. E (1997), “Does It Matter That Madagascar Is an Island?” Human Ecology, Volume 25, No. 3, pp. 481-489 Donque, G. (1965), « Le contexte océanique des anciennes migrations : vents et courants dans l’océan indien », Taloha, Volume 1, pp.43-59. Eastwood, E. B and Smith, B. W (2005), “Fingerprints of the Khoekhoen: Geometric and handprinted Rock Art in the Central Limpopo Basin, Southern Africa”, Goodwin Series, RSA, South African Archaeological Society, Volume 9, pp.63-76 Evans, A. J (1897), “Further Discoveries of Cretan and Aegean Script: With Libyan and ProtoEgyptian Comparisons”, The Journal of Hellenic Studies, The Society for the Promotion of Hellenic Studies, Volume 17, pp.327-395 Farrujia de la Rosa, A., Pichler W., Rodrigue A., Marín S.G. 2010, The Lybico-Berber and Latino-Canarian Scripts and the Colonization of the Canary Islands, African Archaeological Review, vol.27, no.1, pp.13-41 Faublée, J. (1947), Récits Bara, Travaux et mémoires de l’Institut d’ethnologie XLXII, Paris, Université de Paris Fell, B. (1975), America BC: Ancient Settlers in the New World, New York, A Demeter Press Book Grandidier, A., Charles-Roux J., Delhorbe C., Froidevaux H. and Grandidier G. (1902), Collection des ouvrages anciens concernant Madagascar, Tome II, Paris, Comité de Madagascar Grigson, C. (1991), “An African Origin of African Cattle? Some Archaeological Evidence”, The African Archaeological Review, Volume 9, pp.119-144 Hall, S. and Smith, B. (2000), “Empowering Places: Rock Shelters and Ritual Control in FarmerForager Interactions in the Northern Province”, Goodwin Series, RSA, South African Archaeological Society, Volume 8, pp.30-46 Hébert, J. C (1967), « Essai d’interprétation de la stèle indéchiffrée d’Ambilobe », Taloha, Volume 2, pp.XLVII-LVII Hébert, J. C (1971), « Les tatouages de la côte est de Madagascar, d’après Chapelier (1794- 1806) », Taloha, Volume 4, pp.211-223 Huygues-Belrose, V. (1997). Les supports anciens de l’écriture à Madagascar, Etudes Océan Indien, vol.22, pp.7-19, INALCO Janvier, Y. (1975), « La géographie gréco-romaine a-t-elle connu Madagascar? Le point de la question », Omaly sy Anio, No.1-2, pp.11-41 Jones, H. (1960), The Geography of Strabo, Volume 1-2, London, William Heinemann Ltd. Juma, A. (2004), Unguja Ukuu on Zanzibar – An Archaeological Study of Early Urbanism, Sweden, Uppsala University Kessy, E. (2009a), “Analysis of Lithic Artefacts From Mwanampambe and Kuumbi Caves, Zanzibar”, in Chami F. (eds.), Zanzibar and the Swahili Coast from c30,000 years ago, Dar es Salaam, E&D Vision Publishers, pp.131-143 Kessy, E. (2009b), “Stone Artefacts from Male”, in Chami F. (eds.), Zanzibar and the Swahili Coast from c30,000 years ago, Dar es Salaam, E&D Vision Publishers, pp.184-88 Kobishchanov, Y.M. (1965), “On the Problem of Sea Voyages of Ancient Africans in the Indian Ocean”, The Journal of African History, Vol.6, No.2, pp.137-141, Cambridge, Cambridge University Press. Lacroix, W. F (1998), Africa in Antiquity, Nijmegen, Catholic University
Tanambelo V.R. RASOLONDRAINY,
(Centre de Documentation et de Recherche sur l’Art et la Tradition Orale de Madagascar)
Abstract:
This article presents the result of an archaeological study on the first prehistoric rock paintings site found in Madagascar and the rest of islands of Western Indian Ocean south of the equator.
Based on systematic methods and techniques, the study provides groundbreaking archaeological evidences that contribute largely to the understanding of the prehistory of Madagascar which was not archaeologically known thitherto to have prehistoric rock paintings, nor human settlement prior to 8th century AD. A Libyco-Berber inscription was discovered on the rock paintings of Ampasimaiky Rock Shelter, in Southwestern Madagascar. Based on this writing -supported by ancient historical records, circumstantial archaeological and paleontological evidences, linguistic data, and to some extent DNA analysis results -, the present study provides minimum age falling in the bracket time around 500 BC to roughly 8th century AD for the rock paintings of Ampasimaiky rock shelter. Furthermore, the study traces back early contacts between Madagascar and Northern Africa during prehistoric times.
Introduction
With the kind assistance of Dr Chantal Radimilahy, National Coordinator of the Network of African Archaeologists in Madagascar, I researched the ethnographic exhibition on the website of the Museum of Ethnography in Geneva (MEG). The theme was the peoples and cultures of the southwestern region of Madagascar, and was based on an exhibition of photographs taken by the French ethnographer Jacques Faublée in 1939. The pictures included rock paintings found in caves and rock shelters in the Isalo region. After discussions with my teachers at the Archaeology Unit of the University of Dar es Salaam, Professors Felix Chami, Fidelis Masao, and Audax Mabulla, I further investigated the significance of these paintings for the prehistoric period of Madagascar. This study is taken from a Master of Arts in Archaeology entitled
“Archaeologial study of the prehistoric rock paintings of Ampasimaiky rock shelter, in the upper Onilahy, Isalo region, Southwestern Madagascar”.
————–
Africa has many archaeological sites with prehistoric rock paintings (see Coulson and Campbell 2001; Willcox 1984). However, no such sites have yet been scientifically
scientifically reported from Madagascar, which is only 400 kilometres from southern Africa where they are abundant. All rock art previously found in Madagascar has been judged to have been produced in the recent past (Faublée 1947: 164; Hébert 1971: 218). This judgement is based on the theory that Africans were incapable of having navigated the high seas during prehistoric times (Donque 1965: 43; Grandidier et al. 1902; Southall 1975: 194; Vérin 1975, 1981). It was therefore out of the question to think that they had the possibility of transferring the practice of rock painting to the adjacent islands of Africa such as Madagascar. However, recent archaeological findings have demonstrated that the east coast of Africa and adjacent islands were already in contact since the prehistoric period (see Chami and Kwekason 2003; Chami et al. 2009; Juma 2004; Kessy 2009a,
2009b). These prompt a problem that led me to revise the antiquity of the rock paintings in Madagascar.
On the other hand, the name of the Arabic-Malagasy script – Sorabe meaning “great script” – indicates that there was an ancient script in Madagascar that was distinguished from Arabic (Hébert,
1967:XLVII; Huygues-Belrose 1997: 8). Although Hébert (1967:XLVIII) presumed that such a hypothetical script would have to be of Indian origin and would be discovered from
Although Hébert (1967:XLVIII) assumed that such a hypothetical script would have to be of Indian origin and would be discovered from stele and parietal inscriptions, no such inscription has so far been found in Madagascar.
My problem was therefore to find out whether this would appear on the rock paintings reported by the website. What provoked this idea the most was that Chami (2006:84-85; 2008) argued that the African Geometric and Amorphous Schematic (GAS) rock paintings and engravings are ancient writing.
Another problem is that radiocarbon 14 dating placed the earliest human settlement of Madagascar around the 8th century AD (Allibert 2008: 7; Dewar 1996: 474).
Although earlier dates are still being considered, many researchers have assumed that settlement of Madagascar would not go beyond the year W AD (for more information see Dewar 1997). This assumption has made research on Malagasy material culture going back to the Stone Age more than difficult. Any findings of material cultures dating from prehistoric times have been discarded, overlooked or misinterpreted. However, subfossil animal bone remains bearing traces of butchery found in paleontological sites in the southwestern region indicate the presence of human activity
(For more information, see Blench 2007; Perez et al. 2005). Despite this evidence, an archaeological site has not yet been
archaeological site with similar dates has not yet been found on the entire island.
With the discovery of a rock shelter with rock paintings, discovering new information about the prehistory of Madagascar before the 8th century proved to be feasible. All the more so as some of the paintings and engravings are
Especially since some of the SGA-type rock paintings and engravings found on the African continent – similar to those found in the photos on the MEG website – have been radiocarbon dated from the 1st millennium BC to the early 1st millennium AD (Chami, 2008: 54; Hall and Smith 2000; Lynche and Donahue 1980; Willcox 1984: 26, 83.
Fieldwork and results
Despite various challenges and constraints – mainly caused by the lack of archaeological reconnaissance prior to the fieldwork, insecurity in the study area, and the sacredness of the site to be studied – we had only somehow managed to conduct survey, archaeological excavation, and documentation of the paintings in the Ampasimaiky rock shelter. It is only the result of the latter that will be presented in this article.
The rock painting site is a rock shelter located at the foot of a hill immediately next to a small stream called Ampasimaiky (Figure 2), from which the name of the site derives. The presence of this water source may have influenced the painter’s choice of location.
The shelter itself is 26.8m long (south-north axis) and 5m wide (inwards).
Its ceiling begins to slope inwards from 4m high at its entrance to approximately 1.5m high at its irregular wall. The deepest part of the rock shelter – a chamber – is currently used by the royal family “Bara Zafimanely” as a burial place. Dung and ox hoof prints found on the ground indicate that cattle frequent the place. This occurrence should have disturbed the soil.
However, the areas around the site are dominated by rocky hills covered with grass. Only in the deep canyon near the confluence of Ampasimaiky and the Sakamarekely River is a patch of thick vegetation up to 10 m high observed along the stream. This situation would make agricultural activity almost impossible around the site, unless one had to move 1.5 km to the bank, where rice cultivation takes place today.
An archaeological survey of the southern Isalo region (Figure 1) was carried out and a rock painting site was discovered. Professor Chami participated in the fieldwork, advising on the approach to be followed in the treatment of these previously unstudied cave cultures in Madagascar.
Madagascar. History students from the University of Toliara, Mex Tafitasoa, Nadya
Ravololonirina, and Norbert Antilahy, and some villagers (Photo 1) participated in the work. The Network of African Archaeologists in Madagascar, under the direction of Dr Chantal Radimilahy and Prof Barthélemy Manjakahery, provided the logistical equipment used. The various signatures The various signatures of national and local administrative approval, the security provided by the Gendarmerie Brigade of Benenitsy, and above all the kindness and openness of King Zafimanely (Photo 1) – who allowed us access to the (burial) site – contributed enormously to the accomplishment of the expedition.
Madagascar. History students from the University of Toliara, Mex Tafitasoa, Nadya
Ravololonirina, and Norbert Antilahy, and some villagers (Photo 1) participated in the work. The Network of African Archaeologists in Madagascar, under the direction of Dr Chantal Radimilahy and Prof Barthélemy Manjakahery, provided the logistical equipment used. The various signatures The various signatures of national and local administrative approval, the security provided by the Gendarmerie Brigade of Benenitsy, and above all the kindness and openness of King Zafimanely (Photo 1) – who allowed us access to the (burial) site – contributed enormously to the accomplishment of the expedition.
The rock shelter at Ampasimaiky faces east. This arrangement makes the painting panels exposed to the sun in the morning and in the shade beyond noon. This arrangement suggests that painting activities are conducted either early in the morning while the heat of the sun is still moderate, or beyond noon when shade is available below the shelter.
This is based on the fact that the height of the panels does not exceed 130 cm, suggesting that the paintings were intended for people sitting on the floor of the undercroft shelter
Photo 2). Minimal, therefore, is the likelihood of the painting activity – involving seated people – having been conducted during the time when the sun is heating the shelter floor.
However, the view outside the site is limited. The opposing hills – on the other bank of the Ampasimaiky stream – prevent a person standing under the shelter from having a good view of the surroundings (Photo 3). It is much more likely that the view outwards – which implies a defensive strategy (Smith 1995) – was not among the painter’s concerns when he/she chose the site. This demonstrates that the layout of the Ampasimaiky rock shelter favours invaders from unexpectedly attacking its residents. The place would therefore not be suitable for permanent habitation.
Photo 3
Especially since a person standing on the hills of the other bank would be able to look directly at the other one positioned under the rock shelter, unless he hides in the deeper part of the shelter. Since the wall of this “chamber”, according to the local population, is not painted, it is plausible that the activities related to the paintings in the Ampasimaiky rock shelter are not clandestine. However, the paintings were not intended to be seen by people passing close to the site. They are not seen from a distance. Not only are the panels placed below 130 cm in height, but the slope of the ground before the access
to the shelter also prevents anyone from seeing them until they get close to the entrance. Nevertheless, the use of darker coloured paints (red and black) suggests the painter’s intention to contrast the paints he/she was using against the pale colour of the sandstone wall. This implies his concern for the visibility of his paintings.
The walls of the rock shelter have red, reddish orange, burgundy red, black, and white paintings in monochrome, bichrome and polychrome patterns. The images
are painted with the fingertip. They depict zoomorphic (mainly zebu) and anthropomorphic (semi-naturalist and stylised) figures, as well as schematic-geometric-amorphic symbols (Figure 3). The latter dominate and are represented by quadrangular, circular and elliptical signs, dot/dash lines, and alphabet-like signs. In total, the Ampasimaiky rock paintings have 166 recognisable images.
Analysis and Interpretation
Based on comparative studies of the images, it has been found that the images of the Ampasimaiky rock paintings are not different from the images depicted on the prehistoric SGA rock art of the African continent (see Chami, 2008; Chaplin, 1974; Coulson and Campbell 2001;
Eastwood and Smith, 2005; Lynch and Donahue, 1980; Mabulla, 2005; Smith, 1995; Teka 2008; Willcox 1984). The similarities apparently suggest links between the two regions (Madagascar and the mainland) in the
(Madagascar and the continent) in prehistoric times. This suggests that the Ampasimaiky rock paintings are of African origin and could be relatively dated to the same period as those on the continent (Thackeray, 1983: 22). Comparative studies of the images have also shown that more than 30 types of images in the Ampasimaiky rock paintings are closely related to different ancient scripts such as Libyco-Berber, Protosinaitic, Meroitic (cursive), Punic Iberian, Ancient Phoenician, Tifinagh, and Aegean, Cretan and Proto-Egyptian linear signs respectively (Figure 4). These similarities would also suggest that the SGA-type rock paintings of Ampasimaiky are ancient writing.
figure 4
Analysis and Interpretation
Based on comparative studies of the images, it has been found that the images of the Ampasimaiky rock paintings are not different from the images depicted on the prehistoric SGA rock art of the African continent (see Chami, 2008; Chaplin, 1974; Coulson and Campbell 2001;
Eastwood and Smith, 2005; Lynch and Donahue, 1980; Mabulla, 2005; Smith, 1995; Teka 2008; Willcox 1984). The similarities apparently suggest links between the two regions (Madagascar and the mainland) in the
(Madagascar and the continent) in prehistoric times. This suggests that the Ampasimaiky rock paintings are of African origin and could be relatively dated to the same period as those on the continent (Thackeray, 1983: 22). Comparative studies of the images have also shown that more than 30 types of images in the Ampasimaiky rock paintings are closely related to different ancient scripts such as Libyco-Berber, Protosinaitic, Meroitic (cursive), Punic Iberian, Ancient Phoenician, Tifinagh, and Aegean, Cretan and Proto-Egyptian linear signs respectively (Figure 4). These similarities would also suggest that the SGA-type rock paintings of Ampasimaiky are ancient writing. them (Willcox, 1984: 210); the same has been applied to that of Professor Barry Fell, a specialist in Libyan-Berber writing. The same has been applied to that of Professor Barry Fell, a specialist in Libyan-Berber writing, who translated the glyphs of Driekopseiland in southern Africa as Libyan-Berber writing. But it has also been refuted on the grounds that the Bushmen could not write (Willcox 1984: 210); it is important to recall
It is important to recall in this connection that Libyco-Berber inscriptions inscribed on petroglyphs in the Canary Islands were not accepted as such, under the pretext that the Guanches (indigenous people of the Canary Islands) could not have acquired or adopted any writing system in prehistoric times (Cline 1953; Farrujia de la Rosa et al. 2010: 16-17; Willcox 1984: 48). This erroneous model persisted until reputable scholars succeeded in translating an inscription reading “save me from javelin annihilation” (Cline 1953: 273). Today, it is commonly accepted that the groups of aligned glyphs from the Canary Islands are Libyco-Berber inscriptions. Recently, Chami (2006: 84-5) insisted that African SGA rock paintings and engravings are ancient scripts. He has shown that some, such as those at Namoratunga in Kenya, are arranged. For the accounts of those that
are not arranged, he argued that at the time when the ancient scripts of Ethiopia, India, China, Egypt and Mesopotamia were still using ideograms, the characters too were not aligned (see also Casson 1966: 141). This can be amplified by the case of some Sahelian writing systems such as the ‘Nsibidi. This script has no ‘order in the arrangement’, but it is practically recognised as a writing system (Dayrell
1910; Macgregor 1909). Zahan (1950) identified the Dogon and Bambara pictographs as scripts based on the argument that
“writing consists essentially in the communication of thought by means of conventional signs, and the pictographs of both Dogon and Bambara serve to convey thoughts from one writer to another or from writer to lay public” (pp.137) « Une écriture consiste essentiellement de la communication de pensée par le moyen de signes conventionnels, et les pictographes de Dogon et de Bambara servent à transporter des pensées d’un auteur à l’autre ou d’un auteur au public”.
Therefore, it is also possible that without the “order in the arrangement”, the Ampasimaiky SGA-type rock paintings could have served to communicate thoughts and information. This would qualify them as a WRITING SYSTEM! Especially since some groups of symbols are horizontally and vertically aligned (Photo 4). Since “order in arrangement” has been established by some scholars as the sine qua non for a group of symbols/signs to qualify as an inscription (Willcox 1984: 210), one can reasonably assert, based on the above evidence, that the
AMPASIMAIKY RUPESTRATED ROCK SHELTER PAINTINGS ARE INSCRIPTIONS! This assertion is clearly supported by the vertically arranged group of symbols in panel 8 (see Photo 4b). Indeed, the repetition of the first sign in the third character (read downwards) is undoubtedly demonstrating the USE OF THE ALPHABET, and thus the
ALPHABET, so this is WRITING! All the more so as all five characters constituting the inscription are closely related to the Libyan-Berber alphabets (see
www.ancientscripts.com), the proto-Egyptian, Aegean and Cretan linear signs respectively
(Figure 5).
Figure 5 : Le groupe de signes verticalement arrangées d’Ampasimaiky comparé avec des caractères alphabétiques libyco-berbères et des signes linéaires proto-Egyptiens, égéens, et crétois (Tableau : Evans, 1897 :386).
Bien que des signes linéaires proto-Égyptiens, crétois et égéens aient été également verticalement écrits, je favoriserais l’origine libyco-berbère de l’inscription d’Ampasimaiky sous l’influence de la description de Walter Cline concernant les inscriptions libyco-berbères comme
suit).
“They [Libyco-Berber inscriptions] are usually very brief, consisting of personal names and a few phrases … The characters are simple and predominantly rectilinear. Variant alphabet may have existed. Most of inscriptions are to be read upward beginning with the left-hand column, though horizontal rows reading from right to left are not uncommon and some columns read downwards. The words are not separated. Though these signs sometimes occur in groups of four or five, they often stand singly…” (Cline 1953: 273)
All the underlined sentences qualify the vertically arranged inscription of Ampasimaiky as being of Libyco-berberber origin. Although this ancient script was written either horizontally or vertically, it should be remembered that the Libyco- Berber peoples used two different alphabets for each row (see http://www.ancientscripts.com). All five characters of the Ampasimaiky inscription are exclusively related to the vertical Libyco- Berber alphabet (Figure 6). Transliterating them to Latin alphabets – based on the transliteration of www.ancientscripts.com – one could read “ğ r ğ t f” (downwards) or “f t ğ r ğ” (upwards). The phonetic value ‘r’ for the presumed circle/square sign can be read as ‘3’ (= a), if the represented sign is perceived as a triangular symbol (see Fell 1975: 271). In the present state of the study, decoding the meaning of the inscription is not yet within the scope of the author. This task would rather be reserved for palaeographers and epigraphers.
Figure 7: The group of vertically arranged signs of Ampasimaiky transliterated in the Libyco-Berber alphabet
It is important to point out here that certain Libyco-Berber inscriptions brought back from the Canary Islands are not different from that of Ampasimaiky (see Figure 8). You can clearly see their resemblance.
Figure 8: The group of vertically arranged signs from Ampasimaiky compared with a Libyco-Berber inscription from the Canary Islands (Drawings by Farrujia de la Rosa et al. 2010: 29)
Discussion and Conclusion
Some of the images found at Ampasimaiky are similar to those of the rock art of continental Africa. However, the painting tradition of southwestern Madagascar is not
However, the painting tradition of southwestern Madagascar does not seem to originate from the continental region closest to Madagascar, i.e. Mozambique and the rest of the central Limpopo basin. In fact, Professor Benjamin Smith, an internationally recognised authority on the study of SGA rock paintings in sub-Saharan Africa, has taken the view that they are not typically affiliated with the Khoekhoen SGA rock paintings, nor with the geometric art of central southern Africa (pers. comm.). The exclusive representation of humped cattle associated with SGA-type symbols is particular to Ampasimaiky. Since no ancient representation of zebu (Bos indicus) has yet been found in the region between Sudan in the north and the Zambezi River in the south (Magnavita 2006: 62), it can be inferred that the tradition of rock paintings in Ampasimaiky originated in North Africa. Evidence of a bronze zebu figurine (of the type Bos indicus) inscribed with SGA signs dating back to 200 CE at Zeban Kutur, Eritrea (Photo 4) would confirm the early use of SGA signs among pastoral zebu-breeding peoples such as those depicted in the Ampasimaiky rock paintings
Photo 4: Bronze figure of zebu inscribed with SGA type signs in Zeban Kutur, Eritrea (Photo: Magnavita 2006:63)
The Libyan-Berber inscription discovered at Ampasimaiky also reinforces this proposal. The Libyan-Berber script emerged in North Africa around 500 BC and ceased to be used around 800 AD (Cline 1953).
around 800 AD (Cline 1953). This implies that the Libyan-Berber inscription of Ampasimaiky should have been written during the period between this bracket of dates (500 BC – AD 800). It is worth mentioning that Iamboulos, a Greek merchant who lived in the 3rd
century BC, reported about a vertical order of writing that was practised by the inhabitants of the ‘island of the sun’ (Allibert 2010: 111; Kobishschanow, 1965: 138; Oldfather, 1961 vol.2: 65-67; Winston 1976: 221). This island is believed to be geographically located adjacent to the east coast of Africa, south of the Equator. Although some of the geographical and ecological information in Iamboulos’ account is very much in line with that of the Malagasy west coast (for more details see the
Although some of the geographical and ecological information in Iamboulos’ account closely matches that of the Malagasy west coast (for more information see Allibert 2010), some scholars have consistently judged this account to be fictional, imaginary, and fanciful (Cary and Warmington 1963: 233-245; Winston 1976). One of the criteria used to censor this narrative is the use of vertical order writing by the islanders. Certainly, it was difficult for some scholars to accept the existence of a pre-Islamic script used in the adjacent islands of the east coast of Africa such as Madagascar (Janvier 1975: 19). However, with the discovery of the Libyan-Berber inscription at Ampasimaiky, this denial no longer holds. New evidence is uncovered to confirm that the “island of the sun” of Iamboulos could have been Madagascar. In any case, man was already present in the vicinity of Ampasimaiky when the Greek merchant passed through the island. Human activity is attested by subfossil lemur bones bearing cutting marks and radiocarbon-dated from the 5th to the 4th century BC at Taolambiby, in the upper Onilahy valley (see Perez et al., 2005). All the more so since ancient written documents (see Jones 1960: 235, 269, 315; Strabo 17, 1, 13 in Lacroix 1998: 95), archaeology (see Cary and Warmington 1963: 130; Chami 2007: 6; 2009; Wainwright 1940: 165), genetic study (see Ricaut et al. 2009: 5-6), and zooarchaeology (Chami 2006: 137; Grigson 1991: 124) all confirm the ancient contacts that existed during this period between the Onilahy Valley and the Onilahy Valley,
During this period, the Nile Valley and the Mediterranean world on the one hand, and the eastern coast of Africa and the adjacent islands on the other.
Références Allibert, C. (2008), “Austronesian Migration and the Establishment of the Malagasy Civilization: Contrasted Readings in Linguistics, Archaeology, Genetics and Cultural Anthropology”, Diogenes, SAGE, Volume 218, pp.7-16 Allibert, C. (2010), “L’île de Iamboulos est-elle Madagascar? », Pount, no.4, pp.103-119 Blench, R. (2007), “New Palaeozoogeographical Evidence for the Settlement of Madagascar”, Azania, Volume XLII, pp.69-82 Cary M. and Warmington, E. (1963), The Ancient Explorers, Middlesex, Penguin Books Casson, L. (1966), Ancient Egypt, Nederland, Time-life International Chami, F. A (2006), The Unity of African Ancient History, 3000 BC to AD 500, Dar es Salaam, E&D Limited Chami, F. A (2008), “The Great Lakes: a complexity of cultural wellsprings”, in Arnold, M. (eds.) Art in Eastern Africa, pp.47-64, Dar-es-Salaam, Mkuki na Nyota Chami F. A and Kwekason A. (2003), “Neolithic Pottery Traditions from the Islands, the Coast, and the Interior of East Africa”, African Archaeological Review, Volume 20, No. 2, pp.65-80 Chami, F. A, Tabibou, I. A and Abderhamane, B. (2009), “Preliminary Report of Archaeological Work Conducted on the Southern Ngazidja Island”, in Chami, F. (eds.), Zanzibar and the Swahili Coast from c30.000 years ago, pp.115-128, Dar es Salaam, E&D Vision Publishers Chaplin, T. (1974), “The Prehistoric Rock Art of the Lake Victoria Region”, Azania Volume 9, pp.1-50 Cline, W. (1953), “Berber Dialects and Berber Script”, Southwestern Journal of Anthropology, Vol.9, No.3, pp.268-276, New Mexico, New Mexico University Press Coulson D. and Campbell A. (2001), African Rock Art, Paintings and Engravings on Stone, New York, Harry N. Abrams Incorporated Dayrell, E. (1910), “Some “Nsibidi” Signs”, Man, Vol.10, pp.113-114. Dewar, R. E (1996), “The Archaeology of the Early Settlement of Madagascar”, in Reade J. (eds.) The Indian Ocean in Antiquity, pp.471-486 Dewar, R. E (1997), “Does It Matter That Madagascar Is an Island?” Human Ecology, Volume 25, No. 3, pp. 481-489 Donque, G. (1965), « Le contexte océanique des anciennes migrations : vents et courants dans l’océan indien », Taloha, Volume 1, pp.43-59. Eastwood, E. B and Smith, B. W (2005), “Fingerprints of the Khoekhoen: Geometric and handprinted Rock Art in the Central Limpopo Basin, Southern Africa”, Goodwin Series, RSA, South African Archaeological Society, Volume 9, pp.63-76 Evans, A. J (1897), “Further Discoveries of Cretan and Aegean Script: With Libyan and ProtoEgyptian Comparisons”, The Journal of Hellenic Studies, The Society for the Promotion of Hellenic Studies, Volume 17, pp.327-395 Farrujia de la Rosa, A., Pichler W., Rodrigue A., Marín S.G. 2010, The Lybico-Berber and Latino-Canarian Scripts and the Colonization of the Canary Islands, African Archaeological Review, vol.27, no.1, pp.13-41 Faublée, J. (1947), Récits Bara, Travaux et mémoires de l’Institut d’ethnologie XLXII, Paris, Université de Paris Fell, B. (1975), America BC: Ancient Settlers in the New World, New York, A Demeter Press Book Grandidier, A., Charles-Roux J., Delhorbe C., Froidevaux H. and Grandidier G. (1902), Collection des ouvrages anciens concernant Madagascar, Tome II, Paris, Comité de Madagascar Grigson, C. (1991), “An African Origin of African Cattle? Some Archaeological Evidence”, The African Archaeological Review, Volume 9, pp.119-144 Hall, S. and Smith, B. (2000), “Empowering Places: Rock Shelters and Ritual Control in FarmerForager Interactions in the Northern Province”, Goodwin Series, RSA, South African Archaeological Society, Volume 8, pp.30-46 Hébert, J. C (1967), « Essai d’interprétation de la stèle indéchiffrée d’Ambilobe », Taloha, Volume 2, pp.XLVII-LVII Hébert, J. C (1971), « Les tatouages de la côte est de Madagascar, d’après Chapelier (1794- 1806) », Taloha, Volume 4, pp.211-223 Huygues-Belrose, V. (1997). Les supports anciens de l’écriture à Madagascar, Etudes Océan Indien, vol.22, pp.7-19, INALCO Janvier, Y. (1975), « La géographie gréco-romaine a-t-elle connu Madagascar? Le point de la question », Omaly sy Anio, No.1-2, pp.11-41 Jones, H. (1960), The Geography of Strabo, Volume 1-2, London, William Heinemann Ltd. Juma, A. (2004), Unguja Ukuu on Zanzibar – An Archaeological Study of Early Urbanism, Sweden, Uppsala University Kessy, E. (2009a), “Analysis of Lithic Artefacts From Mwanampambe and Kuumbi Caves, Zanzibar”, in Chami F. (eds.), Zanzibar and the Swahili Coast from c30,000 years ago, Dar es Salaam, E&D Vision Publishers, pp.131-143 Kessy, E. (2009b), “Stone Artefacts from Male”, in Chami F. (eds.), Zanzibar and the Swahili Coast from c30,000 years ago, Dar es Salaam, E&D Vision Publishers, pp.184-88 Kobishchanov, Y.M. (1965), “On the Problem of Sea Voyages of Ancient Africans in the Indian Ocean”, The Journal of African History, Vol.6, No.2, pp.137-141, Cambridge, Cambridge University Press. Lacroix, W. F (1998), Africa in Antiquity, Nijmegen, Catholic University Lynch, B. M and Donahue, R. (1980), “A Statistical Analysis of Two Rock-Art Sites in
Northwest Kenya”, Journal of Field Archaeology, Volume 7, pp.75-85
Mabulla, A. Z (2005), “The Rock Art of Mara Region, Tanzania”, Azania, Volume 40, 19-42
Macgregor, J. K (1909), “Some Notes on Nsibidi”, The Journal of the Royal Institute of Great
Britain and Ireland, Volume 39, pp.209-219
MacPhee R. D and Burney D. (1991), “Dating of modified femora of extinct dwarf
Hippopotamus from southern Madagascar: implications for constraining human
colonization and vertebrate extinction events”, Journal of Archaeological Science, No.
18, pp. 695-706
Magnavita, C. (2006), “Ancient Humped Cattle in Africa: A View from the Chad Basin”, The
African Archaeological Review, Volume 23, No. 3/4, pp.55-84
Oldfather, C. H (1961), Diodorus of Sicily, 12 Volumes, London, William Heinemann
Perez, V. R Godfrey L. R, Nowak-Kemp M., Burney D. A, Ratsimbazafy J. and Vasey N.
(2005), “Evidence of Early Butchery of Giant Lemurs in Madagascar”, Journal of
Human Evolution, No. 49, pp.722-742
Radimilahy, C. (2011), « Contribution á l’Archéologie du Sud-Ouest de Madagascar »,
Civilisations des mondes insulaires (Madagascar, îles du canal de Mozambique,
Mascareignes, Polynésie, Guyanes), Paris, Éditions Karthala, pp.825-853
Rasolondrainy, T. (2011), Archaeological Study of the Prehistoric Rock Paintings of
Ampasimaiky Rock Shelter, in the Upper Onilahy, Isalo Region, Southwestern
Madagascar. M.A. Dissertation unpublished, University of Dar es Salaam.
Ricaut, F., Razafindrazaka H., Cox, M., Dugoujon J., Guitard E., Sambo C., Mormina M.,
Mirazon-Larh M., Ludes B., and Crubézy E. 2009, A New Deep Branch of Eurasian
mtDNA Macrohaplogroup M Reveals Additional Complexity Regarding the Settlement
of Madagascar. BMC Genomics, vol.10, pp.1-9
Smith, B. W (1995), Rock Art in South-Central Africa. A Study Based on the Pictographs of
Dedza District, Malawi and Kasama District Zambia, Unpublished Phd Thesis,
University of Cambridge
Southall, A. (1975), “The Problem of Malagasy Origins”, in Chittick H. N and Rotberg R. I
(eds.), East Africa and the Orient, Cultural Syntheses in Pre-Colonial Times, New
York, Africana Publishing Company, pp. 192-215
Teka, Z. (2008). “Distribution and Significance of Ancient Rock Art Sites in Eritrea”, in
Schmidt P. et al (eds.), The Archaeology of Ancient Eritrea. Eritrea: The Red Sea Press,
Inc., pp.49-61
Thackeray, A. (1983), “Dating the Rock Art of Southern Africa”, Goodwin Series, Volume 4,
pp.21-26.
Vérin P., (1975), “Austronesian Contributions to the Culture of Madagascar: Some
Archaeological Problems”, in Chittick H. N and Rotberg R. I (eds.), East Africa and the
Orient, Cultural Syntheses in Pre-Colonial Times, New York, Africana Publishing
Company, pp. 164-191Vérin, P. (1981), “Madagascar”, in Mokhtar G. (eds.) General History of Africa, Volume 2, 693-
717
Wainwright, G. A (1940), “The Egyptian Origin of the New Year’s Sacrifice at Zanzibar”, Man,
Volume 40, pp.164-167
Willcox, A. R (1984), The Rock Art of Africa, New York, Holmes and Meier Publishers
Winston, D. (1976), “Iambulus’ “Island of the Sun” and Hellenistic Literary Utopias”, Science
Fiction Studies, SF-TH Inc., Volume 3, No. 3, pp.219-227
Zahan, D. (1950), “Pictographic Writing in the Western Sudan”, Man, Volume 50, pp.136-138
Northwest Kenya”, Journal of Field Archaeology, Volume 7, pp.75-85
Mabulla, A. Z (2005), “The Rock Art of Mara Region, Tanzania”, Azania, Volume 40, 19-42
Macgregor, J. K (1909), “Some Notes on Nsibidi”, The Journal of the Royal Institute of Great
Britain and Ireland, Volume 39, pp.209-219
MacPhee R. D and Burney D. (1991), “Dating of modified femora of extinct dwarf
Hippopotamus from southern Madagascar: implications for constraining human
colonization and vertebrate extinction events”, Journal of Archaeological Science, No.
18, pp. 695-706
Magnavita, C. (2006), “Ancient Humped Cattle in Africa: A View from the Chad Basin”, The
African Archaeological Review, Volume 23, No. 3/4, pp.55-84
Oldfather, C. H (1961), Diodorus of Sicily, 12 Volumes, London, William Heinemann
Perez, V. R Godfrey L. R, Nowak-Kemp M., Burney D. A, Ratsimbazafy J. and Vasey N.
(2005), “Evidence of Early Butchery of Giant Lemurs in Madagascar”, Journal of
Human Evolution, No. 49, pp.722-742
Radimilahy, C. (2011), « Contribution á l’Archéologie du Sud-Ouest de Madagascar »,
Civilisations des mondes insulaires (Madagascar, îles du canal de Mozambique,
Mascareignes, Polynésie, Guyanes), Paris, Éditions Karthala, pp.825-853
Rasolondrainy, T. (2011), Archaeological Study of the Prehistoric Rock Paintings of
Ampasimaiky Rock Shelter, in the Upper Onilahy, Isalo Region, Southwestern
Madagascar. M.A. Dissertation unpublished, University of Dar es Salaam.
