Roman Presence in India

Muziris, Arikamedu, and Aalgankulam - Windows to Roman Trade

In the previous post, we discussed Indians who ventured abroad and sometimes settled in Red Sea ports and Egypt for short periods while engaged in Roman–Indian trade. While we uncovered some shipping routes, ancient ports, and evidence of their presence, we had planned to examine evidence of Roman presence in South India.

A quick recap of the situation in India at the turn of the Gregorian calendar. South India, more specifically, Tamilakam, was in the midst of the Sangam age (300BCE-300CE), and the Cheras, Pandyas, and Cholas were emerging as established kingdoms. The Chera empire ruled over the Malabar and Kongu areas of South India, stretching across both sides of the Western Ghats, connected only through the Palghat gap, and now split into the modern-day states of Kerala and Tamil Nadu. Tamil was the primary language during this period. The spoken languages were Tamil dialects; the region was known as Tamilakam and was gradually divided into the Chera, Chola, and Pandya dominions. So, whether the trader was from a Malabar port belonging to the Chera kingdom, such as Muziris, or from a port on the east coast controlled by the Cholas or Pandyas, they were all broadly classified as Tamils.

Eastern Chola and Pandyan traders sailed back and forth between the west-coast ports, using local boats (Thonis) or traveling along the ancient (peruvazhis) trade highways that connected to Cheranad through the Palghat gap. There are mentions of Indian, Arab and Roman ships sailing to the East Coast, a topic we will check out as we progress further.

Gurukkal had mentioned: The presumption that the Tamil chieftains had a leading role in overseas trade is hardly realistic. Even the presumption that the chieftain had shipped his goods only up to the Red Sea coast and had depended on intermediaries for the remaining jobs is unbelievable for the same reasons. Continuing, he added: The Mediterranean relics recently unearthed at Pattanam, as in the case of Arikamedu, prompt us to presume that contemporary Indian ports were, in fact, foreign merchant camps and their bazaars." Run under the financial support and military protection of the Roman Emperor all along the highly inhospitable camel/caravan routes (on the last leg, that is) for the safe transport of the precious goods, it was virtually Roman trade, no less, no more. Was he right? Let’s first check the goings-on at ancient Tamil ports and purported Eastern Tamilakam settlements, and then check the present status at Muziris.

Casson notes that after building larger (150’-180’ long) and stronger-hulled ships, Graeco-Roman sailors mastered the Arabian seas, achieving shorter, safer sailing times than Arabs on their dhows. He explains that the routes originated at Myros Hormos and Berenike, followed the monsoon winds, and terminated at Barygaza (Bharuch) and Muziris in India. He adds that goods were then transferred to Indian vessels for delivery to East Coast ports. Thus material disgorged at, say, Muziris reached eastern ports in multiple shipments via land-hugging local ships (open sangras/thonis and other craft).

Exports comprised silks, cottons, pepper, nard (aromatic oils from N India), gingers, ivory, gemstones (beryl, etc.), and some iron ore and semi-finished goods. The meager imports, after gold and silver coins used to pay for exports, were mainly for the consumption of foreigners settled in inland and port locales, as well as the richer clientele in Madurai and Puhar, and included raw glass, copper, tin, lead, wine, olive oil, etc., once a year. Some evidence indicates that glass was shipped from Kanchi to China, and that silk from China arrived at these ports for transfer to Muziris and thence to the West. Interestingly, Roman coins were also imported for use in local trade after being slashed or counter-stamped to supplement locally minted coins.

HP Ray, however, disagrees. She doubts that Graeco-Roman vessels were used on the sea routes to India and emphasizes the superior quality of Indian and Arabian ships. Furthermore, she argues that the archaeological discoveries at Arikamedu don’t amount to evidence of a Roman trade colony there. Ball concurs, adding that Graeco-Roman ships were incapable of sailing the Indian Ocean because their square sails were too primitive to cope with local wind patterns, and that the few western traders who wanted to trade with India would use Arabian or Indian vessels. Fauconnier debates these arguments and clarifies that Graeco-Roman ships were indeed active in the trade networks of the Red Sea and the Indian Ocean. He feels that native craft with sewn hulls were quite adequate, intermingled with Roman vessels, and traveled afar, as we saw in the previous article, which discussed evidence of Indian settlements at Berenike and Quseir al-Qudaim.

