Between Carthage and Zilum: ancient travel and exchanges across the central Sahara. (Patreon)

A thousand years before the Phoenician city of Carthage was founded, a complex society emerged along the southern shores of Lake Chad in what is today North-eastern Nigeria.

The recent discovery of Zilum and other ancient settlements in the lake chad basin, many of which were contemprenous with the flourishing of Carthage and the Garamantian civilization of southern Libya, has prompted a re-consideration of ancient Trans-Saharan exchanges between west Africa and the Mediterranean world.

The exceptional convergence of textural accounts and archeological evidence from the 5th century BC seem to point to a fascinating era of exchanges and movements between Carthage and west Africa. First is the appearance of a Carthaginian force with African auxiliaries on the island of Sicily, followed by descriptions of a daring naval expedition from Carthage to the Atlantic coast of west Africa. Next is the appearance of a mysterious civilization of the Garamantes whose capital has recently been discovered at Germa in Libya, and is dated to about the same era that the proto-urban settlement at Zilum was at its peak.

This article looks at the evidence for ancient contacts between the mediteranean world and west Africa, focusing on the historical links between Carthage and Zilum, inorder to uncover the origins of the aethiopian auxiliaries of Carthage.

Ruins of Carthage

(left) coin with head of an aethiopian, early 3rd century BC, Chiana Valley, Italy, British museum (right) coin with head of an eathiopian and an elephant, early 3rd century BC, Chiana Valley, Italy, private collection. These coins are asscoiated with Carthage's invasion of Rome.

Brief background on the 'black' auxiliaries in Carthaginian armies.

In the year 480 BC, a large Carthaginian force landed on the island Sicily and advanced east towards the city of Himera. The greek tyrant Gelon, who ruled over the island, led his own smaller army to meet the invading force at Himera. The ensuing battle was a decisive victory for Gelon, with many Carthaginian causalities and captives.

Among those captured in battle by Gelon were a group of 'very black' auxiliaries, who according to the roman author Frontinus (d. 103), were employed by the Carthaginians. This was the first of several references and depictions of 'black' Africans (aethiopians) in Carthaginian armies operating in the meditteranean, and such references would increase during Hannibal's invasion of Rome when contigents of 'black' elephant riders left a lasting impression in Roman material culture and memory.

(Blacks in Antiquity pg 4, 131, Before color prejudice pg 31-32)

The Carthaginian army led by general Hannibal, who invaded Roman Italy in the early third century with an army of elephants, is known to have carried several coins depicting  'aethiopian' men and elephants. These coins appear in several sites across the empire, from Spain to Sicily to Italy, but were likely struck at Carthage itself and later distributed during the campaigns.

While these coins have been subject to many interpretations over the years, its more readily accepted that Hannibal mostly utilized the African elephant even though the coins depict an Indian elephant. Theories that Hannibal could have acquired Indian elephants through Ptolemaic Egypt have since been discredited, (and with them, the possibility that the aethiopians were from Kush)  What’s beyond doubt is that the aethiopian figures shown alongside these elephants were from Africa.

(Carthaginian coinage in perspective pg 19, Carthage and the Indian elephant pg 122)

Who were these 'black' Carthaginians and where did they come from?

Roman statuette of a war elephant and its rider , originally from Pompei, Museo Archeologico Nazionale, Naples, 2nd century BC

A brief history of Carthage from its founding to the voyage of Hanno the navigator

The state of Carthage was founded in 814BC by Phoenician settlers in modern Tunisia, and represented a multi-ethnic society of mostly Punic-speaking groups and North-African Berbers, as well as sizeable communities from the African Sahara and southern Europe. Though it developed into a large empire, its nearly total destruction in 146 BCE at the hands of Roman armies left no internal histories of Carthage and only limited Carthaginian (epigraphic) records for historians to study. All reconstructions of Carthaginian history therefore begin with an intellectual framework inherited from Greco-Roman sources.