Ricaut, F., Razafindrazaka H., Cox, M., Dugoujon J., Guitard E., Sambo C., Mormina M.,
Mirazon-Larh M., Ludes B., and Crubézy E. 2009, A New Deep Branch of Eurasian
mtDNA Macrohaplogroup M Reveals Additional Complexity Regarding the Settlement
of Madagascar. BMC Genomics, vol.10, pp.1-9
Smith, B. W (1995), Rock Art in South-Central Africa. A Study Based on the Pictographs of
Dedza District, Malawi and Kasama District Zambia, Unpublished Phd Thesis,
University of Cambridge
Southall, A. (1975), “The Problem of Malagasy Origins”, in Chittick H. N and Rotberg R. I
(eds.), East Africa and the Orient, Cultural Syntheses in Pre-Colonial Times, New
York, Africana Publishing Company, pp. 192-215
Teka, Z. (2008). “Distribution and Significance of Ancient Rock Art Sites in Eritrea”, in
Schmidt P. et al (eds.), The Archaeology of Ancient Eritrea. Eritrea: The Red Sea Press,
Inc., pp.49-61
Thackeray, A. (1983), “Dating the Rock Art of Southern Africa”, Goodwin Series, Volume 4,
pp.21-26.
Vérin P., (1975), “Austronesian Contributions to the Culture of Madagascar: Some
Archaeological Problems”, in Chittick H. N and Rotberg R. I (eds.), East Africa and the
Orient, Cultural Syntheses in Pre-Colonial Times, New York, Africana Publishing
Company, pp. 164-191Vérin, P. (1981), “Madagascar”, in Mokhtar G. (eds.) General History of Africa, Volume 2, 693-
717
Wainwright, G. A (1940), “The Egyptian Origin of the New Year’s Sacrifice at Zanzibar”, Man,
Volume 40, pp.164-167
Willcox, A. R (1984), The Rock Art of Africa, New York, Holmes and Meier Publishers
Winston, D. (1976), “Iambulus’ “Island of the Sun” and Hellenistic Literary Utopias”, Science
Fiction Studies, SF-TH Inc., Volume 3, No. 3, pp.219-227
Zahan, D. (1950), “Pictographic Writing in the Western Sudan”, Man, Volume 50, pp.136-138
Tanambelo V.R. RASOLONDRAINY,
(Centre de Documentation et de Recherche sur l’Art et la Tradition Orale de Madagascar)
Abstract:
This article presents the result of an archaeological study on the first prehistoric rock paintings site found in Madagascar and the rest of islands of Western Indian Ocean south of the equator.
Based on systematic methods and techniques, the study provides groundbreaking archaeological evidences that contribute largely to the understanding of the prehistory of Madagascar which was not archaeologically known thitherto to have prehistoric rock paintings, nor human settlement prior to 8th century AD. A Libyco-Berber inscription was discovered on the rock paintings of Ampasimaiky Rock Shelter, in Southwestern Madagascar. Based on this writing -supported by ancient historical records, circumstantial archaeological and paleontological evidences, linguistic data, and to some extent DNA analysis results -, the present study provides minimum age falling in the bracket time around 500 BC to roughly 8th century AD for the rock paintings of Ampasimaiky rock shelter. Furthermore, the study traces back early contacts between Madagascar and Northern Africa during prehistoric times.
Introduction
With the kind assistance of Dr Chantal Radimilahy, National Coordinator of the Network of African Archaeologists in Madagascar, I researched the ethnographic exhibition on the website of the Museum of Ethnography in Geneva (MEG). The theme was the peoples and cultures of the southwestern region of Madagascar, and was based on an exhibition of photographs taken by the French ethnographer Jacques Faublée in 1939. The pictures included rock paintings found in caves and rock shelters in the Isalo region. After discussions with my teachers at the Archaeology Unit of the University of Dar es Salaam, Professors Felix Chami, Fidelis Masao, and Audax Mabulla, I further investigated the significance of these paintings for the prehistoric period of Madagascar. This study is taken from a Master of Arts in Archaeology entitled
“Archaeologial study of the prehistoric rock paintings of Ampasimaiky rock shelter, in the upper Onilahy, Isalo region, Southwestern Madagascar”.
————–
Africa has many archaeological sites with prehistoric rock paintings (see Coulson and Campbell 2001; Willcox 1984). However, no such sites have yet been scientifically
scientifically reported from Madagascar, which is only 400 kilometres from southern Africa where they are abundant. All rock art previously found in Madagascar has been judged to have been produced in the recent past (Faublée 1947: 164; Hébert 1971: 218). This judgement is based on the theory that Africans were incapable of having navigated the high seas during prehistoric times (Donque 1965: 43; Grandidier et al. 1902; Southall 1975: 194; Vérin 1975, 1981). It was therefore out of the question to think that they had the possibility of transferring the practice of rock painting to the adjacent islands of Africa such as Madagascar. However, recent archaeological findings have demonstrated that the east coast of Africa and adjacent islands were already in contact since the prehistoric period (see Chami and Kwekason 2003; Chami et al. 2009; Juma 2004; Kessy 2009a,
2009b). These prompt a problem that led me to revise the antiquity of the rock paintings in Madagascar.
On the other hand, the name of the Arabic-Malagasy script – Sorabe meaning “great script” – indicates that there was an ancient script in Madagascar that was distinguished from Arabic (Hébert,
1967:XLVII; Huygues-Belrose 1997: 8). Although Hébert (1967:XLVIII) presumed that such a hypothetical script would have to be of Indian origin and would be discovered from
Although Hébert (1967:XLVIII) assumed that such a hypothetical script would have to be of Indian origin and would be discovered from stele and parietal inscriptions, no such inscription has so far been found in Madagascar.
My problem was therefore to find out whether this would appear on the rock paintings reported by the website. What provoked this idea the most was that Chami (2006:84-85; 2008) argued that the African Geometric and Amorphous Schematic (GAS) rock paintings and engravings are ancient writing.
Another problem is that radiocarbon 14 dating placed the earliest human settlement of Madagascar around the 8th century AD (Allibert 2008: 7; Dewar 1996: 474).
Although earlier dates are still being considered, many researchers have assumed that settlement of Madagascar would not go beyond the year W AD (for more information see Dewar 1997). This assumption has made research on Malagasy material culture going back to the Stone Age more than difficult. Any findings of material cultures dating from prehistoric times have been discarded, overlooked or misinterpreted. However, subfossil animal bone remains bearing traces of butchery found in paleontological sites in the southwestern region indicate the presence of human activity
(For more information, see Blench 2007; Perez et al. 2005). Despite this evidence, an archaeological site has not yet been
archaeological site with similar dates has not yet been found on the entire island.
With the discovery of a rock shelter with rock paintings, discovering new information about the prehistory of Madagascar before the 8th century proved to be feasible. All the more so as some of the paintings and engravings are
Especially since some of the SGA-type rock paintings and engravings found on the African continent – similar to those found in the photos on the MEG website – have been radiocarbon dated from the 1st millennium BC to the early 1st millennium AD (Chami, 2008: 54; Hall and Smith 2000; Lynche and Donahue 1980; Willcox 1984: 26, 83.
Fieldwork and results
Despite various challenges and constraints – mainly caused by the lack of archaeological reconnaissance prior to the fieldwork, insecurity in the study area, and the sacredness of the site to be studied – we had only somehow managed to conduct survey, archaeological excavation, and documentation of the paintings in the Ampasimaiky rock shelter. It is only the result of the latter that will be presented in this article.
The rock painting site is a rock shelter located at the foot of a hill immediately next to a small stream called Ampasimaiky (Figure 2), from which the name of the site derives. The presence of this water source may have influenced the painter’s choice of location.
The shelter itself is 26.8m long (south-north axis) and 5m wide (inwards).
Its ceiling begins to slope inwards from 4m high at its entrance to approximately 1.5m high at its irregular wall. The deepest part of the rock shelter – a chamber – is currently used by the royal family “Bara Zafimanely” as a burial place. Dung and ox hoof prints found on the ground indicate that cattle frequent the place. This occurrence should have disturbed the soil.
However, the areas around the site are dominated by rocky hills covered with grass. Only in the deep canyon near the confluence of Ampasimaiky and the Sakamarekely River is a patch of thick vegetation up to 10 m high observed along the stream. This situation would make agricultural activity almost impossible around the site, unless one had to move 1.5 km to the bank, where rice cultivation takes place today.
An archaeological survey of the southern Isalo region (Figure 1) was carried out and a rock painting site was discovered. Professor Chami participated in the fieldwork, advising on the approach to be followed in the treatment of these previously unstudied cave cultures in Madagascar.
Madagascar. History students from the University of Toliara, Mex Tafitasoa, Nadya
Ravololonirina, and Norbert Antilahy, and some villagers (Photo 1) participated in the work. The Network of African Archaeologists in Madagascar, under the direction of Dr Chantal Radimilahy and Prof Barthélemy Manjakahery, provided the logistical equipment used. The various signatures The various signatures of national and local administrative approval, the security provided by the Gendarmerie Brigade of Benenitsy, and above all the kindness and openness of King Zafimanely (Photo 1) – who allowed us access to the (burial) site – contributed enormously to the accomplishment of the expedition.
Madagascar. History students from the University of Toliara, Mex Tafitasoa, Nadya
Ravololonirina, and Norbert Antilahy, and some villagers (Photo 1) participated in the work. The Network of African Archaeologists in Madagascar, under the direction of Dr Chantal Radimilahy and Prof Barthélemy Manjakahery, provided the logistical equipment used. The various signatures The various signatures of national and local administrative approval, the security provided by the Gendarmerie Brigade of Benenitsy, and above all the kindness and openness of King Zafimanely (Photo 1) – who allowed us access to the (burial) site – contributed enormously to the accomplishment of the expedition.
The rock shelter at Ampasimaiky faces east. This arrangement makes the painting panels exposed to the sun in the morning and in the shade beyond noon. This arrangement suggests that painting activities are conducted either early in the morning while the heat of the sun is still moderate, or beyond noon when shade is available below the shelter.
This is based on the fact that the height of the panels does not exceed 130 cm, suggesting that the paintings were intended for people sitting on the floor of the undercroft shelter
Photo 2). Minimal, therefore, is the likelihood of the painting activity – involving seated people – having been conducted during the time when the sun is heating the shelter floor.
However, the view outside the site is limited. The opposing hills – on the other bank of the Ampasimaiky stream – prevent a person standing under the shelter from having a good view of the surroundings (Photo 3). It is much more likely that the view outwards – which implies a defensive strategy (Smith 1995) – was not among the painter’s concerns when he/she chose the site. This demonstrates that the layout of the Ampasimaiky rock shelter favours invaders from unexpectedly attacking its residents. The place would therefore not be suitable for permanent habitation.
Photo 3
Especially since a person standing on the hills of the other bank would be able to look directly at the other one positioned under the rock shelter, unless he hides in the deeper part of the shelter. Since the wall of this “chamber”, according to the local population, is not painted, it is plausible that the activities related to the paintings in the Ampasimaiky rock shelter are not clandestine. However, the paintings were not intended to be seen by people passing close to the site. They are not seen from a distance. Not only are the panels placed below 130 cm in height, but the slope of the ground before the access
to the shelter also prevents anyone from seeing them until they get close to the entrance. Nevertheless, the use of darker coloured paints (red and black) suggests the painter’s intention to contrast the paints he/she was using against the pale colour of the sandstone wall. This implies his concern for the visibility of his paintings.
The walls of the rock shelter have red, reddish orange, burgundy red, black, and white paintings in monochrome, bichrome and polychrome patterns. The images
are painted with the fingertip. They depict zoomorphic (mainly zebu) and anthropomorphic (semi-naturalist and stylised) figures, as well as schematic-geometric-amorphic symbols (Figure 3). The latter dominate and are represented by quadrangular, circular and elliptical signs, dot/dash lines, and alphabet-like signs. In total, the Ampasimaiky rock paintings have 166 recognisable images.
Analysis and Interpretation
Based on comparative studies of the images, it has been found that the images of the Ampasimaiky rock paintings are not different from the images depicted on the prehistoric SGA rock art of the African continent (see Chami, 2008; Chaplin, 1974; Coulson and Campbell 2001;
Eastwood and Smith, 2005; Lynch and Donahue, 1980; Mabulla, 2005; Smith, 1995; Teka 2008; Willcox 1984). The similarities apparently suggest links between the two regions (Madagascar and the mainland) in the
(Madagascar and the continent) in prehistoric times. This suggests that the Ampasimaiky rock paintings are of African origin and could be relatively dated to the same period as those on the continent (Thackeray, 1983: 22). Comparative studies of the images have also shown that more than 30 types of images in the Ampasimaiky rock paintings are closely related to different ancient scripts such as Libyco-Berber, Protosinaitic, Meroitic (cursive), Punic Iberian, Ancient Phoenician, Tifinagh, and Aegean, Cretan and Proto-Egyptian linear signs respectively (Figure 4). These similarities would also suggest that the SGA-type rock paintings of Ampasimaiky are ancient writing.
figure 4
Analysis and Interpretation
Based on comparative studies of the images, it has been found that the images of the Ampasimaiky rock paintings are not different from the images depicted on the prehistoric SGA rock art of the African continent (see Chami, 2008; Chaplin, 1974; Coulson and Campbell 2001;
Eastwood and Smith, 2005; Lynch and Donahue, 1980; Mabulla, 2005; Smith, 1995; Teka 2008; Willcox 1984). The similarities apparently suggest links between the two regions (Madagascar and the mainland) in the
(Madagascar and the continent) in prehistoric times. This suggests that the Ampasimaiky rock paintings are of African origin and could be relatively dated to the same period as those on the continent (Thackeray, 1983: 22). Comparative studies of the images have also shown that more than 30 types of images in the Ampasimaiky rock paintings are closely related to different ancient scripts such as Libyco-Berber, Protosinaitic, Meroitic (cursive), Punic Iberian, Ancient Phoenician, Tifinagh, and Aegean, Cretan and Proto-Egyptian linear signs respectively (Figure 4). These similarities would also suggest that the SGA-type rock paintings of Ampasimaiky are ancient writing. them (Willcox, 1984: 210); the same has been applied to that of Professor Barry Fell, a specialist in Libyan-Berber writing. The same has been applied to that of Professor Barry Fell, a specialist in Libyan-Berber writing, who translated the glyphs of Driekopseiland in southern Africa as Libyan-Berber writing. But it has also been refuted on the grounds that the Bushmen could not write (Willcox 1984: 210); it is important to recall
It is important to recall in this connection that Libyco-Berber inscriptions inscribed on petroglyphs in the Canary Islands were not accepted as such, under the pretext that the Guanches (indigenous people of the Canary Islands) could not have acquired or adopted any writing system in prehistoric times (Cline 1953; Farrujia de la Rosa et al. 2010: 16-17; Willcox 1984: 48). This erroneous model persisted until reputable scholars succeeded in translating an inscription reading “save me from javelin annihilation” (Cline 1953: 273). Today, it is commonly accepted that the groups of aligned glyphs from the Canary Islands are Libyco-Berber inscriptions. Recently, Chami (2006: 84-5) insisted that African SGA rock paintings and engravings are ancient scripts. He has shown that some, such as those at Namoratunga in Kenya, are arranged. For the accounts of those that
are not arranged, he argued that at the time when the ancient scripts of Ethiopia, India, China, Egypt and Mesopotamia were still using ideograms, the characters too were not aligned (see also Casson 1966: 141). This can be amplified by the case of some Sahelian writing systems such as the ‘Nsibidi. This script has no ‘order in the arrangement’, but it is practically recognised as a writing system (Dayrell
1910; Macgregor 1909). Zahan (1950) identified the Dogon and Bambara pictographs as scripts based on the argument that
“writing consists essentially in the communication of thought by means of conventional signs, and the pictographs of both Dogon and Bambara serve to convey thoughts from one writer to another or from writer to lay public” (pp.137) « Une écriture consiste essentiellement de la communication de pensée par le moyen de signes conventionnels, et les pictographes de Dogon et de Bambara servent à transporter des pensées d’un auteur à l’autre ou d’un auteur au public”.
Therefore, it is also possible that without the “order in the arrangement”, the Ampasimaiky SGA-type rock paintings could have served to communicate thoughts and information. This would qualify them as a WRITING SYSTEM! Especially since some groups of symbols are horizontally and vertically aligned (Photo 4). Since “order in arrangement” has been established by some scholars as the sine qua non for a group of symbols/signs to qualify as an inscription (Willcox 1984: 210), one can reasonably assert, based on the above evidence, that the
AMPASIMAIKY RUPESTRATED ROCK SHELTER PAINTINGS ARE INSCRIPTIONS! This assertion is clearly supported by the vertically arranged group of symbols in panel 8 (see Photo 4b). Indeed, the repetition of the first sign in the third character (read downwards) is undoubtedly demonstrating the USE OF THE ALPHABET, and thus the
ALPHABET, so this is WRITING! All the more so as all five characters constituting the inscription are closely related to the Libyan-Berber alphabets (see
www.ancientscripts.com), the proto-Egyptian, Aegean and Cretan linear signs respectively
(Figure 5).
Figure 5 : Le groupe de signes verticalement arrangées d’Ampasimaiky comparé avec des caractères alphabétiques libyco-berbères et des signes linéaires proto-Egyptiens, égéens, et crétois (Tableau : Evans, 1897 :386).
Bien que des signes linéaires proto-Égyptiens, crétois et égéens aient été également verticalement écrits, je favoriserais l’origine libyco-berbère de l’inscription d’Ampasimaiky sous l’influence de la description de Walter Cline concernant les inscriptions libyco-berbères comme
suit).
“They [Libyco-Berber inscriptions] are usually very brief, consisting of personal names and a few phrases … The characters are simple and predominantly rectilinear. Variant alphabet may have existed. Most of inscriptions are to be read upward beginning with the left-hand column, though horizontal rows reading from right to left are not uncommon and some columns read downwards. The words are not separated. Though these signs sometimes occur in groups of four or five, they often stand singly…” (Cline 1953: 273)
All the underlined sentences qualify the vertically arranged inscription of Ampasimaiky as being of Libyco-berberber origin. Although this ancient script was written either horizontally or vertically, it should be remembered that the Libyco- Berber peoples used two different alphabets for each row (see http://www.ancientscripts.com). All five characters of the Ampasimaiky inscription are exclusively related to the vertical Libyco- Berber alphabet (Figure 6). Transliterating them to Latin alphabets – based on the transliteration of www.ancientscripts.com – one could read “ğ r ğ t f” (downwards) or “f t ğ r ğ” (upwards). The phonetic value ‘r’ for the presumed circle/square sign can be read as ‘3’ (= a), if the represented sign is perceived as a triangular symbol (see Fell 1975: 271). In the present state of the study, decoding the meaning of the inscription is not yet within the scope of the author. This task would rather be reserved for palaeographers and epigraphers.
Figure 7: The group of vertically arranged signs of Ampasimaiky transliterated in the Libyco-Berber alphabet
It is important to point out here that certain Libyco-Berber inscriptions brought back from the Canary Islands are not different from that of Ampasimaiky (see Figure 8). You can clearly see their resemblance.
Figure 8: The group of vertically arranged signs from Ampasimaiky compared with a Libyco-Berber inscription from the Canary Islands (Drawings by Farrujia de la Rosa et al. 2010: 29)
Discussion and Conclusion
Some of the images found at Ampasimaiky are similar to those of the rock art of continental Africa. However, the painting tradition of southwestern Madagascar is not
However, the painting tradition of southwestern Madagascar does not seem to originate from the continental region closest to Madagascar, i.e. Mozambique and the rest of the central Limpopo basin. In fact, Professor Benjamin Smith, an internationally recognised authority on the study of SGA rock paintings in sub-Saharan Africa, has taken the view that they are not typically affiliated with the Khoekhoen SGA rock paintings, nor with the geometric art of central southern Africa (pers. comm.). The exclusive representation of humped cattle associated with SGA-type symbols is particular to Ampasimaiky. Since no ancient representation of zebu (Bos indicus) has yet been found in the region between Sudan in the north and the Zambezi River in the south (Magnavita 2006: 62), it can be inferred that the tradition of rock paintings in Ampasimaiky originated in North Africa. Evidence of a bronze zebu figurine (of the type Bos indicus) inscribed with SGA signs dating back to 200 CE at Zeban Kutur, Eritrea (Photo 4) would confirm the early use of SGA signs among pastoral zebu-breeding peoples such as those depicted in the Ampasimaiky rock paintings
Photo 4: Bronze figure of zebu inscribed with SGA type signs in Zeban Kutur, Eritrea (Photo: Magnavita 2006:63)
The Libyan-Berber inscription discovered at Ampasimaiky also reinforces this proposal. The Libyan-Berber script emerged in North Africa around 500 BC and ceased to be used around 800 AD (Cline 1953).
around 800 AD (Cline 1953). This implies that the Libyan-Berber inscription of Ampasimaiky should have been written during the period between this bracket of dates (500 BC – AD 800). It is worth mentioning that Iamboulos, a Greek merchant who lived in the 3rd
century BC, reported about a vertical order of writing that was practised by the inhabitants of the ‘island of the sun’ (Allibert 2010: 111; Kobishschanow, 1965: 138; Oldfather, 1961 vol.2: 65-67; Winston 1976: 221). This island is believed to be geographically located adjacent to the east coast of Africa, south of the Equator. Although some of the geographical and ecological information in Iamboulos’ account is very much in line with that of the Malagasy west coast (for more details see the
Although some of the geographical and ecological information in Iamboulos’ account closely matches that of the Malagasy west coast (for more information see Allibert 2010), some scholars have consistently judged this account to be fictional, imaginary, and fanciful (Cary and Warmington 1963: 233-245; Winston 1976). One of the criteria used to censor this narrative is the use of vertical order writing by the islanders. Certainly, it was difficult for some scholars to accept the existence of a pre-Islamic script used in the adjacent islands of the east coast of Africa such as Madagascar (Janvier 1975: 19). However, with the discovery of the Libyan-Berber inscription at Ampasimaiky, this denial no longer holds. New evidence is uncovered to confirm that the “island of the sun” of Iamboulos could have been Madagascar. In any case, man was already present in the vicinity of Ampasimaiky when the Greek merchant passed through the island. Human activity is attested by subfossil lemur bones bearing cutting marks and radiocarbon-dated from the 5th to the 4th century BC at Taolambiby, in the upper Onilahy valley (see Perez et al., 2005). All the more so since ancient written documents (see Jones 1960: 235, 269, 315; Strabo 17, 1, 13 in Lacroix 1998: 95), archaeology (see Cary and Warmington 1963: 130; Chami 2007: 6; 2009; Wainwright 1940: 165), genetic study (see Ricaut et al. 2009: 5-6), and zooarchaeology (Chami 2006: 137; Grigson 1991: 124) all confirm the ancient contacts that existed during this period between the Onilahy Valley and the Onilahy Valley,
During this period, the Nile Valley and the Mediterranean world on the one hand, and the eastern coast of Africa and the adjacent islands on the other.
Références Allibert, C. (2008), “Austronesian Migration and the Establishment of the Malagasy Civilization: Contrasted Readings in Linguistics, Archaeology, Genetics and Cultural Anthropology”, Diogenes, SAGE, Volume 218, pp.7-16 Allibert, C. (2010), “L’île de Iamboulos est-elle Madagascar? », Pount, no.4, pp.103-119 Blench, R. (2007), “New Palaeozoogeographical Evidence for the Settlement of Madagascar”, Azania, Volume XLII, pp.69-82 Cary M. and Warmington, E. (1963), The Ancient Explorers, Middlesex, Penguin Books Casson, L. (1966), Ancient Egypt, Nederland, Time-life International Chami, F. A (2006), The Unity of African Ancient History, 3000 BC to AD 500, Dar es Salaam, E&D Limited Chami, F. A (2008), “The Great Lakes: a complexity of cultural wellsprings”, in Arnold, M. (eds.) Art in Eastern Africa, pp.47-64, Dar-es-Salaam, Mkuki na Nyota Chami F. A and Kwekason A. (2003), “Neolithic Pottery Traditions from the Islands, the Coast, and the Interior of East Africa”, African Archaeological Review, Volume 20, No. 2, pp.65-80 Chami, F. A, Tabibou, I. A and Abderhamane, B. (2009), “Preliminary Report of Archaeological Work Conducted on the Southern Ngazidja Island”, in Chami, F. (eds.), Zanzibar and the Swahili Coast from c30.000 years ago, pp.115-128, Dar es Salaam, E&D Vision Publishers Chaplin, T. (1974), “The Prehistoric Rock Art of the Lake Victoria Region”, Azania Volume 9, pp.1-50 Cline, W. (1953), “Berber Dialects and Berber Script”, Southwestern Journal of Anthropology, Vol.9, No.3, pp.268-276, New Mexico, New Mexico University Press Coulson D. and Campbell A. (2001), African Rock Art, Paintings and Engravings on Stone, New York, Harry N. Abrams Incorporated Dayrell, E. (1910), “Some “Nsibidi” Signs”, Man, Vol.10, pp.113-114. Dewar, R. E (1996), “The Archaeology of the Early Settlement of Madagascar”, in Reade J. (eds.) The Indian Ocean in Antiquity, pp.471-486 Dewar, R. E (1997), “Does It Matter That Madagascar Is an Island?” Human Ecology, Volume 25, No. 3, pp. 481-489 Donque, G. (1965), « Le contexte océanique des anciennes migrations : vents et courants dans l’océan indien », Taloha, Volume 1, pp.43-59. Eastwood, E. B and Smith, B. W (2005), “Fingerprints of the Khoekhoen: Geometric and handprinted Rock Art in the Central Limpopo Basin, Southern Africa”, Goodwin Series, RSA, South African Archaeological Society, Volume 9, pp.63-76 Evans, A. J (1897), “Further Discoveries of Cretan and Aegean Script: With Libyan and ProtoEgyptian Comparisons”, The Journal of Hellenic Studies, The Society for the Promotion of Hellenic Studies, Volume 17, pp.327-395 Farrujia de la Rosa, A., Pichler W., Rodrigue A., Marín S.G. 2010, The Lybico-Berber and Latino-Canarian Scripts and the Colonization of the Canary Islands, African Archaeological Review, vol.27, no.1, pp.13-41 Faublée, J. (1947), Récits Bara, Travaux et mémoires de l’Institut d’ethnologie XLXII, Paris, Université de Paris Fell, B. (1975), America BC: Ancient Settlers in the New World, New York, A Demeter Press Book Grandidier, A., Charles-Roux J., Delhorbe C., Froidevaux H. and Grandidier G. (1902), Collection des ouvrages anciens concernant Madagascar, Tome II, Paris, Comité de Madagascar Grigson, C. (1991), “An African Origin of African Cattle? Some Archaeological Evidence”, The African Archaeological Review, Volume 9, pp.119-144 Hall, S. and Smith, B. (2000), “Empowering Places: Rock Shelters and Ritual Control in FarmerForager Interactions in the Northern Province”, Goodwin Series, RSA, South African Archaeological Society, Volume 8, pp.30-46 Hébert, J. C (1967), « Essai d’interprétation de la stèle indéchiffrée d’Ambilobe », Taloha, Volume 2, pp.XLVII-LVII Hébert, J. C (1971), « Les tatouages de la côte est de Madagascar, d’après Chapelier (1794- 1806) », Taloha, Volume 4, pp.211-223 Huygues-Belrose, V. (1997). Les supports anciens de l’écriture à Madagascar, Etudes Océan Indien, vol.22, pp.7-19, INALCO Janvier, Y. (1975), « La géographie gréco-romaine a-t-elle connu Madagascar? Le point de la question », Omaly sy Anio, No.1-2, pp.11-41 Jones, H. (1960), The Geography of Strabo, Volume 1-2, London, William Heinemann Ltd. Juma, A. (2004), Unguja Ukuu on Zanzibar – An Archaeological Study of Early Urbanism, Sweden, Uppsala University Kessy, E. (2009a), “Analysis of Lithic Artefacts From Mwanampambe and Kuumbi Caves, Zanzibar”, in Chami F. (eds.), Zanzibar and the Swahili Coast from c30,000 years ago, Dar es Salaam, E&D Vision Publishers, pp.131-143 Kessy, E. (2009b), “Stone Artefacts from Male”, in Chami F. (eds.), Zanzibar and the Swahili Coast from c30,000 years ago, Dar es Salaam, E&D Vision Publishers, pp.184-88 Kobishchanov, Y.M. (1965), “On the Problem of Sea Voyages of Ancient Africans in the Indian Ocean”, The Journal of African History, Vol.6, No.2, pp.137-141, Cambridge, Cambridge University Press. Lacroix, W. F (1998), Africa in Antiquity, Nijmegen, Catholic University
Tanambelo V.R. RASOLONDRAINY,
(Centre de Documentation et de Recherche sur l’Art et la Tradition Orale de Madagascar)
Abstract:
This article presents the result of an archaeological study on the first prehistoric rock paintings site found in Madagascar and the rest of islands of Western Indian Ocean south of the equator.
Based on systematic methods and techniques, the study provides groundbreaking archaeological evidences that contribute largely to the understanding of the prehistory of Madagascar which was not archaeologically known thitherto to have prehistoric rock paintings, nor human settlement prior to 8th century AD. A Libyco-Berber inscription was discovered on the rock paintings of Ampasimaiky Rock Shelter, in Southwestern Madagascar. Based on this writing -supported by ancient historical records, circumstantial archaeological and paleontological evidences, linguistic data, and to some extent DNA analysis results -, the present study provides minimum age falling in the bracket time around 500 BC to roughly 8th century AD for the rock paintings of Ampasimaiky rock shelter. Furthermore, the study traces back early contacts between Madagascar and Northern Africa during prehistoric times.
Introduction
With the kind assistance of Dr Chantal Radimilahy, National Coordinator of the Network of African Archaeologists in Madagascar, I researched the ethnographic exhibition on the website of the Museum of Ethnography in Geneva (MEG). The theme was the peoples and cultures of the southwestern region of Madagascar, and was based on an exhibition of photographs taken by the French ethnographer Jacques Faublée in 1939. The pictures included rock paintings found in caves and rock shelters in the Isalo region. After discussions with my teachers at the Archaeology Unit of the University of Dar es Salaam, Professors Felix Chami, Fidelis Masao, and Audax Mabulla, I further investigated the significance of these paintings for the prehistoric period of Madagascar. This study is taken from a Master of Arts in Archaeology entitled
“Archaeologial study of the prehistoric rock paintings of Ampasimaiky rock shelter, in the upper Onilahy, Isalo region, Southwestern Madagascar”.
————–
Africa has many archaeological sites with prehistoric rock paintings (see Coulson and Campbell 2001; Willcox 1984). However, no such sites have yet been scientifically
scientifically reported from Madagascar, which is only 400 kilometres from southern Africa where they are abundant. All rock art previously found in Madagascar has been judged to have been produced in the recent past (Faublée 1947: 164; Hébert 1971: 218). This judgement is based on the theory that Africans were incapable of having navigated the high seas during prehistoric times (Donque 1965: 43; Grandidier et al. 1902; Southall 1975: 194; Vérin 1975, 1981). It was therefore out of the question to think that they had the possibility of transferring the practice of rock painting to the adjacent islands of Africa such as Madagascar. However, recent archaeological findings have demonstrated that the east coast of Africa and adjacent islands were already in contact since the prehistoric period (see Chami and Kwekason 2003; Chami et al. 2009; Juma 2004; Kessy 2009a,
2009b). These prompt a problem that led me to revise the antiquity of the rock paintings in Madagascar.
On the other hand, the name of the Arabic-Malagasy script – Sorabe meaning “great script” – indicates that there was an ancient script in Madagascar that was distinguished from Arabic (Hébert,
1967:XLVII; Huygues-Belrose 1997: 8). Although Hébert (1967:XLVIII) presumed that such a hypothetical script would have to be of Indian origin and would be discovered from
Although Hébert (1967:XLVIII) assumed that such a hypothetical script would have to be of Indian origin and would be discovered from stele and parietal inscriptions, no such inscription has so far been found in Madagascar.
My problem was therefore to find out whether this would appear on the rock paintings reported by the website. What provoked this idea the most was that Chami (2006:84-85; 2008) argued that the African Geometric and Amorphous Schematic (GAS) rock paintings and engravings are ancient writing.
Another problem is that radiocarbon 14 dating placed the earliest human settlement of Madagascar around the 8th century AD (Allibert 2008: 7; Dewar 1996: 474).
Although earlier dates are still being considered, many researchers have assumed that settlement of Madagascar would not go beyond the year W AD (for more information see Dewar 1997). This assumption has made research on Malagasy material culture going back to the Stone Age more than difficult. Any findings of material cultures dating from prehistoric times have been discarded, overlooked or misinterpreted. However, subfossil animal bone remains bearing traces of butchery found in paleontological sites in the southwestern region indicate the presence of human activity
(For more information, see Blench 2007; Perez et al. 2005). Despite this evidence, an archaeological site has not yet been
archaeological site with similar dates has not yet been found on the entire island.
With the discovery of a rock shelter with rock paintings, discovering new information about the prehistory of Madagascar before the 8th century proved to be feasible. All the more so as some of the paintings and engravings are
Especially since some of the SGA-type rock paintings and engravings found on the African continent – similar to those found in the photos on the MEG website – have been radiocarbon dated from the 1st millennium BC to the early 1st millennium AD (Chami, 2008: 54; Hall and Smith 2000; Lynche and Donahue 1980; Willcox 1984: 26, 83.
Fieldwork and results
Despite various challenges and constraints – mainly caused by the lack of archaeological reconnaissance prior to the fieldwork, insecurity in the study area, and the sacredness of the site to be studied – we had only somehow managed to conduct survey, archaeological excavation, and documentation of the paintings in the Ampasimaiky rock shelter. It is only the result of the latter that will be presented in this article.
The rock painting site is a rock shelter located at the foot of a hill immediately next to a small stream called Ampasimaiky (Figure 2), from which the name of the site derives. The presence of this water source may have influenced the painter’s choice of location.
The shelter itself is 26.8m long (south-north axis) and 5m wide (inwards).
Its ceiling begins to slope inwards from 4m high at its entrance to approximately 1.5m high at its irregular wall. The deepest part of the rock shelter – a chamber – is currently used by the royal family “Bara Zafimanely” as a burial place. Dung and ox hoof prints found on the ground indicate that cattle frequent the place. This occurrence should have disturbed the soil.
However, the areas around the site are dominated by rocky hills covered with grass. Only in the deep canyon near the confluence of Ampasimaiky and the Sakamarekely River is a patch of thick vegetation up to 10 m high observed along the stream. This situation would make agricultural activity almost impossible around the site, unless one had to move 1.5 km to the bank, where rice cultivation takes place today.
An archaeological survey of the southern Isalo region (Figure 1) was carried out and a rock painting site was discovered. Professor Chami participated in the fieldwork, advising on the approach to be followed in the treatment of these previously unstudied cave cultures in Madagascar.
