Before going on explore these topics in more depth, it is worth briefly reflecting on the challenges faced by those who wish to study the historical connections between these regions. These challenges include, not only the availability of evidence, but how it might be interpreted. This is particularly relevant when it comes to issues of identify, agency and whether material evidence can be viewed as proxies for wider socio-economic or cultural phenomena.

In terms of the nature of the source material we have available to work with, it is important to note that the accounts of many Hellenistic era authors who wrote about Alexander, as well as those that discuss cross-cultural interaction in the Red Sea and western Indian Ocean (like Agatharchides), do not survive in their own right but often appear as fragments in later Roman Imperial era and Late Antique texts. Similarly a lot of South Asian poetic, philosophical, and political traditions (including various Buddhist cannons), sometimes based on much older (oral) accounts, often do not appear in written (manuscript) form until around the first millennium CE. The details these sources provide are useful from many different

vantage points, but they can convey information which is variously contemporary, anachronistic, a fossilisation of earlier traditions and potentially slanted by the diverse interests and prejudices of the authors (genre, socio-cultural biases, etc). Besides literary sources, increasing archaeological work at a range of international sites has greatly enhanced our understanding of connectivity across the ancient Indian Ocean in recent decades. But challenges remain for a variety of reasons, including accessing material. For example, at the important site of Bhambore, Piachentini Forani has noted the difficulties of excavating below the Indo-Scythians and Indo-Parthian layers due to the water table.

Even when material can be unearthed from different excavations sites, its interpretation may prove fraught. For instance, even if we can identify “foreign” material culture (that is items produced or originating from another region) in specific contexts, it does not necessarily tell us who moved this material (multiple historical actors may have been involved), how it was achieved (i.e., the methods and routes) and why (i.e., what motives lay behind it). The old adage that pots do not equal people remains relevant. For example, Tomber identified the presence of Torpedo jars at the site of Koteshwar in the Indus Delta, a ware previously interpreted as coming from the Roman dominated Mediterranean, but in fact likely to have been produced in Parthian/Sassanian territory. Its presence is indicative of the links between the Persian Gulf region and the Indus Delta, but it isolation we cannot make assumptions about who was responsible for its transit in this individual instance, or what its transit might represent—an object of trade or an item of personal consumption for a merchant or sailor, or perhaps something else entirely?

The collection of 400 odd objects of international provenance in the so-called Begram hoard is a good example of controversies of using material culture as proxy evidence for wider phenomena. Over the decades there has been much debate about whether the objects represent royally acquired treasures or a merchant’s storehouse, and whether they are indicative of cosmopolitan consumption with the Kushan Empire or a reflection of a supposed “middleman” role in the “Indus” branch of the Silk Road. Of relevance for us, is Cambon’s analysis of the copious glassware from the site. An Egyptian origin for these pieces is argued on the basis of chemical analysis, perhaps giving credence to the idea that the Mediterranean objects, at least,were transported via a port like the Barbarikon at the mouth of the Indus, and then taken up country through various networks to Begram.

A further challenge is the legacy of Romano-centrism, which was a prominent feature in much (western) scholarship from the late nineteenth and twentieth centuries, and which sometimes still features in more recent publications. This often involved the reductive and mistaken assumption that it was primarily the “Romans” who were proactively responsible for shipping goods across the western Indian Ocean, with other communities supposedly being more passively or reactively engaged. Of relevance for us, is the fact that some scholars have questioned whether there is evidence for Arabian and South Asian cross-oceanic voyaging prior to the early centuries CE, while others have assumed that Arabian and South Asian sailors and traders confined themselves to tramping in lower value, bulk commodities, with Roman ships engaging in cross-oceanic voyaging and the transit of “luxuries”. These highly questionable assumptions have partly been advanced on the basis of the limited evidence for Arabian and South Asian shipping traditions prior to the first millennium CE, and the notion that Graeco-Roman ships were larger and better suited to deal with the sometime ferocious monsoon winds.Ksạtrapas). As we shall now see, we need not imagine that there was an absence of technological or nautical competence in the region prior to the early centuries CE.