In any case, there was robust maritime traffic between the Red Sea and West Indian ports. Regarding the identity of local Indian sailors on this route, one could conclude that they were Gujaratis as well as the forerunners of the Marakkars.

Yavanas

Many Sangam-era poems and literary works allude to fair-skinned, blue-eyed, tall, and fierce-looking foreigners, known as Yavanas. The term derives from Ionia, a region in Greece, and it is believed that over time, Ionia became Yonaa and Yavanaa. By the first half of the first millennium CE, the term Yavana denoted a foreigner from the west, encompassing Graeco-Romans, Arabs, Persians, and others. They were not only merchants but also skilled sailors, glass, metal, and wood craftsmen, mercenaries, and guards. The Yavanas are described as carefree, enjoying life, drinking, walking along the streets at night, and living in well-made houses.

The Ports

The ports used by Roman traders were Muziris (near Kodungallur) and Tyndis (Tondi-Ponnani or Pantalayani), controlled by the Cheras. Road transport was through Chera country, through the Palghat Pass (touching Coimbatore, Erode, Salem, and Karur). Nelcynda (Niranam, or around Kottayam) and Bakare (probably Purakkad), both under Pandyan control, served as hubs for transshipment onto local vessels to Madurai consumers and Eastern Roman settlements. Other Pandya/Chola ports, such as Alagankulam (Ramanathapuram) and Arikamedu (Virampatnam or Poduke), were perhaps Indo-Roman stations at the other end of the local trade route, suggesting strong trade ties. Northerly ports, such as Puhar and Kanchi, as well as Manthai (Mannar – Ceylon) in the South, are also frequently mentioned in texts.

Roman Presence

Kanakalatha Mukund opines that Arikamedu and Alagankulam, near Rameswaram, were ports where Roman ships traded. While the archaeological finds at Arikamedu are well known, the beautiful graffito of a Roman ship on a piece of pottery at Alagankulam is less well known and further proof that Roman ships regularly visited the eastern ports. These trading ports, Arikamedu, Alagankulam, and Puhar, as well as the inland capital Madurai, had settled Roman colonies.

Let’s now turn to the East Coast to assess these conclusions. While many excavations in the Tamilakam have uncovered Roman remains, a couple of sites, such as Arikamedu and Alagankulam, provide evidence of settlements. Poems mention beachfront Yavana quarters in Puhar, and both barter and currency-based trade were prevalent. Soon enough, foreigners were found traveling and trading in inland cities like Madurai. They were also employed as guards, perhaps as mercenaries, and as metalworkers or carpenters. Puhar, like Muziris, faded over time and, together with Madurai, is well described in the ancient works Silapathikaram and Manimekalai.

One reason for a possible Roman presence on the East coast was the problems with the ancient Silk Road, after which silk was transported by sea to Eastern Indian ports and thence via Western ports.

Kanchi

Kanchipuram, a major capital of the Pallava dynasty and an inland trade center, had significant commercial, non-colonial interactions with the Roman Empire (1st century BCE – 3rd/4th century CE). Romans visited this hub for spices and textiles, and finds of Mediterranean amphorae mark evidence of their presence.

Mahabalipuram

A few Roman coins of Theodosius (4th century AD) found in the region suggest that Mahabalipuram also had trade contact with the Roman world around the Christian era. Amphorae shards have also been found in the area. However, nothing in the findings thus far suggests there was a settlement there.

Puhar - Kaveripattinam

Puhar (Kaveripattinam) was a thriving Chola port city with a strong Roman presence, driven by maritime trade. Archaeological finds, including Roman pottery and coins, indicate commerce and possibly the presence of residential buildings and warehouses in the harbor area.