Madagascar. History students from the University of Toliara, Mex Tafitasoa, Nadya
Ravololonirina, and Norbert Antilahy, and some villagers (Photo 1) participated in the work. The Network of African Archaeologists in Madagascar, under the direction of Dr Chantal Radimilahy and Prof Barthélemy Manjakahery, provided the logistical equipment used. The various signatures The various signatures of national and local administrative approval, the security provided by the Gendarmerie Brigade of Benenitsy, and above all the kindness and openness of King Zafimanely (Photo 1) – who allowed us access to the (burial) site – contributed enormously to the accomplishment of the expedition.
Madagascar. History students from the University of Toliara, Mex Tafitasoa, Nadya
Ravololonirina, and Norbert Antilahy, and some villagers (Photo 1) participated in the work. The Network of African Archaeologists in Madagascar, under the direction of Dr Chantal Radimilahy and Prof Barthélemy Manjakahery, provided the logistical equipment used. The various signatures The various signatures of national and local administrative approval, the security provided by the Gendarmerie Brigade of Benenitsy, and above all the kindness and openness of King Zafimanely (Photo 1) – who allowed us access to the (burial) site – contributed enormously to the accomplishment of the expedition.
The rock shelter at Ampasimaiky faces east. This arrangement makes the painting panels exposed to the sun in the morning and in the shade beyond noon. This arrangement suggests that painting activities are conducted either early in the morning while the heat of the sun is still moderate, or beyond noon when shade is available below the shelter.
This is based on the fact that the height of the panels does not exceed 130 cm, suggesting that the paintings were intended for people sitting on the floor of the undercroft shelter
Photo 2). Minimal, therefore, is the likelihood of the painting activity – involving seated people – having been conducted during the time when the sun is heating the shelter floor.
However, the view outside the site is limited. The opposing hills – on the other bank of the Ampasimaiky stream – prevent a person standing under the shelter from having a good view of the surroundings (Photo 3). It is much more likely that the view outwards – which implies a defensive strategy (Smith 1995) – was not among the painter’s concerns when he/she chose the site. This demonstrates that the layout of the Ampasimaiky rock shelter favours invaders from unexpectedly attacking its residents. The place would therefore not be suitable for permanent habitation.
Photo 3
Especially since a person standing on the hills of the other bank would be able to look directly at the other one positioned under the rock shelter, unless he hides in the deeper part of the shelter. Since the wall of this “chamber”, according to the local population, is not painted, it is plausible that the activities related to the paintings in the Ampasimaiky rock shelter are not clandestine. However, the paintings were not intended to be seen by people passing close to the site. They are not seen from a distance. Not only are the panels placed below 130 cm in height, but the slope of the ground before the access
to the shelter also prevents anyone from seeing them until they get close to the entrance. Nevertheless, the use of darker coloured paints (red and black) suggests the painter’s intention to contrast the paints he/she was using against the pale colour of the sandstone wall. This implies his concern for the visibility of his paintings.
The walls of the rock shelter have red, reddish orange, burgundy red, black, and white paintings in monochrome, bichrome and polychrome patterns. The images
are painted with the fingertip. They depict zoomorphic (mainly zebu) and anthropomorphic (semi-naturalist and stylised) figures, as well as schematic-geometric-amorphic symbols (Figure 3). The latter dominate and are represented by quadrangular, circular and elliptical signs, dot/dash lines, and alphabet-like signs. In total, the Ampasimaiky rock paintings have 166 recognisable images.
Analysis and Interpretation
Based on comparative studies of the images, it has been found that the images of the Ampasimaiky rock paintings are not different from the images depicted on the prehistoric SGA rock art of the African continent (see Chami, 2008; Chaplin, 1974; Coulson and Campbell 2001;
Eastwood and Smith, 2005; Lynch and Donahue, 1980; Mabulla, 2005; Smith, 1995; Teka 2008; Willcox 1984). The similarities apparently suggest links between the two regions (Madagascar and the mainland) in the
(Madagascar and the continent) in prehistoric times. This suggests that the Ampasimaiky rock paintings are of African origin and could be relatively dated to the same period as those on the continent (Thackeray, 1983: 22). Comparative studies of the images have also shown that more than 30 types of images in the Ampasimaiky rock paintings are closely related to different ancient scripts such as Libyco-Berber, Protosinaitic, Meroitic (cursive), Punic Iberian, Ancient Phoenician, Tifinagh, and Aegean, Cretan and Proto-Egyptian linear signs respectively (Figure 4). These similarities would also suggest that the SGA-type rock paintings of Ampasimaiky are ancient writing.
figure 4
Analysis and Interpretation
Based on comparative studies of the images, it has been found that the images of the Ampasimaiky rock paintings are not different from the images depicted on the prehistoric SGA rock art of the African continent (see Chami, 2008; Chaplin, 1974; Coulson and Campbell 2001;
Eastwood and Smith, 2005; Lynch and Donahue, 1980; Mabulla, 2005; Smith, 1995; Teka 2008; Willcox 1984). The similarities apparently suggest links between the two regions (Madagascar and the mainland) in the
(Madagascar and the continent) in prehistoric times. This suggests that the Ampasimaiky rock paintings are of African origin and could be relatively dated to the same period as those on the continent (Thackeray, 1983: 22). Comparative studies of the images have also shown that more than 30 types of images in the Ampasimaiky rock paintings are closely related to different ancient scripts such as Libyco-Berber, Protosinaitic, Meroitic (cursive), Punic Iberian, Ancient Phoenician, Tifinagh, and Aegean, Cretan and Proto-Egyptian linear signs respectively (Figure 4). These similarities would also suggest that the SGA-type rock paintings of Ampasimaiky are ancient writing. them (Willcox, 1984: 210); the same has been applied to that of Professor Barry Fell, a specialist in Libyan-Berber writing. The same has been applied to that of Professor Barry Fell, a specialist in Libyan-Berber writing, who translated the glyphs of Driekopseiland in southern Africa as Libyan-Berber writing. But it has also been refuted on the grounds that the Bushmen could not write (Willcox 1984: 210); it is important to recall
It is important to recall in this connection that Libyco-Berber inscriptions inscribed on petroglyphs in the Canary Islands were not accepted as such, under the pretext that the Guanches (indigenous people of the Canary Islands) could not have acquired or adopted any writing system in prehistoric times (Cline 1953; Farrujia de la Rosa et al. 2010: 16-17; Willcox 1984: 48). This erroneous model persisted until reputable scholars succeeded in translating an inscription reading “save me from javelin annihilation” (Cline 1953: 273). Today, it is commonly accepted that the groups of aligned glyphs from the Canary Islands are Libyco-Berber inscriptions. Recently, Chami (2006: 84-5) insisted that African SGA rock paintings and engravings are ancient scripts. He has shown that some, such as those at Namoratunga in Kenya, are arranged. For the accounts of those that
are not arranged, he argued that at the time when the ancient scripts of Ethiopia, India, China, Egypt and Mesopotamia were still using ideograms, the characters too were not aligned (see also Casson 1966: 141). This can be amplified by the case of some Sahelian writing systems such as the ‘Nsibidi. This script has no ‘order in the arrangement’, but it is practically recognised as a writing system (Dayrell
1910; Macgregor 1909). Zahan (1950) identified the Dogon and Bambara pictographs as scripts based on the argument that
“writing consists essentially in the communication of thought by means of conventional signs, and the pictographs of both Dogon and Bambara serve to convey thoughts from one writer to another or from writer to lay public” (pp.137) « Une écriture consiste essentiellement de la communication de pensée par le moyen de signes conventionnels, et les pictographes de Dogon et de Bambara servent à transporter des pensées d’un auteur à l’autre ou d’un auteur au public”.
Therefore, it is also possible that without the “order in the arrangement”, the Ampasimaiky SGA-type rock paintings could have served to communicate thoughts and information. This would qualify them as a WRITING SYSTEM! Especially since some groups of symbols are horizontally and vertically aligned (Photo 4). Since “order in arrangement” has been established by some scholars as the sine qua non for a group of symbols/signs to qualify as an inscription (Willcox 1984: 210), one can reasonably assert, based on the above evidence, that the
AMPASIMAIKY RUPESTRATED ROCK SHELTER PAINTINGS ARE INSCRIPTIONS! This assertion is clearly supported by the vertically arranged group of symbols in panel 8 (see Photo 4b). Indeed, the repetition of the first sign in the third character (read downwards) is undoubtedly demonstrating the USE OF THE ALPHABET, and thus the
ALPHABET, so this is WRITING! All the more so as all five characters constituting the inscription are closely related to the Libyan-Berber alphabets (see
www.ancientscripts.com), the proto-Egyptian, Aegean and Cretan linear signs respectively
(Figure 5).
Figure 5 : Le groupe de signes verticalement arrangées d’Ampasimaiky comparé avec des caractères alphabétiques libyco-berbères et des signes linéaires proto-Egyptiens, égéens, et crétois (Tableau : Evans, 1897 :386).
Bien que des signes linéaires proto-Égyptiens, crétois et égéens aient été également verticalement écrits, je favoriserais l’origine libyco-berbère de l’inscription d’Ampasimaiky sous l’influence de la description de Walter Cline concernant les inscriptions libyco-berbères comme
suit).
“They [Libyco-Berber inscriptions] are usually very brief, consisting of personal names and a few phrases … The characters are simple and predominantly rectilinear. Variant alphabet may have existed. Most of inscriptions are to be read upward beginning with the left-hand column, though horizontal rows reading from right to left are not uncommon and some columns read downwards. The words are not separated. Though these signs sometimes occur in groups of four or five, they often stand singly…” (Cline 1953: 273)
All the underlined sentences qualify the vertically arranged inscription of Ampasimaiky as being of Libyco-berberber origin. Although this ancient script was written either horizontally or vertically, it should be remembered that the Libyco- Berber peoples used two different alphabets for each row (see http://www.ancientscripts.com). All five characters of the Ampasimaiky inscription are exclusively related to the vertical Libyco- Berber alphabet (Figure 6). Transliterating them to Latin alphabets – based on the transliteration of www.ancientscripts.com – one could read “ğ r ğ t f” (downwards) or “f t ğ r ğ” (upwards). The phonetic value ‘r’ for the presumed circle/square sign can be read as ‘3’ (= a), if the represented sign is perceived as a triangular symbol (see Fell 1975: 271). In the present state of the study, decoding the meaning of the inscription is not yet within the scope of the author. This task would rather be reserved for palaeographers and epigraphers.
Figure 7: The group of vertically arranged signs of Ampasimaiky transliterated in the Libyco-Berber alphabet
It is important to point out here that certain Libyco-Berber inscriptions brought back from the Canary Islands are not different from that of Ampasimaiky (see Figure 8). You can clearly see their resemblance.
Figure 8: The group of vertically arranged signs from Ampasimaiky compared with a Libyco-Berber inscription from the Canary Islands (Drawings by Farrujia de la Rosa et al. 2010: 29)
Discussion and Conclusion
Some of the images found at Ampasimaiky are similar to those of the rock art of continental Africa. However, the painting tradition of southwestern Madagascar is not
However, the painting tradition of southwestern Madagascar does not seem to originate from the continental region closest to Madagascar, i.e. Mozambique and the rest of the central Limpopo basin. In fact, Professor Benjamin Smith, an internationally recognised authority on the study of SGA rock paintings in sub-Saharan Africa, has taken the view that they are not typically affiliated with the Khoekhoen SGA rock paintings, nor with the geometric art of central southern Africa (pers. comm.). The exclusive representation of humped cattle associated with SGA-type symbols is particular to Ampasimaiky. Since no ancient representation of zebu (Bos indicus) has yet been found in the region between Sudan in the north and the Zambezi River in the south (Magnavita 2006: 62), it can be inferred that the tradition of rock paintings in Ampasimaiky originated in North Africa. Evidence of a bronze zebu figurine (of the type Bos indicus) inscribed with SGA signs dating back to 200 CE at Zeban Kutur, Eritrea (Photo 4) would confirm the early use of SGA signs among pastoral zebu-breeding peoples such as those depicted in the Ampasimaiky rock paintings
Photo 4: Bronze figure of zebu inscribed with SGA type signs in Zeban Kutur, Eritrea (Photo: Magnavita 2006:63)
The Libyan-Berber inscription discovered at Ampasimaiky also reinforces this proposal. The Libyan-Berber script emerged in North Africa around 500 BC and ceased to be used around 800 AD (Cline 1953).
around 800 AD (Cline 1953). This implies that the Libyan-Berber inscription of Ampasimaiky should have been written during the period between this bracket of dates (500 BC – AD 800). It is worth mentioning that Iamboulos, a Greek merchant who lived in the 3rd
century BC, reported about a vertical order of writing that was practised by the inhabitants of the ‘island of the sun’ (Allibert 2010: 111; Kobishschanow, 1965: 138; Oldfather, 1961 vol.2: 65-67; Winston 1976: 221). This island is believed to be geographically located adjacent to the east coast of Africa, south of the Equator. Although some of the geographical and ecological information in Iamboulos’ account is very much in line with that of the Malagasy west coast (for more details see the
Although some of the geographical and ecological information in Iamboulos’ account closely matches that of the Malagasy west coast (for more information see Allibert 2010), some scholars have consistently judged this account to be fictional, imaginary, and fanciful (Cary and Warmington 1963: 233-245; Winston 1976). One of the criteria used to censor this narrative is the use of vertical order writing by the islanders. Certainly, it was difficult for some scholars to accept the existence of a pre-Islamic script used in the adjacent islands of the east coast of Africa such as Madagascar (Janvier 1975: 19). However, with the discovery of the Libyan-Berber inscription at Ampasimaiky, this denial no longer holds. New evidence is uncovered to confirm that the “island of the sun” of Iamboulos could have been Madagascar. In any case, man was already present in the vicinity of Ampasimaiky when the Greek merchant passed through the island. Human activity is attested by subfossil lemur bones bearing cutting marks and radiocarbon-dated from the 5th to the 4th century BC at Taolambiby, in the upper Onilahy valley (see Perez et al., 2005). All the more so since ancient written documents (see Jones 1960: 235, 269, 315; Strabo 17, 1, 13 in Lacroix 1998: 95), archaeology (see Cary and Warmington 1963: 130; Chami 2007: 6; 2009; Wainwright 1940: 165), genetic study (see Ricaut et al. 2009: 5-6), and zooarchaeology (Chami 2006: 137; Grigson 1991: 124) all confirm the ancient contacts that existed during this period between the Onilahy Valley and the Onilahy Valley,
During this period, the Nile Valley and the Mediterranean world on the one hand, and the eastern coast of Africa and the adjacent islands on the other.
Références Allibert, C. (2008), “Austronesian Migration and the Establishment of the Malagasy Civilization: Contrasted Readings in Linguistics, Archaeology, Genetics and Cultural Anthropology”, Diogenes, SAGE, Volume 218, pp.7-16 Allibert, C. (2010), “L’île de Iamboulos est-elle Madagascar? », Pount, no.4, pp.103-119 Blench, R. (2007), “New Palaeozoogeographical Evidence for the Settlement of Madagascar”, Azania, Volume XLII, pp.69-82 Cary M. and Warmington, E. (1963), The Ancient Explorers, Middlesex, Penguin Books Casson, L. (1966), Ancient Egypt, Nederland, Time-life International Chami, F. A (2006), The Unity of African Ancient History, 3000 BC to AD 500, Dar es Salaam, E&D Limited Chami, F. A (2008), “The Great Lakes: a complexity of cultural wellsprings”, in Arnold, M. (eds.) Art in Eastern Africa, pp.47-64, Dar-es-Salaam, Mkuki na Nyota Chami F. A and Kwekason A. (2003), “Neolithic Pottery Traditions from the Islands, the Coast, and the Interior of East Africa”, African Archaeological Review, Volume 20, No. 2, pp.65-80 Chami, F. A, Tabibou, I. A and Abderhamane, B. (2009), “Preliminary Report of Archaeological Work Conducted on the Southern Ngazidja Island”, in Chami, F. (eds.), Zanzibar and the Swahili Coast from c30.000 years ago, pp.115-128, Dar es Salaam, E&D Vision Publishers Chaplin, T. (1974), “The Prehistoric Rock Art of the Lake Victoria Region”, Azania Volume 9, pp.1-50 Cline, W. (1953), “Berber Dialects and Berber Script”, Southwestern Journal of Anthropology, Vol.9, No.3, pp.268-276, New Mexico, New Mexico University Press Coulson D. and Campbell A. (2001), African Rock Art, Paintings and Engravings on Stone, New York, Harry N. Abrams Incorporated Dayrell, E. (1910), “Some “Nsibidi” Signs”, Man, Vol.10, pp.113-114. Dewar, R. E (1996), “The Archaeology of the Early Settlement of Madagascar”, in Reade J. (eds.) The Indian Ocean in Antiquity, pp.471-486 Dewar, R. E (1997), “Does It Matter That Madagascar Is an Island?” Human Ecology, Volume 25, No. 3, pp. 481-489 Donque, G. (1965), « Le contexte océanique des anciennes migrations : vents et courants dans l’océan indien », Taloha, Volume 1, pp.43-59. Eastwood, E. B and Smith, B. W (2005), “Fingerprints of the Khoekhoen: Geometric and handprinted Rock Art in the Central Limpopo Basin, Southern Africa”, Goodwin Series, RSA, South African Archaeological Society, Volume 9, pp.63-76 Evans, A. J (1897), “Further Discoveries of Cretan and Aegean Script: With Libyan and ProtoEgyptian Comparisons”, The Journal of Hellenic Studies, The Society for the Promotion of Hellenic Studies, Volume 17, pp.327-395 Farrujia de la Rosa, A., Pichler W., Rodrigue A., Marín S.G. 2010, The Lybico-Berber and Latino-Canarian Scripts and the Colonization of the Canary Islands, African Archaeological Review, vol.27, no.1, pp.13-41 Faublée, J. (1947), Récits Bara, Travaux et mémoires de l’Institut d’ethnologie XLXII, Paris, Université de Paris Fell, B. (1975), America BC: Ancient Settlers in the New World, New York, A Demeter Press Book Grandidier, A., Charles-Roux J., Delhorbe C., Froidevaux H. and Grandidier G. (1902), Collection des ouvrages anciens concernant Madagascar, Tome II, Paris, Comité de Madagascar Grigson, C. (1991), “An African Origin of African Cattle? Some Archaeological Evidence”, The African Archaeological Review, Volume 9, pp.119-144 Hall, S. and Smith, B. (2000), “Empowering Places: Rock Shelters and Ritual Control in FarmerForager Interactions in the Northern Province”, Goodwin Series, RSA, South African Archaeological Society, Volume 8, pp.30-46 Hébert, J. C (1967), « Essai d’interprétation de la stèle indéchiffrée d’Ambilobe », Taloha, Volume 2, pp.XLVII-LVII Hébert, J. C (1971), « Les tatouages de la côte est de Madagascar, d’après Chapelier (1794- 1806) », Taloha, Volume 4, pp.211-223 Huygues-Belrose, V. (1997). Les supports anciens de l’écriture à Madagascar, Etudes Océan Indien, vol.22, pp.7-19, INALCO Janvier, Y. (1975), « La géographie gréco-romaine a-t-elle connu Madagascar? Le point de la question », Omaly sy Anio, No.1-2, pp.11-41 Jones, H. (1960), The Geography of Strabo, Volume 1-2, London, William Heinemann Ltd. Juma, A. (2004), Unguja Ukuu on Zanzibar – An Archaeological Study of Early Urbanism, Sweden, Uppsala University Kessy, E. (2009a), “Analysis of Lithic Artefacts From Mwanampambe and Kuumbi Caves, Zanzibar”, in Chami F. (eds.), Zanzibar and the Swahili Coast from c30,000 years ago, Dar es Salaam, E&D Vision Publishers, pp.131-143 Kessy, E. (2009b), “Stone Artefacts from Male”, in Chami F. (eds.), Zanzibar and the Swahili Coast from c30,000 years ago, Dar es Salaam, E&D Vision Publishers, pp.184-88 Kobishchanov, Y.M. (1965), “On the Problem of Sea Voyages of Ancient Africans in the Indian Ocean”, The Journal of African History, Vol.6, No.2, pp.137-141, Cambridge, Cambridge University Press. Lacroix, W. F (1998), Africa in Antiquity, Nijmegen, Catholic University Lynch, B. M and Donahue, R. (1980), “A Statistical Analysis of Two Rock-Art Sites in
Northwest Kenya”, Journal of Field Archaeology, Volume 7, pp.75-85
Mabulla, A. Z (2005), “The Rock Art of Mara Region, Tanzania”, Azania, Volume 40, 19-42
Macgregor, J. K (1909), “Some Notes on Nsibidi”, The Journal of the Royal Institute of Great
Britain and Ireland, Volume 39, pp.209-219
MacPhee R. D and Burney D. (1991), “Dating of modified femora of extinct dwarf
Hippopotamus from southern Madagascar: implications for constraining human
colonization and vertebrate extinction events”, Journal of Archaeological Science, No.
18, pp. 695-706
Magnavita, C. (2006), “Ancient Humped Cattle in Africa: A View from the Chad Basin”, The
African Archaeological Review, Volume 23, No. 3/4, pp.55-84
Oldfather, C. H (1961), Diodorus of Sicily, 12 Volumes, London, William Heinemann
Perez, V. R Godfrey L. R, Nowak-Kemp M., Burney D. A, Ratsimbazafy J. and Vasey N.
(2005), “Evidence of Early Butchery of Giant Lemurs in Madagascar”, Journal of
Human Evolution, No. 49, pp.722-742
Radimilahy, C. (2011), « Contribution á l’Archéologie du Sud-Ouest de Madagascar »,
Civilisations des mondes insulaires (Madagascar, îles du canal de Mozambique,
Mascareignes, Polynésie, Guyanes), Paris, Éditions Karthala, pp.825-853
Rasolondrainy, T. (2011), Archaeological Study of the Prehistoric Rock Paintings of
Ampasimaiky Rock Shelter, in the Upper Onilahy, Isalo Region, Southwestern
Madagascar. M.A. Dissertation unpublished, University of Dar es Salaam.
Ricaut, F., Razafindrazaka H., Cox, M., Dugoujon J., Guitard E., Sambo C., Mormina M.,
Mirazon-Larh M., Ludes B., and Crubézy E. 2009, A New Deep Branch of Eurasian
mtDNA Macrohaplogroup M Reveals Additional Complexity Regarding the Settlement
of Madagascar. BMC Genomics, vol.10, pp.1-9
Smith, B. W (1995), Rock Art in South-Central Africa. A Study Based on the Pictographs of
Dedza District, Malawi and Kasama District Zambia, Unpublished Phd Thesis,
University of Cambridge
Southall, A. (1975), “The Problem of Malagasy Origins”, in Chittick H. N and Rotberg R. I
(eds.), East Africa and the Orient, Cultural Syntheses in Pre-Colonial Times, New
York, Africana Publishing Company, pp. 192-215
Teka, Z. (2008). “Distribution and Significance of Ancient Rock Art Sites in Eritrea”, in
Schmidt P. et al (eds.), The Archaeology of Ancient Eritrea. Eritrea: The Red Sea Press,
Inc., pp.49-61
Thackeray, A. (1983), “Dating the Rock Art of Southern Africa”, Goodwin Series, Volume 4,
pp.21-26.
Vérin P., (1975), “Austronesian Contributions to the Culture of Madagascar: Some
Archaeological Problems”, in Chittick H. N and Rotberg R. I (eds.), East Africa and the
Orient, Cultural Syntheses in Pre-Colonial Times, New York, Africana Publishing
Company, pp. 164-191
Northwest Kenya”, Journal of Field Archaeology, Volume 7, pp.75-85
Mabulla, A. Z (2005), “The Rock Art of Mara Region, Tanzania”, Azania, Volume 40, 19-42
Macgregor, J. K (1909), “Some Notes on Nsibidi”, The Journal of the Royal Institute of Great
Britain and Ireland, Volume 39, pp.209-219
MacPhee R. D and Burney D. (1991), “Dating of modified femora of extinct dwarf
Hippopotamus from southern Madagascar: implications for constraining human
colonization and vertebrate extinction events”, Journal of Archaeological Science, No.
18, pp. 695-706
Magnavita, C. (2006), “Ancient Humped Cattle in Africa: A View from the Chad Basin”, The
African Archaeological Review, Volume 23, No. 3/4, pp.55-84
Oldfather, C. H (1961), Diodorus of Sicily, 12 Volumes, London, William Heinemann
Perez, V. R Godfrey L. R, Nowak-Kemp M., Burney D. A, Ratsimbazafy J. and Vasey N.
(2005), “Evidence of Early Butchery of Giant Lemurs in Madagascar”, Journal of
Human Evolution, No. 49, pp.722-742
Radimilahy, C. (2011), « Contribution á l’Archéologie du Sud-Ouest de Madagascar »,
Civilisations des mondes insulaires (Madagascar, îles du canal de Mozambique,
Mascareignes, Polynésie, Guyanes), Paris, Éditions Karthala, pp.825-853
Rasolondrainy, T. (2011), Archaeological Study of the Prehistoric Rock Paintings of
Ampasimaiky Rock Shelter, in the Upper Onilahy, Isalo Region, Southwestern
Madagascar. M.A. Dissertation unpublished, University of Dar es Salaam.
Ricaut, F., Razafindrazaka H., Cox, M., Dugoujon J., Guitard E., Sambo C., Mormina M.,
Mirazon-Larh M., Ludes B., and Crubézy E. 2009, A New Deep Branch of Eurasian
mtDNA Macrohaplogroup M Reveals Additional Complexity Regarding the Settlement
of Madagascar. BMC Genomics, vol.10, pp.1-9
Smith, B. W (1995), Rock Art in South-Central Africa. A Study Based on the Pictographs of
Dedza District, Malawi and Kasama District Zambia, Unpublished Phd Thesis,
University of Cambridge
Southall, A. (1975), “The Problem of Malagasy Origins”, in Chittick H. N and Rotberg R. I
(eds.), East Africa and the Orient, Cultural Syntheses in Pre-Colonial Times, New
York, Africana Publishing Company, pp. 192-215
Teka, Z. (2008). “Distribution and Significance of Ancient Rock Art Sites in Eritrea”, in
Schmidt P. et al (eds.), The Archaeology of Ancient Eritrea. Eritrea: The Red Sea Press,
Inc., pp.49-61
Thackeray, A. (1983), “Dating the Rock Art of Southern Africa”, Goodwin Series, Volume 4,
pp.21-26.
Vérin P., (1975), “Austronesian Contributions to the Culture of Madagascar: Some
Archaeological Problems”, in Chittick H. N and Rotberg R. I (eds.), East Africa and the
Orient, Cultural Syntheses in Pre-Colonial Times, New York, Africana Publishing
Company, pp. 164-191
Tanambelo V.R. RASOLONDRAINY,
(Centre de Documentation et de Recherche sur l’Art et la Tradition Orale de Madagascar)
Abstract:
This article presents the result of an archaeological study on the first prehistoric rock paintings site found in Madagascar and the rest of islands of Western Indian Ocean south of the equator.
Based on systematic methods and techniques, the study provides groundbreaking archaeological evidences that contribute largely to the understanding of the prehistory of Madagascar which was not archaeologically known thitherto to have prehistoric rock paintings, nor human settlement prior to 8th century AD. A Libyco-Berber inscription was discovered on the rock paintings of Ampasimaiky Rock Shelter, in Southwestern Madagascar. Based on this writing -supported by ancient historical records, circumstantial archaeological and paleontological evidences, linguistic data, and to some extent DNA analysis results -, the present study provides minimum age falling in the bracket time around 500 BC to roughly 8th century AD for the rock paintings of Ampasimaiky rock shelter. Furthermore, the study traces back early contacts between Madagascar and Northern Africa during prehistoric times.
Introduction
With the kind assistance of Dr Chantal Radimilahy, National Coordinator of the Network of African Archaeologists in Madagascar, I researched the ethnographic exhibition on the website of the Museum of Ethnography in Geneva (MEG). The theme was the peoples and cultures of the southwestern region of Madagascar, and was based on an exhibition of photographs taken by the French ethnographer Jacques Faublée in 1939. The pictures included rock paintings found in caves and rock shelters in the Isalo region. After discussions with my teachers at the Archaeology Unit of the University of Dar es Salaam, Professors Felix Chami, Fidelis Masao, and Audax Mabulla, I further investigated the significance of these paintings for the prehistoric period of Madagascar. This study is taken from a Master of Arts in Archaeology entitled
“Archaeologial study of the prehistoric rock paintings of Ampasimaiky rock shelter, in the upper Onilahy, Isalo region, Southwestern Madagascar”.
————–
Africa has many archaeological sites with prehistoric rock paintings (see Coulson and Campbell 2001; Willcox 1984). However, no such sites have yet been scientifically
scientifically reported from Madagascar, which is only 400 kilometres from southern Africa where they are abundant. All rock art previously found in Madagascar has been judged to have been produced in the recent past (Faublée 1947: 164; Hébert 1971: 218). This judgement is based on the theory that Africans were incapable of having navigated the high seas during prehistoric times (Donque 1965: 43; Grandidier et al. 1902; Southall 1975: 194; Vérin 1975, 1981). It was therefore out of the question to think that they had the possibility of transferring the practice of rock painting to the adjacent islands of Africa such as Madagascar. However, recent archaeological findings have demonstrated that the east coast of Africa and adjacent islands were already in contact since the prehistoric period (see Chami and Kwekason 2003; Chami et al. 2009; Juma 2004; Kessy 2009a,
2009b). These prompt a problem that led me to revise the antiquity of the rock paintings in Madagascar.
On the other hand, the name of the Arabic-Malagasy script – Sorabe meaning “great script” – indicates that there was an ancient script in Madagascar that was distinguished from Arabic (Hébert,
1967:XLVII; Huygues-Belrose 1997: 8). Although Hébert (1967:XLVIII) presumed that such a hypothetical script would have to be of Indian origin and would be discovered from
Although Hébert (1967:XLVIII) assumed that such a hypothetical script would have to be of Indian origin and would be discovered from stele and parietal inscriptions, no such inscription has so far been found in Madagascar.
My problem was therefore to find out whether this would appear on the rock paintings reported by the website. What provoked this idea the most was that Chami (2006:84-85; 2008) argued that the African Geometric and Amorphous Schematic (GAS) rock paintings and engravings are ancient writing.
Another problem is that radiocarbon 14 dating placed the earliest human settlement of Madagascar around the 8th century AD (Allibert 2008: 7; Dewar 1996: 474).
Although earlier dates are still being considered, many researchers have assumed that settlement of Madagascar would not go beyond the year W AD (for more information see Dewar 1997). This assumption has made research on Malagasy material culture going back to the Stone Age more than difficult. Any findings of material cultures dating from prehistoric times have been discarded, overlooked or misinterpreted. However, subfossil animal bone remains bearing traces of butchery found in paleontological sites in the southwestern region indicate the presence of human activity
(For more information, see Blench 2007; Perez et al. 2005). Despite this evidence, an archaeological site has not yet been
archaeological site with similar dates has not yet been found on the entire island.
With the discovery of a rock shelter with rock paintings, discovering new information about the prehistory of Madagascar before the 8th century proved to be feasible. All the more so as some of the paintings and engravings are
Especially since some of the SGA-type rock paintings and engravings found on the African continent – similar to those found in the photos on the MEG website – have been radiocarbon dated from the 1st millennium BC to the early 1st millennium AD (Chami, 2008: 54; Hall and Smith 2000; Lynche and Donahue 1980; Willcox 1984: 26, 83.
Fieldwork and results
Despite various challenges and constraints – mainly caused by the lack of archaeological reconnaissance prior to the fieldwork, insecurity in the study area, and the sacredness of the site to be studied – we had only somehow managed to conduct survey, archaeological excavation, and documentation of the paintings in the Ampasimaiky rock shelter. It is only the result of the latter that will be presented in this article.
The rock painting site is a rock shelter located at the foot of a hill immediately next to a small stream called Ampasimaiky (Figure 2), from which the name of the site derives. The presence of this water source may have influenced the painter’s choice of location.
The shelter itself is 26.8m long (south-north axis) and 5m wide (inwards).
Its ceiling begins to slope inwards from 4m high at its entrance to approximately 1.5m high at its irregular wall. The deepest part of the rock shelter – a chamber – is currently used by the royal family “Bara Zafimanely” as a burial place. Dung and ox hoof prints found on the ground indicate that cattle frequent the place. This occurrence should have disturbed the soil.
However, the areas around the site are dominated by rocky hills covered with grass. Only in the deep canyon near the confluence of Ampasimaiky and the Sakamarekely River is a patch of thick vegetation up to 10 m high observed along the stream. This situation would make agricultural activity almost impossible around the site, unless one had to move 1.5 km to the bank, where rice cultivation takes place today.
An archaeological survey of the southern Isalo region (Figure 1) was carried out and a rock painting site was discovered. Professor Chami participated in the fieldwork, advising on the approach to be followed in the treatment of these previously unstudied cave cultures in Madagascar.
Madagascar. History students from the University of Toliara, Mex Tafitasoa, Nadya
Ravololonirina, and Norbert Antilahy, and some villagers (Photo 1) participated in the work. The Network of African Archaeologists in Madagascar, under the direction of Dr Chantal Radimilahy and Prof Barthélemy Manjakahery, provided the logistical equipment used. The various signatures The various signatures of national and local administrative approval, the security provided by the Gendarmerie Brigade of Benenitsy, and above all the kindness and openness of King Zafimanely (Photo 1) – who allowed us access to the (burial) site – contributed enormously to the accomplishment of the expedition.
Madagascar. History students from the University of Toliara, Mex Tafitasoa, Nadya
Ravololonirina, and Norbert Antilahy, and some villagers (Photo 1) participated in the work. The Network of African Archaeologists in Madagascar, under the direction of Dr Chantal Radimilahy and Prof Barthélemy Manjakahery, provided the logistical equipment used. The various signatures The various signatures of national and local administrative approval, the security provided by the Gendarmerie Brigade of Benenitsy, and above all the kindness and openness of King Zafimanely (Photo 1) – who allowed us access to the (burial) site – contributed enormously to the accomplishment of the expedition.
The rock shelter at Ampasimaiky faces east. This arrangement makes the painting panels exposed to the sun in the morning and in the shade beyond noon. This arrangement suggests that painting activities are conducted either early in the morning while the heat of the sun is still moderate, or beyond noon when shade is available below the shelter.
This is based on the fact that the height of the panels does not exceed 130 cm, suggesting that the paintings were intended for people sitting on the floor of the undercroft shelter
Photo 2). Minimal, therefore, is the likelihood of the painting activity – involving seated people – having been conducted during the time when the sun is heating the shelter floor.
However, the view outside the site is limited. The opposing hills – on the other bank of the Ampasimaiky stream – prevent a person standing under the shelter from having a good view of the surroundings (Photo 3). It is much more likely that the view outwards – which implies a defensive strategy (Smith 1995) – was not among the painter’s concerns when he/she chose the site. This demonstrates that the layout of the Ampasimaiky rock shelter favours invaders from unexpectedly attacking its residents. The place would therefore not be suitable for permanent habitation.