However, these assumptions are erroneous. As Whitewright has observed, the sewn-plank shipping tradition common in Arabia and South Asia was actually very well-suited to both coasting and cross-oceanic voyaging, was sufficiently flexible to deal with rough conditions, and could be fairly easily repaired. It is true that epigraphic and pictographic evidence for South Asian shipping technology and sailing practices are more clearly discernible in the early to mid-first millennium CE, such as the evidence from the Hoq cave on Socotra, depictions at Ajanta (paintings and a relief at Aurangabad) in the western Deccan, and images on Satavahana coins. Of additional interest here is the epigraph recorded in the cave by an Ajitivarman who refers to himself as a Śaka. Similarly, two other individuals designate themselves as ksạtrapas (which could indicate that they derived from the territory of the Śaka

In the immediate centuries following the death of Alexander, it is clear that trade links between the Indus Delta, Persian Gulf and Red Sea regions continued to grow. Indeed, in spite of the scepticism of some scholars (critiqued above) about the existence of sailing traditions in this period, it is quite clear from our sources that Alexander’s admiral Nearchus was able to find local pilots and sailors while voyaging along the coast of Gedrosia (and the Persian Gulf). Also, the polymath Posidonius (ca. early first century BCE) provides us with an account about a shipwrecked Indian sailor on the Red Sea coast (discussed further below), a claim that further points to pre-existing South Asian nautical knowledge. Moreover, the availability of goods from Arabia, South Asia and possibly even the Far East in Ptolemaic Egypt and other parts of the Hellenistic Mediterranean seem to indirectly demonstrate the existence of exchange networks. But, for the sake of brevity, I will not dissect the scattered pieces of material, papyrological, epigraphic and literary testimony that hint at this Hellenistic period exchange.

Instead, I will focus my attention on the different narratives about the how Graeco-Egyptian mariners became increasingly involved in cross-oceanic voyaging. There are two broad traditions surrounding these developments in Classical literature, one connected to the individual called Eudoxos of Cyzicus, the other to a possibly invented figure called Hippalos. In the Eudoxus narrative, we are told by Posidonius that a shipwrecked Indian sailor was discovered by Ptolemaic coastal guards. Having been taught Greek in Alexandria, the sailor promised to help navigate an expeditionary party to India, of which Eudoxus was a member. An initial voyage took place under Ptolemy VIII, and then a subsequent expedition was sent out by Cleopatra III—placing both events in the last two decades of the second century BCE. The Hippalos narrative is briefly relayed by the author of the Periplus Maris Erythraei. The author makes the following statement: ‘[t]he helmsman Hippalos, having observed the position of the emporia and the characteristics of the sea, was the first to voyage over open sea.’ This (invented) figure cannot be chronologically placed, beyond the fact that the myth surrounding him obviously pre-dates the mid-first century CE, when the author of the Periplus was active. Pliny also refers to a Hip(p)alus wind which merchants would use to sail from Syagrus on the promontory of Arabia to Patala.

Much debate has surrounded whether Eudoxus or Hippalos were real Columbus-esque figures, or if we are really looking at long-termer processes of developing navigational techniques and knowledge of the western shoreline of South Asia; an increase in knowledge that would allow for confidence that one would reach their destination after sailing the open ocean. The latter idea is more plausible to my mind. But of primary interest for out purposes is how these processes developed and the importance of intermediary hubs in this period. One such hub was Socotra. In the second century BCE, Agatharchides comments that the Fortunate Islands, that is Socotra, acted as a meeting point for merchants from Patala, Persia, and Carmania; likely also merchants operating from Egypt if we can accept Cosmas’s much later comments (sixth century CE) about the Ptolemies sending colonist to the island. The southern coast of Arabia, and particularly the site of Aden, similarly acted as an intermediary hub. This settlement, called Eudaimon Arabia in the Periplus, is said to have received cargos from those coming from Egypt and India, but that their merchants did not travel further. No explicit date is provided for this site’s heyday, beyond a reference it its former prosperity prior to its sack by Kaīsar (Caesar). All kinds of arguments have raged about how to interpret this claim. The salient one for our purposes is it had led some to subscribe to a dubious view that an Arab monopoly existed until the Augustan period when this was broken, and a flood gate of direct trade opened.

