When Cosmas Indicopleustes introduced his account of Adulis and its throne he said that he had gone there as a trader together with other traders from both Alexandria and Elath (at the head of today’s Gulf of ‘Aqaba): “In Adulis, which is the name for the city of the Ethiopians that lies about two miles from the coast and serves as the port for the people of Axum, where those of us from Alexandria and Elath were engaged in commerce, there is a throne.…”1 This is not the only time when Cosmas identifies himself as a merchant in his Christian Topography.2 His other allusions to personal experiences leave no doubt that he operated in the area of the Red Sea and the Indian Ocean as far as the Persian Gulf and Ceylon, although there is not the slightest reason to believe that he ever went to the Indian subcontinent. The word Indicopleustes, which means “sailor to India,” was attached to him in the Middle Ages. Cosmas, as we have seen, is a banal Christian name that was certainly not his real name. It suggests the kosmos (“world”), for which the design was of enormous interest to Cosmas.
In fact, the full name Cosmas Indicopleustes does not appear anywhere in the work he actually wrote. The author identifies himself simply as “a Christian,” and it was not until the eleventh century that the manuscripts equipped him with the sonorous name by which we know him today, “Cosmas the sailor to India.” The reference to India in the epithet Indicopleustes can be no more than a reflection of the use of the geographical term “India,” both in Cosmas and elsewhere, to refer to a much wider region than the subcontinent that bears that name. Ancient writers applied it freely both to the east coast of Africa and to the southwest corner of the Arabian peninsula. Cosmas reports that in sailing to the Horn of Africa he went to “inner India,” and he uses the same expression for the land of Ḥimyar in the Arabian peninsula.3 He sailed as far as Ceylon, which he correctly designates by its ancient name, Taprobanê, while locating the island as lying in “inner India.”4 The use of “inner” for “more remote” or “outlying” territories in relation to a designated region had a long tradition in ancient geographical writing in both Greek and Semitic languages.5
It is clear that Cosmas viewed the base of his operations as lying well to the west of Ceylon in what would have been understood to be a still more remote part of inner India. This can easily be seen from the title of a work ascribed to Palladius, “On the Peoples of India and the Brahmans,” in which the author has much to say about the Ethiopians in eastern Africa.6 Cosmas’ world lay in the waters on either side of Arabia, in the Persian Gulf and the Red Sea, as well as in the expanse of the Indian Ocean into which the Red Sea debouches through the Gulf of Aden. This part of the Indian Ocean touched both Somalia in Africa and the south coast of the peninsula from Aden eastwards.
Not surprisingly, Cosmas passed by Socotra on his way between the Red Sea and the Persian Gulf. That island, together with its archipelago, lying off the coast of Yemen to the east of Cape Guardafui, had long been a station for merchants, as can be seen from the reference to it under its ancient name of Dioscourides in the Periplus of the Red Sea, which, as we remarked earlier, is now securely anchored to the mid-first century by its reference to the Nabataean king Malichus II (40–70 AD).7 Cosmas was under the impression that the island had been colonized by the Ptolemies in the Hellenistic period because, although he never put in there himself, he encountered inhabitants of the island when he was in Ethiopia and found that they still spoke Greek.8 Nevertheless, archaeological remains have failed to indicate any trace of Hellenic culture on Socotra, despite some extraordinary recent discoveries in the cave of Ḥôq on the island. These have revealed other settlements than Greek, including a colony of Palmyrenes in 258 AD, as well as, at an indeterminate date, settlements of Ethiopians and South Arabians. Graffiti in Ethiopic and South Arabian scripts and an inscription on a wooden tablet in Palmyrene Aramaic have been deciphered.9
Cosmas was something of an autodidact. He claims to have learned everything he knew from a Christian, Mar Aba, called Patrikios in Greek, with whom he had once studied in Alexandria. That learned cleric ultimately went on to become the archbishop of all the Nestorian communities in Persia. He served in his post from 540 to 552, a time frame that fits perfectly with the apparent period of composition of the Christian Topography. Fortunately, this can be determined by Cosmas’ assertion that he was asked to furnish the negus with copies of the throne inscriptions at Adulis during the visit he made to the site twenty-five years earlier. The imperialist claims of these texts strongly imply that the negus was contemplating the great Axumite expedition to Ḥimyar of 525, and this ought to mean that Cosmas was present at Adulis between 523, when the first provocations for war occurred in Arabia, and 525, when the expedition was launched. It is true that Cosmas dates his visit to the “beginning” (archê) of the reign of Justin, who ruled as emperor at Constantinople from 518 to 527. But after a quarter-century Cosmas can hardly be expected to be very precise in recalling exactly when Justin came to power. We have to recognize that his work includes references to two solar eclipses that occurred in 547.10 Hence he could not possibly have been in Adulis before late 522 or early 523, which is the period when the provocations that aroused the Ethiopians began to occur. Accordingly, the various details in Cosmas deliver a date of composition of between 548 and 550, which falls precisely during the episcopate of Cosmas’ teacher Patrikios.