(Carthage's other wars pg 1-6, The Carthaginian empire pg xiii)

Map of the Mediterranean world in the 3rd century BC

Textural records indicate that Carthage experienced an intensive phase of growth from the mid-sixth century until about 400 BCE, but the resources necessary to maintain its territories reduced the Carthaginians’ ability to open up new territories. As Carthaginian citizens begun to resist overseas campaigns, ancient sources reflect an increasing dependence on hired mercenaries rather than citizen-soldiers. There are indeed a few epigraphic inscriptions which attest to indigenous Berber populations serving in the Carthaginian army in the 3rd century BC.

(Carthage's other wars pg 18-21,  The Carthaginian empire pg 8-14, 167)

The extent to which Carthaginian power reached into Africa is however minimal. Archeological evidence indicates Carthaginian expansion into North-Africa begun in the 5th century, but was mostly concentrated within modern Tunisia in a region covering no more than 30,000 sqkm (ie: less than 20% of the modern country). The surrounding Berber communities in north-eastern Algeria and north-western Libya were neverthless in close contact with Carthage and their populations were partially incorporated as dependancies. The epigraphic evidence shows that while Carthage's presence in Tunisia was expansive, it was nothing of the scale implied by Greco-Roman sources.

(The Carthaginian empire pg 142-149)

However, the Carthaginians did establish far flung colonies along the southern coast of the mediteranean as far as the Atlantic coast in Morocco. An ancient text said to have been written by the Carthaginian navigator Hanno, titled the 'Periplus of Hanno', supplies the names of landmarks and colonies all lying along the northern and central coasts of Morocco upto the island of Cerne (possibly Mogador in Morocco). Hanno reportedly sailed further south into west Africa, he writes that:

"Later, we again sailed from Cerne for twelve days along the coast, all of which was inhabited by aethiopians, who ran away from us. Even the Lixites with us could not understand their language. [12] On the twelfth day, we anchored under a high wooded mountain range, the trees on which were fragrant and of many different kinds. We passed this mountain range in two days' sail, and arrived at an immense bay on either side of which was low-lying land. From here we saw at night fires flaring up on all sides at irregular intervals. [14] Taking on water, we sailed along the coast for five days until we reached a great gulf, which the interpreters said was called West Horn. In it was a large island, and in this island a marine lake itself containing an island.

Landing on it we saw nothing but forest, and at night many fires being kindled; we heard the noise of pipes, cymbals, and drums, and the shouts of a large crowd. We were seized with fear, and the interpreters advised us to leave the island. [15] We sailed away quickly and coasted along a region with a flagrant smell of burning timber, from which steams of fire plunged into the sea. We could not approach the land because of the heat. [16] We therefore sailed quickly on in some fear, and in four days' time we saw the land ablaze at night; in the middle of this area one fire towered above the others and appeared to touch the stars; this was the highest mountain which we saw and was called the Chariot of the Gods. [17]

Following the river of fire for three days we came to a gulf called the Southern Horn. [18] In this gulf was an island like the one last mentioned, with a lake in which was another island. This was full of savages; by far the greater number were women with hairy bodies, called by our interpreters 'gorillas'. We gave chase to the men but could not catch any for they climbed up steep rocks and pelted us with stones. However we captured three women who bit and scratched their captors. We killed and flayed them and brought their skins back to Carthage. This was as far as we could sail owing to lack of provisions."

Unfortunately, this section of the voyage is considered by most west Africanists as fictional, with one writing that "the information in the Periplus is so vague that even if the account were true it would be valueless as a source on west Africa." Afterall, the topography of the region, the animals described and the 'barbarous' strangers were all known to any Mediterranean navigator and frequently used as literary tropes.

(The Periplus of Hanno pg 49-50, 54-58)

While its unlikely that Hanno ever sailed south of Morocco, for reasons already explained here  its more plausible that the Carthaginians did have some contact with their more southerly African neighbors; the kingdom of the Garamantes.