In contrast to this idea, Pliny the Elder presents a more progressive picture. He comments that in the period after the voyage of Nearchus, traders operating from Egypt would set out from Cape Syagros in order to cross over to Patale, using the favonius or Hippalus wind (that is, a westerly wind). Subsequently, a shorter and safer crossing was made from Cape Syagros to Sigerus, a location that could be Gujarat or perhaps the western coast of India. Pliny finally goes into a detailed account of how merchants in his own time would travel from Alexandria and sail to emporia like Muziris and Nelcynda in southwest India. Pliny does not give precise chronological parameters for these apparent “stages” of navigation, beyond the fact that it happened after the time of Alexander. But I would argue from documentary evidence, notably the appearance of Ptolemaic officials with oversight for Eythraean and Indian seas, the account about Eudoxus, and even Strabo’s acerbic comments about Augustan era trading ventures to India being on a larger scale than seen under the late Ptolemies, that cross-oceanic voyaging was already taking place by the late Ptolemaic period.

We do not need to imagine from Pliny’s account that all trade links between Roman Egypt and South Asia centred on the Malabar coast by his time (24-79 CE). It is true that there has been some debate surrounding the relative intensity of Red Sea links with the northern and southern subcontinent, based on finds from the Red Sea ports and issues like the comparative dearth of Roman coins in the northwest. But the testimony of the Periplus makes it undeniably clear that trade continued between Egypt and the Indus Delta. Indeed, it is explicitly stated that those who set out from Egypt for Skythia and the port of Barbarikon would do so in July. The author gives a fairly vivid description of the seaboard of Skythia—that is roughly modern Balochistan—leading up to the river Sinthos (Σίνθος). Refence is made to the colouration of the sea caused by the rivers discharge, the appearance of snakes called graai, and details about the seven river mouths, of which only the middle one is navigable.

Vessels anchored at Barbarikon would have their cargos taken upriver to the ‘metropolis of Skythia’ called Minnagar; to which we are told the ‘Parthians’, are constantly chasing each other off the throne. At Minnagar a market existed for a range of textiles, a gemstone called ‘chrysolithon’, which is likely peridot from St John’s Island (Zabargad Island) in the Red Sea, silverware, money, coral, glassware, wine, storax, and frankincense. We the exception of frankincense, most of these goods likely come from Roman territory. In return one could pick up certain plant products, like bdellium and nard, Seric skins, turquoise, lapis lazuli, cloth and indigo. Some of the plant-based products may have come from the Indus area or Skythia, while turquoise and lapis lazuli will have come from further north. Seric skins, also mentioned by Pliny, presumably came from either the Central Asia or the Far East.
Other evidence may also point to direct or indirect trade links. This includes red coral beads from at a stupa at Mirpur Khas and five surface finds of aurei of Antoninus Pius from Manikyala (near Rewat). In the other direction, it has been argued that a Gallo-Roman emerald from Miribel may derive from the Swat-Mingora area, based on oxygen isotope analysis. A comment by Pliny about the availability of smaragdi (green stones) from Scythia may lend credence to this hypothesis. These scattered pieces of evidence and the testimony of the Periplus all suggest that trade took place, though it is difficult to determine its intensity.

It is important to briefly note that links between the Indus Delta and the Persian Gulf region remained strong, and probably were more significant than those with the Red Sea area. A point apparent from the testimony of the Periplus. Moreover, epigraphic and pictorial evidence from Palmyra allude to voyages to “Skythia”. Certainly, a rich array of material finds have been recovered from Palmyrene tombs, including pearls, agate/carnelian beads, and perhaps of most significance silks—possibly coming via the Indus branch of the Silk Road? Jean-François Salles has even suggested that the appearance of some Mediterranean material at sites like ed-Dur in the Persian Gulf could have come indirectly via trade with ports such as the Barbarikon.