As a merchant, Cosmas would naturally have called at the port of Adulis, where he was able to observe and describe the famous throne. Already in the middle of the first century AD this lay along the route of traders in the Red Sea, as the itinerary of the Periplus of the Red Sea makes plain. That anonymous work, cited for the location of Adulis in the foregoing chapter, provides a detailed, if not topographically precise account of the town, but its writer was clearly aware of its dependence upon the city of Axum in the highlands to the south. In addition, he alludes to an otherwise unknown king in a broad area that extended southwards another eighty miles or so along the coast from the Gulf of Zula. The king of that entire region in the Horn of Africa, which the Periplus calls generally Barbaria, greatly impressed the merchant author of the manual. Although the king’s realm probably included the Axumite territory, there is no reason to think that his capital was Axum. Barbaria seems to have been a much more extensive area, probably incorporating modern Djibouti and northern Somalia. The name itself had nothing to do with “barbarians” (the Greek term for non-Greeks) but reflects an indigenous appellation either for the people of Barbaria or conceivably for a divinity such as Barbar who had a temple on the island of Bahrain. The king, we are told, was named Zoskales. Although he is otherwise unattested, we learn from the anonymous merchant that he had a far from superficial knowledge of Greek. His mastery of the written language became evident to the writer in ways we cannot ascertain.11 But it is worth remembering this early attestation of Greek in the region when we consider the appearance of that language on the Axumite throne inscriptions and on other stones that survive to this day. It is reasonable to assume that Greek was conspicuously used, at least in the upper levels of local administration in East Africa, from the beginning of the Roman Empire and probably before.
Only a decade or two after the Periplus was written, the Roman polymath Pliny the Elder also singled out Adulis (oppidum Adulitarum) as a great emporium frequented by traders in the region.12 His subsequent reference in the same passage to a sinus Abalitu, in connection with a Diodorus Island, seems not to have been properly understood. This bay (sinus) can only indicate the Gulf of Zula. The Periplus had explicitly connected it with an island of the same name, and Abalitu must therefore be a deformation of Adulitou. By his own acknowledgment Pliny was drawing on the scholarly writings of King Juba II of Mauretania for his information: “In this part I have decided to follow … King Juba in the volumes he wrote for Gaius Caesar concerning his Arabian expedition.”13 The allusion is to preparation for the campaigns of Augustus’ grandson in the East at the end of the first century BC.
Since Juba, who was no less a scholar than a monarch, ruled in North Africa during Augustus’ reign and made use of earlier Hellenistic sources for a work that was designed to instruct the young prince, we can safely assume that Adulis was already important as an emporium in the first century BC. This is pertinent for the earlier of the two inscriptions on the throne with its record of the overseas exploits of Ptolemy III. But of course Cosmas, who remains our sole source for these documents, naturally had no sense of the chronological implications of what he had seen and transcribed. His mistaken belief that the two inscribed texts he saw were all part of a single document leaves no doubt about that.