A corridor between Carthage and west-Africa?: a brief history of the Garamantian polity in southern Libya

About 1,500km south-east of Carthage was the civilization of the Garamantes, an ancient polity which emerged in the Fezzan region of southern Libya around 1,000BC. The Garamantes were a highly complex oasis kingdom centered at their capital Germa/Jarma, controlling several important oasis towns such as Zuwila, Idri, Sabha, Brak, Qasr ash Sharraba and Tijrhi. The development of Garamantes settlements are divided into four phases; the Early Garamantian period (1000–500 BC), the Proto-Garamantian period (500–1 BC), the Classic Garamantian period (AD 1–300) and the Late Garamantian period (AD 300–700)

(Urbanisation and state formation pg 53-58)

The Garamantes were a predominantly agricultural society growing wheat, barley and the grape vine alongside the date palm. The early buildings at Garamantian sites were oval and sub-rectangular huts with one to three rooms located in undefended settlements at hillfort sites. By the late 1st millennium BC, the sites took on a more proto-urban character with mudbrick construction of complex multi-roomed buildings with surrounding walls and embankments. Many of these would increasingly acquired a more meditterean architecture in the classic period. The most prominent of these was the 9ha site of Jarma, the 15-18 ha site of Qasr ash-Sharraba, and the 4.5ha site of Zuwila.

(Urbanisation and state formation pg 71-75, 79, 82, 99-100)

Botanical remains attest to the oasis agricultural economy, which become more intensive by the mid to late 1st millennium with the introduction of summer crops like sorghum and pearl millet, as well as Cotton. This coincided with a significant increase in the numbers of villages that was accompanied by the construction of major irrigation works (foggaras). The foggaras and the increased settlements appeared in the Garamantian landscape at the same time as a number of new crops from the west African Sahel, indicating a substantial increase in the population of the Garamantian heartland.

(Urbanisation and state formation pg 76-77, pg 98)

There's extensive evidence of local craft production including hand-made pottery, grinding stones, polishers, bone tools, vesicular basalt lamps, leather products, and iron tools indicates the presence of specialists among the Garamantian population. There's also extensive evidence for trade between the Garamantes and the mediteranean world indicated, however, much of this evidence comes from the classic and late periods during the common era. The first imported ceramics in a Garamantian site appear at Saniat Jibril around 100AD.

(Urbanisation and state formation pg 93, 73, 99, The Garamantes and the Origins of Saharan Trade pg 25)

Ruins of the Garamatian capital of Jarma and the royal necropolis. most of these ruins date from the classic and late period in the common era.

The Garamantians were classified by some authors as aethiopians but were distinguished from aethiopians by others. They first appear in Hedrotus' account from 430 BC, who writes that the Garamntes "avoid all men and the society of any others; they do not posses any war-like weapon nor do they know how to defend themselves". In another section, he writes that the Garamantes are a "powerful nation, they lay earth upon the salt, and then saw their ground… These Garamantes hunt the aethiopian troglodytes in four-horse chariots; for the aethiopian troglodytes are the swiftest of foot of all men of whom we had any account given."

Herodotus didn't visit the region but obtained his information from second hand accounts, stating the "he only repeats what the 'lybians' say". He locates the Garamantian state about 19 days from Aethiopia (Kush) and 13 days from Egypt, but north of another Libyan group called the Nasamonians who are said to be the Lixites of Hanno's account.

(Herodotus: A New and Literal Version with a Geographical and General Index by Henry Cary pg 296, 299, Herodotus on the Garamantes pg 203-215 )

Later descriptions of Garamantians come from Roman authors, many of whom obtained their information from Roman visits to the region and its subsquent military conquest. These include the geographers Strabo (d. 20) and Pilny (79AD), both of whom provide detailed descriptions of the Garamantean state and its capital, as well as locating the Garamantians among the aethiopians. The Garamantians were also described by Lucan (d. 65) as sunburned/scorched; by Arnobius (d. 330AD) as swarthy/tawny; and by Ptolemy as moderately black. Ptolemy also seems to consider the Garamantes as aethiopians when he states that they had the same king as their neighbors, the inhabitants of inner aethiopia, but elsewhere he seems to distinguish between aethiopians and Garamantes.