Besides trade connections, there is some evidence for diplomatic contact between the Augustan court and envoys from Scythia and “India”. These accounts also stress the importance of the memory of Alexander. The first incident is related by Paulus Orosius, who claims Indian and Scythian ambassadors met the Emperor Augustus when he was in Tarraco, so around 25 BCE. A second subsequent embassy arrived at Antioch and then went to Samos, placing this event around 20 BCE. A comment by Cassius Dio may imply the second embassy was a follow up from the previous one. Much controversy has surrounded the goals of the second embassy and the identity of the ruler that sent it. The polities of the Śakas, Indo-Parthians and Satavahanas have all been suggested. I suspect, that either the Indo-Parthians or Satavahanas may be the most plausible candidates for the polity from which the “Indian” embassy derived, whereas when our Graeco-Roman sources speak of embassies from Scythia/ Σκυθια this could be an allusion to the Śakas. As we have already discussed, the seaboard of Skythia described in the Periplus seems to refer to the area along the coast of modern Balochistan, going up to the river Sinthos/Σίνθος; likely the same Skythia alluded to in our Palmyrene inscriptions. In any case, the reference to Zarmanochegas, or Zamaros, who is said to come from Bargosa, that is Barygaza, at the very least points to north-western polity of the “Indian” embassy.

I will leave aside the different suggestions surrounding the goals of the embassy or embassies, and the various (ultimately speculative) notions that the envoys were sent for either commercial reasons or to establish some kind of military alliance. Instead, I focus on the allusions to Alexander that are made in connection with the embassies that arrived in 25 BCE and 20 BCE. In the first episode (25 BCE), Paulus Orosius makes a direct comparison by paralleling Augustus’ receiving of Indians and Scythians at the far western city of Tarraco, to the almost certainly apocryphal claim that Alexander received envoys from Spain and Gaul when he had returned to Babylon. The ambassadors are also supposed to have extoled the glories of Alexander to the Augustus. Strabo’s reference to Pandion as another “Porus”, and the parallels between Zarmanochegas and Calanus, the former who is said to have burnt himself alive at Athens, would require little imaginative effort on the part of contemporaries.

There are two potential interpretations that one can make. The first is to view these comparisons to Alexander as largely fabrications by our Graeco-Roman authors, who were either consciously trying to glorify Augustus or picking up on Augustan messaging (perhaps reflected in Nicolaus of Damascus’ original account). The second possibility is that the envoys were highly conscious of the significant of Alexander’s legacy and tried to exploit this cultural memory with the aim of gaining a more favourable reception. We cannot be definitive about either possibility.

This usefully brings us to the final point of consideration for this paper. Did commercial contacts with the Indus lead to the spread of new knowledge about the region in the Roman Imperial period? How easily could ancient authors escape from the shadow of Alexander’s legacy and the “scientific” reports of those that had accompanied his expedition? The simple answer to the latter question is, it depends where you look. Unsurprisingly Alexander’s campaign down the Indus, the march across the Gedrosia desert and the voyage of Nearchus loom large in the historical, biographical, and ethnographic discourse. However, this is not to say that more contemporary information failed to intrude. Eggermont, for example, argues that the impression given by Curtius Rufus (first century CE?) that Alexander only sailed down one branch of the Indus Delta could reflect an anachronism indicative of more contemporary trade with the Barbarikon.

But it is in the geographic, topographic, and encyclopedic literature where we can most readily see interplay between the intellectual tradition that can be traced back to Alexander’s time and the following generations, versus newer geographic and topographical information deriving from merchant sources. Strabo (ca. end of the first century BCE to very beginning of the first century CE) nicely exemplifies these potential tensions. At one point in his Geography he apologizes for having to rely on the accounts of those who accompanied Alexander and on later authors like Apollodoros of Artemita. And yet, when more contemporary information was available from merchant reports, he expresses disdain. For example, he dismisses the testimony of those who sail from Egypt to India and the Ganges, since they are merely private citizens and of no use regarding the histories of the places they have seen.

A little later, Pliny appears willing to draw upon the accounts of Alexandrian merchants and financiers when discussing trade routes, as has been argued for his description of the trade routes to India and his allusions to the Alexandrian calendar (alongside the Julian calendar). Claudius Ptolemy (second century CE) also makes use of topographic data likely derived from merchants. For example, it has been argued that his reference to the sea known either as Hippodos or Hippalon (around the coast of Somalia), adjoining the Indian Ocean on its eastern side, implies a reliance on an oral report rather than written source for this information. But even if both authors did make use of merchant testimony, they remained heavily influenced by Hellenistic era literature. Notably, one can see this in Ptolemy’s orientation of South Asia along a west-east axis, as opposed to the more accurate north-south orientation given in the Periplus; potentially hinting at resistance to move away from more established “intellectual” models.