After his career as a merchant Cosmas went on to become, a quarter-century later, the Christian apologist that we know from his book. His travels had clearly instilled in him a profound interest in geography that bore fruit in a work, regrettably lost, in which he had described the entire known world. He refers to it in the prologue to his Christian Topography as a work of reference that any reader of the Topography should consult. He reports that it encompassed the entire earth and all countries:
[It is] the volume we wrote for the Christ-loving Constantine, in which the whole earth has been fully described, both this one and what lies beyond the Ocean, as well as all countries: the southern regions from Alexandria to the southern ocean (by which I mean the Nile River and the adjacent regions and peoples of all Egypt and Ethiopia), and the Arabian Gulf with adjacent regions and peoples as far as this same ocean. Equally I include the land between the river and the Gulf, its cities, countries, and peoples.14
Assuming his readers would have access to this earlier book, Cosmas devoted himself in his surviving Topography to a pious refutation of various cosmological claims advanced by pagans or supposedly misguided Christians, in particular the representation of the universe as a sphere. He had already written a book, now lost like his Geography, to describe the movement of the stars and, as he says in his prologue, his objective was “to destroy the error of pagan hypotheses.” The ninth book of the Topography is also devoted to this subject and presumably resumes what Cosmas had written in his earlier treatise. Most remarkably in the Topography Cosmas resumed his spirited defense of the idea that the Jewish tabernacle is an image of the world (kosmos). The supposedly apostolic author of the Epistle to the Hebrews, Paul or someone else, had declared that Moses, after creating on Sinai the tabernacle that reproduced what he had actually beheld from God, transformed the single tabernacle into two by the interposition of a curtain or veil: one representing the world of mankind and the other the world to come. For Cosmas it was important to establish, from his Christian perspective, that the tabernacle as a whole was a representation of the world.15
This meant that for him the universe was a rectangular solid, which he believed to be longer horizontally, from east to west, than it was wide, from north to south. He also believed that the section on the other side of the veil in the Mosaic image of the tabernacle comprehended the heavens, with a cylindrical cap on top. Cosmas considered the entire box that constituted the kosmos a kind of house (oikos) in the shape of a cube (kubos)—despite the differing length and width. The notion of a cube was borrowed from the Septuagint Greek text of Job, for which the original Hebrew offered no equivalent term. As the most knowledgeable modern exegete of the Topography has prudently declared, “Cosmas is not rich in geometrical accuracy.”16 Although it is hard to take his idea of the world seriously, particularly for an age in which the spherical representation of the world was commonly accepted, we have to be grateful that Cosmas’ preoccupation with this issue led him to write his Topography, including all the precious details of his travels. Even if Cosmas’ book was by design more a work of theological cosmology than geography, we may be grateful that he felt obliged to illustrate his argument with his geographical knowledge and his practical experience in commerce.
This anonymous person, bearing a banal Christian name that was not his own and obsessed by representations of the world (kosmos), could hardly have been more different from the austere merchant-captain who put together the Periplus of the Red Sea some five centuries before. For Cosmas Indicopleustes, Adulis was no more than the place where an inscribed throne and an adjacent inscribed stele happened to pique the curiosity of an Ethiopian negus who happened to be planning to launch an overseas war. The fact that this negus was himself a Christian doubtless moved Cosmas to comply with his request for copies of the inscriptions. On the other hand, he showed no curiosity whatever about the provocations for that Ethiopian war or indeed about the Axumite kingdom itself.