(Before color prejudice pg 9)

Detail of a mosaic depicting the execution of Damnatio ad Bestias (condemnation to beasts) with two Garamantian captives, similar to the account of the defeat of Garamantes in AD 70 detailed by Tacitus. Jamahiriya Museum, Tripoli, Libya. From Dar Buc Ammera villa (Zliten)

However, by the time of the Roman expeditions south of the Garamantian lands between 87-92AD, the land of aethiopia was considered to be much further south, perhaps near lake chad. A closer reading of the Garamantians' descriptions suggests that while they were similar to the aethiopians, they were distinguished from the particular ethnotype found in Roman depictions of the aethiopians of carthage.

(Romans and Blacks pg 61-62)

We therefore turn to the the Lake chad basin, 1,500 km south of the Garamantian capital, to find out whether it was the home of the aethiopians of Carthage.

The Gajiganna complex and the ancient settlement at Zilum

Recent archeological digs in the south-western margins of Lake Chad have revealed a number of large ancient fortified settlements that were part of the Neolithic culture known as the Gajigana complex. Known from its distinctive material culture and settlement patterns, the hierachical cultural groups of the Gajigana complex (ca. 1800-400BC) stand in sharp contrast to the preceding family-based life in small pastoral groups that had lived in the lake chad basin.

The early Gajigana phase from 1800-1000BC was characterized by an increased reliance on domesticated plants rather than pastoralism, permanent settlements in village communities with clay houses, and the creation of large settlement mounds. This phase was followed by an intermediate period where settlements declined and the population fell possibly due to climatic changes in the lake chad basin.

(Urbanization and state formation pg 564-565)

Settlements recovered rapidly during the third phase from 600-400BC with atleast 14 settlement sites, some as large as 14ha in size. The two largest settlements were Galgalkura A (12–14 ha) and Zilum (12–13 ha), and their clustering alongside smaller nucleated settlements suggests site hierarchies possibly denoting functional differences, and political hierarchies.

The most studied Gajiganna site is Zilum, a flat settlement of 12-13 ha lying at the edge of such a sand ridge. Unlike the relatively-well preserved Tichitt sites of which it is a contemporary, Zilum has been subject to intense erosion and was thus largely uncovered through magnetometry. Zilum was a highly nucleated site occupied entirely by many round structures built with clay, and surrounded by a monumental ditch and wall system that was upto 6 meters wide and 2 meters deep. It was estimated that the whole settlement had about 2,500 inhabitants at its height.

(Zilum: a mid-first millennium BC fortified settlement pg 154-158)

Heavily eroded settlement of Zilum strewn with fragments of ancient pottery and outlines of domestic structures

Magnetogram showing the outlines and arrangements of some sub-surface features as well as the location of an old ditch segment

The existance of an extensive ditch and wall system at Zilum, as well as at each of the Gajiganna sites indicates that the inhabitants of the complex sought protection by erecting earthen walls. The sheer length of these moats adds up to hundreds of meters, and suggests the need for large groups of labourers, and since construction work of this scale would have required organizing and directing, it suggests the presence of higher-ranked members of the community to commission and coordinate the efforts.

(Urbanisation and state formation pg 572)

There is archaeological evidence for craft production at Zilum that hints at some occupational differentiation. Pottery in the form of large storage vessels, as well as stone beads, stone tools and possibly leather were produced by specialists among Zilum's residential population.  Zilum was however not an iron-age settlement, unlike the contemprenous Nok and Tichitt sites to its west, or even the Garamantes to the north. Its also likely that regional trade existed between the Gajiganna settlements, but there's very little evidence for long-distance trade at Zilum or any other sites.