By comparison, a striking feature of the Periplus is the willingness to incorporate information deriving from Arabian and South Asian merchants, among others, alongside the author’s own personal experience. This includes topographic details and ethnonyms that can often be absent from the “scientific tradition” of Graeco-Roman literature. Of particular interest for us is the use of the word Σίνθος to allude to the Indus River, in contrast to the more typical Ἰνδός (or in Latin Indus); though Ptolemy does also employ the former term. The description of the Indus in the Periplus also stands out. As already noted, the author refers to seven river mouths, of which only the middle one is navigable. In contrast, other writers like Strabo, following Hellenistic tradition, tend to emphasize the perspective of those travelling down to the shore, with the Indus emptying into the sea by two mouths, and encompassing the so-called island of Patalenê.

Interestingly, Ptolemy also breaks the mold by mentioning the seven branches of the Indus (in a similar vein to the Periplus, though offering more detail on the names of the branches). Eggermont assumed that he was mistakenly including other rivers like the Hab River as one of the branches, and that this idea derived from a dubious Ionian tradition of drawing parallels with geographic equivalents; in this case assuming that because the Nile had seven mouths so must the Indus. However, I would argue that he may have been borrowing from merchant sources, where, from a coastal perspective, it might look like other streams or rivers were outlets of the Indus.

Indeed, a major distinction between the intellectual tradition and the merchant tradition is one of vantage point. For those interested the natural features, flora, fauna, and ethnography in Central and Southern Asia, the accounts of “big men”, like those who accompanied Alexander, was clearly of paramount utility. Whereas, for the author of the Periplus topographic features and natural phenomena were primarily of interest when they offered useful reference points by which a sailor could locate themselves, or when it was necessary to be mindful of dangerous sailing conditions. This was largely from a coastal vantage point, as is true of many earlier periploi. References to political and security circumstances were also made, as well as the goods desired by rulers and courtiers. These differing perspectives may be even more pronounced if we accept Bukharin’s argument that the author of the Periplus (our main written testimony for the “merchant tradition”) was largely indifferent to the “scientific tradition” which developed post-Alexander. He asserts that the author was more engaged with “popular” Archaic and Classical ideas about regions like Arabia and South Asia, appearing to be uninterested Hellenistic era geographic and ethnographic ideas.

The juxtaposition of these two perspectives can nicely be encapsulated in Pliny discussion of Patala and the region of Patalis. When talking about the Indus River region, Pliny includes information on distances, tribes of people, mythic traditions linked to Hercules and Father Liber, cities founded by Alexander, and geographic features, such as the apparent triangular shape of island of Patalene—ideas which are tied to Alexander’s campaigns and their aftermath. In contrast, when talking about the development of voyages from Egypt to South Asia, of which trade with Patala was significant from an early stage, merchant sources seem to have been utilized. The latter did not superseded the former as a source of information, but clearly it was sometimes used as a supplement.

The Indus Delta region, known in Graeco-Roman sources as Patalene, had long been connected into maritime networks in the western Indian Ocean, though the nature and intensity of these networks shifted over time. This was already a reality when Alexander’s army arrived, and continued to be so after his departure. This is not to suggest that Alexander’s presence had no impact, as the establishment of a naval station and fortifications at Patala, and the foundation of Alexander’s Harbour, attest. But it is difficult to appreciate the direct, long-term practical significance of these actions from the surviving evidence. Nevertheless, there is little doubt that the actions of Alexander and the reports of those that accompanied him would loom large in the imaginations of later Graeco-Roman authors. This continued to be the case even in the early to mid-first millennium CE, a time of much greater connectivity (direct and indirect) between the Mediterranean world and the Indus region.

In light of this, the aim of this paper has been to consider the following question: did new information deriving from merchants and travellers ever supplant the information from earlier Alexander-era authors and Hellenistic intellectuals, which continued to dominate Imperial era Indography? The answer seems to be largely no. However, it does appear that in some geographic and encyclopaedic literature there was a willingness to make use of information from the “merchant tradition” (reports deriving from merchants and travellers) as a supplement. New information from merchant accounts was a useful for garnering topographic data, toponyms and sailing schedules. However, the ethnographic and political information potentially gleaned by these merchants and travellers does not seem to have filtered into the writings of many Graeco-Roman intellectuals. In this respect, the legacy of Alexander had too firm a grip on the imagination.