Photo 3
Especially since a person standing on the hills of the other bank would be able to look directly at the other one positioned under the rock shelter, unless he hides in the deeper part of the shelter. Since the wall of this “chamber”, according to the local population, is not painted, it is plausible that the activities related to the paintings in the Ampasimaiky rock shelter are not clandestine. However, the paintings were not intended to be seen by people passing close to the site. They are not seen from a distance. Not only are the panels placed below 130 cm in height, but the slope of the ground before the access
to the shelter also prevents anyone from seeing them until they get close to the entrance. Nevertheless, the use of darker coloured paints (red and black) suggests the painter’s intention to contrast the paints he/she was using against the pale colour of the sandstone wall. This implies his concern for the visibility of his paintings.
The walls of the rock shelter have red, reddish orange, burgundy red, black, and white paintings in monochrome, bichrome and polychrome patterns. The images
are painted with the fingertip. They depict zoomorphic (mainly zebu) and anthropomorphic (semi-naturalist and stylised) figures, as well as schematic-geometric-amorphic symbols (Figure 3). The latter dominate and are represented by quadrangular, circular and elliptical signs, dot/dash lines, and alphabet-like signs. In total, the Ampasimaiky rock paintings have 166 recognisable images.
Analysis and Interpretation
Based on comparative studies of the images, it has been found that the images of the Ampasimaiky rock paintings are not different from the images depicted on the prehistoric SGA rock art of the African continent (see Chami, 2008; Chaplin, 1974; Coulson and Campbell 2001;
Eastwood and Smith, 2005; Lynch and Donahue, 1980; Mabulla, 2005; Smith, 1995; Teka 2008; Willcox 1984). The similarities apparently suggest links between the two regions (Madagascar and the mainland) in the
(Madagascar and the continent) in prehistoric times. This suggests that the Ampasimaiky rock paintings are of African origin and could be relatively dated to the same period as those on the continent (Thackeray, 1983: 22). Comparative studies of the images have also shown that more than 30 types of images in the Ampasimaiky rock paintings are closely related to different ancient scripts such as Libyco-Berber, Protosinaitic, Meroitic (cursive), Punic Iberian, Ancient Phoenician, Tifinagh, and Aegean, Cretan and Proto-Egyptian linear signs respectively (Figure 4). These similarities would also suggest that the SGA-type rock paintings of Ampasimaiky are ancient writing.
figure 4
Analysis and Interpretation
Based on comparative studies of the images, it has been found that the images of the Ampasimaiky rock paintings are not different from the images depicted on the prehistoric SGA rock art of the African continent (see Chami, 2008; Chaplin, 1974; Coulson and Campbell 2001;
Eastwood and Smith, 2005; Lynch and Donahue, 1980; Mabulla, 2005; Smith, 1995; Teka 2008; Willcox 1984). The similarities apparently suggest links between the two regions (Madagascar and the mainland) in the
(Madagascar and the continent) in prehistoric times. This suggests that the Ampasimaiky rock paintings are of African origin and could be relatively dated to the same period as those on the continent (Thackeray, 1983: 22). Comparative studies of the images have also shown that more than 30 types of images in the Ampasimaiky rock paintings are closely related to different ancient scripts such as Libyco-Berber, Protosinaitic, Meroitic (cursive), Punic Iberian, Ancient Phoenician, Tifinagh, and Aegean, Cretan and Proto-Egyptian linear signs respectively (Figure 4). These similarities would also suggest that the SGA-type rock paintings of Ampasimaiky are ancient writing. them (Willcox, 1984: 210); the same has been applied to that of Professor Barry Fell, a specialist in Libyan-Berber writing. The same has been applied to that of Professor Barry Fell, a specialist in Libyan-Berber writing, who translated the glyphs of Driekopseiland in southern Africa as Libyan-Berber writing. But it has also been refuted on the grounds that the Bushmen could not write (Willcox 1984: 210); it is important to recall
It is important to recall in this connection that Libyco-Berber inscriptions inscribed on petroglyphs in the Canary Islands were not accepted as such, under the pretext that the Guanches (indigenous people of the Canary Islands) could not have acquired or adopted any writing system in prehistoric times (Cline 1953; Farrujia de la Rosa et al. 2010: 16-17; Willcox 1984: 48). This erroneous model persisted until reputable scholars succeeded in translating an inscription reading “save me from javelin annihilation” (Cline 1953: 273). Today, it is commonly accepted that the groups of aligned glyphs from the Canary Islands are Libyco-Berber inscriptions. Recently, Chami (2006: 84-5) insisted that African SGA rock paintings and engravings are ancient scripts. He has shown that some, such as those at Namoratunga in Kenya, are arranged. For the accounts of those that
are not arranged, he argued that at the time when the ancient scripts of Ethiopia, India, China, Egypt and Mesopotamia were still using ideograms, the characters too were not aligned (see also Casson 1966: 141). This can be amplified by the case of some Sahelian writing systems such as the ‘Nsibidi. This script has no ‘order in the arrangement’, but it is practically recognised as a writing system (Dayrell
1910; Macgregor 1909). Zahan (1950) identified the Dogon and Bambara pictographs as scripts based on the argument that
“writing consists essentially in the communication of thought by means of conventional signs, and the pictographs of both Dogon and Bambara serve to convey thoughts from one writer to another or from writer to lay public” (pp.137) « Une écriture consiste essentiellement de la communication de pensée par le moyen de signes conventionnels, et les pictographes de Dogon et de Bambara servent à transporter des pensées d’un auteur à l’autre ou d’un auteur au public”.
Therefore, it is also possible that without the “order in the arrangement”, the Ampasimaiky SGA-type rock paintings could have served to communicate thoughts and information. This would qualify them as a WRITING SYSTEM! Especially since some groups of symbols are horizontally and vertically aligned (Photo 4). Since “order in arrangement” has been established by some scholars as the sine qua non for a group of symbols/signs to qualify as an inscription (Willcox 1984: 210), one can reasonably assert, based on the above evidence, that the
AMPASIMAIKY RUPESTRATED ROCK SHELTER PAINTINGS ARE INSCRIPTIONS! This assertion is clearly supported by the vertically arranged group of symbols in panel 8 (see Photo 4b). Indeed, the repetition of the first sign in the third character (read downwards) is undoubtedly demonstrating the USE OF THE ALPHABET, and thus the
ALPHABET, so this is WRITING! All the more so as all five characters constituting the inscription are closely related to the Libyan-Berber alphabets (see
www.ancientscripts.com), the proto-Egyptian, Aegean and Cretan linear signs respectively
(Figure 5).
Figure 5 : Le groupe de signes verticalement arrangées d’Ampasimaiky comparé avec des caractères alphabétiques libyco-berbères et des signes linéaires proto-Egyptiens, égéens, et crétois (Tableau : Evans, 1897 :386).
Bien que des signes linéaires proto-Égyptiens, crétois et égéens aient été également verticalement écrits, je favoriserais l’origine libyco-berbère de l’inscription d’Ampasimaiky sous l’influence de la description de Walter Cline concernant les inscriptions libyco-berbères comme
suit).
“They [Libyco-Berber inscriptions] are usually very brief, consisting of personal names and a few phrases … The characters are simple and predominantly rectilinear. Variant alphabet may have existed. Most of inscriptions are to be read upward beginning with the left-hand column, though horizontal rows reading from right to left are not uncommon and some columns read downwards. The words are not separated. Though these signs sometimes occur in groups of four or five, they often stand singly…” (Cline 1953: 273)
All the underlined sentences qualify the vertically arranged inscription of Ampasimaiky as being of Libyco-berberber origin. Although this ancient script was written either horizontally or vertically, it should be remembered that the Libyco- Berber peoples used two different alphabets for each row (see http://www.ancientscripts.com). All five characters of the Ampasimaiky inscription are exclusively related to the vertical Libyco- Berber alphabet (Figure 6). Transliterating them to Latin alphabets – based on the transliteration of www.ancientscripts.com – one could read “ğ r ğ t f” (downwards) or “f t ğ r ğ” (upwards). The phonetic value ‘r’ for the presumed circle/square sign can be read as ‘3’ (= a), if the represented sign is perceived as a triangular symbol (see Fell 1975: 271). In the present state of the study, decoding the meaning of the inscription is not yet within the scope of the author. This task would rather be reserved for palaeographers and epigraphers.
Figure 7: The group of vertically arranged signs of Ampasimaiky transliterated in the Libyco-Berber alphabet
It is important to point out here that certain Libyco-Berber inscriptions brought back from the Canary Islands are not different from that of Ampasimaiky (see Figure 8). You can clearly see their resemblance.
Figure 8: The group of vertically arranged signs from Ampasimaiky compared with a Libyco-Berber inscription from the Canary Islands (Drawings by Farrujia de la Rosa et al. 2010: 29)
Discussion and Conclusion
Some of the images found at Ampasimaiky are similar to those of the rock art of continental Africa. However, the painting tradition of southwestern Madagascar is not
However, the painting tradition of southwestern Madagascar does not seem to originate from the continental region closest to Madagascar, i.e. Mozambique and the rest of the central Limpopo basin. In fact, Professor Benjamin Smith, an internationally recognised authority on the study of SGA rock paintings in sub-Saharan Africa, has taken the view that they are not typically affiliated with the Khoekhoen SGA rock paintings, nor with the geometric art of central southern Africa (pers. comm.). The exclusive representation of humped cattle associated with SGA-type symbols is particular to Ampasimaiky. Since no ancient representation of zebu (Bos indicus) has yet been found in the region between Sudan in the north and the Zambezi River in the south (Magnavita 2006: 62), it can be inferred that the tradition of rock paintings in Ampasimaiky originated in North Africa. Evidence of a bronze zebu figurine (of the type Bos indicus) inscribed with SGA signs dating back to 200 CE at Zeban Kutur, Eritrea (Photo 4) would confirm the early use of SGA signs among pastoral zebu-breeding peoples such as those depicted in the Ampasimaiky rock paintings
Photo 4: Bronze figure of zebu inscribed with SGA type signs in Zeban Kutur, Eritrea (Photo: Magnavita 2006:63)
The Libyan-Berber inscription discovered at Ampasimaiky also reinforces this proposal. The Libyan-Berber script emerged in North Africa around 500 BC and ceased to be used around 800 AD (Cline 1953).
around 800 AD (Cline 1953). This implies that the Libyan-Berber inscription of Ampasimaiky should have been written during the period between this bracket of dates (500 BC – AD 800). It is worth mentioning that Iamboulos, a Greek merchant who lived in the 3rd
century BC, reported about a vertical order of writing that was practised by the inhabitants of the ‘island of the sun’ (Allibert 2010: 111; Kobishschanow, 1965: 138; Oldfather, 1961 vol.2: 65-67; Winston 1976: 221). This island is believed to be geographically located adjacent to the east coast of Africa, south of the Equator. Although some of the geographical and ecological information in Iamboulos’ account is very much in line with that of the Malagasy west coast (for more details see the
Although some of the geographical and ecological information in Iamboulos’ account closely matches that of the Malagasy west coast (for more information see Allibert 2010), some scholars have consistently judged this account to be fictional, imaginary, and fanciful (Cary and Warmington 1963: 233-245; Winston 1976). One of the criteria used to censor this narrative is the use of vertical order writing by the islanders. Certainly, it was difficult for some scholars to accept the existence of a pre-Islamic script used in the adjacent islands of the east coast of Africa such as Madagascar (Janvier 1975: 19). However, with the discovery of the Libyan-Berber inscription at Ampasimaiky, this denial no longer holds. New evidence is uncovered to confirm that the “island of the sun” of Iamboulos could have been Madagascar. In any case, man was already present in the vicinity of Ampasimaiky when the Greek merchant passed through the island. Human activity is attested by subfossil lemur bones bearing cutting marks and radiocarbon-dated from the 5th to the 4th century BC at Taolambiby, in the upper Onilahy valley (see Perez et al., 2005). All the more so since ancient written documents (see Jones 1960: 235, 269, 315; Strabo 17, 1, 13 in Lacroix 1998: 95), archaeology (see Cary and Warmington 1963: 130; Chami 2007: 6; 2009; Wainwright 1940: 165), genetic study (see Ricaut et al. 2009: 5-6), and zooarchaeology (Chami 2006: 137; Grigson 1991: 124) all confirm the ancient contacts that existed during this period between the Onilahy Valley and the Onilahy Valley,
During this period, the Nile Valley and the Mediterranean world on the one hand, and the eastern coast of Africa and the adjacent islands on the other.
Références Allibert, C. (2008), “Austronesian Migration and the Establishment of the Malagasy Civilization: Contrasted Readings in Linguistics, Archaeology, Genetics and Cultural Anthropology”, Diogenes, SAGE, Volume 218, pp.7-16 Allibert, C. (2010), “L’île de Iamboulos est-elle Madagascar? », Pount, no.4, pp.103-119 Blench, R. (2007), “New Palaeozoogeographical Evidence for the Settlement of Madagascar”, Azania, Volume XLII, pp.69-82 Cary M. and Warmington, E. (1963), The Ancient Explorers, Middlesex, Penguin Books Casson, L. (1966), Ancient Egypt, Nederland, Time-life International Chami, F. A (2006), The Unity of African Ancient History, 3000 BC to AD 500, Dar es Salaam, E&D Limited Chami, F. A (2008), “The Great Lakes: a complexity of cultural wellsprings”, in Arnold, M. (eds.) Art in Eastern Africa, pp.47-64, Dar-es-Salaam, Mkuki na Nyota Chami F. A and Kwekason A. (2003), “Neolithic Pottery Traditions from the Islands, the Coast, and the Interior of East Africa”, African Archaeological Review, Volume 20, No. 2, pp.65-80 Chami, F. A, Tabibou, I. A and Abderhamane, B. (2009), “Preliminary Report of Archaeological Work Conducted on the Southern Ngazidja Island”, in Chami, F. (eds.), Zanzibar and the Swahili Coast from c30.000 years ago, pp.115-128, Dar es Salaam, E&D Vision Publishers Chaplin, T. (1974), “The Prehistoric Rock Art of the Lake Victoria Region”, Azania Volume 9, pp.1-50 Cline, W. (1953), “Berber Dialects and Berber Script”, Southwestern Journal of Anthropology, Vol.9, No.3, pp.268-276, New Mexico, New Mexico University Press Coulson D. and Campbell A. (2001), African Rock Art, Paintings and Engravings on Stone, New York, Harry N. Abrams Incorporated Dayrell, E. (1910), “Some “Nsibidi” Signs”, Man, Vol.10, pp.113-114. Dewar, R. E (1996), “The Archaeology of the Early Settlement of Madagascar”, in Reade J. (eds.) The Indian Ocean in Antiquity, pp.471-486 Dewar, R. E (1997), “Does It Matter That Madagascar Is an Island?” Human Ecology, Volume 25, No. 3, pp. 481-489 Donque, G. (1965), « Le contexte océanique des anciennes migrations : vents et courants dans l’océan indien », Taloha, Volume 1, pp.43-59. Eastwood, E. B and Smith, B. W (2005), “Fingerprints of the Khoekhoen: Geometric and handprinted Rock Art in the Central Limpopo Basin, Southern Africa”, Goodwin Series, RSA, South African Archaeological Society, Volume 9, pp.63-76 Evans, A. J (1897), “Further Discoveries of Cretan and Aegean Script: With Libyan and ProtoEgyptian Comparisons”, The Journal of Hellenic Studies, The Society for the Promotion of Hellenic Studies, Volume 17, pp.327-395 Farrujia de la Rosa, A., Pichler W., Rodrigue A., Marín S.G. 2010, The Lybico-Berber and Latino-Canarian Scripts and the Colonization of the Canary Islands, African Archaeological Review, vol.27, no.1, pp.13-41 Faublée, J. (1947), Récits Bara, Travaux et mémoires de l’Institut d’ethnologie XLXII, Paris, Université de Paris Fell, B. (1975), America BC: Ancient Settlers in the New World, New York, A Demeter Press Book Grandidier, A., Charles-Roux J., Delhorbe C., Froidevaux H. and Grandidier G. (1902), Collection des ouvrages anciens concernant Madagascar, Tome II, Paris, Comité de Madagascar Grigson, C. (1991), “An African Origin of African Cattle? Some Archaeological Evidence”, The African Archaeological Review, Volume 9, pp.119-144 Hall, S. and Smith, B. (2000), “Empowering Places: Rock Shelters and Ritual Control in FarmerForager Interactions in the Northern Province”, Goodwin Series, RSA, South African Archaeological Society, Volume 8, pp.30-46 Hébert, J. C (1967), « Essai d’interprétation de la stèle indéchiffrée d’Ambilobe », Taloha, Volume 2, pp.XLVII-LVII Hébert, J. C (1971), « Les tatouages de la côte est de Madagascar, d’après Chapelier (1794- 1806) », Taloha, Volume 4, pp.211-223 Huygues-Belrose, V. (1997). Les supports anciens de l’écriture à Madagascar, Etudes Océan Indien, vol.22, pp.7-19, INALCO Janvier, Y. (1975), « La géographie gréco-romaine a-t-elle connu Madagascar? Le point de la question », Omaly sy Anio, No.1-2, pp.11-41 Jones, H. (1960), The Geography of Strabo, Volume 1-2, London, William Heinemann Ltd. Juma, A. (2004), Unguja Ukuu on Zanzibar – An Archaeological Study of Early Urbanism, Sweden, Uppsala University Kessy, E. (2009a), “Analysis of Lithic Artefacts From Mwanampambe and Kuumbi Caves, Zanzibar”, in Chami F. (eds.), Zanzibar and the Swahili Coast from c30,000 years ago, Dar es Salaam, E&D Vision Publishers, pp.131-143 Kessy, E. (2009b), “Stone Artefacts from Male”, in Chami F. (eds.), Zanzibar and the Swahili Coast from c30,000 years ago, Dar es Salaam, E&D Vision Publishers, pp.184-88 Kobishchanov, Y.M. (1965), “On the Problem of Sea Voyages of Ancient Africans in the Indian Ocean”, The Journal of African History, Vol.6, No.2, pp.137-141, Cambridge, Cambridge University Press. Lacroix, W. F (1998), Africa in Antiquity, Nijmegen, Catholic University
Tanambelo V.R. RASOLONDRAINY,
(Centre de Documentation et de Recherche sur l’Art et la Tradition Orale de Madagascar)
Abstract:
This article presents the result of an archaeological study on the first prehistoric rock paintings site found in Madagascar and the rest of islands of Western Indian Ocean south of the equator.
Based on systematic methods and techniques, the study provides groundbreaking archaeological evidences that contribute largely to the understanding of the prehistory of Madagascar which was not archaeologically known thitherto to have prehistoric rock paintings, nor human settlement prior to 8th century AD. A Libyco-Berber inscription was discovered on the rock paintings of Ampasimaiky Rock Shelter, in Southwestern Madagascar. Based on this writing -supported by ancient historical records, circumstantial archaeological and paleontological evidences, linguistic data, and to some extent DNA analysis results -, the present study provides minimum age falling in the bracket time around 500 BC to roughly 8th century AD for the rock paintings of Ampasimaiky rock shelter. Furthermore, the study traces back early contacts between Madagascar and Northern Africa during prehistoric times.
Introduction
With the kind assistance of Dr Chantal Radimilahy, National Coordinator of the Network of African Archaeologists in Madagascar, I researched the ethnographic exhibition on the website of the Museum of Ethnography in Geneva (MEG). The theme was the peoples and cultures of the southwestern region of Madagascar, and was based on an exhibition of photographs taken by the French ethnographer Jacques Faublée in 1939. The pictures included rock paintings found in caves and rock shelters in the Isalo region. After discussions with my teachers at the Archaeology Unit of the University of Dar es Salaam, Professors Felix Chami, Fidelis Masao, and Audax Mabulla, I further investigated the significance of these paintings for the prehistoric period of Madagascar. This study is taken from a Master of Arts in Archaeology entitled
“Archaeologial study of the prehistoric rock paintings of Ampasimaiky rock shelter, in the upper Onilahy, Isalo region, Southwestern Madagascar”.
————–
Africa has many archaeological sites with prehistoric rock paintings (see Coulson and Campbell 2001; Willcox 1984). However, no such sites have yet been scientifically
scientifically reported from Madagascar, which is only 400 kilometres from southern Africa where they are abundant. All rock art previously found in Madagascar has been judged to have been produced in the recent past (Faublée 1947: 164; Hébert 1971: 218). This judgement is based on the theory that Africans were incapable of having navigated the high seas during prehistoric times (Donque 1965: 43; Grandidier et al. 1902; Southall 1975: 194; Vérin 1975, 1981). It was therefore out of the question to think that they had the possibility of transferring the practice of rock painting to the adjacent islands of Africa such as Madagascar. However, recent archaeological findings have demonstrated that the east coast of Africa and adjacent islands were already in contact since the prehistoric period (see Chami and Kwekason 2003; Chami et al. 2009; Juma 2004; Kessy 2009a,
2009b). These prompt a problem that led me to revise the antiquity of the rock paintings in Madagascar.
On the other hand, the name of the Arabic-Malagasy script – Sorabe meaning “great script” – indicates that there was an ancient script in Madagascar that was distinguished from Arabic (Hébert,
1967:XLVII; Huygues-Belrose 1997: 8). Although Hébert (1967:XLVIII) presumed that such a hypothetical script would have to be of Indian origin and would be discovered from
Although Hébert (1967:XLVIII) assumed that such a hypothetical script would have to be of Indian origin and would be discovered from stele and parietal inscriptions, no such inscription has so far been found in Madagascar.
My problem was therefore to find out whether this would appear on the rock paintings reported by the website. What provoked this idea the most was that Chami (2006:84-85; 2008) argued that the African Geometric and Amorphous Schematic (GAS) rock paintings and engravings are ancient writing.
Another problem is that radiocarbon 14 dating placed the earliest human settlement of Madagascar around the 8th century AD (Allibert 2008: 7; Dewar 1996: 474).
Although earlier dates are still being considered, many researchers have assumed that settlement of Madagascar would not go beyond the year W AD (for more information see Dewar 1997). This assumption has made research on Malagasy material culture going back to the Stone Age more than difficult. Any findings of material cultures dating from prehistoric times have been discarded, overlooked or misinterpreted. However, subfossil animal bone remains bearing traces of butchery found in paleontological sites in the southwestern region indicate the presence of human activity
(For more information, see Blench 2007; Perez et al. 2005). Despite this evidence, an archaeological site has not yet been
archaeological site with similar dates has not yet been found on the entire island.
With the discovery of a rock shelter with rock paintings, discovering new information about the prehistory of Madagascar before the 8th century proved to be feasible. All the more so as some of the paintings and engravings are
Especially since some of the SGA-type rock paintings and engravings found on the African continent – similar to those found in the photos on the MEG website – have been radiocarbon dated from the 1st millennium BC to the early 1st millennium AD (Chami, 2008: 54; Hall and Smith 2000; Lynche and Donahue 1980; Willcox 1984: 26, 83.
Fieldwork and results
Despite various challenges and constraints – mainly caused by the lack of archaeological reconnaissance prior to the fieldwork, insecurity in the study area, and the sacredness of the site to be studied – we had only somehow managed to conduct survey, archaeological excavation, and documentation of the paintings in the Ampasimaiky rock shelter. It is only the result of the latter that will be presented in this article.
The rock painting site is a rock shelter located at the foot of a hill immediately next to a small stream called Ampasimaiky (Figure 2), from which the name of the site derives. The presence of this water source may have influenced the painter’s choice of location.
The shelter itself is 26.8m long (south-north axis) and 5m wide (inwards).
Its ceiling begins to slope inwards from 4m high at its entrance to approximately 1.5m high at its irregular wall. The deepest part of the rock shelter – a chamber – is currently used by the royal family “Bara Zafimanely” as a burial place. Dung and ox hoof prints found on the ground indicate that cattle frequent the place. This occurrence should have disturbed the soil.
However, the areas around the site are dominated by rocky hills covered with grass. Only in the deep canyon near the confluence of Ampasimaiky and the Sakamarekely River is a patch of thick vegetation up to 10 m high observed along the stream. This situation would make agricultural activity almost impossible around the site, unless one had to move 1.5 km to the bank, where rice cultivation takes place today.
An archaeological survey of the southern Isalo region (Figure 1) was carried out and a rock painting site was discovered. Professor Chami participated in the fieldwork, advising on the approach to be followed in the treatment of these previously unstudied cave cultures in Madagascar.
Madagascar. History students from the University of Toliara, Mex Tafitasoa, Nadya
Ravololonirina, and Norbert Antilahy, and some villagers (Photo 1) participated in the work. The Network of African Archaeologists in Madagascar, under the direction of Dr Chantal Radimilahy and Prof Barthélemy Manjakahery, provided the logistical equipment used. The various signatures The various signatures of national and local administrative approval, the security provided by the Gendarmerie Brigade of Benenitsy, and above all the kindness and openness of King Zafimanely (Photo 1) – who allowed us access to the (burial) site – contributed enormously to the accomplishment of the expedition.
Madagascar. History students from the University of Toliara, Mex Tafitasoa, Nadya
Ravololonirina, and Norbert Antilahy, and some villagers (Photo 1) participated in the work. The Network of African Archaeologists in Madagascar, under the direction of Dr Chantal Radimilahy and Prof Barthélemy Manjakahery, provided the logistical equipment used. The various signatures The various signatures of national and local administrative approval, the security provided by the Gendarmerie Brigade of Benenitsy, and above all the kindness and openness of King Zafimanely (Photo 1) – who allowed us access to the (burial) site – contributed enormously to the accomplishment of the expedition.
The rock shelter at Ampasimaiky faces east. This arrangement makes the painting panels exposed to the sun in the morning and in the shade beyond noon. This arrangement suggests that painting activities are conducted either early in the morning while the heat of the sun is still moderate, or beyond noon when shade is available below the shelter.
This is based on the fact that the height of the panels does not exceed 130 cm, suggesting that the paintings were intended for people sitting on the floor of the undercroft shelter
Photo 2). Minimal, therefore, is the likelihood of the painting activity – involving seated people – having been conducted during the time when the sun is heating the shelter floor.
However, the view outside the site is limited. The opposing hills – on the other bank of the Ampasimaiky stream – prevent a person standing under the shelter from having a good view of the surroundings (Photo 3). It is much more likely that the view outwards – which implies a defensive strategy (Smith 1995) – was not among the painter’s concerns when he/she chose the site. This demonstrates that the layout of the Ampasimaiky rock shelter favours invaders from unexpectedly attacking its residents. The place would therefore not be suitable for permanent habitation.
Photo 3
Especially since a person standing on the hills of the other bank would be able to look directly at the other one positioned under the rock shelter, unless he hides in the deeper part of the shelter. Since the wall of this “chamber”, according to the local population, is not painted, it is plausible that the activities related to the paintings in the Ampasimaiky rock shelter are not clandestine. However, the paintings were not intended to be seen by people passing close to the site. They are not seen from a distance. Not only are the panels placed below 130 cm in height, but the slope of the ground before the access
to the shelter also prevents anyone from seeing them until they get close to the entrance. Nevertheless, the use of darker coloured paints (red and black) suggests the painter’s intention to contrast the paints he/she was using against the pale colour of the sandstone wall. This implies his concern for the visibility of his paintings.
The walls of the rock shelter have red, reddish orange, burgundy red, black, and white paintings in monochrome, bichrome and polychrome patterns. The images
are painted with the fingertip. They depict zoomorphic (mainly zebu) and anthropomorphic (semi-naturalist and stylised) figures, as well as schematic-geometric-amorphic symbols (Figure 3). The latter dominate and are represented by quadrangular, circular and elliptical signs, dot/dash lines, and alphabet-like signs. In total, the Ampasimaiky rock paintings have 166 recognisable images.
Analysis and Interpretation
Based on comparative studies of the images, it has been found that the images of the Ampasimaiky rock paintings are not different from the images depicted on the prehistoric SGA rock art of the African continent (see Chami, 2008; Chaplin, 1974; Coulson and Campbell 2001;
Eastwood and Smith, 2005; Lynch and Donahue, 1980; Mabulla, 2005; Smith, 1995; Teka 2008; Willcox 1984). The similarities apparently suggest links between the two regions (Madagascar and the mainland) in the
(Madagascar and the continent) in prehistoric times. This suggests that the Ampasimaiky rock paintings are of African origin and could be relatively dated to the same period as those on the continent (Thackeray, 1983: 22). Comparative studies of the images have also shown that more than 30 types of images in the Ampasimaiky rock paintings are closely related to different ancient scripts such as Libyco-Berber, Protosinaitic, Meroitic (cursive), Punic Iberian, Ancient Phoenician, Tifinagh, and Aegean, Cretan and Proto-Egyptian linear signs respectively (Figure 4). These similarities would also suggest that the SGA-type rock paintings of Ampasimaiky are ancient writing.
figure 4
Analysis and Interpretation
Based on comparative studies of the images, it has been found that the images of the Ampasimaiky rock paintings are not different from the images depicted on the prehistoric SGA rock art of the African continent (see Chami, 2008; Chaplin, 1974; Coulson and Campbell 2001;
Eastwood and Smith, 2005; Lynch and Donahue, 1980; Mabulla, 2005; Smith, 1995; Teka 2008; Willcox 1984). The similarities apparently suggest links between the two regions (Madagascar and the mainland) in the
(Madagascar and the continent) in prehistoric times. This suggests that the Ampasimaiky rock paintings are of African origin and could be relatively dated to the same period as those on the continent (Thackeray, 1983: 22). Comparative studies of the images have also shown that more than 30 types of images in the Ampasimaiky rock paintings are closely related to different ancient scripts such as Libyco-Berber, Protosinaitic, Meroitic (cursive), Punic Iberian, Ancient Phoenician, Tifinagh, and Aegean, Cretan and Proto-Egyptian linear signs respectively (Figure 4). These similarities would also suggest that the SGA-type rock paintings of Ampasimaiky are ancient writing. them (Willcox, 1984: 210); the same has been applied to that of Professor Barry Fell, a specialist in Libyan-Berber writing. The same has been applied to that of Professor Barry Fell, a specialist in Libyan-Berber writing, who translated the glyphs of Driekopseiland in southern Africa as Libyan-Berber writing. But it has also been refuted on the grounds that the Bushmen could not write (Willcox 1984: 210); it is important to recall
It is important to recall in this connection that Libyco-Berber inscriptions inscribed on petroglyphs in the Canary Islands were not accepted as such, under the pretext that the Guanches (indigenous people of the Canary Islands) could not have acquired or adopted any writing system in prehistoric times (Cline 1953; Farrujia de la Rosa et al. 2010: 16-17; Willcox 1984: 48). This erroneous model persisted until reputable scholars succeeded in translating an inscription reading “save me from javelin annihilation” (Cline 1953: 273). Today, it is commonly accepted that the groups of aligned glyphs from the Canary Islands are Libyco-Berber inscriptions. Recently, Chami (2006: 84-5) insisted that African SGA rock paintings and engravings are ancient scripts. He has shown that some, such as those at Namoratunga in Kenya, are arranged. For the accounts of those that
are not arranged, he argued that at the time when the ancient scripts of Ethiopia, India, China, Egypt and Mesopotamia were still using ideograms, the characters too were not aligned (see also Casson 1966: 141). This can be amplified by the case of some Sahelian writing systems such as the ‘Nsibidi. This script has no ‘order in the arrangement’, but it is practically recognised as a writing system (Dayrell
1910; Macgregor 1909). Zahan (1950) identified the Dogon and Bambara pictographs as scripts based on the argument that
“writing consists essentially in the communication of thought by means of conventional signs, and the pictographs of both Dogon and Bambara serve to convey thoughts from one writer to another or from writer to lay public” (pp.137) « Une écriture consiste essentiellement de la communication de pensée par le moyen de signes conventionnels, et les pictographes de Dogon et de Bambara servent à transporter des pensées d’un auteur à l’autre ou d’un auteur au public”.
Therefore, it is also possible that without the “order in the arrangement”, the Ampasimaiky SGA-type rock paintings could have served to communicate thoughts and information. This would qualify them as a WRITING SYSTEM! Especially since some groups of symbols are horizontally and vertically aligned (Photo 4). Since “order in arrangement” has been established by some scholars as the sine qua non for a group of symbols/signs to qualify as an inscription (Willcox 1984: 210), one can reasonably assert, based on the above evidence, that the
AMPASIMAIKY RUPESTRATED ROCK SHELTER PAINTINGS ARE INSCRIPTIONS! This assertion is clearly supported by the vertically arranged group of symbols in panel 8 (see Photo 4b). Indeed, the repetition of the first sign in the third character (read downwards) is undoubtedly demonstrating the USE OF THE ALPHABET, and thus the
ALPHABET, so this is WRITING! All the more so as all five characters constituting the inscription are closely related to the Libyan-Berber alphabets (see
www.ancientscripts.com), the proto-Egyptian, Aegean and Cretan linear signs respectively
(Figure 5).
Figure 5 : Le groupe de signes verticalement arrangées d’Ampasimaiky comparé avec des caractères alphabétiques libyco-berbères et des signes linéaires proto-Egyptiens, égéens, et crétois (Tableau : Evans, 1897 :386).
Bien que des signes linéaires proto-Égyptiens, crétois et égéens aient été également verticalement écrits, je favoriserais l’origine libyco-berbère de l’inscription d’Ampasimaiky sous l’influence de la description de Walter Cline concernant les inscriptions libyco-berbères comme
suit).
“They [Libyco-Berber inscriptions] are usually very brief, consisting of personal names and a few phrases … The characters are simple and predominantly rectilinear. Variant alphabet may have existed. Most of inscriptions are to be read upward beginning with the left-hand column, though horizontal rows reading from right to left are not uncommon and some columns read downwards. The words are not separated. Though these signs sometimes occur in groups of four or five, they often stand singly…” (Cline 1953: 273)
All the underlined sentences qualify the vertically arranged inscription of Ampasimaiky as being of Libyco-berberber origin. Although this ancient script was written either horizontally or vertically, it should be remembered that the Libyco- Berber peoples used two different alphabets for each row (see http://www.ancientscripts.com). All five characters of the Ampasimaiky inscription are exclusively related to the vertical Libyco- Berber alphabet (Figure 6). Transliterating them to Latin alphabets – based on the transliteration of www.ancientscripts.com – one could read “ğ r ğ t f” (downwards) or “f t ğ r ğ” (upwards). The phonetic value ‘r’ for the presumed circle/square sign can be read as ‘3’ (= a), if the represented sign is perceived as a triangular symbol (see Fell 1975: 271). In the present state of the study, decoding the meaning of the inscription is not yet within the scope of the author. This task would rather be reserved for palaeographers and epigraphers.
Figure 7: The group of vertically arranged signs of Ampasimaiky transliterated in the Libyco-Berber alphabet
It is important to point out here that certain Libyco-Berber inscriptions brought back from the Canary Islands are not different from that of Ampasimaiky (see Figure 8). You can clearly see their resemblance.