By contrast, the author of the Periplus was clearly attentive to the culture he encountered and delivered a precious, if laconic account of it. His observations serve to illuminate the chronological space between the two inscriptions on the Adulis throne. The author’s own world was the Roman Empire of the Julio-Claudians, later than the Hellenistic Ptolemies and earlier than the unidentified pagan king at Axum. Adulis, according to the Periplus, was a legally recognized trading center, an emporion nomimon (legal emporium), on a deep bay extending southwards for about 200 stades (ca. 20 nautical miles) from the open sea. As we have already seen, this is the Gulf of Zula, the former Annesley Bay. What exactly a “legal emporium” might have been has long been subject to debate, but since the author applies the phrase to only three trading centers out of the thirty-seven that are named in the manual the odds are, as Lionel Casson has argued, that it was a place in which trade was allowed and regulated under the authority of the local ruler and was not simply an open souk or bazaar regulated by some kind of international law.17 In other words, it was neither a market town legally established as such by the Roman government, nor was it a place that was simply “law-abiding,” which would have presumably applied just as well to the thirty-four other ports.
The author of the Periplus states clearly that the metropolis of Adulis was Axum, which he calls Axômitês. This certainly implies that what went on there was subject to the authority of its metropolis. Ivory was the chief commodity in Adulis’ trade, and the Periplus reports that all the ivory “from beyond the Nile” came into Axum and was transported from there to Adulis. The writer goes on to add, “The mass of elephants and rhinoceroses that are slaughtered all inhabit the upland regions, although on rare occasions they are also seen along the shore around Adulis itself.”18 It is therefore not surprising that, as we shall see, the earlier inscription on the site, dating from the reign of Ptolemy III in the third century BC, explicitly refers to elephants in the area.
The vast territory of East Africa, which would appear to have included modern Ethiopia, Eritrea, Djibouti, and Somalia, had been under the rule of the king Zoskales, whose knowledge of the Greek language had so much impressed the author of the Periplus. Hence it is not unlikely that this was the ruler who controlled the trade at the “legal emporium” of Adulis in the first century AD because the Periplus describes him as “fussy about his possessions and always enlarging them.” But, for all that, he was said to have been otherwise a fine person and, of course, steeped in Greek culture. The Hellenic character of this local monarchy in the Julio-Claudian age of the Roman Empire may well have been rooted in prior contact with the Ptolemies, to which the earlier Adulis inscription bears witness. It certainly underlies and explains the continuing use of Greek in the region across the five centuries between the Periplus and Cosmas.
Even as the Ethiopic language began to be used in its classical form of Ge‘ez, the rulers continued to use Greek to advertise their exploits on the inscriptions they set up alongside parallel texts in Ethiopic. But in the centuries down to Cosmas’ day the Ethiopic texts were, on occasion, also inscribed in the alien script of South Arabian (Sabaic), which had to be read from right to left rather than, as in Ethiopic, from left to right. Curiously, it was Greek that served as the link to Ethiopia’s more distant past while, at the same time, it provided access to the lingua franca of the entire eastern Mediterranean world. Even on the royal coinage of Axum Greek normally appeared together with Ethiopic.
Cosmas’ transcription of the Ptolemaic stele therefore carries us back to an era of Greek and ivory, long before the notices of Juba II of Mauretania that were subsequently picked up by the elder Pliny or the visit of the author of the Periplus to the port city of Adulis. If Zoskales’ realm was as extensive as it seems to have been, the Ethiopian component, with Axum as the metropolis of Adulis, must have been only part of a much larger territory. The rise of the Axumite kings of Ethiopia and their expansion across the Red Sea into the Arabian peninsula would have altered significantly the distribution of power in the lands of East Africa to the south, where Djibouti and Somalia are today. But, as the Ethiopians enlarged their conquests in Africa, they looked increasingly to the north, as well as to the south, and they had good reasons for doing so. A vibrant Nubian civilization centered at Meroë in the territory of the middle Nile posed a major threat to Ethiopia’s northern frontier and potentially controlled both the sources of the Nile and access to Egypt, not to mention the ivory that the Periplus trader saw at Adulis. This region was destined to threaten and provoke the Ethiopians for several centuries until they brought an end to the Meroitic kingdom.
But the Ptolemaic inscription that Cosmas found and transcribed, although clearly connected in some way with his later text, took his readers much farther back in time than either he or his first readers could possibly have imagined. It evoked a vanished world, the Horn of Africa in the Hellenistic Age, several centuries before Zoskales.