Only two exceptional finds of lip-plugs made of amazonite may potentially indicate long distance contacts, possibly from the Mandara Mountains (c.140 km away) or the Biu Plateau (c.190 km away) or the Air Mountains (c.900 km away) or the (Tibesti c.1600 km away).

(Zilum: a mid-first millennium BC fortified settlement pg 166-167)

Zilum and other Gajiganna sites were abandoned around 400BC and there would be no archeological evidence for settlements in the region until the beginning of the common era when a different cultural group emerged. By this time, the empire of Carthage had been subsumed by Rome, and the Garamantian state was at its greatest extent before it too would be incoporated into the Roman empire as a client state.

the characteristic pot-burials of the 'sao' (Holouf) culture which occupied the Gajiganna area in the long after the fall of Zilum

Resolving conflicting accounts of ancient trade and population movements across the Sahara

Herodotus' exceptional reference to Garamantian charioters 'hunting' trogodlye aethiopians has for long fueled debates about the ancient exchanges between the northern and southern parts of the central Sahara. Combined with the same author's mention of the 'silent trade' in gold between the Carthaginians and interior groups of North Africa, it has been taken by some early scholars as evidence of Trans-Saharan trade during antiquity, who then draw parallels between it and the better-known medieval trade across the Sahara.

However, the existance of a 'silent trade' was the first to be discredited by west Africanists, particulary in a monograph by Paulo de Moraes Farias, who shows it to be nothing more than a 'mythical stereotype' that was repeated by some writers over the centuries and uncritically reproduced by modern scholars as a factual description of exchanges.

Additionally, John Swanson argued that the presumed gold trade between Carthage and the North African interior doesn't appear in any classical sources besides Herodotus, and appears to infer a type of trade secrecy which Herodotus' account would counter-intuitively obliviate. And finally, as explained above, there's considerable doubt regarding the authenticity and interpretation of Hanno's purported voyage to west Africa.

(Silent Trade: Myth and Historical Evidence, The Myth of Trans-Saharan Trade during the Roman Era, The "Periplus of Hanno" )

However, the recent archeological studies of the Garamantian sites in the Fezzan has reignited the debate about the extent of ancient Trans-Saharan trade. The archeologist David Mattingly who has worked extensively on Garamantian sites, argues that "the Trans-Saharan zone was more connected than has been previously realised or widely admitted" and has argued scholars to reconsider the debate on ancient Trans-Saharan linkages.

Mattingly suggests that during the Garamantian era, north–south routes were established across the Sahara, connecting Fezzan to the Mediterranean and Sub-Saharan regions. While he contends that the debate on ancient Trans-Saharan trade is one that is "based on relatively little hard evidence" he supports arguments for its early existance, because in his opinion; the Saharan oases "do not make economic sense if solely dependent on irrigated agriculture." His response to critics such as Raulph Austen, who have found no evidence of meditterean goods in the southern Sahara, is that the Garamantians absorbed Roman goods and instead exported their own manufactures such as beads and textiles in exchange for gold and possibly captives.

(The Garamantes and the Origins of Saharan Trade)

However, specialists who have worked on the southern Saharan settlements --which would have received such Garamantian products-- maintain that there is little evidence for any long distance trade nor of Garamantian items south of the Sahara. The archeologist Carlos Magnavita who discovered the Gajiganna site of Zilum has found little evidence for external contact with the northern Sahara, and even less evidence for long distance trade. Magnavita has also collected analytical results from tests of gold samples from North Africa and west Africa, showing that the latter only appeared in the former region during the Islamic era.

(Urbanisation and state formation pg 516, Landscapes, Sources and Intellectual Projects pg 25-32)

Finally, there's still little definitive evidence for the movement of people from the southern Sahara into the Garamantian world beyond the presence of Saharan crops such as millet and sorghum. While future discoveries of artefacts and practices from the Southern Sahara in Garamantian contexts may indicate contacts between the two regions, the conclusions that can be drawn from such are far from clear. For example, the fascinating discovery of a female burial with a lip-plug in a Garamatian cemetary, hints at Southern-Saharan affinities but provides little information beyond that and is rather exceptional.