Arikamedu

Arikamedu – The site of Arikamedu, also known as Virampatnam and identified with the ancient Poduke mentioned in the Periplus Maris Erythraei, is located on the Ariyankuppam river, about 3 km south of Pondicherry, in a lagoon formed by the Ariyankuppam and the Gingee rivers. The sheltered location of Arikamedu was undoubtedly a crucial factor in its founding, since the southeastern coast of India is almost devoid of natural harbors. According to the PME and PNH, it served as a local harbor, receiving goods from other parts of India. Wheeler’s excavations at Arikamedu in 1945 were significant in identifying that the trading post was active during the first two centuries AD. Following the ASI and Wheeler excavations, Jean-Marie Casal investigated the site in 1947-8.

Bagley maintains that the discovery of amphorae, Arretine ware, glass fragments, and other artifacts indicates that the site functioned as a maritime trading center long before Augustus. In her paper ‘Arikamedu rediscovered’, she establishes that the site is unique as the only one in South India to have yielded evidence of continuous trade with the West. It had a robust bead-making industry, exporting to distant customers, including Romans. She then connects the Tamil Brahmi inscriptions at Quseir al-Qadim to those at Arikamedu, suggesting that this ancient port also flourished as an Indo-Roman emporium. She points to evidence that Roman garum (fish sauce), wine, olive oil, and a handful of Italian lamps prove Roman presence, or even residence, at Arikamedu. 

Van Arede and Zampierin argue that this inference rests on scant evidence and that, based on excavations to date, the evidence is far from sufficient to support the view of Arikamedu as a Roman trade emporium, or even as home to a substantial settlement of Roman traders. Other historians conclude that the samples may have come from a single Roman visit and that the site's decline may have been due to rising sea levels. After reviewing the excavations and reports, we can conclude that Arikamedu does not presently provide evidence that it was a trading outpost.

Alagankulam

Archaeological surveys in Alangankulam have identified Roman pottery and coins from the late fourth and fifth centuries AD. Alangankulam was the official harbor of the Pandya Kingdom in Madurai. Key evidence of this presence includes Roman coins of Emperor Valentinian II (c. 375 CE), Roman amphorae (storage jars), rouletted pottery, and Roman-style glass unguentaria (perfume bottles). The site lies at the confluence of the Vaigai River and the Bay of Bengal in Ramanathapuram district. The Archaeological Department found a brick structure and evidence of a shell-bangle-making industry, an iron-smelting workshop, terracotta ring wells, Roman coins, pottery, and a cross-inscribed seal. They also discovered a structure made of burnt bricks, indicating that people living at this site used such bricks during the Sangam period in Tamil Nadu. Perhaps it was a silk emporium (together with Puhar and Kaveripatnam).

One of the most interesting finds has been a graffito of a three-masted Roman ship at Alagankulam. Casson compares the inscription with two well-known representations of typical large Roman ships of the period: one in a mosaic found at Ostia, Rome's port, and the other in a third-century CE mosaic from a house in Rome. The graffito portrays a three-masted sailing ship meant for long voyages and may have been just a depiction of one such ship the artist saw in an Indian port (not necessarily at Alagankulam). Thus, this graffito does not establish that this ship or one of its kind sailed to Alagankulam, for it would have been difficult for it to sail around Ceylon.

Keeladi

Keeladi, on the south bank of the Vaigai River, about 12 km east of Madurai, lies on the same ancient trade route from Madurai to Rameswaram. Beginning in 2013, KA Ramakrishna began excavations at Keeladi after many potsherds were discovered in the locality. Excavations revealed sophisticated dwellings, drainage systems, tanks and wells, brick walls, and other features. Archaeologists conjectured that some of the finds resembled dyeing vats, suggesting that Keeladi might have been a textile unit. Many of the features – open channels, ring wells, and tanks – resembled structures discovered at Arikamedu. From a trench at Keeladi, the bones of domesticated Arabian horses were retrieved. More than a hundred Brahmi inscriptions similar to those found at Arikamedu were found. The inscribed shards were dated to between the fifth and first centuries BCE.