Figure 8: The group of vertically arranged signs from Ampasimaiky compared with a Libyco-Berber inscription from the Canary Islands (Drawings by Farrujia de la Rosa et al. 2010: 29)
Discussion and Conclusion
Some of the images found at Ampasimaiky are similar to those of the rock art of continental Africa. However, the painting tradition of southwestern Madagascar is not
However, the painting tradition of southwestern Madagascar does not seem to originate from the continental region closest to Madagascar, i.e. Mozambique and the rest of the central Limpopo basin. In fact, Professor Benjamin Smith, an internationally recognised authority on the study of SGA rock paintings in sub-Saharan Africa, has taken the view that they are not typically affiliated with the Khoekhoen SGA rock paintings, nor with the geometric art of central southern Africa (pers. comm.). The exclusive representation of humped cattle associated with SGA-type symbols is particular to Ampasimaiky. Since no ancient representation of zebu (Bos indicus) has yet been found in the region between Sudan in the north and the Zambezi River in the south (Magnavita 2006: 62), it can be inferred that the tradition of rock paintings in Ampasimaiky originated in North Africa. Evidence of a bronze zebu figurine (of the type Bos indicus) inscribed with SGA signs dating back to 200 CE at Zeban Kutur, Eritrea (Photo 4) would confirm the early use of SGA signs among pastoral zebu-breeding peoples such as those depicted in the Ampasimaiky rock paintings
Photo 4: Bronze figure of zebu inscribed with SGA type signs in Zeban Kutur, Eritrea (Photo: Magnavita 2006:63)
The Libyan-Berber inscription discovered at Ampasimaiky also reinforces this proposal. The Libyan-Berber script emerged in North Africa around 500 BC and ceased to be used around 800 AD (Cline 1953).
around 800 AD (Cline 1953). This implies that the Libyan-Berber inscription of Ampasimaiky should have been written during the period between this bracket of dates (500 BC – AD 800). It is worth mentioning that Iamboulos, a Greek merchant who lived in the 3rd
century BC, reported about a vertical order of writing that was practised by the inhabitants of the ‘island of the sun’ (Allibert 2010: 111; Kobishschanow, 1965: 138; Oldfather, 1961 vol.2: 65-67; Winston 1976: 221). This island is believed to be geographically located adjacent to the east coast of Africa, south of the Equator. Although some of the geographical and ecological information in Iamboulos’ account is very much in line with that of the Malagasy west coast (for more details see the
Although some of the geographical and ecological information in Iamboulos’ account closely matches that of the Malagasy west coast (for more information see Allibert 2010), some scholars have consistently judged this account to be fictional, imaginary, and fanciful (Cary and Warmington 1963: 233-245; Winston 1976). One of the criteria used to censor this narrative is the use of vertical order writing by the islanders. Certainly, it was difficult for some scholars to accept the existence of a pre-Islamic script used in the adjacent islands of the east coast of Africa such as Madagascar (Janvier 1975: 19). However, with the discovery of the Libyan-Berber inscription at Ampasimaiky, this denial no longer holds. New evidence is uncovered to confirm that the “island of the sun” of Iamboulos could have been Madagascar. In any case, man was already present in the vicinity of Ampasimaiky when the Greek merchant passed through the island. Human activity is attested by subfossil lemur bones bearing cutting marks and radiocarbon-dated from the 5th to the 4th century BC at Taolambiby, in the upper Onilahy valley (see Perez et al., 2005). All the more so since ancient written documents (see Jones 1960: 235, 269, 315; Strabo 17, 1, 13 in Lacroix 1998: 95), archaeology (see Cary and Warmington 1963: 130; Chami 2007: 6; 2009; Wainwright 1940: 165), genetic study (see Ricaut et al. 2009: 5-6), and zooarchaeology (Chami 2006: 137; Grigson 1991: 124) all confirm the ancient contacts that existed during this period between the Onilahy Valley and the Onilahy Valley,
During this period, the Nile Valley and the Mediterranean world on the one hand, and the eastern coast of Africa and the adjacent islands on the other.
Références Allibert, C. (2008), “Austronesian Migration and the Establishment of the Malagasy Civilization: Contrasted Readings in Linguistics, Archaeology, Genetics and Cultural Anthropology”, Diogenes, SAGE, Volume 218, pp.7-16 Allibert, C. (2010), “L’île de Iamboulos est-elle Madagascar? », Pount, no.4, pp.103-119 Blench, R. (2007), “New Palaeozoogeographical Evidence for the Settlement of Madagascar”, Azania, Volume XLII, pp.69-82 Cary M. and Warmington, E. (1963), The Ancient Explorers, Middlesex, Penguin Books Casson, L. (1966), Ancient Egypt, Nederland, Time-life International Chami, F. A (2006), The Unity of African Ancient History, 3000 BC to AD 500, Dar es Salaam, E&D Limited Chami, F. A (2008), “The Great Lakes: a complexity of cultural wellsprings”, in Arnold, M. (eds.) Art in Eastern Africa, pp.47-64, Dar-es-Salaam, Mkuki na Nyota Chami F. A and Kwekason A. (2003), “Neolithic Pottery Traditions from the Islands, the Coast, and the Interior of East Africa”, African Archaeological Review, Volume 20, No. 2, pp.65-80 Chami, F. A, Tabibou, I. A and Abderhamane, B. (2009), “Preliminary Report of Archaeological Work Conducted on the Southern Ngazidja Island”, in Chami, F. (eds.), Zanzibar and the Swahili Coast from c30.000 years ago, pp.115-128, Dar es Salaam, E&D Vision Publishers Chaplin, T. (1974), “The Prehistoric Rock Art of the Lake Victoria Region”, Azania Volume 9, pp.1-50 Cline, W. (1953), “Berber Dialects and Berber Script”, Southwestern Journal of Anthropology, Vol.9, No.3, pp.268-276, New Mexico, New Mexico University Press Coulson D. and Campbell A. (2001), African Rock Art, Paintings and Engravings on Stone, New York, Harry N. Abrams Incorporated Dayrell, E. (1910), “Some “Nsibidi” Signs”, Man, Vol.10, pp.113-114. Dewar, R. E (1996), “The Archaeology of the Early Settlement of Madagascar”, in Reade J. (eds.) The Indian Ocean in Antiquity, pp.471-486 Dewar, R. E (1997), “Does It Matter That Madagascar Is an Island?” Human Ecology, Volume 25, No. 3, pp. 481-489 Donque, G. (1965), « Le contexte océanique des anciennes migrations : vents et courants dans l’océan indien », Taloha, Volume 1, pp.43-59. Eastwood, E. B and Smith, B. W (2005), “Fingerprints of the Khoekhoen: Geometric and handprinted Rock Art in the Central Limpopo Basin, Southern Africa”, Goodwin Series, RSA, South African Archaeological Society, Volume 9, pp.63-76 Evans, A. J (1897), “Further Discoveries of Cretan and Aegean Script: With Libyan and ProtoEgyptian Comparisons”, The Journal of Hellenic Studies, The Society for the Promotion of Hellenic Studies, Volume 17, pp.327-395 Farrujia de la Rosa, A., Pichler W., Rodrigue A., Marín S.G. 2010, The Lybico-Berber and Latino-Canarian Scripts and the Colonization of the Canary Islands, African Archaeological Review, vol.27, no.1, pp.13-41 Faublée, J. (1947), Récits Bara, Travaux et mémoires de l’Institut d’ethnologie XLXII, Paris, Université de Paris Fell, B. (1975), America BC: Ancient Settlers in the New World, New York, A Demeter Press Book Grandidier, A., Charles-Roux J., Delhorbe C., Froidevaux H. and Grandidier G. (1902), Collection des ouvrages anciens concernant Madagascar, Tome II, Paris, Comité de Madagascar Grigson, C. (1991), “An African Origin of African Cattle? Some Archaeological Evidence”, The African Archaeological Review, Volume 9, pp.119-144 Hall, S. and Smith, B. (2000), “Empowering Places: Rock Shelters and Ritual Control in FarmerForager Interactions in the Northern Province”, Goodwin Series, RSA, South African Archaeological Society, Volume 8, pp.30-46 Hébert, J. C (1967), « Essai d’interprétation de la stèle indéchiffrée d’Ambilobe », Taloha, Volume 2, pp.XLVII-LVII Hébert, J. C (1971), « Les tatouages de la côte est de Madagascar, d’après Chapelier (1794- 1806) », Taloha, Volume 4, pp.211-223 Huygues-Belrose, V. (1997). Les supports anciens de l’écriture à Madagascar, Etudes Océan Indien, vol.22, pp.7-19, INALCO Janvier, Y. (1975), « La géographie gréco-romaine a-t-elle connu Madagascar? Le point de la question », Omaly sy Anio, No.1-2, pp.11-41 Jones, H. (1960), The Geography of Strabo, Volume 1-2, London, William Heinemann Ltd. Juma, A. (2004), Unguja Ukuu on Zanzibar – An Archaeological Study of Early Urbanism, Sweden, Uppsala University Kessy, E. (2009a), “Analysis of Lithic Artefacts From Mwanampambe and Kuumbi Caves, Zanzibar”, in Chami F. (eds.), Zanzibar and the Swahili Coast from c30,000 years ago, Dar es Salaam, E&D Vision Publishers, pp.131-143 Kessy, E. (2009b), “Stone Artefacts from Male”, in Chami F. (eds.), Zanzibar and the Swahili Coast from c30,000 years ago, Dar es Salaam, E&D Vision Publishers, pp.184-88 Kobishchanov, Y.M. (1965), “On the Problem of Sea Voyages of Ancient Africans in the Indian Ocean”, The Journal of African History, Vol.6, No.2, pp.137-141, Cambridge, Cambridge University Press. Lacroix, W. F (1998), Africa in Antiquity, Nijmegen, Catholic University
Tanambelo V.R. RASOLONDRAINY,
(Centre de Documentation et de Recherche sur l’Art et la Tradition Orale de Madagascar)
Abstract:
This article presents the result of an archaeological study on the first prehistoric rock paintings site found in Madagascar and the rest of islands of Western Indian Ocean south of the equator.
Based on systematic methods and techniques, the study provides groundbreaking archaeological evidences that contribute largely to the understanding of the prehistory of Madagascar which was not archaeologically known thitherto to have prehistoric rock paintings, nor human settlement prior to 8th century AD. A Libyco-Berber inscription was discovered on the rock paintings of Ampasimaiky Rock Shelter, in Southwestern Madagascar. Based on this writing -supported by ancient historical records, circumstantial archaeological and paleontological evidences, linguistic data, and to some extent DNA analysis results -, the present study provides minimum age falling in the bracket time around 500 BC to roughly 8th century AD for the rock paintings of Ampasimaiky rock shelter. Furthermore, the study traces back early contacts between Madagascar and Northern Africa during prehistoric times.
Introduction
With the kind assistance of Dr Chantal Radimilahy, National Coordinator of the Network of African Archaeologists in Madagascar, I researched the ethnographic exhibition on the website of the Museum of Ethnography in Geneva (MEG). The theme was the peoples and cultures of the southwestern region of Madagascar, and was based on an exhibition of photographs taken by the French ethnographer Jacques Faublée in 1939. The pictures included rock paintings found in caves and rock shelters in the Isalo region. After discussions with my teachers at the Archaeology Unit of the University of Dar es Salaam, Professors Felix Chami, Fidelis Masao, and Audax Mabulla, I further investigated the significance of these paintings for the prehistoric period of Madagascar. This study is taken from a Master of Arts in Archaeology entitled
“Archaeologial study of the prehistoric rock paintings of Ampasimaiky rock shelter, in the upper Onilahy, Isalo region, Southwestern Madagascar”.
————–
Africa has many archaeological sites with prehistoric rock paintings (see Coulson and Campbell 2001; Willcox 1984). However, no such sites have yet been scientifically
scientifically reported from Madagascar, which is only 400 kilometres from southern Africa where they are abundant. All rock art previously found in Madagascar has been judged to have been produced in the recent past (Faublée 1947: 164; Hébert 1971: 218). This judgement is based on the theory that Africans were incapable of having navigated the high seas during prehistoric times (Donque 1965: 43; Grandidier et al. 1902; Southall 1975: 194; Vérin 1975, 1981). It was therefore out of the question to think that they had the possibility of transferring the practice of rock painting to the adjacent islands of Africa such as Madagascar. However, recent archaeological findings have demonstrated that the east coast of Africa and adjacent islands were already in contact since the prehistoric period (see Chami and Kwekason 2003; Chami et al. 2009; Juma 2004; Kessy 2009a,
2009b). These prompt a problem that led me to revise the antiquity of the rock paintings in Madagascar.
On the other hand, the name of the Arabic-Malagasy script – Sorabe meaning “great script” – indicates that there was an ancient script in Madagascar that was distinguished from Arabic (Hébert,
1967:XLVII; Huygues-Belrose 1997: 8). Although Hébert (1967:XLVIII) presumed that such a hypothetical script would have to be of Indian origin and would be discovered from
Although Hébert (1967:XLVIII) assumed that such a hypothetical script would have to be of Indian origin and would be discovered from stele and parietal inscriptions, no such inscription has so far been found in Madagascar.
My problem was therefore to find out whether this would appear on the rock paintings reported by the website. What provoked this idea the most was that Chami (2006:84-85; 2008) argued that the African Geometric and Amorphous Schematic (GAS) rock paintings and engravings are ancient writing.
Another problem is that radiocarbon 14 dating placed the earliest human settlement of Madagascar around the 8th century AD (Allibert 2008: 7; Dewar 1996: 474).
Although earlier dates are still being considered, many researchers have assumed that settlement of Madagascar would not go beyond the year W AD (for more information see Dewar 1997). This assumption has made research on Malagasy material culture going back to the Stone Age more than difficult. Any findings of material cultures dating from prehistoric times have been discarded, overlooked or misinterpreted. However, subfossil animal bone remains bearing traces of butchery found in paleontological sites in the southwestern region indicate the presence of human activity
(For more information, see Blench 2007; Perez et al. 2005). Despite this evidence, an archaeological site has not yet been
archaeological site with similar dates has not yet been found on the entire island.
With the discovery of a rock shelter with rock paintings, discovering new information about the prehistory of Madagascar before the 8th century proved to be feasible. All the more so as some of the paintings and engravings are
Especially since some of the SGA-type rock paintings and engravings found on the African continent – similar to those found in the photos on the MEG website – have been radiocarbon dated from the 1st millennium BC to the early 1st millennium AD (Chami, 2008: 54; Hall and Smith 2000; Lynche and Donahue 1980; Willcox 1984: 26, 83.
Fieldwork and results
Despite various challenges and constraints – mainly caused by the lack of archaeological reconnaissance prior to the fieldwork, insecurity in the study area, and the sacredness of the site to be studied – we had only somehow managed to conduct survey, archaeological excavation, and documentation of the paintings in the Ampasimaiky rock shelter. It is only the result of the latter that will be presented in this article.
The rock painting site is a rock shelter located at the foot of a hill immediately next to a small stream called Ampasimaiky (Figure 2), from which the name of the site derives. The presence of this water source may have influenced the painter’s choice of location.
The shelter itself is 26.8m long (south-north axis) and 5m wide (inwards).
Its ceiling begins to slope inwards from 4m high at its entrance to approximately 1.5m high at its irregular wall. The deepest part of the rock shelter – a chamber – is currently used by the royal family “Bara Zafimanely” as a burial place. Dung and ox hoof prints found on the ground indicate that cattle frequent the place. This occurrence should have disturbed the soil.
However, the areas around the site are dominated by rocky hills covered with grass. Only in the deep canyon near the confluence of Ampasimaiky and the Sakamarekely River is a patch of thick vegetation up to 10 m high observed along the stream. This situation would make agricultural activity almost impossible around the site, unless one had to move 1.5 km to the bank, where rice cultivation takes place today.
An archaeological survey of the southern Isalo region (Figure 1) was carried out and a rock painting site was discovered. Professor Chami participated in the fieldwork, advising on the approach to be followed in the treatment of these previously unstudied cave cultures in Madagascar.
Madagascar. History students from the University of Toliara, Mex Tafitasoa, Nadya
Ravololonirina, and Norbert Antilahy, and some villagers (Photo 1) participated in the work. The Network of African Archaeologists in Madagascar, under the direction of Dr Chantal Radimilahy and Prof Barthélemy Manjakahery, provided the logistical equipment used. The various signatures The various signatures of national and local administrative approval, the security provided by the Gendarmerie Brigade of Benenitsy, and above all the kindness and openness of King Zafimanely (Photo 1) – who allowed us access to the (burial) site – contributed enormously to the accomplishment of the expedition.
Madagascar. History students from the University of Toliara, Mex Tafitasoa, Nadya
Ravololonirina, and Norbert Antilahy, and some villagers (Photo 1) participated in the work. The Network of African Archaeologists in Madagascar, under the direction of Dr Chantal Radimilahy and Prof Barthélemy Manjakahery, provided the logistical equipment used. The various signatures The various signatures of national and local administrative approval, the security provided by the Gendarmerie Brigade of Benenitsy, and above all the kindness and openness of King Zafimanely (Photo 1) – who allowed us access to the (burial) site – contributed enormously to the accomplishment of the expedition.
The rock shelter at Ampasimaiky faces east. This arrangement makes the painting panels exposed to the sun in the morning and in the shade beyond noon. This arrangement suggests that painting activities are conducted either early in the morning while the heat of the sun is still moderate, or beyond noon when shade is available below the shelter.
This is based on the fact that the height of the panels does not exceed 130 cm, suggesting that the paintings were intended for people sitting on the floor of the undercroft shelter
Photo 2). Minimal, therefore, is the likelihood of the painting activity – involving seated people – having been conducted during the time when the sun is heating the shelter floor.
However, the view outside the site is limited. The opposing hills – on the other bank of the Ampasimaiky stream – prevent a person standing under the shelter from having a good view of the surroundings (Photo 3). It is much more likely that the view outwards – which implies a defensive strategy (Smith 1995) – was not among the painter’s concerns when he/she chose the site. This demonstrates that the layout of the Ampasimaiky rock shelter favours invaders from unexpectedly attacking its residents. The place would therefore not be suitable for permanent habitation.
Photo 3
Especially since a person standing on the hills of the other bank would be able to look directly at the other one positioned under the rock shelter, unless he hides in the deeper part of the shelter. Since the wall of this “chamber”, according to the local population, is not painted, it is plausible that the activities related to the paintings in the Ampasimaiky rock shelter are not clandestine. However, the paintings were not intended to be seen by people passing close to the site. They are not seen from a distance. Not only are the panels placed below 130 cm in height, but the slope of the ground before the access
to the shelter also prevents anyone from seeing them until they get close to the entrance. Nevertheless, the use of darker coloured paints (red and black) suggests the painter’s intention to contrast the paints he/she was using against the pale colour of the sandstone wall. This implies his concern for the visibility of his paintings.
The walls of the rock shelter have red, reddish orange, burgundy red, black, and white paintings in monochrome, bichrome and polychrome patterns. The images
are painted with the fingertip. They depict zoomorphic (mainly zebu) and anthropomorphic (semi-naturalist and stylised) figures, as well as schematic-geometric-amorphic symbols (Figure 3). The latter dominate and are represented by quadrangular, circular and elliptical signs, dot/dash lines, and alphabet-like signs. In total, the Ampasimaiky rock paintings have 166 recognisable images.
Analysis and Interpretation
Based on comparative studies of the images, it has been found that the images of the Ampasimaiky rock paintings are not different from the images depicted on the prehistoric SGA rock art of the African continent (see Chami, 2008; Chaplin, 1974; Coulson and Campbell 2001;
Eastwood and Smith, 2005; Lynch and Donahue, 1980; Mabulla, 2005; Smith, 1995; Teka 2008; Willcox 1984). The similarities apparently suggest links between the two regions (Madagascar and the mainland) in the
(Madagascar and the continent) in prehistoric times. This suggests that the Ampasimaiky rock paintings are of African origin and could be relatively dated to the same period as those on the continent (Thackeray, 1983: 22). Comparative studies of the images have also shown that more than 30 types of images in the Ampasimaiky rock paintings are closely related to different ancient scripts such as Libyco-Berber, Protosinaitic, Meroitic (cursive), Punic Iberian, Ancient Phoenician, Tifinagh, and Aegean, Cretan and Proto-Egyptian linear signs respectively (Figure 4). These similarities would also suggest that the SGA-type rock paintings of Ampasimaiky are ancient writing.
figure 4
Analysis and Interpretation
Based on comparative studies of the images, it has been found that the images of the Ampasimaiky rock paintings are not different from the images depicted on the prehistoric SGA rock art of the African continent (see Chami, 2008; Chaplin, 1974; Coulson and Campbell 2001;
Eastwood and Smith, 2005; Lynch and Donahue, 1980; Mabulla, 2005; Smith, 1995; Teka 2008; Willcox 1984). The similarities apparently suggest links between the two regions (Madagascar and the mainland) in the
(Madagascar and the continent) in prehistoric times. This suggests that the Ampasimaiky rock paintings are of African origin and could be relatively dated to the same period as those on the continent (Thackeray, 1983: 22). Comparative studies of the images have also shown that more than 30 types of images in the Ampasimaiky rock paintings are closely related to different ancient scripts such as Libyco-Berber, Protosinaitic, Meroitic (cursive), Punic Iberian, Ancient Phoenician, Tifinagh, and Aegean, Cretan and Proto-Egyptian linear signs respectively (Figure 4). These similarities would also suggest that the SGA-type rock paintings of Ampasimaiky are ancient writing. them (Willcox, 1984: 210); the same has been applied to that of Professor Barry Fell, a specialist in Libyan-Berber writing. The same has been applied to that of Professor Barry Fell, a specialist in Libyan-Berber writing, who translated the glyphs of Driekopseiland in southern Africa as Libyan-Berber writing. But it has also been refuted on the grounds that the Bushmen could not write (Willcox 1984: 210); it is important to recall
It is important to recall in this connection that Libyco-Berber inscriptions inscribed on petroglyphs in the Canary Islands were not accepted as such, under the pretext that the Guanches (indigenous people of the Canary Islands) could not have acquired or adopted any writing system in prehistoric times (Cline 1953; Farrujia de la Rosa et al. 2010: 16-17; Willcox 1984: 48). This erroneous model persisted until reputable scholars succeeded in translating an inscription reading “save me from javelin annihilation” (Cline 1953: 273). Today, it is commonly accepted that the groups of aligned glyphs from the Canary Islands are Libyco-Berber inscriptions. Recently, Chami (2006: 84-5) insisted that African SGA rock paintings and engravings are ancient scripts. He has shown that some, such as those at Namoratunga in Kenya, are arranged. For the accounts of those that
are not arranged, he argued that at the time when the ancient scripts of Ethiopia, India, China, Egypt and Mesopotamia were still using ideograms, the characters too were not aligned (see also Casson 1966: 141). This can be amplified by the case of some Sahelian writing systems such as the ‘Nsibidi. This script has no ‘order in the arrangement’, but it is practically recognised as a writing system (Dayrell
1910; Macgregor 1909). Zahan (1950) identified the Dogon and Bambara pictographs as scripts based on the argument that
“writing consists essentially in the communication of thought by means of conventional signs, and the pictographs of both Dogon and Bambara serve to convey thoughts from one writer to another or from writer to lay public” (pp.137) « Une écriture consiste essentiellement de la communication de pensée par le moyen de signes conventionnels, et les pictographes de Dogon et de Bambara servent à transporter des pensées d’un auteur à l’autre ou d’un auteur au public”.
Therefore, it is also possible that without the “order in the arrangement”, the Ampasimaiky SGA-type rock paintings could have served to communicate thoughts and information. This would qualify them as a WRITING SYSTEM! Especially since some groups of symbols are horizontally and vertically aligned (Photo 4). Since “order in arrangement” has been established by some scholars as the sine qua non for a group of symbols/signs to qualify as an inscription (Willcox 1984: 210), one can reasonably assert, based on the above evidence, that the
AMPASIMAIKY RUPESTRATED ROCK SHELTER PAINTINGS ARE INSCRIPTIONS! This assertion is clearly supported by the vertically arranged group of symbols in panel 8 (see Photo 4b). Indeed, the repetition of the first sign in the third character (read downwards) is undoubtedly demonstrating the USE OF THE ALPHABET, and thus the
ALPHABET, so this is WRITING! All the more so as all five characters constituting the inscription are closely related to the Libyan-Berber alphabets (see
www.ancientscripts.com), the proto-Egyptian, Aegean and Cretan linear signs respectively
(Figure 5).
Figure 5 : Le groupe de signes verticalement arrangées d’Ampasimaiky comparé avec des caractères alphabétiques libyco-berbères et des signes linéaires proto-Egyptiens, égéens, et crétois (Tableau : Evans, 1897 :386).
Bien que des signes linéaires proto-Égyptiens, crétois et égéens aient été également verticalement écrits, je favoriserais l’origine libyco-berbère de l’inscription d’Ampasimaiky sous l’influence de la description de Walter Cline concernant les inscriptions libyco-berbères comme
suit).
“They [Libyco-Berber inscriptions] are usually very brief, consisting of personal names and a few phrases … The characters are simple and predominantly rectilinear. Variant alphabet may have existed. Most of inscriptions are to be read upward beginning with the left-hand column, though horizontal rows reading from right to left are not uncommon and some columns read downwards. The words are not separated. Though these signs sometimes occur in groups of four or five, they often stand singly…” (Cline 1953: 273)
All the underlined sentences qualify the vertically arranged inscription of Ampasimaiky as being of Libyco-berberber origin. Although this ancient script was written either horizontally or vertically, it should be remembered that the Libyco- Berber peoples used two different alphabets for each row (see http://www.ancientscripts.com). All five characters of the Ampasimaiky inscription are exclusively related to the vertical Libyco- Berber alphabet (Figure 6). Transliterating them to Latin alphabets – based on the transliteration of www.ancientscripts.com – one could read “ğ r ğ t f” (downwards) or “f t ğ r ğ” (upwards). The phonetic value ‘r’ for the presumed circle/square sign can be read as ‘3’ (= a), if the represented sign is perceived as a triangular symbol (see Fell 1975: 271). In the present state of the study, decoding the meaning of the inscription is not yet within the scope of the author. This task would rather be reserved for palaeographers and epigraphers.
Figure 7: The group of vertically arranged signs of Ampasimaiky transliterated in the Libyco-Berber alphabet
It is important to point out here that certain Libyco-Berber inscriptions brought back from the Canary Islands are not different from that of Ampasimaiky (see Figure 8). You can clearly see their resemblance.
Figure 8: The group of vertically arranged signs from Ampasimaiky compared with a Libyco-Berber inscription from the Canary Islands (Drawings by Farrujia de la Rosa et al. 2010: 29)
Discussion and Conclusion
Some of the images found at Ampasimaiky are similar to those of the rock art of continental Africa. However, the painting tradition of southwestern Madagascar is not
However, the painting tradition of southwestern Madagascar does not seem to originate from the continental region closest to Madagascar, i.e. Mozambique and the rest of the central Limpopo basin. In fact, Professor Benjamin Smith, an internationally recognised authority on the study of SGA rock paintings in sub-Saharan Africa, has taken the view that they are not typically affiliated with the Khoekhoen SGA rock paintings, nor with the geometric art of central southern Africa (pers. comm.). The exclusive representation of humped cattle associated with SGA-type symbols is particular to Ampasimaiky. Since no ancient representation of zebu (Bos indicus) has yet been found in the region between Sudan in the north and the Zambezi River in the south (Magnavita 2006: 62), it can be inferred that the tradition of rock paintings in Ampasimaiky originated in North Africa. Evidence of a bronze zebu figurine (of the type Bos indicus) inscribed with SGA signs dating back to 200 CE at Zeban Kutur, Eritrea (Photo 4) would confirm the early use of SGA signs among pastoral zebu-breeding peoples such as those depicted in the Ampasimaiky rock paintings
Photo 4: Bronze figure of zebu inscribed with SGA type signs in Zeban Kutur, Eritrea (Photo: Magnavita 2006:63)
The Libyan-Berber inscription discovered at Ampasimaiky also reinforces this proposal. The Libyan-Berber script emerged in North Africa around 500 BC and ceased to be used around 800 AD (Cline 1953).
around 800 AD (Cline 1953). This implies that the Libyan-Berber inscription of Ampasimaiky should have been written during the period between this bracket of dates (500 BC – AD 800). It is worth mentioning that Iamboulos, a Greek merchant who lived in the 3rd
century BC, reported about a vertical order of writing that was practised by the inhabitants of the ‘island of the sun’ (Allibert 2010: 111; Kobishschanow, 1965: 138; Oldfather, 1961 vol.2: 65-67; Winston 1976: 221). This island is believed to be geographically located adjacent to the east coast of Africa, south of the Equator. Although some of the geographical and ecological information in Iamboulos’ account is very much in line with that of the Malagasy west coast (for more details see the
Although some of the geographical and ecological information in Iamboulos’ account closely matches that of the Malagasy west coast (for more information see Allibert 2010), some scholars have consistently judged this account to be fictional, imaginary, and fanciful (Cary and Warmington 1963: 233-245; Winston 1976). One of the criteria used to censor this narrative is the use of vertical order writing by the islanders. Certainly, it was difficult for some scholars to accept the existence of a pre-Islamic script used in the adjacent islands of the east coast of Africa such as Madagascar (Janvier 1975: 19). However, with the discovery of the Libyan-Berber inscription at Ampasimaiky, this denial no longer holds. New evidence is uncovered to confirm that the “island of the sun” of Iamboulos could have been Madagascar. In any case, man was already present in the vicinity of Ampasimaiky when the Greek merchant passed through the island. Human activity is attested by subfossil lemur bones bearing cutting marks and radiocarbon-dated from the 5th to the 4th century BC at Taolambiby, in the upper Onilahy valley (see Perez et al., 2005). All the more so since ancient written documents (see Jones 1960: 235, 269, 315; Strabo 17, 1, 13 in Lacroix 1998: 95), archaeology (see Cary and Warmington 1963: 130; Chami 2007: 6; 2009; Wainwright 1940: 165), genetic study (see Ricaut et al. 2009: 5-6), and zooarchaeology (Chami 2006: 137; Grigson 1991: 124) all confirm the ancient contacts that existed during this period between the Onilahy Valley and the Onilahy Valley,
During this period, the Nile Valley and the Mediterranean world on the one hand, and the eastern coast of Africa and the adjacent islands on the other.
Références Allibert, C. (2008), “Austronesian Migration and the Establishment of the Malagasy Civilization: Contrasted Readings in Linguistics, Archaeology, Genetics and Cultural Anthropology”, Diogenes, SAGE, Volume 218, pp.7-16 Allibert, C. (2010), “L’île de Iamboulos est-elle Madagascar? », Pount, no.4, pp.103-119 Blench, R. (2007), “New Palaeozoogeographical Evidence for the Settlement of Madagascar”, Azania, Volume XLII, pp.69-82 Cary M. and Warmington, E. (1963), The Ancient Explorers, Middlesex, Penguin Books Casson, L. (1966), Ancient Egypt, Nederland, Time-life International Chami, F. A (2006), The Unity of African Ancient History, 3000 BC to AD 500, Dar es Salaam, E&D Limited Chami, F. A (2008), “The Great Lakes: a complexity of cultural wellsprings”, in Arnold, M. (eds.) Art in Eastern Africa, pp.47-64, Dar-es-Salaam, Mkuki na Nyota Chami F. A and Kwekason A. (2003), “Neolithic Pottery Traditions from the Islands, the Coast, and the Interior of East Africa”, African Archaeological Review, Volume 20, No. 2, pp.65-80 Chami, F. A, Tabibou, I. A and Abderhamane, B. (2009), “Preliminary Report of Archaeological Work Conducted on the Southern Ngazidja Island”, in Chami, F. (eds.), Zanzibar and the Swahili Coast from c30.000 years ago, pp.115-128, Dar es Salaam, E&D Vision Publishers Chaplin, T. (1974), “The Prehistoric Rock Art of the Lake Victoria Region”, Azania Volume 9, pp.1-50 Cline, W. (1953), “Berber Dialects and Berber Script”, Southwestern Journal of Anthropology, Vol.9, No.3, pp.268-276, New Mexico, New Mexico University Press Coulson D. and Campbell A. (2001), African Rock Art, Paintings and Engravings on Stone, New York, Harry N. Abrams Incorporated Dayrell, E. (1910), “Some “Nsibidi” Signs”, Man, Vol.10, pp.113-114. Dewar, R. E (1996), “The Archaeology of the Early Settlement of Madagascar”, in Reade J. (eds.) The Indian Ocean in Antiquity, pp.471-486 Dewar, R. E (1997), “Does It Matter That Madagascar Is an Island?” Human Ecology, Volume 25, No. 3, pp. 481-489 Donque, G. (1965), « Le contexte océanique des anciennes migrations : vents et courants dans l’océan indien », Taloha, Volume 1, pp.43-59. Eastwood, E. B and Smith, B. W (2005), “Fingerprints of the Khoekhoen: Geometric and handprinted Rock Art in the Central Limpopo Basin, Southern Africa”, Goodwin Series, RSA, South African Archaeological Society, Volume 9, pp.63-76 Evans, A. J (1897), “Further Discoveries of Cretan and Aegean Script: With Libyan and ProtoEgyptian Comparisons”, The Journal of Hellenic Studies, The Society for the Promotion of Hellenic Studies, Volume 17, pp.327-395 Farrujia de la Rosa, A., Pichler W., Rodrigue A., Marín S.G. 2010, The Lybico-Berber and Latino-Canarian Scripts and the Colonization of the Canary Islands, African Archaeological Review, vol.27, no.1, pp.13-41 Faublée, J. (1947), Récits Bara, Travaux et mémoires de l’Institut d’ethnologie XLXII, Paris, Université de Paris Fell, B. (1975), America BC: Ancient Settlers in the New World, New York, A Demeter Press Book Grandidier, A., Charles-Roux J., Delhorbe C., Froidevaux H. and Grandidier G. (1902), Collection des ouvrages anciens concernant Madagascar, Tome II, Paris, Comité de Madagascar Grigson, C. (1991), “An African Origin of African Cattle? Some Archaeological Evidence”, The African Archaeological Review, Volume 9, pp.119-144 Hall, S. and Smith, B. (2000), “Empowering Places: Rock Shelters and Ritual Control in FarmerForager Interactions in the Northern Province”, Goodwin Series, RSA, South African Archaeological Society, Volume 8, pp.30-46 Hébert, J. C (1967), « Essai d’interprétation de la stèle indéchiffrée d’Ambilobe », Taloha, Volume 2, pp.XLVII-LVII Hébert, J. C (1971), « Les tatouages de la côte est de Madagascar, d’après Chapelier (1794- 1806) », Taloha, Volume 4, pp.211-223 Huygues-Belrose, V. (1997). Les supports anciens de l’écriture à Madagascar, Etudes Océan Indien, vol.22, pp.7-19, INALCO Janvier, Y. (1975), « La géographie gréco-romaine a-t-elle connu Madagascar? Le point de la question », Omaly sy Anio, No.1-2, pp.11-41 Jones, H. (1960), The Geography of Strabo, Volume 1-2, London, William Heinemann Ltd. Juma, A. (2004), Unguja Ukuu on Zanzibar – An Archaeological Study of Early Urbanism, Sweden, Uppsala University Kessy, E. (2009a), “Analysis of Lithic Artefacts From Mwanampambe and Kuumbi Caves, Zanzibar”, in Chami F. (eds.), Zanzibar and the Swahili Coast from c30,000 years ago, Dar es Salaam, E&D Vision Publishers, pp.131-143 Kessy, E. (2009b), “Stone Artefacts from Male”, in Chami F. (eds.), Zanzibar and the Swahili Coast from c30,000 years ago, Dar es Salaam, E&D Vision Publishers, pp.184-88 Kobishchanov, Y.M. (1965), “On the Problem of Sea Voyages of Ancient Africans in the Indian Ocean”, The Journal of African History, Vol.6, No.2, pp.137-141, Cambridge, Cambridge University Press. Lacroix, W. F (1998), Africa in Antiquity, Nijmegen, Catholic University
Tanambelo V.R. RASOLONDRAINY,
(Centre de Documentation et de Recherche sur l’Art et la Tradition Orale de Madagascar)
Abstract:
This article presents the result of an archaeological study on the first prehistoric rock paintings site found in Madagascar and the rest of islands of Western Indian Ocean south of the equator.