(Burials, Migration and Identity in the Ancient Sahara and Beyond pg 419)

Based on textural records of a joint Roman-Garamantian expedition into the south during the 1st century AD, some scholars speculated that "slave raids" from the north stimulated the founding of cities such as Jenne-Jeno and Zilum. But this argument has several obvious shortfalls, the first being that both settlements at Jenne-jeno and Zilum were products of ancient cultural developments in the Mema region of Mali and the Lake Chad basin, which long predate the existance of the Garamantians and the Romans.

(Landscapes, Sources and Intellectual Projects pg 33-34, Zilum: a mid-first millennium BC fortified settlement pg 168)

The second, more potent criticism is that the presence of fortifications does not by itself infer protection from slave raiding, if at all, but is often instead pursued for a variety of reasons including farmer-herder disputes, local conflict, robbery, or simply as monuments of political power. The fact that each of the Gajiganna sites was once surrounded by ditches and earthen walls indicates that protection seems to have been a major factor in their settlement clusters.

However, as with the case for the medieval fortifications of Holouf, its much more probable that local, rather than long distance factors, were the major contributor in instigating the mixture of nucleation, fortification and site clustering in the Gajiganna area.

(Urbanisation and state formation pg 582)

Aerial photo of Gulfey, a fortified Kotoko town near the southern margins of Lake Chad.

Conclusion: Ancient trans-saharan exchanges and the origins of the aethiopians of Carthage.

The above outline casts considerable doubt on the existance of ancient links between the Mediterranean and lake Chad region before the common era. Neither the textural records, nor the archeological studies provide any sufficient evidence for trade or population movements within the central Sahara, compared for example to the better-known links in the eastern Sahara as well as the medieval links between the Fezzan and Kanem-Bornu.

While the Garamantian state was doubtlessly an important Saharan polity that may have played a significant role in regional exchanges, the economic history of similar states in the same region such as the desert kingdom of Wadai, and the Oasis towns of Kawar, shows that the Saharan states need not depend on long distance trade to flourish in their relatively harsh environment.

It is therefore likely that the aethiopian auxiliaries frequently encountered in classical descriptions of Carthaginian armies were drawn from autochtonous populations of North Africa, rather than west Africans who had moved, voluntarily or not, into Carthage's territory.

References.

The Carthaginian empire by Nathan Pilkington

Carthage's Other Wars: Carthaginian Warfare Outside the 'Punic Wars' Against Rome by Dexter Hoyos

Urbanisation and State Formation in the Ancient Sahara and Beyond edited by Martin Sterry, David J. Mattingly

Burials, Migration and Identity in the Ancient Sahara and Beyond by M. C. Gatto, D. J. Mattingly, N. Ray, M. Sterr

Landscapes, Sources and Intellectual Projects of the West African Past: Essays in Honour of Paulo Fernando de Moraes Farias

Herodotus: A New and Literal Version with a Geographical and General Index by Henry Cary

Before Color Prejudice: The Ancient View of Blacks By Frank M. Snowden

Romans and Blacks by Lloyd A. Thompson

Herodotus on the Garamantes: A Problem in Protohistory by Daniel F. McCall

The Garamantes and the Origins of Saharan Trade by David J. Mattingly

Zilum: a mid-first millennium BC fortified settlement by Carlos Magnavita

Silent Trade: Myth and Historical Evidence by P. F. de Moraes Farias

Carthaginian coinage in perspective by Paolo Visonà

Carthage and the Indian Elephant by Michael B. Charles

The Myth of Trans-Saharan Trade during the Roman Era by John T. Swanson

The "Periplus of Hanno" in the History and Historiography of Black Africa by P. E. H. Hair

link