Though the findings at Keeladi show no Roman connections, Keeladi was part of the Sangam-era complex around Madurai, which points to a robust consumer culture. However, it is an ancient site of importance, perhaps connected to Arikamedu and Alagankulam, and part of a larger trade complex, the extent of which will be determined after further excavations, which, it seems, are marred by a lot of political wranglings and whatnot – all beautifully explained in Sowmiya Ashok’s book.

Muziris & Nelcynda

Connected to Madurai is Vanchi, the alternate Chera capital near Kodungallur. This was where the Sangam heroine Kannagi fled from Madurai and stayed, and it is near Pattanam, purportedly the location of Muziris. We discussed the overland trade routes connecting West Coast ports via the Vaniyamkulam – Madurai peruvazhi, through the Palghat gap, and the many trade centers around Madurai, from which inland goods were transported to Tyndis, Muziris, and Nelcynda on bullock-cart caravans. There are hardly any mentions of Yavanas in Tyndis, but Sangam poems do mention Yavanas in Muziris.

The town of Muchiri and the neighboring region were under Chera control. The poems state that ships of the Yavanaas came through the Chulliam Periyaaru (the Periyar River). The golden prize (material wealth) brought by the ships was carried ashore by small 'backwater boats' called kazhittoni. Thus, Roman ships were stationed a short distance from the port, and the smaller boats that plied the backwaters carried the goods they brought. Documentary evidence, such as the Muziris papyrus, provides clues to a Roman trading settlement in Muziris, but because the entire topography of the area was altered over time and by natural disasters, the exact location of Muziris remains disputed, or at the very least, hotly debated.

We had previously discussed the location of Muziris and its activities in some detail, but let us take a fresh look from a settlement perspective. Though I have used the term colony previously, I need to define it – it is not a colony in the sense of people being colonized, but rather a quasi-settlement. We do not know for how long, but let us assume that it remained so until direct Roman trade with India petered off for various reasons. When the trade with Muziris began is unknown; however, a document discovered in 1985, the Muziris Papyrus, dates it to the 2nd century, by which time trade seems to have been well established. 

Pattanam excavations

Pattanam is located near Vadakkekara village, north of Paravur and south of the river Periyar. The Peutingerian Table lists a place called 'Patinae' near Muziris. The presence of amphorae, terra sigillata, and other ceramic and non-ceramic materials indicates that Yavana ships arrived at Pattanam. Perhaps it was merely a trading market. The layout of the brick structures suggests the entire town was planned, with streets oriented in a particular direction.

Cobb notes that excavations at Pattanam (Muziris) reveal its integration into broader Indian and international trade networks. The presence of early historic jars and Rouletted Ware demonstrates the port’s connection to Arikamedu and southeast India. At the same time, Yemenite, Parthian, and Sasanian ceramics attest to links with western Asia. Pattanam also had connections with the Mediterranean world, as evidenced by Roman amphora sherds, including those from the Bay of Naples, as well as sherds from oil and garum containers and Roman glassware. Botanical remains unearthed include pepper, cardamom, frankincense, and grape seeds.

Although there has been some discussion of the temple of Augustus, it has been dismissed because he was not a king worthy of worship. Most historians conclude that it was an Agastya temple in Cape Comorin.

Nelcynda & Becare

Schoff identified Nelkynda with Kottayam on the Meenachil River and Becare with Purakkad. Casson equated it with Niranam based on the distance mentioned in the texts. These locales have not been excavated, though it is noted that Nakkida, near Tiruvalla, may have been the Nelcynda of the past. This town was under the control of the Pandiyas, whose center was at Madurai, but they took control of the port for its commercial significance.

Bead trail

To make glass beads, one first needs to have access to glass. Arikamedu was a bead manufacturing site and different formulae were used to produce the opaque red glass and glass of other colors. Peter Francis, an expert on this topic, believes that Arikamedu pioneered production techniques that are still used today. He believes that its exports were more toward the East than the West. We discussed the road link between the East and West Coasts of Tamilakam many years ago; see the article on the Palghat Gap listed under references.