Based on systematic methods and techniques, the study provides groundbreaking archaeological evidences that contribute largely to the understanding of the prehistory of Madagascar which was not archaeologically known thitherto to have prehistoric rock paintings, nor human settlement prior to 8th century AD. A Libyco-Berber inscription was discovered on the rock paintings of Ampasimaiky Rock Shelter, in Southwestern Madagascar. Based on this writing -supported by ancient historical records, circumstantial archaeological and paleontological evidences, linguistic data, and to some extent DNA analysis results -, the present study provides minimum age falling in the bracket time around 500 BC to roughly 8th century AD for the rock paintings of Ampasimaiky rock shelter. Furthermore, the study traces back early contacts between Madagascar and Northern Africa during prehistoric times.
Introduction
With the kind assistance of Dr Chantal Radimilahy, National Coordinator of the Network of African Archaeologists in Madagascar, I researched the ethnographic exhibition on the website of the Museum of Ethnography in Geneva (MEG). The theme was the peoples and cultures of the southwestern region of Madagascar, and was based on an exhibition of photographs taken by the French ethnographer Jacques Faublée in 1939. The pictures included rock paintings found in caves and rock shelters in the Isalo region. After discussions with my teachers at the Archaeology Unit of the University of Dar es Salaam, Professors Felix Chami, Fidelis Masao, and Audax Mabulla, I further investigated the significance of these paintings for the prehistoric period of Madagascar. This study is taken from a Master of Arts in Archaeology entitled
“Archaeologial study of the prehistoric rock paintings of Ampasimaiky rock shelter, in the upper Onilahy, Isalo region, Southwestern Madagascar”.
————–
Africa has many archaeological sites with prehistoric rock paintings (see Coulson and Campbell 2001; Willcox 1984). However, no such sites have yet been scientifically
scientifically reported from Madagascar, which is only 400 kilometres from southern Africa where they are abundant. All rock art previously found in Madagascar has been judged to have been produced in the recent past (Faublée 1947: 164; Hébert 1971: 218). This judgement is based on the theory that Africans were incapable of having navigated the high seas during prehistoric times (Donque 1965: 43; Grandidier et al. 1902; Southall 1975: 194; Vérin 1975, 1981). It was therefore out of the question to think that they had the possibility of transferring the practice of rock painting to the adjacent islands of Africa such as Madagascar. However, recent archaeological findings have demonstrated that the east coast of Africa and adjacent islands were already in contact since the prehistoric period (see Chami and Kwekason 2003; Chami et al. 2009; Juma 2004; Kessy 2009a,
2009b). These prompt a problem that led me to revise the antiquity of the rock paintings in Madagascar.
On the other hand, the name of the Arabic-Malagasy script – Sorabe meaning “great script” – indicates that there was an ancient script in Madagascar that was distinguished from Arabic (Hébert,
1967:XLVII; Huygues-Belrose 1997: 8). Although Hébert (1967:XLVIII) presumed that such a hypothetical script would have to be of Indian origin and would be discovered from
Although Hébert (1967:XLVIII) assumed that such a hypothetical script would have to be of Indian origin and would be discovered from stele and parietal inscriptions, no such inscription has so far been found in Madagascar.
My problem was therefore to find out whether this would appear on the rock paintings reported by the website. What provoked this idea the most was that Chami (2006:84-85; 2008) argued that the African Geometric and Amorphous Schematic (GAS) rock paintings and engravings are ancient writing.
Another problem is that radiocarbon 14 dating placed the earliest human settlement of Madagascar around the 8th century AD (Allibert 2008: 7; Dewar 1996: 474).
Although earlier dates are still being considered, many researchers have assumed that settlement of Madagascar would not go beyond the year W AD (for more information see Dewar 1997). This assumption has made research on Malagasy material culture going back to the Stone Age more than difficult. Any findings of material cultures dating from prehistoric times have been discarded, overlooked or misinterpreted. However, subfossil animal bone remains bearing traces of butchery found in paleontological sites in the southwestern region indicate the presence of human activity
(For more information, see Blench 2007; Perez et al. 2005). Despite this evidence, an archaeological site has not yet been
archaeological site with similar dates has not yet been found on the entire island.
With the discovery of a rock shelter with rock paintings, discovering new information about the prehistory of Madagascar before the 8th century proved to be feasible. All the more so as some of the paintings and engravings are
Especially since some of the SGA-type rock paintings and engravings found on the African continent – similar to those found in the photos on the MEG website – have been radiocarbon dated from the 1st millennium BC to the early 1st millennium AD (Chami, 2008: 54; Hall and Smith 2000; Lynche and Donahue 1980; Willcox 1984: 26, 83.
Fieldwork and results
Despite various challenges and constraints – mainly caused by the lack of archaeological reconnaissance prior to the fieldwork, insecurity in the study area, and the sacredness of the site to be studied – we had only somehow managed to conduct survey, archaeological excavation, and documentation of the paintings in the Ampasimaiky rock shelter. It is only the result of the latter that will be presented in this article.
The rock painting site is a rock shelter located at the foot of a hill immediately next to a small stream called Ampasimaiky (Figure 2), from which the name of the site derives. The presence of this water source may have influenced the painter’s choice of location.
The shelter itself is 26.8m long (south-north axis) and 5m wide (inwards).
Its ceiling begins to slope inwards from 4m high at its entrance to approximately 1.5m high at its irregular wall. The deepest part of the rock shelter – a chamber – is currently used by the royal family “Bara Zafimanely” as a burial place. Dung and ox hoof prints found on the ground indicate that cattle frequent the place. This occurrence should have disturbed the soil.
However, the areas around the site are dominated by rocky hills covered with grass. Only in the deep canyon near the confluence of Ampasimaiky and the Sakamarekely River is a patch of thick vegetation up to 10 m high observed along the stream. This situation would make agricultural activity almost impossible around the site, unless one had to move 1.5 km to the bank, where rice cultivation takes place today.
An archaeological survey of the southern Isalo region (Figure 1) was carried out and a rock painting site was discovered. Professor Chami participated in the fieldwork, advising on the approach to be followed in the treatment of these previously unstudied cave cultures in Madagascar.
Madagascar. History students from the University of Toliara, Mex Tafitasoa, Nadya
Ravololonirina, and Norbert Antilahy, and some villagers (Photo 1) participated in the work. The Network of African Archaeologists in Madagascar, under the direction of Dr Chantal Radimilahy and Prof Barthélemy Manjakahery, provided the logistical equipment used. The various signatures The various signatures of national and local administrative approval, the security provided by the Gendarmerie Brigade of Benenitsy, and above all the kindness and openness of King Zafimanely (Photo 1) – who allowed us access to the (burial) site – contributed enormously to the accomplishment of the expedition.
Madagascar. History students from the University of Toliara, Mex Tafitasoa, Nadya
Ravololonirina, and Norbert Antilahy, and some villagers (Photo 1) participated in the work. The Network of African Archaeologists in Madagascar, under the direction of Dr Chantal Radimilahy and Prof Barthélemy Manjakahery, provided the logistical equipment used. The various signatures The various signatures of national and local administrative approval, the security provided by the Gendarmerie Brigade of Benenitsy, and above all the kindness and openness of King Zafimanely (Photo 1) – who allowed us access to the (burial) site – contributed enormously to the accomplishment of the expedition.
The rock shelter at Ampasimaiky faces east. This arrangement makes the painting panels exposed to the sun in the morning and in the shade beyond noon. This arrangement suggests that painting activities are conducted either early in the morning while the heat of the sun is still moderate, or beyond noon when shade is available below the shelter.
This is based on the fact that the height of the panels does not exceed 130 cm, suggesting that the paintings were intended for people sitting on the floor of the undercroft shelter
Photo 2). Minimal, therefore, is the likelihood of the painting activity – involving seated people – having been conducted during the time when the sun is heating the shelter floor.
However, the view outside the site is limited. The opposing hills – on the other bank of the Ampasimaiky stream – prevent a person standing under the shelter from having a good view of the surroundings (Photo 3). It is much more likely that the view outwards – which implies a defensive strategy (Smith 1995) – was not among the painter’s concerns when he/she chose the site. This demonstrates that the layout of the Ampasimaiky rock shelter favours invaders from unexpectedly attacking its residents. The place would therefore not be suitable for permanent habitation.
Photo 3
Especially since a person standing on the hills of the other bank would be able to look directly at the other one positioned under the rock shelter, unless he hides in the deeper part of the shelter. Since the wall of this “chamber”, according to the local population, is not painted, it is plausible that the activities related to the paintings in the Ampasimaiky rock shelter are not clandestine. However, the paintings were not intended to be seen by people passing close to the site. They are not seen from a distance. Not only are the panels placed below 130 cm in height, but the slope of the ground before the access
to the shelter also prevents anyone from seeing them until they get close to the entrance. Nevertheless, the use of darker coloured paints (red and black) suggests the painter’s intention to contrast the paints he/she was using against the pale colour of the sandstone wall. This implies his concern for the visibility of his paintings.
The walls of the rock shelter have red, reddish orange, burgundy red, black, and white paintings in monochrome, bichrome and polychrome patterns. The images
are painted with the fingertip. They depict zoomorphic (mainly zebu) and anthropomorphic (semi-naturalist and stylised) figures, as well as schematic-geometric-amorphic symbols (Figure 3). The latter dominate and are represented by quadrangular, circular and elliptical signs, dot/dash lines, and alphabet-like signs. In total, the Ampasimaiky rock paintings have 166 recognisable images.
Analysis and Interpretation
Based on comparative studies of the images, it has been found that the images of the Ampasimaiky rock paintings are not different from the images depicted on the prehistoric SGA rock art of the African continent (see Chami, 2008; Chaplin, 1974; Coulson and Campbell 2001;
Eastwood and Smith, 2005; Lynch and Donahue, 1980; Mabulla, 2005; Smith, 1995; Teka 2008; Willcox 1984). The similarities apparently suggest links between the two regions (Madagascar and the mainland) in the
(Madagascar and the continent) in prehistoric times. This suggests that the Ampasimaiky rock paintings are of African origin and could be relatively dated to the same period as those on the continent (Thackeray, 1983: 22). Comparative studies of the images have also shown that more than 30 types of images in the Ampasimaiky rock paintings are closely related to different ancient scripts such as Libyco-Berber, Protosinaitic, Meroitic (cursive), Punic Iberian, Ancient Phoenician, Tifinagh, and Aegean, Cretan and Proto-Egyptian linear signs respectively (Figure 4). These similarities would also suggest that the SGA-type rock paintings of Ampasimaiky are ancient writing.
figure 4
Analysis and Interpretation
Based on comparative studies of the images, it has been found that the images of the Ampasimaiky rock paintings are not different from the images depicted on the prehistoric SGA rock art of the African continent (see Chami, 2008; Chaplin, 1974; Coulson and Campbell 2001;
Eastwood and Smith, 2005; Lynch and Donahue, 1980; Mabulla, 2005; Smith, 1995; Teka 2008; Willcox 1984). The similarities apparently suggest links between the two regions (Madagascar and the mainland) in the
(Madagascar and the continent) in prehistoric times. This suggests that the Ampasimaiky rock paintings are of African origin and could be relatively dated to the same period as those on the continent (Thackeray, 1983: 22). Comparative studies of the images have also shown that more than 30 types of images in the Ampasimaiky rock paintings are closely related to different ancient scripts such as Libyco-Berber, Protosinaitic, Meroitic (cursive), Punic Iberian, Ancient Phoenician, Tifinagh, and Aegean, Cretan and Proto-Egyptian linear signs respectively (Figure 4). These similarities would also suggest that the SGA-type rock paintings of Ampasimaiky are ancient writing. them (Willcox, 1984: 210); the same has been applied to that of Professor Barry Fell, a specialist in Libyan-Berber writing. The same has been applied to that of Professor Barry Fell, a specialist in Libyan-Berber writing, who translated the glyphs of Driekopseiland in southern Africa as Libyan-Berber writing. But it has also been refuted on the grounds that the Bushmen could not write (Willcox 1984: 210); it is important to recall
It is important to recall in this connection that Libyco-Berber inscriptions inscribed on petroglyphs in the Canary Islands were not accepted as such, under the pretext that the Guanches (indigenous people of the Canary Islands) could not have acquired or adopted any writing system in prehistoric times (Cline 1953; Farrujia de la Rosa et al. 2010: 16-17; Willcox 1984: 48). This erroneous model persisted until reputable scholars succeeded in translating an inscription reading “save me from javelin annihilation” (Cline 1953: 273). Today, it is commonly accepted that the groups of aligned glyphs from the Canary Islands are Libyco-Berber inscriptions. Recently, Chami (2006: 84-5) insisted that African SGA rock paintings and engravings are ancient scripts. He has shown that some, such as those at Namoratunga in Kenya, are arranged. For the accounts of those that
are not arranged, he argued that at the time when the ancient scripts of Ethiopia, India, China, Egypt and Mesopotamia were still using ideograms, the characters too were not aligned (see also Casson 1966: 141). This can be amplified by the case of some Sahelian writing systems such as the ‘Nsibidi. This script has no ‘order in the arrangement’, but it is practically recognised as a writing system (Dayrell
1910; Macgregor 1909). Zahan (1950) identified the Dogon and Bambara pictographs as scripts based on the argument that
“writing consists essentially in the communication of thought by means of conventional signs, and the pictographs of both Dogon and Bambara serve to convey thoughts from one writer to another or from writer to lay public” (pp.137) « Une écriture consiste essentiellement de la communication de pensée par le moyen de signes conventionnels, et les pictographes de Dogon et de Bambara servent à transporter des pensées d’un auteur à l’autre ou d’un auteur au public”.
Therefore, it is also possible that without the “order in the arrangement”, the Ampasimaiky SGA-type rock paintings could have served to communicate thoughts and information. This would qualify them as a WRITING SYSTEM! Especially since some groups of symbols are horizontally and vertically aligned (Photo 4). Since “order in arrangement” has been established by some scholars as the sine qua non for a group of symbols/signs to qualify as an inscription (Willcox 1984: 210), one can reasonably assert, based on the above evidence, that the
AMPASIMAIKY RUPESTRATED ROCK SHELTER PAINTINGS ARE INSCRIPTIONS! This assertion is clearly supported by the vertically arranged group of symbols in panel 8 (see Photo 4b). Indeed, the repetition of the first sign in the third character (read downwards) is undoubtedly demonstrating the USE OF THE ALPHABET, and thus the
ALPHABET, so this is WRITING! All the more so as all five characters constituting the inscription are closely related to the Libyan-Berber alphabets (see
www.ancientscripts.com), the proto-Egyptian, Aegean and Cretan linear signs respectively
(Figure 5).
Figure 5 : Le groupe de signes verticalement arrangées d’Ampasimaiky comparé avec des caractères alphabétiques libyco-berbères et des signes linéaires proto-Egyptiens, égéens, et crétois (Tableau : Evans, 1897 :386).
Bien que des signes linéaires proto-Égyptiens, crétois et égéens aient été également verticalement écrits, je favoriserais l’origine libyco-berbère de l’inscription d’Ampasimaiky sous l’influence de la description de Walter Cline concernant les inscriptions libyco-berbères comme
suit).
“They [Libyco-Berber inscriptions] are usually very brief, consisting of personal names and a few phrases … The characters are simple and predominantly rectilinear. Variant alphabet may have existed. Most of inscriptions are to be read upward beginning with the left-hand column, though horizontal rows reading from right to left are not uncommon and some columns read downwards. The words are not separated. Though these signs sometimes occur in groups of four or five, they often stand singly…” (Cline 1953: 273)
All the underlined sentences qualify the vertically arranged inscription of Ampasimaiky as being of Libyco-berberber origin. Although this ancient script was written either horizontally or vertically, it should be remembered that the Libyco- Berber peoples used two different alphabets for each row (see http://www.ancientscripts.com). All five characters of the Ampasimaiky inscription are exclusively related to the vertical Libyco- Berber alphabet (Figure 6). Transliterating them to Latin alphabets – based on the transliteration of www.ancientscripts.com – one could read “ğ r ğ t f” (downwards) or “f t ğ r ğ” (upwards). The phonetic value ‘r’ for the presumed circle/square sign can be read as ‘3’ (= a), if the represented sign is perceived as a triangular symbol (see Fell 1975: 271). In the present state of the study, decoding the meaning of the inscription is not yet within the scope of the author. This task would rather be reserved for palaeographers and epigraphers.
Figure 7: The group of vertically arranged signs of Ampasimaiky transliterated in the Libyco-Berber alphabet
It is important to point out here that certain Libyco-Berber inscriptions brought back from the Canary Islands are not different from that of Ampasimaiky (see Figure 8). You can clearly see their resemblance.
Figure 8: The group of vertically arranged signs from Ampasimaiky compared with a Libyco-Berber inscription from the Canary Islands (Drawings by Farrujia de la Rosa et al. 2010: 29)
Discussion and Conclusion
Some of the images found at Ampasimaiky are similar to those of the rock art of continental Africa. However, the painting tradition of southwestern Madagascar is not
However, the painting tradition of southwestern Madagascar does not seem to originate from the continental region closest to Madagascar, i.e. Mozambique and the rest of the central Limpopo basin. In fact, Professor Benjamin Smith, an internationally recognised authority on the study of SGA rock paintings in sub-Saharan Africa, has taken the view that they are not typically affiliated with the Khoekhoen SGA rock paintings, nor with the geometric art of central southern Africa (pers. comm.). The exclusive representation of humped cattle associated with SGA-type symbols is particular to Ampasimaiky. Since no ancient representation of zebu (Bos indicus) has yet been found in the region between Sudan in the north and the Zambezi River in the south (Magnavita 2006: 62), it can be inferred that the tradition of rock paintings in Ampasimaiky originated in North Africa. Evidence of a bronze zebu figurine (of the type Bos indicus) inscribed with SGA signs dating back to 200 CE at Zeban Kutur, Eritrea (Photo 4) would confirm the early use of SGA signs among pastoral zebu-breeding peoples such as those depicted in the Ampasimaiky rock paintings
Photo 4: Bronze figure of zebu inscribed with SGA type signs in Zeban Kutur, Eritrea (Photo: Magnavita 2006:63)
The Libyan-Berber inscription discovered at Ampasimaiky also reinforces this proposal. The Libyan-Berber script emerged in North Africa around 500 BC and ceased to be used around 800 AD (Cline 1953).
around 800 AD (Cline 1953). This implies that the Libyan-Berber inscription of Ampasimaiky should have been written during the period between this bracket of dates (500 BC – AD 800). It is worth mentioning that Iamboulos, a Greek merchant who lived in the 3rd
century BC, reported about a vertical order of writing that was practised by the inhabitants of the ‘island of the sun’ (Allibert 2010: 111; Kobishschanow, 1965: 138; Oldfather, 1961 vol.2: 65-67; Winston 1976: 221). This island is believed to be geographically located adjacent to the east coast of Africa, south of the Equator. Although some of the geographical and ecological information in Iamboulos’ account is very much in line with that of the Malagasy west coast (for more details see the
Although some of the geographical and ecological information in Iamboulos’ account closely matches that of the Malagasy west coast (for more information see Allibert 2010), some scholars have consistently judged this account to be fictional, imaginary, and fanciful (Cary and Warmington 1963: 233-245; Winston 1976). One of the criteria used to censor this narrative is the use of vertical order writing by the islanders. Certainly, it was difficult for some scholars to accept the existence of a pre-Islamic script used in the adjacent islands of the east coast of Africa such as Madagascar (Janvier 1975: 19). However, with the discovery of the Libyan-Berber inscription at Ampasimaiky, this denial no longer holds. New evidence is uncovered to confirm that the “island of the sun” of Iamboulos could have been Madagascar. In any case, man was already present in the vicinity of Ampasimaiky when the Greek merchant passed through the island. Human activity is attested by subfossil lemur bones bearing cutting marks and radiocarbon-dated from the 5th to the 4th century BC at Taolambiby, in the upper Onilahy valley (see Perez et al., 2005). All the more so since ancient written documents (see Jones 1960: 235, 269, 315; Strabo 17, 1, 13 in Lacroix 1998: 95), archaeology (see Cary and Warmington 1963: 130; Chami 2007: 6; 2009; Wainwright 1940: 165), genetic study (see Ricaut et al. 2009: 5-6), and zooarchaeology (Chami 2006: 137; Grigson 1991: 124) all confirm the ancient contacts that existed during this period between the Onilahy Valley and the Onilahy Valley,
During this period, the Nile Valley and the Mediterranean world on the one hand, and the eastern coast of Africa and the adjacent islands on the other.
Références Allibert, C. (2008), “Austronesian Migration and the Establishment of the Malagasy Civilization: Contrasted Readings in Linguistics, Archaeology, Genetics and Cultural Anthropology”, Diogenes, SAGE, Volume 218, pp.7-16 Allibert, C. (2010), “L’île de Iamboulos est-elle Madagascar? », Pount, no.4, pp.103-119 Blench, R. (2007), “New Palaeozoogeographical Evidence for the Settlement of Madagascar”, Azania, Volume XLII, pp.69-82 Cary M. and Warmington, E. (1963), The Ancient Explorers, Middlesex, Penguin Books Casson, L. (1966), Ancient Egypt, Nederland, Time-life International Chami, F. A (2006), The Unity of African Ancient History, 3000 BC to AD 500, Dar es Salaam, E&D Limited Chami, F. A (2008), “The Great Lakes: a complexity of cultural wellsprings”, in Arnold, M. (eds.) Art in Eastern Africa, pp.47-64, Dar-es-Salaam, Mkuki na Nyota Chami F. A and Kwekason A. (2003), “Neolithic Pottery Traditions from the Islands, the Coast, and the Interior of East Africa”, African Archaeological Review, Volume 20, No. 2, pp.65-80 Chami, F. A, Tabibou, I. A and Abderhamane, B. (2009), “Preliminary Report of Archaeological Work Conducted on the Southern Ngazidja Island”, in Chami, F. (eds.), Zanzibar and the Swahili Coast from c30.000 years ago, pp.115-128, Dar es Salaam, E&D Vision Publishers Chaplin, T. (1974), “The Prehistoric Rock Art of the Lake Victoria Region”, Azania Volume 9, pp.1-50 Cline, W. (1953), “Berber Dialects and Berber Script”, Southwestern Journal of Anthropology, Vol.9, No.3, pp.268-276, New Mexico, New Mexico University Press Coulson D. and Campbell A. (2001), African Rock Art, Paintings and Engravings on Stone, New York, Harry N. Abrams Incorporated Dayrell, E. (1910), “Some “Nsibidi” Signs”, Man, Vol.10, pp.113-114. Dewar, R. E (1996), “The Archaeology of the Early Settlement of Madagascar”, in Reade J. (eds.) The Indian Ocean in Antiquity, pp.471-486 Dewar, R. E (1997), “Does It Matter That Madagascar Is an Island?” Human Ecology, Volume 25, No. 3, pp. 481-489 Donque, G. (1965), « Le contexte océanique des anciennes migrations : vents et courants dans l’océan indien », Taloha, Volume 1, pp.43-59. Eastwood, E. B and Smith, B. W (2005), “Fingerprints of the Khoekhoen: Geometric and handprinted Rock Art in the Central Limpopo Basin, Southern Africa”, Goodwin Series, RSA, South African Archaeological Society, Volume 9, pp.63-76 Evans, A. J (1897), “Further Discoveries of Cretan and Aegean Script: With Libyan and ProtoEgyptian Comparisons”, The Journal of Hellenic Studies, The Society for the Promotion of Hellenic Studies, Volume 17, pp.327-395 Farrujia de la Rosa, A., Pichler W., Rodrigue A., Marín S.G. 2010, The Lybico-Berber and Latino-Canarian Scripts and the Colonization of the Canary Islands, African Archaeological Review, vol.27, no.1, pp.13-41 Faublée, J. (1947), Récits Bara, Travaux et mémoires de l’Institut d’ethnologie XLXII, Paris, Université de Paris Fell, B. (1975), America BC: Ancient Settlers in the New World, New York, A Demeter Press Book Grandidier, A., Charles-Roux J., Delhorbe C., Froidevaux H. and Grandidier G. (1902), Collection des ouvrages anciens concernant Madagascar, Tome II, Paris, Comité de Madagascar Grigson, C. (1991), “An African Origin of African Cattle? Some Archaeological Evidence”, The African Archaeological Review, Volume 9, pp.119-144 Hall, S. and Smith, B. (2000), “Empowering Places: Rock Shelters and Ritual Control in FarmerForager Interactions in the Northern Province”, Goodwin Series, RSA, South African Archaeological Society, Volume 8, pp.30-46 Hébert, J. C (1967), « Essai d’interprétation de la stèle indéchiffrée d’Ambilobe », Taloha, Volume 2, pp.XLVII-LVII Hébert, J. C (1971), « Les tatouages de la côte est de Madagascar, d’après Chapelier (1794- 1806) », Taloha, Volume 4, pp.211-223 Huygues-Belrose, V. (1997). Les supports anciens de l’écriture à Madagascar, Etudes Océan Indien, vol.22, pp.7-19, INALCO Janvier, Y. (1975), « La géographie gréco-romaine a-t-elle connu Madagascar? Le point de la question », Omaly sy Anio, No.1-2, pp.11-41 Jones, H. (1960), The Geography of Strabo, Volume 1-2, London, William Heinemann Ltd. Juma, A. (2004), Unguja Ukuu on Zanzibar – An Archaeological Study of Early Urbanism, Sweden, Uppsala University Kessy, E. (2009a), “Analysis of Lithic Artefacts From Mwanampambe and Kuumbi Caves, Zanzibar”, in Chami F. (eds.), Zanzibar and the Swahili Coast from c30,000 years ago, Dar es Salaam, E&D Vision Publishers, pp.131-143 Kessy, E. (2009b), “Stone Artefacts from Male”, in Chami F. (eds.), Zanzibar and the Swahili Coast from c30,000 years ago, Dar es Salaam, E&D Vision Publishers, pp.184-88 Kobishchanov, Y.M. (1965), “On the Problem of Sea Voyages of Ancient Africans in the Indian Ocean”, The Journal of African History, Vol.6, No.2, pp.137-141, Cambridge, Cambridge University Press. Lacroix, W. F (1998), Africa in Antiquity, Nijmegen, Catholic University
Tanambelo V.R. RASOLONDRAINY,
(Centre de Documentation et de Recherche sur l’Art et la Tradition Orale de Madagascar)
Abstract:
This article presents the result of an archaeological study on the first prehistoric rock paintings site found in Madagascar and the rest of islands of Western Indian Ocean south of the equator.
Based on systematic methods and techniques, the study provides groundbreaking archaeological evidences that contribute largely to the understanding of the prehistory of Madagascar which was not archaeologically known thitherto to have prehistoric rock paintings, nor human settlement prior to 8th century AD. A Libyco-Berber inscription was discovered on the rock paintings of Ampasimaiky Rock Shelter, in Southwestern Madagascar. Based on this writing -supported by ancient historical records, circumstantial archaeological and paleontological evidences, linguistic data, and to some extent DNA analysis results -, the present study provides minimum age falling in the bracket time around 500 BC to roughly 8th century AD for the rock paintings of Ampasimaiky rock shelter. Furthermore, the study traces back early contacts between Madagascar and Northern Africa during prehistoric times.
Introduction
With the kind assistance of Dr Chantal Radimilahy, National Coordinator of the Network of African Archaeologists in Madagascar, I researched the ethnographic exhibition on the website of the Museum of Ethnography in Geneva (MEG). The theme was the peoples and cultures of the southwestern region of Madagascar, and was based on an exhibition of photographs taken by the French ethnographer Jacques Faublée in 1939. The pictures included rock paintings found in caves and rock shelters in the Isalo region. After discussions with my teachers at the Archaeology Unit of the University of Dar es Salaam, Professors Felix Chami, Fidelis Masao, and Audax Mabulla, I further investigated the significance of these paintings for the prehistoric period of Madagascar. This study is taken from a Master of Arts in Archaeology entitled
“Archaeologial study of the prehistoric rock paintings of Ampasimaiky rock shelter, in the upper Onilahy, Isalo region, Southwestern Madagascar”.
————–
Africa has many archaeological sites with prehistoric rock paintings (see Coulson and Campbell 2001; Willcox 1984). However, no such sites have yet been scientifically
scientifically reported from Madagascar, which is only 400 kilometres from southern Africa where they are abundant. All rock art previously found in Madagascar has been judged to have been produced in the recent past (Faublée 1947: 164; Hébert 1971: 218). This judgement is based on the theory that Africans were incapable of having navigated the high seas during prehistoric times (Donque 1965: 43; Grandidier et al. 1902; Southall 1975: 194; Vérin 1975, 1981). It was therefore out of the question to think that they had the possibility of transferring the practice of rock painting to the adjacent islands of Africa such as Madagascar. However, recent archaeological findings have demonstrated that the east coast of Africa and adjacent islands were already in contact since the prehistoric period (see Chami and Kwekason 2003; Chami et al. 2009; Juma 2004; Kessy 2009a,
2009b). These prompt a problem that led me to revise the antiquity of the rock paintings in Madagascar.
On the other hand, the name of the Arabic-Malagasy script – Sorabe meaning “great script” – indicates that there was an ancient script in Madagascar that was distinguished from Arabic (Hébert,
1967:XLVII; Huygues-Belrose 1997: 8). Although Hébert (1967:XLVIII) presumed that such a hypothetical script would have to be of Indian origin and would be discovered from
Although Hébert (1967:XLVIII) assumed that such a hypothetical script would have to be of Indian origin and would be discovered from stele and parietal inscriptions, no such inscription has so far been found in Madagascar.
My problem was therefore to find out whether this would appear on the rock paintings reported by the website. What provoked this idea the most was that Chami (2006:84-85; 2008) argued that the African Geometric and Amorphous Schematic (GAS) rock paintings and engravings are ancient writing.
Another problem is that radiocarbon 14 dating placed the earliest human settlement of Madagascar around the 8th century AD (Allibert 2008: 7; Dewar 1996: 474).
Although earlier dates are still being considered, many researchers have assumed that settlement of Madagascar would not go beyond the year W AD (for more information see Dewar 1997). This assumption has made research on Malagasy material culture going back to the Stone Age more than difficult. Any findings of material cultures dating from prehistoric times have been discarded, overlooked or misinterpreted. However, subfossil animal bone remains bearing traces of butchery found in paleontological sites in the southwestern region indicate the presence of human activity
(For more information, see Blench 2007; Perez et al. 2005). Despite this evidence, an archaeological site has not yet been
archaeological site with similar dates has not yet been found on the entire island.
With the discovery of a rock shelter with rock paintings, discovering new information about the prehistory of Madagascar before the 8th century proved to be feasible. All the more so as some of the paintings and engravings are
Especially since some of the SGA-type rock paintings and engravings found on the African continent – similar to those found in the photos on the MEG website – have been radiocarbon dated from the 1st millennium BC to the early 1st millennium AD (Chami, 2008: 54; Hall and Smith 2000; Lynche and Donahue 1980; Willcox 1984: 26, 83.
Fieldwork and results
Despite various challenges and constraints – mainly caused by the lack of archaeological reconnaissance prior to the fieldwork, insecurity in the study area, and the sacredness of the site to be studied – we had only somehow managed to conduct survey, archaeological excavation, and documentation of the paintings in the Ampasimaiky rock shelter. It is only the result of the latter that will be presented in this article.
The rock painting site is a rock shelter located at the foot of a hill immediately next to a small stream called Ampasimaiky (Figure 2), from which the name of the site derives. The presence of this water source may have influenced the painter’s choice of location.
The shelter itself is 26.8m long (south-north axis) and 5m wide (inwards).
Its ceiling begins to slope inwards from 4m high at its entrance to approximately 1.5m high at its irregular wall. The deepest part of the rock shelter – a chamber – is currently used by the royal family “Bara Zafimanely” as a burial place. Dung and ox hoof prints found on the ground indicate that cattle frequent the place. This occurrence should have disturbed the soil.
However, the areas around the site are dominated by rocky hills covered with grass. Only in the deep canyon near the confluence of Ampasimaiky and the Sakamarekely River is a patch of thick vegetation up to 10 m high observed along the stream. This situation would make agricultural activity almost impossible around the site, unless one had to move 1.5 km to the bank, where rice cultivation takes place today.
An archaeological survey of the southern Isalo region (Figure 1) was carried out and a rock painting site was discovered. Professor Chami participated in the fieldwork, advising on the approach to be followed in the treatment of these previously unstudied cave cultures in Madagascar.
Madagascar. History students from the University of Toliara, Mex Tafitasoa, Nadya
Ravololonirina, and Norbert Antilahy, and some villagers (Photo 1) participated in the work. The Network of African Archaeologists in Madagascar, under the direction of Dr Chantal Radimilahy and Prof Barthélemy Manjakahery, provided the logistical equipment used. The various signatures The various signatures of national and local administrative approval, the security provided by the Gendarmerie Brigade of Benenitsy, and above all the kindness and openness of King Zafimanely (Photo 1) – who allowed us access to the (burial) site – contributed enormously to the accomplishment of the expedition.
Madagascar. History students from the University of Toliara, Mex Tafitasoa, Nadya
Ravololonirina, and Norbert Antilahy, and some villagers (Photo 1) participated in the work. The Network of African Archaeologists in Madagascar, under the direction of Dr Chantal Radimilahy and Prof Barthélemy Manjakahery, provided the logistical equipment used. The various signatures The various signatures of national and local administrative approval, the security provided by the Gendarmerie Brigade of Benenitsy, and above all the kindness and openness of King Zafimanely (Photo 1) – who allowed us access to the (burial) site – contributed enormously to the accomplishment of the expedition.
The rock shelter at Ampasimaiky faces east. This arrangement makes the painting panels exposed to the sun in the morning and in the shade beyond noon. This arrangement suggests that painting activities are conducted either early in the morning while the heat of the sun is still moderate, or beyond noon when shade is available below the shelter.
This is based on the fact that the height of the panels does not exceed 130 cm, suggesting that the paintings were intended for people sitting on the floor of the undercroft shelter
Photo 2). Minimal, therefore, is the likelihood of the painting activity – involving seated people – having been conducted during the time when the sun is heating the shelter floor.
However, the view outside the site is limited. The opposing hills – on the other bank of the Ampasimaiky stream – prevent a person standing under the shelter from having a good view of the surroundings (Photo 3). It is much more likely that the view outwards – which implies a defensive strategy (Smith 1995) – was not among the painter’s concerns when he/she chose the site. This demonstrates that the layout of the Ampasimaiky rock shelter favours invaders from unexpectedly attacking its residents. The place would therefore not be suitable for permanent habitation.
Photo 3
Especially since a person standing on the hills of the other bank would be able to look directly at the other one positioned under the rock shelter, unless he hides in the deeper part of the shelter. Since the wall of this “chamber”, according to the local population, is not painted, it is plausible that the activities related to the paintings in the Ampasimaiky rock shelter are not clandestine. However, the paintings were not intended to be seen by people passing close to the site. They are not seen from a distance. Not only are the panels placed below 130 cm in height, but the slope of the ground before the access
to the shelter also prevents anyone from seeing them until they get close to the entrance. Nevertheless, the use of darker coloured paints (red and black) suggests the painter’s intention to contrast the paints he/she was using against the pale colour of the sandstone wall. This implies his concern for the visibility of his paintings.
The walls of the rock shelter have red, reddish orange, burgundy red, black, and white paintings in monochrome, bichrome and polychrome patterns. The images
are painted with the fingertip. They depict zoomorphic (mainly zebu) and anthropomorphic (semi-naturalist and stylised) figures, as well as schematic-geometric-amorphic symbols (Figure 3). The latter dominate and are represented by quadrangular, circular and elliptical signs, dot/dash lines, and alphabet-like signs. In total, the Ampasimaiky rock paintings have 166 recognisable images.
Analysis and Interpretation
Based on comparative studies of the images, it has been found that the images of the Ampasimaiky rock paintings are not different from the images depicted on the prehistoric SGA rock art of the African continent (see Chami, 2008; Chaplin, 1974; Coulson and Campbell 2001;
Eastwood and Smith, 2005; Lynch and Donahue, 1980; Mabulla, 2005; Smith, 1995; Teka 2008; Willcox 1984). The similarities apparently suggest links between the two regions (Madagascar and the mainland) in the
(Madagascar and the continent) in prehistoric times. This suggests that the Ampasimaiky rock paintings are of African origin and could be relatively dated to the same period as those on the continent (Thackeray, 1983: 22). Comparative studies of the images have also shown that more than 30 types of images in the Ampasimaiky rock paintings are closely related to different ancient scripts such as Libyco-Berber, Protosinaitic, Meroitic (cursive), Punic Iberian, Ancient Phoenician, Tifinagh, and Aegean, Cretan and Proto-Egyptian linear signs respectively (Figure 4). These similarities would also suggest that the SGA-type rock paintings of Ampasimaiky are ancient writing.
figure 4
Analysis and Interpretation
Based on comparative studies of the images, it has been found that the images of the Ampasimaiky rock paintings are not different from the images depicted on the prehistoric SGA rock art of the African continent (see Chami, 2008; Chaplin, 1974; Coulson and Campbell 2001;
Eastwood and Smith, 2005; Lynch and Donahue, 1980; Mabulla, 2005; Smith, 1995; Teka 2008; Willcox 1984). The similarities apparently suggest links between the two regions (Madagascar and the mainland) in the
(Madagascar and the continent) in prehistoric times. This suggests that the Ampasimaiky rock paintings are of African origin and could be relatively dated to the same period as those on the continent (Thackeray, 1983: 22). Comparative studies of the images have also shown that more than 30 types of images in the Ampasimaiky rock paintings are closely related to different ancient scripts such as Libyco-Berber, Protosinaitic, Meroitic (cursive), Punic Iberian, Ancient Phoenician, Tifinagh, and Aegean, Cretan and Proto-Egyptian linear signs respectively (Figure 4). These similarities would also suggest that the SGA-type rock paintings of Ampasimaiky are ancient writing. them (Willcox, 1984: 210); the same has been applied to that of Professor Barry Fell, a specialist in Libyan-Berber writing. The same has been applied to that of Professor Barry Fell, a specialist in Libyan-Berber writing, who translated the glyphs of Driekopseiland in southern Africa as Libyan-Berber writing. But it has also been refuted on the grounds that the Bushmen could not write (Willcox 1984: 210); it is important to recall
It is important to recall in this connection that Libyco-Berber inscriptions inscribed on petroglyphs in the Canary Islands were not accepted as such, under the pretext that the Guanches (indigenous people of the Canary Islands) could not have acquired or adopted any writing system in prehistoric times (Cline 1953; Farrujia de la Rosa et al. 2010: 16-17; Willcox 1984: 48). This erroneous model persisted until reputable scholars succeeded in translating an inscription reading “save me from javelin annihilation” (Cline 1953: 273). Today, it is commonly accepted that the groups of aligned glyphs from the Canary Islands are Libyco-Berber inscriptions. Recently, Chami (2006: 84-5) insisted that African SGA rock paintings and engravings are ancient scripts. He has shown that some, such as those at Namoratunga in Kenya, are arranged. For the accounts of those that
are not arranged, he argued that at the time when the ancient scripts of Ethiopia, India, China, Egypt and Mesopotamia were still using ideograms, the characters too were not aligned (see also Casson 1966: 141). This can be amplified by the case of some Sahelian writing systems such as the ‘Nsibidi. This script has no ‘order in the arrangement’, but it is practically recognised as a writing system (Dayrell
1910; Macgregor 1909). Zahan (1950) identified the Dogon and Bambara pictographs as scripts based on the argument that
“writing consists essentially in the communication of thought by means of conventional signs, and the pictographs of both Dogon and Bambara serve to convey thoughts from one writer to another or from writer to lay public” (pp.137) « Une écriture consiste essentiellement de la communication de pensée par le moyen de signes conventionnels, et les pictographes de Dogon et de Bambara servent à transporter des pensées d’un auteur à l’autre ou d’un auteur au public”.
Therefore, it is also possible that without the “order in the arrangement”, the Ampasimaiky SGA-type rock paintings could have served to communicate thoughts and information. This would qualify them as a WRITING SYSTEM! Especially since some groups of symbols are horizontally and vertically aligned (Photo 4). Since “order in arrangement” has been established by some scholars as the sine qua non for a group of symbols/signs to qualify as an inscription (Willcox 1984: 210), one can reasonably assert, based on the above evidence, that the
AMPASIMAIKY RUPESTRATED ROCK SHELTER PAINTINGS ARE INSCRIPTIONS! This assertion is clearly supported by the vertically arranged group of symbols in panel 8 (see Photo 4b). Indeed, the repetition of the first sign in the third character (read downwards) is undoubtedly demonstrating the USE OF THE ALPHABET, and thus the
ALPHABET, so this is WRITING! All the more so as all five characters constituting the inscription are closely related to the Libyan-Berber alphabets (see
www.ancientscripts.com), the proto-Egyptian, Aegean and Cretan linear signs respectively
(Figure 5).
Figure 5 : Le groupe de signes verticalement arrangées d’Ampasimaiky comparé avec des caractères alphabétiques libyco-berbères et des signes linéaires proto-Egyptiens, égéens, et crétois (Tableau : Evans, 1897 :386).
Bien que des signes linéaires proto-Égyptiens, crétois et égéens aient été également verticalement écrits, je favoriserais l’origine libyco-berbère de l’inscription d’Ampasimaiky sous l’influence de la description de Walter Cline concernant les inscriptions libyco-berbères comme
suit).
“They [Libyco-Berber inscriptions] are usually very brief, consisting of personal names and a few phrases … The characters are simple and predominantly rectilinear. Variant alphabet may have existed. Most of inscriptions are to be read upward beginning with the left-hand column, though horizontal rows reading from right to left are not uncommon and some columns read downwards. The words are not separated. Though these signs sometimes occur in groups of four or five, they often stand singly…” (Cline 1953: 273)
All the underlined sentences qualify the vertically arranged inscription of Ampasimaiky as being of Libyco-berberber origin. Although this ancient script was written either horizontally or vertically, it should be remembered that the Libyco- Berber peoples used two different alphabets for each row (see http://www.ancientscripts.com). All five characters of the Ampasimaiky inscription are exclusively related to the vertical Libyco- Berber alphabet (Figure 6). Transliterating them to Latin alphabets – based on the transliteration of www.ancientscripts.com – one could read “ğ r ğ t f” (downwards) or “f t ğ r ğ” (upwards). The phonetic value ‘r’ for the presumed circle/square sign can be read as ‘3’ (= a), if the represented sign is perceived as a triangular symbol (see Fell 1975: 271). In the present state of the study, decoding the meaning of the inscription is not yet within the scope of the author. This task would rather be reserved for palaeographers and epigraphers.
Figure 7: The group of vertically arranged signs of Ampasimaiky transliterated in the Libyco-Berber alphabet
It is important to point out here that certain Libyco-Berber inscriptions brought back from the Canary Islands are not different from that of Ampasimaiky (see Figure 8). You can clearly see their resemblance.
Figure 8: The group of vertically arranged signs from Ampasimaiky compared with a Libyco-Berber inscription from the Canary Islands (Drawings by Farrujia de la Rosa et al. 2010: 29)
Discussion and Conclusion
Some of the images found at Ampasimaiky are similar to those of the rock art of continental Africa. However, the painting tradition of southwestern Madagascar is not
However, the painting tradition of southwestern Madagascar does not seem to originate from the continental region closest to Madagascar, i.e. Mozambique and the rest of the central Limpopo basin. In fact, Professor Benjamin Smith, an internationally recognised authority on the study of SGA rock paintings in sub-Saharan Africa, has taken the view that they are not typically affiliated with the Khoekhoen SGA rock paintings, nor with the geometric art of central southern Africa (pers. comm.). The exclusive representation of humped cattle associated with SGA-type symbols is particular to Ampasimaiky. Since no ancient representation of zebu (Bos indicus) has yet been found in the region between Sudan in the north and the Zambezi River in the south (Magnavita 2006: 62), it can be inferred that the tradition of rock paintings in Ampasimaiky originated in North Africa. Evidence of a bronze zebu figurine (of the type Bos indicus) inscribed with SGA signs dating back to 200 CE at Zeban Kutur, Eritrea (Photo 4) would confirm the early use of SGA signs among pastoral zebu-breeding peoples such as those depicted in the Ampasimaiky rock paintings
Photo 4: Bronze figure of zebu inscribed with SGA type signs in Zeban Kutur, Eritrea (Photo: Magnavita 2006:63)
The Libyan-Berber inscription discovered at Ampasimaiky also reinforces this proposal. The Libyan-Berber script emerged in North Africa around 500 BC and ceased to be used around 800 AD (Cline 1953).
around 800 AD (Cline 1953). This implies that the Libyan-Berber inscription of Ampasimaiky should have been written during the period between this bracket of dates (500 BC – AD 800). It is worth mentioning that Iamboulos, a Greek merchant who lived in the 3rd
century BC, reported about a vertical order of writing that was practised by the inhabitants of the ‘island of the sun’ (Allibert 2010: 111; Kobishschanow, 1965: 138; Oldfather, 1961 vol.2: 65-67; Winston 1976: 221). This island is believed to be geographically located adjacent to the east coast of Africa, south of the Equator. Although some of the geographical and ecological information in Iamboulos’ account is very much in line with that of the Malagasy west coast (for more details see the
Although some of the geographical and ecological information in Iamboulos’ account closely matches that of the Malagasy west coast (for more information see Allibert 2010), some scholars have consistently judged this account to be fictional, imaginary, and fanciful (Cary and Warmington 1963: 233-245; Winston 1976). One of the criteria used to censor this narrative is the use of vertical order writing by the islanders. Certainly, it was difficult for some scholars to accept the existence of a pre-Islamic script used in the adjacent islands of the east coast of Africa such as Madagascar (Janvier 1975: 19). However, with the discovery of the Libyan-Berber inscription at Ampasimaiky, this denial no longer holds. New evidence is uncovered to confirm that the “island of the sun” of Iamboulos could have been Madagascar. In any case, man was already present in the vicinity of Ampasimaiky when the Greek merchant passed through the island. Human activity is attested by subfossil lemur bones bearing cutting marks and radiocarbon-dated from the 5th to the 4th century BC at Taolambiby, in the upper Onilahy valley (see Perez et al., 2005). All the more so since ancient written documents (see Jones 1960: 235, 269, 315; Strabo 17, 1, 13 in Lacroix 1998: 95), archaeology (see Cary and Warmington 1963: 130; Chami 2007: 6; 2009; Wainwright 1940: 165), genetic study (see Ricaut et al. 2009: 5-6), and zooarchaeology (Chami 2006: 137; Grigson 1991: 124) all confirm the ancient contacts that existed during this period between the Onilahy Valley and the Onilahy Valley,
During this period, the Nile Valley and the Mediterranean world on the one hand, and the eastern coast of Africa and the adjacent islands on the other.
Références Allibert, C. (2008), “Austronesian Migration and the Establishment of the Malagasy Civilization: Contrasted Readings in Linguistics, Archaeology, Genetics and Cultural Anthropology”, Diogenes, SAGE, Volume 218, pp.7-16 Allibert, C. (2010), “L’île de Iamboulos est-elle Madagascar? », Pount, no.4, pp.103-119 Blench, R. (2007), “New Palaeozoogeographical Evidence for the Settlement of Madagascar”, Azania, Volume XLII, pp.69-82 Cary M. and Warmington, E. (1963), The Ancient Explorers, Middlesex, Penguin Books Casson, L. (1966), Ancient Egypt, Nederland, Time-life International Chami, F. A (2006), The Unity of African Ancient History, 3000 BC to AD 500, Dar es Salaam, E&D Limited Chami, F. A (2008), “The Great Lakes: a complexity of cultural wellsprings”, in Arnold, M. (eds.) Art in Eastern Africa, pp.47-64, Dar-es-Salaam, Mkuki na Nyota Chami F. A and Kwekason A. (2003), “Neolithic Pottery Traditions from the Islands, the Coast, and the Interior of East Africa”, African Archaeological Review, Volume 20, No. 2, pp.65-80 Chami, F. A, Tabibou, I. A and Abderhamane, B. (2009), “Preliminary Report of Archaeological Work Conducted on the Southern Ngazidja Island”, in Chami, F. (eds.), Zanzibar and the Swahili Coast from c30.000 years ago, pp.115-128, Dar es Salaam, E&D Vision Publishers Chaplin, T. (1974), “The Prehistoric Rock Art of the Lake Victoria Region”, Azania Volume 9, pp.1-50 Cline, W. (1953), “Berber Dialects and Berber Script”, Southwestern Journal of Anthropology, Vol.9, No.3, pp.268-276, New Mexico, New Mexico University Press Coulson D. and Campbell A. (2001), African Rock Art, Paintings and Engravings on Stone, New York, Harry N. Abrams Incorporated Dayrell, E. (1910), “Some “Nsibidi” Signs”, Man, Vol.10, pp.113-114. Dewar, R. E (1996), “The Archaeology of the Early Settlement of Madagascar”, in Reade J. (eds.) The Indian Ocean in Antiquity, pp.471-486 Dewar, R. E (1997), “Does It Matter That Madagascar Is an Island?” Human Ecology, Volume 25, No. 3, pp. 481-489 Donque, G. (1965), « Le contexte océanique des anciennes migrations : vents et courants dans l’océan indien », Taloha, Volume 1, pp.43-59. Eastwood, E. B and Smith, B. W (2005), “Fingerprints of the Khoekhoen: Geometric and handprinted Rock Art in the Central Limpopo Basin, Southern Africa”, Goodwin Series, RSA, South African Archaeological Society, Volume 9, pp.63-76 Evans, A. J (1897), “Further Discoveries of Cretan and Aegean Script: With Libyan and ProtoEgyptian Comparisons”, The Journal of Hellenic Studies, The Society for the Promotion of Hellenic Studies, Volume 17, pp.327-395 Farrujia de la Rosa, A., Pichler W., Rodrigue A., Marín S.G. 2010, The Lybico-Berber and Latino-Canarian Scripts and the Colonization of the Canary Islands, African Archaeological Review, vol.27, no.1, pp.13-41 Faublée, J. (1947), Récits Bara, Travaux et mémoires de l’Institut d’ethnologie XLXII, Paris, Université de Paris Fell, B. (1975), America BC: Ancient Settlers in the New World, New York, A Demeter Press Book Grandidier, A., Charles-Roux J., Delhorbe C., Froidevaux H. and Grandidier G. (1902), Collection des ouvrages anciens concernant Madagascar, Tome II, Paris, Comité de Madagascar Grigson, C. (1991), “An African Origin of African Cattle? Some Archaeological Evidence”, The African Archaeological Review, Volume 9, pp.119-144 Hall, S. and Smith, B. (2000), “Empowering Places: Rock Shelters and Ritual Control in FarmerForager Interactions in the Northern Province”, Goodwin Series, RSA, South African Archaeological Society, Volume 8, pp.30-46 Hébert, J. C (1967), « Essai d’interprétation de la stèle indéchiffrée d’Ambilobe », Taloha, Volume 2, pp.XLVII-LVII Hébert, J. C (1971), « Les tatouages de la côte est de Madagascar, d’après Chapelier (1794- 1806) », Taloha, Volume 4, pp.211-223 Huygues-Belrose, V. (1997). Les supports anciens de l’écriture à Madagascar, Etudes Océan Indien, vol.22, pp.7-19, INALCO Janvier, Y. (1975), « La géographie gréco-romaine a-t-elle connu Madagascar? Le point de la question », Omaly sy Anio, No.1-2, pp.11-41 Jones, H. (1960), The Geography of Strabo, Volume 1-2, London, William Heinemann Ltd. Juma, A. (2004), Unguja Ukuu on Zanzibar – An Archaeological Study of Early Urbanism, Sweden, Uppsala University Kessy, E. (2009a), “Analysis of Lithic Artefacts From Mwanampambe and Kuumbi Caves, Zanzibar”, in Chami F. (eds.), Zanzibar and the Swahili Coast from c30,000 years ago, Dar es Salaam, E&D Vision Publishers, pp.131-143 Kessy, E. (2009b), “Stone Artefacts from Male”, in Chami F. (eds.), Zanzibar and the Swahili Coast from c30,000 years ago, Dar es Salaam, E&D Vision Publishers, pp.184-88 Kobishchanov, Y.M. (1965), “On the Problem of Sea Voyages of Ancient Africans in the Indian Ocean”, The Journal of African History, Vol.6, No.2, pp.137-141, Cambridge, Cambridge University Press. Lacroix, W. F (1998), Africa in Antiquity, Nijmegen, Catholic University
Tanambelo V.R. RASOLONDRAINY,
(Centre de Documentation et de Recherche sur l’Art et la Tradition Orale de Madagascar)
Abstract:
This article presents the result of an archaeological study on the first prehistoric rock paintings site found in Madagascar and the rest of islands of Western Indian Ocean south of the equator.
Based on systematic methods and techniques, the study provides groundbreaking archaeological evidences that contribute largely to the understanding of the prehistory of Madagascar which was not archaeologically known thitherto to have prehistoric rock paintings, nor human settlement prior to 8th century AD. A Libyco-Berber inscription was discovered on the rock paintings of Ampasimaiky Rock Shelter, in Southwestern Madagascar. Based on this writing -supported by ancient historical records, circumstantial archaeological and paleontological evidences, linguistic data, and to some extent DNA analysis results -, the present study provides minimum age falling in the bracket time around 500 BC to roughly 8th century AD for the rock paintings of Ampasimaiky rock shelter. Furthermore, the study traces back early contacts between Madagascar and Northern Africa during prehistoric times.
Introduction
With the kind assistance of Dr Chantal Radimilahy, National Coordinator of the Network of African Archaeologists in Madagascar, I researched the ethnographic exhibition on the website of the Museum of Ethnography in Geneva (MEG). The theme was the peoples and cultures of the southwestern region of Madagascar, and was based on an exhibition of photographs taken by the French ethnographer Jacques Faublée in 1939. The pictures included rock paintings found in caves and rock shelters in the Isalo region. After discussions with my teachers at the Archaeology Unit of the University of Dar es Salaam, Professors Felix Chami, Fidelis Masao, and Audax Mabulla, I further investigated the significance of these paintings for the prehistoric period of Madagascar. This study is taken from a Master of Arts in Archaeology entitled
“Archaeologial study of the prehistoric rock paintings of Ampasimaiky rock shelter, in the upper Onilahy, Isalo region, Southwestern Madagascar”.
————–
Africa has many archaeological sites with prehistoric rock paintings (see Coulson and Campbell 2001; Willcox 1984). However, no such sites have yet been scientifically
scientifically reported from Madagascar, which is only 400 kilometres from southern Africa where they are abundant. All rock art previously found in Madagascar has been judged to have been produced in the recent past (Faublée 1947: 164; Hébert 1971: 218). This judgement is based on the theory that Africans were incapable of having navigated the high seas during prehistoric times (Donque 1965: 43; Grandidier et al. 1902; Southall 1975: 194; Vérin 1975, 1981). It was therefore out of the question to think that they had the possibility of transferring the practice of rock painting to the adjacent islands of Africa such as Madagascar. However, recent archaeological findings have demonstrated that the east coast of Africa and adjacent islands were already in contact since the prehistoric period (see Chami and Kwekason 2003; Chami et al. 2009; Juma 2004; Kessy 2009a,
2009b). These prompt a problem that led me to revise the antiquity of the rock paintings in Madagascar.
On the other hand, the name of the Arabic-Malagasy script – Sorabe meaning “great script” – indicates that there was an ancient script in Madagascar that was distinguished from Arabic (Hébert,
1967:XLVII; Huygues-Belrose 1997: 8). Although Hébert (1967:XLVIII) presumed that such a hypothetical script would have to be of Indian origin and would be discovered from
Although Hébert (1967:XLVIII) assumed that such a hypothetical script would have to be of Indian origin and would be discovered from stele and parietal inscriptions, no such inscription has so far been found in Madagascar.
My problem was therefore to find out whether this would appear on the rock paintings reported by the website. What provoked this idea the most was that Chami (2006:84-85; 2008) argued that the African Geometric and Amorphous Schematic (GAS) rock paintings and engravings are ancient writing.
Another problem is that radiocarbon 14 dating placed the earliest human settlement of Madagascar around the 8th century AD (Allibert 2008: 7; Dewar 1996: 474).
Although earlier dates are still being considered, many researchers have assumed that settlement of Madagascar would not go beyond the year W AD (for more information see Dewar 1997). This assumption has made research on Malagasy material culture going back to the Stone Age more than difficult. Any findings of material cultures dating from prehistoric times have been discarded, overlooked or misinterpreted. However, subfossil animal bone remains bearing traces of butchery found in paleontological sites in the southwestern region indicate the presence of human activity
(For more information, see Blench 2007; Perez et al. 2005). Despite this evidence, an archaeological site has not yet been
archaeological site with similar dates has not yet been found on the entire island.
With the discovery of a rock shelter with rock paintings, discovering new information about the prehistory of Madagascar before the 8th century proved to be feasible. All the more so as some of the paintings and engravings are
Especially since some of the SGA-type rock paintings and engravings found on the African continent – similar to those found in the photos on the MEG website – have been radiocarbon dated from the 1st millennium BC to the early 1st millennium AD (Chami, 2008: 54; Hall and Smith 2000; Lynche and Donahue 1980; Willcox 1984: 26, 83.
Fieldwork and results
Despite various challenges and constraints – mainly caused by the lack of archaeological reconnaissance prior to the fieldwork, insecurity in the study area, and the sacredness of the site to be studied – we had only somehow managed to conduct survey, archaeological excavation, and documentation of the paintings in the Ampasimaiky rock shelter. It is only the result of the latter that will be presented in this article.
The rock painting site is a rock shelter located at the foot of a hill immediately next to a small stream called Ampasimaiky (Figure 2), from which the name of the site derives. The presence of this water source may have influenced the painter’s choice of location.
The shelter itself is 26.8m long (south-north axis) and 5m wide (inwards).
Its ceiling begins to slope inwards from 4m high at its entrance to approximately 1.5m high at its irregular wall. The deepest part of the rock shelter – a chamber – is currently used by the royal family “Bara Zafimanely” as a burial place. Dung and ox hoof prints found on the ground indicate that cattle frequent the place. This occurrence should have disturbed the soil.
However, the areas around the site are dominated by rocky hills covered with grass. Only in the deep canyon near the confluence of Ampasimaiky and the Sakamarekely River is a patch of thick vegetation up to 10 m high observed along the stream. This situation would make agricultural activity almost impossible around the site, unless one had to move 1.5 km to the bank, where rice cultivation takes place today.
An archaeological survey of the southern Isalo region (Figure 1) was carried out and a rock painting site was discovered. Professor Chami participated in the fieldwork, advising on the approach to be followed in the treatment of these previously unstudied cave cultures in Madagascar.
Madagascar. History students from the University of Toliara, Mex Tafitasoa, Nadya
Ravololonirina, and Norbert Antilahy, and some villagers (Photo 1) participated in the work. The Network of African Archaeologists in Madagascar, under the direction of Dr Chantal Radimilahy and Prof Barthélemy Manjakahery, provided the logistical equipment used. The various signatures The various signatures of national and local administrative approval, the security provided by the Gendarmerie Brigade of Benenitsy, and above all the kindness and openness of King Zafimanely (Photo 1) – who allowed us access to the (burial) site – contributed enormously to the accomplishment of the expedition.
Madagascar. History students from the University of Toliara, Mex Tafitasoa, Nadya
Ravololonirina, and Norbert Antilahy, and some villagers (Photo 1) participated in the work. The Network of African Archaeologists in Madagascar, under the direction of Dr Chantal Radimilahy and Prof Barthélemy Manjakahery, provided the logistical equipment used. The various signatures The various signatures of national and local administrative approval, the security provided by the Gendarmerie Brigade of Benenitsy, and above all the kindness and openness of King Zafimanely (Photo 1) – who allowed us access to the (burial) site – contributed enormously to the accomplishment of the expedition.
The rock shelter at Ampasimaiky faces east. This arrangement makes the painting panels exposed to the sun in the morning and in the shade beyond noon. This arrangement suggests that painting activities are conducted either early in the morning while the heat of the sun is still moderate, or beyond noon when shade is available below the shelter.
This is based on the fact that the height of the panels does not exceed 130 cm, suggesting that the paintings were intended for people sitting on the floor of the undercroft shelter
Photo 2). Minimal, therefore, is the likelihood of the painting activity – involving seated people – having been conducted during the time when the sun is heating the shelter floor.
However, the view outside the site is limited. The opposing hills – on the other bank of the Ampasimaiky stream – prevent a person standing under the shelter from having a good view of the surroundings (Photo 3). It is much more likely that the view outwards – which implies a defensive strategy (Smith 1995) – was not among the painter’s concerns when he/she chose the site. This demonstrates that the layout of the Ampasimaiky rock shelter favours invaders from unexpectedly attacking its residents. The place would therefore not be suitable for permanent habitation.
Photo 3
Especially since a person standing on the hills of the other bank would be able to look directly at the other one positioned under the rock shelter, unless he hides in the deeper part of the shelter. Since the wall of this “chamber”, according to the local population, is not painted, it is plausible that the activities related to the paintings in the Ampasimaiky rock shelter are not clandestine. However, the paintings were not intended to be seen by people passing close to the site. They are not seen from a distance. Not only are the panels placed below 130 cm in height, but the slope of the ground before the access
to the shelter also prevents anyone from seeing them until they get close to the entrance. Nevertheless, the use of darker coloured paints (red and black) suggests the painter’s intention to contrast the paints he/she was using against the pale colour of the sandstone wall. This implies his concern for the visibility of his paintings.
The walls of the rock shelter have red, reddish orange, burgundy red, black, and white paintings in monochrome, bichrome and polychrome patterns. The images
are painted with the fingertip. They depict zoomorphic (mainly zebu) and anthropomorphic (semi-naturalist and stylised) figures, as well as schematic-geometric-amorphic symbols (Figure 3). The latter dominate and are represented by quadrangular, circular and elliptical signs, dot/dash lines, and alphabet-like signs. In total, the Ampasimaiky rock paintings have 166 recognisable images.
Analysis and Interpretation
Based on comparative studies of the images, it has been found that the images of the Ampasimaiky rock paintings are not different from the images depicted on the prehistoric SGA rock art of the African continent (see Chami, 2008; Chaplin, 1974; Coulson and Campbell 2001;
Eastwood and Smith, 2005; Lynch and Donahue, 1980; Mabulla, 2005; Smith, 1995; Teka 2008; Willcox 1984). The similarities apparently suggest links between the two regions (Madagascar and the mainland) in the
(Madagascar and the continent) in prehistoric times. This suggests that the Ampasimaiky rock paintings are of African origin and could be relatively dated to the same period as those on the continent (Thackeray, 1983: 22). Comparative studies of the images have also shown that more than 30 types of images in the Ampasimaiky rock paintings are closely related to different ancient scripts such as Libyco-Berber, Protosinaitic, Meroitic (cursive), Punic Iberian, Ancient Phoenician, Tifinagh, and Aegean, Cretan and Proto-Egyptian linear signs respectively (Figure 4). These similarities would also suggest that the SGA-type rock paintings of Ampasimaiky are ancient writing.
figure 4
Analysis and Interpretation
Based on comparative studies of the images, it has been found that the images of the Ampasimaiky rock paintings are not different from the images depicted on the prehistoric SGA rock art of the African continent (see Chami, 2008; Chaplin, 1974; Coulson and Campbell 2001;
Eastwood and Smith, 2005; Lynch and Donahue, 1980; Mabulla, 2005; Smith, 1995; Teka 2008; Willcox 1984). The similarities apparently suggest links between the two regions (Madagascar and the mainland) in the
(Madagascar and the continent) in prehistoric times. This suggests that the Ampasimaiky rock paintings are of African origin and could be relatively dated to the same period as those on the continent (Thackeray, 1983: 22). Comparative studies of the images have also shown that more than 30 types of images in the Ampasimaiky rock paintings are closely related to different ancient scripts such as Libyco-Berber, Protosinaitic, Meroitic (cursive), Punic Iberian, Ancient Phoenician, Tifinagh, and Aegean, Cretan and Proto-Egyptian linear signs respectively (Figure 4). These similarities would also suggest that the SGA-type rock paintings of Ampasimaiky are ancient writing. them (Willcox, 1984: 210); the same has been applied to that of Professor Barry Fell, a specialist in Libyan-Berber writing. The same has been applied to that of Professor Barry Fell, a specialist in Libyan-Berber writing, who translated the glyphs of Driekopseiland in southern Africa as Libyan-Berber writing. But it has also been refuted on the grounds that the Bushmen could not write (Willcox 1984: 210); it is important to recall
It is important to recall in this connection that Libyco-Berber inscriptions inscribed on petroglyphs in the Canary Islands were not accepted as such, under the pretext that the Guanches (indigenous people of the Canary Islands) could not have acquired or adopted any writing system in prehistoric times (Cline 1953; Farrujia de la Rosa et al. 2010: 16-17; Willcox 1984: 48). This erroneous model persisted until reputable scholars succeeded in translating an inscription reading “save me from javelin annihilation” (Cline 1953: 273). Today, it is commonly accepted that the groups of aligned glyphs from the Canary Islands are Libyco-Berber inscriptions. Recently, Chami (2006: 84-5) insisted that African SGA rock paintings and engravings are ancient scripts. He has shown that some, such as those at Namoratunga in Kenya, are arranged. For the accounts of those that
are not arranged, he argued that at the time when the ancient scripts of Ethiopia, India, China, Egypt and Mesopotamia were still using ideograms, the characters too were not aligned (see also Casson 1966: 141). This can be amplified by the case of some Sahelian writing systems such as the ‘Nsibidi. This script has no ‘order in the arrangement’, but it is practically recognised as a writing system (Dayrell
1910; Macgregor 1909). Zahan (1950) identified the Dogon and Bambara pictographs as scripts based on the argument that
“writing consists essentially in the communication of thought by means of conventional signs, and the pictographs of both Dogon and Bambara serve to convey thoughts from one writer to another or from writer to lay public” (pp.137) « Une écriture consiste essentiellement de la communication de pensée par le moyen de signes conventionnels, et les pictographes de Dogon et de Bambara servent à transporter des pensées d’un auteur à l’autre ou d’un auteur au public”.
Therefore, it is also possible that without the “order in the arrangement”, the Ampasimaiky SGA-type rock paintings could have served to communicate thoughts and information. This would qualify them as a WRITING SYSTEM! Especially since some groups of symbols are horizontally and vertically aligned (Photo 4). Since “order in arrangement” has been established by some scholars as the sine qua non for a group of symbols/signs to qualify as an inscription (Willcox 1984: 210), one can reasonably assert, based on the above evidence, that the
AMPASIMAIKY RUPESTRATED ROCK SHELTER PAINTINGS ARE INSCRIPTIONS! This assertion is clearly supported by the vertically arranged group of symbols in panel 8 (see Photo 4b). Indeed, the repetition of the first sign in the third character (read downwards) is undoubtedly demonstrating the USE OF THE ALPHABET, and thus the
ALPHABET, so this is WRITING! All the more so as all five characters constituting the inscription are closely related to the Libyan-Berber alphabets (see
www.ancientscripts.com), the proto-Egyptian, Aegean and Cretan linear signs respectively
(Figure 5).
Figure 5 : Le groupe de signes verticalement arrangées d’Ampasimaiky comparé avec des caractères alphabétiques libyco-berbères et des signes linéaires proto-Egyptiens, égéens, et crétois (Tableau : Evans, 1897 :386).
Bien que des signes linéaires proto-Égyptiens, crétois et égéens aient été également verticalement écrits, je favoriserais l’origine libyco-berbère de l’inscription d’Ampasimaiky sous l’influence de la description de Walter Cline concernant les inscriptions libyco-berbères comme
suit).
“They [Libyco-Berber inscriptions] are usually very brief, consisting of personal names and a few phrases … The characters are simple and predominantly rectilinear. Variant alphabet may have existed. Most of inscriptions are to be read upward beginning with the left-hand column, though horizontal rows reading from right to left are not uncommon and some columns read downwards. The words are not separated. Though these signs sometimes occur in groups of four or five, they often stand singly…” (Cline 1953: 273)
All the underlined sentences qualify the vertically arranged inscription of Ampasimaiky as being of Libyco-berberber origin. Although this ancient script was written either horizontally or vertically, it should be remembered that the Libyco- Berber peoples used two different alphabets for each row (see http://www.ancientscripts.com). All five characters of the Ampasimaiky inscription are exclusively related to the vertical Libyco- Berber alphabet (Figure 6). Transliterating them to Latin alphabets – based on the transliteration of www.ancientscripts.com – one could read “ğ r ğ t f” (downwards) or “f t ğ r ğ” (upwards). The phonetic value ‘r’ for the presumed circle/square sign can be read as ‘3’ (= a), if the represented sign is perceived as a triangular symbol (see Fell 1975: 271). In the present state of the study, decoding the meaning of the inscription is not yet within the scope of the author. This task would rather be reserved for palaeographers and epigraphers.
Figure 7: The group of vertically arranged signs of Ampasimaiky transliterated in the Libyco-Berber alphabet
It is important to point out here that certain Libyco-Berber inscriptions brought back from the Canary Islands are not different from that of Ampasimaiky (see Figure 8). You can clearly see their resemblance.