Conclusions

Sailing to the East Coast

Roman merchants were unlikely to have regularly sailed to the east coast of India because doing so carried a greater risk of missing the window (December to early January) for a safe return voyage within the year’s trading season. Western ships never ventured into the waters between Kanyakumari and Ceylon (Taprobane) because the sea was most turbulent and perhaps had tricky mudbanks. The shallow channels between the southern tip of India and Sri Lanka made it difficult for deep-hulled vessels to traverse. In contrast, the weather conditions made it problematic to sail around the southern tip of India until January, the period in which, according to Pliny, most merchants would have already set off for Egypt.

It was possible to travel overland via the Palghat Gap (a mountain pass) to conduct or oversee trade. Still, it is unlikely that Roman merchants undertook such an arduous journey after a difficult sailing trip to Muziris. On the other hand, why would Romans go to the East Coast? Possible reasons include piracy on the west coast, attested by PME and PNH; silting problems at Muziris; reduced costs at manufacturing centers and among craftsmen in the East; and flourishing, richer Pandyan and Chola markets.

In summary, there is strong evidence of a robust Roman-Indian trade, as R Gurukkal maintained, but little evidence of Roman settlements. True, there is better evidence of Arab Yavana (Jonkan) settlements all over the west coast of India, and we have thousands of their descendants amongst us, but hardly any blue-eyed Roman Yavanas. Without a doubt, many of these Romans were short-term residents who left an impact on poets' minds. They may have visited and even worked for brief periods in East Coast ports or towns as mercenaries, guards, craftsmen, etc., but there are no indications of any broader settlements as yet. I will, however, not be surprised if evidence of such a port town on the west coast is discovered one of these days.

The Muziris excavations had petered out for lack of permissions. As I drove along the Kodungallur road to Calicut and then to Cannanore, the one sight that troubled me no end was the ongoing excavations for a massive highway project. Huge earthmoving equipment dug and tossed out earth, and the workers were almost all from Assam, Bengal, or Bihar, with little supervision. Most of the work was near the coastline and the ancient port locales. God knows what got tossed around with the earth – pot shards, amphorae, and whatnot. You can take a guess. Anyway, let’s hope the Pattanam excavations restart and that some of these port towns are finally discovered. All these studies point to one fact: the presence of a more advanced culture in Tamilakam before and after the 1st century AD.

References
Rome & India – the Ancient Sea Trade – Ed Vimala Bagley, R D De Puma
Imperial Rome, Indian Ocean regions & Muziris – Ed K S Mathew
The Ancient Port of Arikamedu Vol 1 – Vimala Bagley
Graeco-Roman merchants in the Indian Ocean - Fauconnier Bram
Indo-Roman Pepper Trade and the Muziris Papyrus – Federico De Romanis
Arikamedu Reconsidered – Vimala Bagley
The Stone Bead Industry of Southern India – Peter Francis
A lot of pepper and a Little Garum – Marike Van Aerde, Daniele Zampierin
In Search of Muziris – Rajan Gurukkal, Dick Whittaker
The Chronology of Roman Trade in the Indian Ocean from Augustus to Early Third Century CE - Matthew Cobb
The Yavana Presence in Ancient India - Himanshu P. Ray
Aagankulam, an ancient Roman port city of Tamil Nadu - Ed TS Sridhar

Historic Alleys – Trade through the Palghat Gap

Maddy’s Ramblings – Early Indians in Egypt

Note:

1.      As you study these books, you will encounter two acronyms: PME and PNH. PME is Periplus Maris Erythraei, and PNH is Pliny the Elder's Naturalis Historia  

2.      An ancient trade highway connected the west and east coasts from Vaniyamkulam (Kerala) to Alagankulam, Arikamedu, Puhar, etc. via the Palakkad Gap. It was later called Rajakesari Peruvazhi. Whether it went all the way to Muziris way back in the past, is yet to be established (PAMA Idukki Exploration Project)