Figure 8: The group of vertically arranged signs from Ampasimaiky compared with a Libyco-Berber inscription from the Canary Islands (Drawings by Farrujia de la Rosa et al. 2010: 29)
Discussion and Conclusion
Some of the images found at Ampasimaiky are similar to those of the rock art of continental Africa. However, the painting tradition of southwestern Madagascar is not
However, the painting tradition of southwestern Madagascar does not seem to originate from the continental region closest to Madagascar, i.e. Mozambique and the rest of the central Limpopo basin. In fact, Professor Benjamin Smith, an internationally recognised authority on the study of SGA rock paintings in sub-Saharan Africa, has taken the view that they are not typically affiliated with the Khoekhoen SGA rock paintings, nor with the geometric art of central southern Africa (pers. comm.). The exclusive representation of humped cattle associated with SGA-type symbols is particular to Ampasimaiky. Since no ancient representation of zebu (Bos indicus) has yet been found in the region between Sudan in the north and the Zambezi River in the south (Magnavita 2006: 62), it can be inferred that the tradition of rock paintings in Ampasimaiky originated in North Africa. Evidence of a bronze zebu figurine (of the type Bos indicus) inscribed with SGA signs dating back to 200 CE at Zeban Kutur, Eritrea (Photo 4) would confirm the early use of SGA signs among pastoral zebu-breeding peoples such as those depicted in the Ampasimaiky rock paintings
Photo 4: Bronze figure of zebu inscribed with SGA type signs in Zeban Kutur, Eritrea (Photo: Magnavita 2006:63)
The Libyan-Berber inscription discovered at Ampasimaiky also reinforces this proposal. The Libyan-Berber script emerged in North Africa around 500 BC and ceased to be used around 800 AD (Cline 1953).
around 800 AD (Cline 1953). This implies that the Libyan-Berber inscription of Ampasimaiky should have been written during the period between this bracket of dates (500 BC – AD 800). It is worth mentioning that Iamboulos, a Greek merchant who lived in the 3rd
century BC, reported about a vertical order of writing that was practised by the inhabitants of the ‘island of the sun’ (Allibert 2010: 111; Kobishschanow, 1965: 138; Oldfather, 1961 vol.2: 65-67; Winston 1976: 221). This island is believed to be geographically located adjacent to the east coast of Africa, south of the Equator. Although some of the geographical and ecological information in Iamboulos’ account is very much in line with that of the Malagasy west coast (for more details see the
Although some of the geographical and ecological information in Iamboulos’ account closely matches that of the Malagasy west coast (for more information see Allibert 2010), some scholars have consistently judged this account to be fictional, imaginary, and fanciful (Cary and Warmington 1963: 233-245; Winston 1976). One of the criteria used to censor this narrative is the use of vertical order writing by the islanders. Certainly, it was difficult for some scholars to accept the existence of a pre-Islamic script used in the adjacent islands of the east coast of Africa such as Madagascar (Janvier 1975: 19). However, with the discovery of the Libyan-Berber inscription at Ampasimaiky, this denial no longer holds. New evidence is uncovered to confirm that the “island of the sun” of Iamboulos could have been Madagascar. In any case, man was already present in the vicinity of Ampasimaiky when the Greek merchant passed through the island. Human activity is attested by subfossil lemur bones bearing cutting marks and radiocarbon-dated from the 5th to the 4th century BC at Taolambiby, in the upper Onilahy valley (see Perez et al., 2005). All the more so since ancient written documents (see Jones 1960: 235, 269, 315; Strabo 17, 1, 13 in Lacroix 1998: 95), archaeology (see Cary and Warmington 1963: 130; Chami 2007: 6; 2009; Wainwright 1940: 165), genetic study (see Ricaut et al. 2009: 5-6), and zooarchaeology (Chami 2006: 137; Grigson 1991: 124) all confirm the ancient contacts that existed during this period between the Onilahy Valley and the Onilahy Valley,
During this period, the Nile Valley and the Mediterranean world on the one hand, and the eastern coast of Africa and the adjacent islands on the other.
Références Allibert, C. (2008), “Austronesian Migration and the Establishment of the Malagasy Civilization: Contrasted Readings in Linguistics, Archaeology, Genetics and Cultural Anthropology”, Diogenes, SAGE, Volume 218, pp.7-16 Allibert, C. (2010), “L’île de Iamboulos est-elle Madagascar? », Pount, no.4, pp.103-119 Blench, R. (2007), “New Palaeozoogeographical Evidence for the Settlement of Madagascar”, Azania, Volume XLII, pp.69-82 Cary M. and Warmington, E. (1963), The Ancient Explorers, Middlesex, Penguin Books Casson, L. (1966), Ancient Egypt, Nederland, Time-life International Chami, F. A (2006), The Unity of African Ancient History, 3000 BC to AD 500, Dar es Salaam, E&D Limited Chami, F. A (2008), “The Great Lakes: a complexity of cultural wellsprings”, in Arnold, M. (eds.) Art in Eastern Africa, pp.47-64, Dar-es-Salaam, Mkuki na Nyota Chami F. A and Kwekason A. (2003), “Neolithic Pottery Traditions from the Islands, the Coast, and the Interior of East Africa”, African Archaeological Review, Volume 20, No. 2, pp.65-80 Chami, F. A, Tabibou, I. A and Abderhamane, B. (2009), “Preliminary Report of Archaeological Work Conducted on the Southern Ngazidja Island”, in Chami, F. (eds.), Zanzibar and the Swahili Coast from c30.000 years ago, pp.115-128, Dar es Salaam, E&D Vision Publishers Chaplin, T. (1974), “The Prehistoric Rock Art of the Lake Victoria Region”, Azania Volume 9, pp.1-50 Cline, W. (1953), “Berber Dialects and Berber Script”, Southwestern Journal of Anthropology, Vol.9, No.3, pp.268-276, New Mexico, New Mexico University Press Coulson D. and Campbell A. (2001), African Rock Art, Paintings and Engravings on Stone, New York, Harry N. Abrams Incorporated Dayrell, E. (1910), “Some “Nsibidi” Signs”, Man, Vol.10, pp.113-114. Dewar, R. E (1996), “The Archaeology of the Early Settlement of Madagascar”, in Reade J. (eds.) The Indian Ocean in Antiquity, pp.471-486 Dewar, R. E (1997), “Does It Matter That Madagascar Is an Island?” Human Ecology, Volume 25, No. 3, pp. 481-489 Donque, G. (1965), « Le contexte océanique des anciennes migrations : vents et courants dans l’océan indien », Taloha, Volume 1, pp.43-59. Eastwood, E. B and Smith, B. W (2005), “Fingerprints of the Khoekhoen: Geometric and handprinted Rock Art in the Central Limpopo Basin, Southern Africa”, Goodwin Series, RSA, South African Archaeological Society, Volume 9, pp.63-76 Evans, A. J (1897), “Further Discoveries of Cretan and Aegean Script: With Libyan and ProtoEgyptian Comparisons”, The Journal of Hellenic Studies, The Society for the Promotion of Hellenic Studies, Volume 17, pp.327-395 Farrujia de la Rosa, A., Pichler W., Rodrigue A., Marín S.G. 2010, The Lybico-Berber and Latino-Canarian Scripts and the Colonization of the Canary Islands, African Archaeological Review, vol.27, no.1, pp.13-41 Faublée, J. (1947), Récits Bara, Travaux et mémoires de l’Institut d’ethnologie XLXII, Paris, Université de Paris Fell, B. (1975), America BC: Ancient Settlers in the New World, New York, A Demeter Press Book Grandidier, A., Charles-Roux J., Delhorbe C., Froidevaux H. and Grandidier G. (1902), Collection des ouvrages anciens concernant Madagascar, Tome II, Paris, Comité de Madagascar Grigson, C. (1991), “An African Origin of African Cattle? Some Archaeological Evidence”, The African Archaeological Review, Volume 9, pp.119-144 Hall, S. and Smith, B. (2000), “Empowering Places: Rock Shelters and Ritual Control in FarmerForager Interactions in the Northern Province”, Goodwin Series, RSA, South African Archaeological Society, Volume 8, pp.30-46 Hébert, J. C (1967), « Essai d’interprétation de la stèle indéchiffrée d’Ambilobe », Taloha, Volume 2, pp.XLVII-LVII Hébert, J. C (1971), « Les tatouages de la côte est de Madagascar, d’après Chapelier (1794- 1806) », Taloha, Volume 4, pp.211-223 Huygues-Belrose, V. (1997). Les supports anciens de l’écriture à Madagascar, Etudes Océan Indien, vol.22, pp.7-19, INALCO Janvier, Y. (1975), « La géographie gréco-romaine a-t-elle connu Madagascar? Le point de la question », Omaly sy Anio, No.1-2, pp.11-41 Jones, H. (1960), The Geography of Strabo, Volume 1-2, London, William Heinemann Ltd. Juma, A. (2004), Unguja Ukuu on Zanzibar – An Archaeological Study of Early Urbanism, Sweden, Uppsala University Kessy, E. (2009a), “Analysis of Lithic Artefacts From Mwanampambe and Kuumbi Caves, Zanzibar”, in Chami F. (eds.), Zanzibar and the Swahili Coast from c30,000 years ago, Dar es Salaam, E&D Vision Publishers, pp.131-143 Kessy, E. (2009b), “Stone Artefacts from Male”, in Chami F. (eds.), Zanzibar and the Swahili Coast from c30,000 years ago, Dar es Salaam, E&D Vision Publishers, pp.184-88 Kobishchanov, Y.M. (1965), “On the Problem of Sea Voyages of Ancient Africans in the Indian Ocean”, The Journal of African History, Vol.6, No.2, pp.137-141, Cambridge, Cambridge University Press. Lacroix, W. F (1998), Africa in Antiquity, Nijmegen, Catholic University
Tanambelo V.R. RASOLONDRAINY,
(Centre de Documentation et de Recherche sur l’Art et la Tradition Orale de Madagascar)
Abstract:
This article presents the result of an archaeological study on the first prehistoric rock paintings site found in Madagascar and the rest of islands of Western Indian Ocean south of the equator.
Based on systematic methods and techniques, the study provides groundbreaking archaeological evidences that contribute largely to the understanding of the prehistory of Madagascar which was not archaeologically known thitherto to have prehistoric rock paintings, nor human settlement prior to 8th century AD. A Libyco-Berber inscription was discovered on the rock paintings of Ampasimaiky Rock Shelter, in Southwestern Madagascar. Based on this writing -supported by ancient historical records, circumstantial archaeological and paleontological evidences, linguistic data, and to some extent DNA analysis results -, the present study provides minimum age falling in the bracket time around 500 BC to roughly 8th century AD for the rock paintings of Ampasimaiky rock shelter. Furthermore, the study traces back early contacts between Madagascar and Northern Africa during prehistoric times.
Introduction
With the kind assistance of Dr Chantal Radimilahy, National Coordinator of the Network of African Archaeologists in Madagascar, I researched the ethnographic exhibition on the website of the Museum of Ethnography in Geneva (MEG). The theme was the peoples and cultures of the southwestern region of Madagascar, and was based on an exhibition of photographs taken by the French ethnographer Jacques Faublée in 1939. The pictures included rock paintings found in caves and rock shelters in the Isalo region. After discussions with my teachers at the Archaeology Unit of the University of Dar es Salaam, Professors Felix Chami, Fidelis Masao, and Audax Mabulla, I further investigated the significance of these paintings for the prehistoric period of Madagascar. This study is taken from a Master of Arts in Archaeology entitled
“Archaeologial study of the prehistoric rock paintings of Ampasimaiky rock shelter, in the upper Onilahy, Isalo region, Southwestern Madagascar”.
————–
Africa has many archaeological sites with prehistoric rock paintings (see Coulson and Campbell 2001; Willcox 1984). However, no such sites have yet been scientifically
scientifically reported from Madagascar, which is only 400 kilometres from southern Africa where they are abundant. All rock art previously found in Madagascar has been judged to have been produced in the recent past (Faublée 1947: 164; Hébert 1971: 218). This judgement is based on the theory that Africans were incapable of having navigated the high seas during prehistoric times (Donque 1965: 43; Grandidier et al. 1902; Southall 1975: 194; Vérin 1975, 1981). It was therefore out of the question to think that they had the possibility of transferring the practice of rock painting to the adjacent islands of Africa such as Madagascar. However, recent archaeological findings have demonstrated that the east coast of Africa and adjacent islands were already in contact since the prehistoric period (see Chami and Kwekason 2003; Chami et al. 2009; Juma 2004; Kessy 2009a,
2009b). These prompt a problem that led me to revise the antiquity of the rock paintings in Madagascar.
On the other hand, the name of the Arabic-Malagasy script – Sorabe meaning “great script” – indicates that there was an ancient script in Madagascar that was distinguished from Arabic (Hébert,
1967:XLVII; Huygues-Belrose 1997: 8). Although Hébert (1967:XLVIII) presumed that such a hypothetical script would have to be of Indian origin and would be discovered from
Although Hébert (1967:XLVIII) assumed that such a hypothetical script would have to be of Indian origin and would be discovered from stele and parietal inscriptions, no such inscription has so far been found in Madagascar.
My problem was therefore to find out whether this would appear on the rock paintings reported by the website. What provoked this idea the most was that Chami (2006:84-85; 2008) argued that the African Geometric and Amorphous Schematic (GAS) rock paintings and engravings are ancient writing.
Another problem is that radiocarbon 14 dating placed the earliest human settlement of Madagascar around the 8th century AD (Allibert 2008: 7; Dewar 1996: 474).
Although earlier dates are still being considered, many researchers have assumed that settlement of Madagascar would not go beyond the year W AD (for more information see Dewar 1997). This assumption has made research on Malagasy material culture going back to the Stone Age more than difficult. Any findings of material cultures dating from prehistoric times have been discarded, overlooked or misinterpreted. However, subfossil animal bone remains bearing traces of butchery found in paleontological sites in the southwestern region indicate the presence of human activity
(For more information, see Blench 2007; Perez et al. 2005). Despite this evidence, an archaeological site has not yet been
archaeological site with similar dates has not yet been found on the entire island.
With the discovery of a rock shelter with rock paintings, discovering new information about the prehistory of Madagascar before the 8th century proved to be feasible. All the more so as some of the paintings and engravings are
Especially since some of the SGA-type rock paintings and engravings found on the African continent – similar to those found in the photos on the MEG website – have been radiocarbon dated from the 1st millennium BC to the early 1st millennium AD (Chami, 2008: 54; Hall and Smith 2000; Lynche and Donahue 1980; Willcox 1984: 26, 83.
Fieldwork and results
Despite various challenges and constraints – mainly caused by the lack of archaeological reconnaissance prior to the fieldwork, insecurity in the study area, and the sacredness of the site to be studied – we had only somehow managed to conduct survey, archaeological excavation, and documentation of the paintings in the Ampasimaiky rock shelter. It is only the result of the latter that will be presented in this article.
The rock painting site is a rock shelter located at the foot of a hill immediately next to a small stream called Ampasimaiky (Figure 2), from which the name of the site derives. The presence of this water source may have influenced the painter’s choice of location.
The shelter itself is 26.8m long (south-north axis) and 5m wide (inwards).
Its ceiling begins to slope inwards from 4m high at its entrance to approximately 1.5m high at its irregular wall. The deepest part of the rock shelter – a chamber – is currently used by the royal family “Bara Zafimanely” as a burial place. Dung and ox hoof prints found on the ground indicate that cattle frequent the place. This occurrence should have disturbed the soil.
However, the areas around the site are dominated by rocky hills covered with grass. Only in the deep canyon near the confluence of Ampasimaiky and the Sakamarekely River is a patch of thick vegetation up to 10 m high observed along the stream. This situation would make agricultural activity almost impossible around the site, unless one had to move 1.5 km to the bank, where rice cultivation takes place today.
An archaeological survey of the southern Isalo region (Figure 1) was carried out and a rock painting site was discovered. Professor Chami participated in the fieldwork, advising on the approach to be followed in the treatment of these previously unstudied cave cultures in Madagascar.
Madagascar. History students from the University of Toliara, Mex Tafitasoa, Nadya
Ravololonirina, and Norbert Antilahy, and some villagers (Photo 1) participated in the work. The Network of African Archaeologists in Madagascar, under the direction of Dr Chantal Radimilahy and Prof Barthélemy Manjakahery, provided the logistical equipment used. The various signatures The various signatures of national and local administrative approval, the security provided by the Gendarmerie Brigade of Benenitsy, and above all the kindness and openness of King Zafimanely (Photo 1) – who allowed us access to the (burial) site – contributed enormously to the accomplishment of the expedition.
Madagascar. History students from the University of Toliara, Mex Tafitasoa, Nadya
Ravololonirina, and Norbert Antilahy, and some villagers (Photo 1) participated in the work. The Network of African Archaeologists in Madagascar, under the direction of Dr Chantal Radimilahy and Prof Barthélemy Manjakahery, provided the logistical equipment used. The various signatures The various signatures of national and local administrative approval, the security provided by the Gendarmerie Brigade of Benenitsy, and above all the kindness and openness of King Zafimanely (Photo 1) – who allowed us access to the (burial) site – contributed enormously to the accomplishment of the expedition.
The rock shelter at Ampasimaiky faces east. This arrangement makes the painting panels exposed to the sun in the morning and in the shade beyond noon. This arrangement suggests that painting activities are conducted either early in the morning while the heat of the sun is still moderate, or beyond noon when shade is available below the shelter.
This is based on the fact that the height of the panels does not exceed 130 cm, suggesting that the paintings were intended for people sitting on the floor of the undercroft shelter
Photo 2). Minimal, therefore, is the likelihood of the painting activity – involving seated people – having been conducted during the time when the sun is heating the shelter floor.
However, the view outside the site is limited. The opposing hills – on the other bank of the Ampasimaiky stream – prevent a person standing under the shelter from having a good view of the surroundings (Photo 3). It is much more likely that the view outwards – which implies a defensive strategy (Smith 1995) – was not among the painter’s concerns when he/she chose the site. This demonstrates that the layout of the Ampasimaiky rock shelter favours invaders from unexpectedly attacking its residents. The place would therefore not be suitable for permanent habitation.
Photo 3
Especially since a person standing on the hills of the other bank would be able to look directly at the other one positioned under the rock shelter, unless he hides in the deeper part of the shelter. Since the wall of this “chamber”, according to the local population, is not painted, it is plausible that the activities related to the paintings in the Ampasimaiky rock shelter are not clandestine. However, the paintings were not intended to be seen by people passing close to the site. They are not seen from a distance. Not only are the panels placed below 130 cm in height, but the slope of the ground before the access
to the shelter also prevents anyone from seeing them until they get close to the entrance. Nevertheless, the use of darker coloured paints (red and black) suggests the painter’s intention to contrast the paints he/she was using against the pale colour of the sandstone wall. This implies his concern for the visibility of his paintings.
The walls of the rock shelter have red, reddish orange, burgundy red, black, and white paintings in monochrome, bichrome and polychrome patterns. The images
are painted with the fingertip. They depict zoomorphic (mainly zebu) and anthropomorphic (semi-naturalist and stylised) figures, as well as schematic-geometric-amorphic symbols (Figure 3). The latter dominate and are represented by quadrangular, circular and elliptical signs, dot/dash lines, and alphabet-like signs. In total, the Ampasimaiky rock paintings have 166 recognisable images.
Analysis and Interpretation
Based on comparative studies of the images, it has been found that the images of the Ampasimaiky rock paintings are not different from the images depicted on the prehistoric SGA rock art of the African continent (see Chami, 2008; Chaplin, 1974; Coulson and Campbell 2001;
Eastwood and Smith, 2005; Lynch and Donahue, 1980; Mabulla, 2005; Smith, 1995; Teka 2008; Willcox 1984). The similarities apparently suggest links between the two regions (Madagascar and the mainland) in the
(Madagascar and the continent) in prehistoric times. This suggests that the Ampasimaiky rock paintings are of African origin and could be relatively dated to the same period as those on the continent (Thackeray, 1983: 22). Comparative studies of the images have also shown that more than 30 types of images in the Ampasimaiky rock paintings are closely related to different ancient scripts such as Libyco-Berber, Protosinaitic, Meroitic (cursive), Punic Iberian, Ancient Phoenician, Tifinagh, and Aegean, Cretan and Proto-Egyptian linear signs respectively (Figure 4). These similarities would also suggest that the SGA-type rock paintings of Ampasimaiky are ancient writing.
figure 4
Analysis and Interpretation
Based on comparative studies of the images, it has been found that the images of the Ampasimaiky rock paintings are not different from the images depicted on the prehistoric SGA rock art of the African continent (see Chami, 2008; Chaplin, 1974; Coulson and Campbell 2001;
Eastwood and Smith, 2005; Lynch and Donahue, 1980; Mabulla, 2005; Smith, 1995; Teka 2008; Willcox 1984). The similarities apparently suggest links between the two regions (Madagascar and the mainland) in the
(Madagascar and the continent) in prehistoric times. This suggests that the Ampasimaiky rock paintings are of African origin and could be relatively dated to the same period as those on the continent (Thackeray, 1983: 22). Comparative studies of the images have also shown that more than 30 types of images in the Ampasimaiky rock paintings are closely related to different ancient scripts such as Libyco-Berber, Protosinaitic, Meroitic (cursive), Punic Iberian, Ancient Phoenician, Tifinagh, and Aegean, Cretan and Proto-Egyptian linear signs respectively (Figure 4). These similarities would also suggest that the SGA-type rock paintings of Ampasimaiky are ancient writing. them (Willcox, 1984: 210); the same has been applied to that of Professor Barry Fell, a specialist in Libyan-Berber writing. The same has been applied to that of Professor Barry Fell, a specialist in Libyan-Berber writing, who translated the glyphs of Driekopseiland in southern Africa as Libyan-Berber writing. But it has also been refuted on the grounds that the Bushmen could not write (Willcox 1984: 210); it is important to recall
It is important to recall in this connection that Libyco-Berber inscriptions inscribed on petroglyphs in the Canary Islands were not accepted as such, under the pretext that the Guanches (indigenous people of the Canary Islands) could not have acquired or adopted any writing system in prehistoric times (Cline 1953; Farrujia de la Rosa et al. 2010: 16-17; Willcox 1984: 48). This erroneous model persisted until reputable scholars succeeded in translating an inscription reading “save me from javelin annihilation” (Cline 1953: 273). Today, it is commonly accepted that the groups of aligned glyphs from the Canary Islands are Libyco-Berber inscriptions. Recently, Chami (2006: 84-5) insisted that African SGA rock paintings and engravings are ancient scripts. He has shown that some, such as those at Namoratunga in Kenya, are arranged. For the accounts of those that
are not arranged, he argued that at the time when the ancient scripts of Ethiopia, India, China, Egypt and Mesopotamia were still using ideograms, the characters too were not aligned (see also Casson 1966: 141). This can be amplified by the case of some Sahelian writing systems such as the ‘Nsibidi. This script has no ‘order in the arrangement’, but it is practically recognised as a writing system (Dayrell
1910; Macgregor 1909). Zahan (1950) identified the Dogon and Bambara pictographs as scripts based on the argument that
“writing consists essentially in the communication of thought by means of conventional signs, and the pictographs of both Dogon and Bambara serve to convey thoughts from one writer to another or from writer to lay public” (pp.137) « Une écriture consiste essentiellement de la communication de pensée par le moyen de signes conventionnels, et les pictographes de Dogon et de Bambara servent à transporter des pensées d’un auteur à l’autre ou d’un auteur au public”.
Therefore, it is also possible that without the “order in the arrangement”, the Ampasimaiky SGA-type rock paintings could have served to communicate thoughts and information. This would qualify them as a WRITING SYSTEM! Especially since some groups of symbols are horizontally and vertically aligned (Photo 4). Since “order in arrangement” has been established by some scholars as the sine qua non for a group of symbols/signs to qualify as an inscription (Willcox 1984: 210), one can reasonably assert, based on the above evidence, that the
AMPASIMAIKY RUPESTRATED ROCK SHELTER PAINTINGS ARE INSCRIPTIONS! This assertion is clearly supported by the vertically arranged group of symbols in panel 8 (see Photo 4b). Indeed, the repetition of the first sign in the third character (read downwards) is undoubtedly demonstrating the USE OF THE ALPHABET, and thus the
ALPHABET, so this is WRITING! All the more so as all five characters constituting the inscription are closely related to the Libyan-Berber alphabets (see
www.ancientscripts.com), the proto-Egyptian, Aegean and Cretan linear signs respectively
(Figure 5).
Figure 5 : Le groupe de signes verticalement arrangées d’Ampasimaiky comparé avec des caractères alphabétiques libyco-berbères et des signes linéaires proto-Egyptiens, égéens, et crétois (Tableau : Evans, 1897 :386).
Bien que des signes linéaires proto-Égyptiens, crétois et égéens aient été également verticalement écrits, je favoriserais l’origine libyco-berbère de l’inscription d’Ampasimaiky sous l’influence de la description de Walter Cline concernant les inscriptions libyco-berbères comme
suit).
“They [Libyco-Berber inscriptions] are usually very brief, consisting of personal names and a few phrases … The characters are simple and predominantly rectilinear. Variant alphabet may have existed. Most of inscriptions are to be read upward beginning with the left-hand column, though horizontal rows reading from right to left are not uncommon and some columns read downwards. The words are not separated. Though these signs sometimes occur in groups of four or five, they often stand singly…” (Cline 1953: 273)
All the underlined sentences qualify the vertically arranged inscription of Ampasimaiky as being of Libyco-berberber origin. Although this ancient script was written either horizontally or vertically, it should be remembered that the Libyco- Berber peoples used two different alphabets for each row (see http://www.ancientscripts.com). All five characters of the Ampasimaiky inscription are exclusively related to the vertical Libyco- Berber alphabet (Figure 6). Transliterating them to Latin alphabets – based on the transliteration of www.ancientscripts.com – one could read “ğ r ğ t f” (downwards) or “f t ğ r ğ” (upwards). The phonetic value ‘r’ for the presumed circle/square sign can be read as ‘3’ (= a), if the represented sign is perceived as a triangular symbol (see Fell 1975: 271). In the present state of the study, decoding the meaning of the inscription is not yet within the scope of the author. This task would rather be reserved for palaeographers and epigraphers.
Figure 7: The group of vertically arranged signs of Ampasimaiky transliterated in the Libyco-Berber alphabet
It is important to point out here that certain Libyco-Berber inscriptions brought back from the Canary Islands are not different from that of Ampasimaiky (see Figure 8). You can clearly see their resemblance.
Figure 8: The group of vertically arranged signs from Ampasimaiky compared with a Libyco-Berber inscription from the Canary Islands (Drawings by Farrujia de la Rosa et al. 2010: 29)
Discussion and Conclusion
Some of the images found at Ampasimaiky are similar to those of the rock art of continental Africa. However, the painting tradition of southwestern Madagascar is not
However, the painting tradition of southwestern Madagascar does not seem to originate from the continental region closest to Madagascar, i.e. Mozambique and the rest of the central Limpopo basin. In fact, Professor Benjamin Smith, an internationally recognised authority on the study of SGA rock paintings in sub-Saharan Africa, has taken the view that they are not typically affiliated with the Khoekhoen SGA rock paintings, nor with the geometric art of central southern Africa (pers. comm.). The exclusive representation of humped cattle associated with SGA-type symbols is particular to Ampasimaiky. Since no ancient representation of zebu (Bos indicus) has yet been found in the region between Sudan in the north and the Zambezi River in the south (Magnavita 2006: 62), it can be inferred that the tradition of rock paintings in Ampasimaiky originated in North Africa. Evidence of a bronze zebu figurine (of the type Bos indicus) inscribed with SGA signs dating back to 200 CE at Zeban Kutur, Eritrea (Photo 4) would confirm the early use of SGA signs among pastoral zebu-breeding peoples such as those depicted in the Ampasimaiky rock paintings
Photo 4: Bronze figure of zebu inscribed with SGA type signs in Zeban Kutur, Eritrea (Photo: Magnavita 2006:63)
The Libyan-Berber inscription discovered at Ampasimaiky also reinforces this proposal. The Libyan-Berber script emerged in North Africa around 500 BC and ceased to be used around 800 AD (Cline 1953).
around 800 AD (Cline 1953). This implies that the Libyan-Berber inscription of Ampasimaiky should have been written during the period between this bracket of dates (500 BC – AD 800). It is worth mentioning that Iamboulos, a Greek merchant who lived in the 3rd
century BC, reported about a vertical order of writing that was practised by the inhabitants of the ‘island of the sun’ (Allibert 2010: 111; Kobishschanow, 1965: 138; Oldfather, 1961 vol.2: 65-67; Winston 1976: 221). This island is believed to be geographically located adjacent to the east coast of Africa, south of the Equator. Although some of the geographical and ecological information in Iamboulos’ account is very much in line with that of the Malagasy west coast (for more details see the
Although some of the geographical and ecological information in Iamboulos’ account closely matches that of the Malagasy west coast (for more information see Allibert 2010), some scholars have consistently judged this account to be fictional, imaginary, and fanciful (Cary and Warmington 1963: 233-245; Winston 1976). One of the criteria used to censor this narrative is the use of vertical order writing by the islanders. Certainly, it was difficult for some scholars to accept the existence of a pre-Islamic script used in the adjacent islands of the east coast of Africa such as Madagascar (Janvier 1975: 19). However, with the discovery of the Libyan-Berber inscription at Ampasimaiky, this denial no longer holds. New evidence is uncovered to confirm that the “island of the sun” of Iamboulos could have been Madagascar. In any case, man was already present in the vicinity of Ampasimaiky when the Greek merchant passed through the island. Human activity is attested by subfossil lemur bones bearing cutting marks and radiocarbon-dated from the 5th to the 4th century BC at Taolambiby, in the upper Onilahy valley (see Perez et al., 2005). All the more so since ancient written documents (see Jones 1960: 235, 269, 315; Strabo 17, 1, 13 in Lacroix 1998: 95), archaeology (see Cary and Warmington 1963: 130; Chami 2007: 6; 2009; Wainwright 1940: 165), genetic study (see Ricaut et al. 2009: 5-6), and zooarchaeology (Chami 2006: 137; Grigson 1991: 124) all confirm the ancient contacts that existed during this period between the Onilahy Valley and the Onilahy Valley,
During this period, the Nile Valley and the Mediterranean world on the one hand, and the eastern coast of Africa and the adjacent islands on the other.
Références Allibert, C. (2008), “Austronesian Migration and the Establishment of the Malagasy Civilization: Contrasted Readings in Linguistics, Archaeology, Genetics and Cultural Anthropology”, Diogenes, SAGE, Volume 218, pp.7-16 Allibert, C. (2010), “L’île de Iamboulos est-elle Madagascar? », Pount, no.4, pp.103-119 Blench, R. (2007), “New Palaeozoogeographical Evidence for the Settlement of Madagascar”, Azania, Volume XLII, pp.69-82 Cary M. and Warmington, E. (1963), The Ancient Explorers, Middlesex, Penguin Books Casson, L. (1966), Ancient Egypt, Nederland, Time-life International Chami, F. A (2006), The Unity of African Ancient History, 3000 BC to AD 500, Dar es Salaam, E&D Limited Chami, F. A (2008), “The Great Lakes: a complexity of cultural wellsprings”, in Arnold, M. (eds.) Art in Eastern Africa, pp.47-64, Dar-es-Salaam, Mkuki na Nyota Chami F. A and Kwekason A. (2003), “Neolithic Pottery Traditions from the Islands, the Coast, and the Interior of East Africa”, African Archaeological Review, Volume 20, No. 2, pp.65-80 Chami, F. A, Tabibou, I. A and Abderhamane, B. (2009), “Preliminary Report of Archaeological Work Conducted on the Southern Ngazidja Island”, in Chami, F. (eds.), Zanzibar and the Swahili Coast from c30.000 years ago, pp.115-128, Dar es Salaam, E&D Vision Publishers Chaplin, T. (1974), “The Prehistoric Rock Art of the Lake Victoria Region”, Azania Volume 9, pp.1-50 Cline, W. (1953), “Berber Dialects and Berber Script”, Southwestern Journal of Anthropology, Vol.9, No.3, pp.268-276, New Mexico, New Mexico University Press Coulson D. and Campbell A. (2001), African Rock Art, Paintings and Engravings on Stone, New York, Harry N. Abrams Incorporated Dayrell, E. (1910), “Some “Nsibidi” Signs”, Man, Vol.10, pp.113-114. Dewar, R. E (1996), “The Archaeology of the Early Settlement of Madagascar”, in Reade J. (eds.) The Indian Ocean in Antiquity, pp.471-486 Dewar, R. E (1997), “Does It Matter That Madagascar Is an Island?” Human Ecology, Volume 25, No. 3, pp. 481-489 Donque, G. (1965), « Le contexte océanique des anciennes migrations : vents et courants dans l’océan indien », Taloha, Volume 1, pp.43-59. Eastwood, E. B and Smith, B. W (2005), “Fingerprints of the Khoekhoen: Geometric and handprinted Rock Art in the Central Limpopo Basin, Southern Africa”, Goodwin Series, RSA, South African Archaeological Society, Volume 9, pp.63-76 Evans, A. J (1897), “Further Discoveries of Cretan and Aegean Script: With Libyan and ProtoEgyptian Comparisons”, The Journal of Hellenic Studies, The Society for the Promotion of Hellenic Studies, Volume 17, pp.327-395 Farrujia de la Rosa, A., Pichler W., Rodrigue A., Marín S.G. 2010, The Lybico-Berber and Latino-Canarian Scripts and the Colonization of the Canary Islands, African Archaeological Review, vol.27, no.1, pp.13-41 Faublée, J. (1947), Récits Bara, Travaux et mémoires de l’Institut d’ethnologie XLXII, Paris, Université de Paris Fell, B. (1975), America BC: Ancient Settlers in the New World, New York, A Demeter Press Book Grandidier, A., Charles-Roux J., Delhorbe C., Froidevaux H. and Grandidier G. (1902), Collection des ouvrages anciens concernant Madagascar, Tome II, Paris, Comité de Madagascar Grigson, C. (1991), “An African Origin of African Cattle? Some Archaeological Evidence”, The African Archaeological Review, Volume 9, pp.119-144 Hall, S. and Smith, B. (2000), “Empowering Places: Rock Shelters and Ritual Control in FarmerForager Interactions in the Northern Province”, Goodwin Series, RSA, South African Archaeological Society, Volume 8, pp.30-46 Hébert, J. C (1967), « Essai d’interprétation de la stèle indéchiffrée d’Ambilobe », Taloha, Volume 2, pp.XLVII-LVII Hébert, J. C (1971), « Les tatouages de la côte est de Madagascar, d’après Chapelier (1794- 1806) », Taloha, Volume 4, pp.211-223 Huygues-Belrose, V. (1997). Les supports anciens de l’écriture à Madagascar, Etudes Océan Indien, vol.22, pp.7-19, INALCO Janvier, Y. (1975), « La géographie gréco-romaine a-t-elle connu Madagascar? Le point de la question », Omaly sy Anio, No.1-2, pp.11-41 Jones, H. (1960), The Geography of Strabo, Volume 1-2, London, William Heinemann Ltd. Juma, A. (2004), Unguja Ukuu on Zanzibar – An Archaeological Study of Early Urbanism, Sweden, Uppsala University Kessy, E. (2009a), “Analysis of Lithic Artefacts From Mwanampambe and Kuumbi Caves, Zanzibar”, in Chami F. (eds.), Zanzibar and the Swahili Coast from c30,000 years ago, Dar es Salaam, E&D Vision Publishers, pp.131-143 Kessy, E. (2009b), “Stone Artefacts from Male”, in Chami F. (eds.), Zanzibar and the Swahili Coast from c30,000 years ago, Dar es Salaam, E&D Vision Publishers, pp.184-88 Kobishchanov, Y.M. (1965), “On the Problem of Sea Voyages of Ancient Africans in the Indian Ocean”, The Journal of African History, Vol.6, No.2, pp.137-141, Cambridge, Cambridge University Press. Lacroix, W. F (1998), Africa in Antiquity, Nijmegen, Catholic University
Leave a Reply