In what ways are the accounts of the Graeco-Romans and the Chinese helpful in reconstructing the social history of India? How far is their information corroborated by other contemporary sources? (2009)
Indigenous Indian textual tradition is overwhelmingly normative — the Dharmashastras prescribe how society ought to be ordered, not how it actually functioned — and that is exactly the gap outside observers, unburdened by prescriptive intent, can help fill. Graeco-Roman envoys, geographers and traders, and Chinese Buddhist pilgrims left descriptive, often incidental testimony on caste practice, slavery, trade and urban life, and religious institutions that indigenous sources rarely state so plainly. But because these were outsiders working with limited language access and their own preoccupations, the value of what they report depends entirely on how far independent Indian evidence — literature, inscriptions, archaeology — corroborates it.
Graeco-Roman Accounts: Court, Trade and Coast
- Megasthenes, the Seleucid ambassador at Chandragupta Maurya’s court, wrote the Indica — now lost but reconstructed through later citations in Arrian, Strabo, Diodorus and Pliny — describing a sevenfold occupational division of society (philosophers, cultivators, herdsmen, artisans, soldiers, overseers, councillors), the reported absence of slavery, and the reported absence of famine in the Mauryan realm.
- Pliny’s Natural History records the Indo-Roman trade route to India’s west coast and Rome’s drain of gold to pay for Indian luxury goods — a rare quantitative window onto the scale of India’s export economy that no Indian text of the period offers.
- The anonymous Periplus of the Erythraean Sea and Ptolemy’s Geography describe west- and east-coast ports such as Barygaza and Muziris and the goods moving through them, useful for reconstructing the social world of merchants and port towns that the Gangetic-focused Dharmashastras largely ignore.
- Herodotus, though he never visited India, preserves the earliest surviving foreign notice of the subcontinent — valuable mainly as a marker of how early Graeco-Indian contact began, rather than as social evidence in its own right.
Chinese Accounts: The Pilgrim’s Eye
- Fa-hien (early 5th century, Gupta period), in A Record of Buddhist Kingdoms, describes a relatively peaceful realm with light criminal punishment, and — significant for social history — the residential segregation of Chandalas outside towns, an outsider confirmation of practised untouchability that Brahmanical prescriptive texts state as a rule but rarely describe as lived reality.
- Hiuen Tsang (7th century, Harsha’s reign), in the Si-Yu-Ki, gives the most detailed foreign description of contemporary caste practice, monastic education at Nalanda, judicial procedure including trial by ordeal, and court life — composed only after he left India, which insulates it somewhat from royal pressure, though not from his own devotional bias toward a Buddhist-patronising king.
- I-tsing, slightly later, concentrated narrowly on monastic discipline and the Nalanda curriculum, useful chiefly for Buddhist institutional history rather than society at large.
Together, this material supplies exactly the texture indigenous prescriptive texts omit: how caste actually operated on the ground, what a Mauryan or Gupta port town looked like, and what an outsider thought worth noting about crime, punishment and religious life.
Corroboration, Contradiction, and the Limits of Outsider Testimony
The real test of this material is whether independent Indian evidence backs it up, and the record is mixed.
- Megasthenes’s claim that India had no slavery is directly contradicted by the Arthashastra’s detailed provisions on dasas and by later Dharmashastra references to bonded labour — most likely a confusion between Indian dasa-status and Greek chattel slavery, rather than evidence that bondage did not exist.
- His sevenfold division corresponds only loosely to the four-varna framework of the Dharmashastras, and not at all to the far finer jati stratification visible in inscriptions — a simplification typical of an observer working through translators, unfamiliar with the underlying social grammar.
- Fa-hien’s report on untouchability, by contrast, is well corroborated: Manusmriti’s rules confining Chandalas to habitation outside the village, and later epigraphic references to outcaste groups, align closely with what he describes.
- The Indo-Roman trade picture from Pliny and the Periplus is unusually well corroborated, and from more than one independent direction at once: excavations at Arikamedu have yielded Roman amphora sherds and rouletted ware, Roman gold and silver coin hoards cluster precisely in Tamil Nadu and Kerala, and Sangam poems such as the Purananuru and Akananuru independently describe Yavana traders, Yavana wine, and Yavana mercenary guards at Tamil courts. A foreign text, the archaeological record, and indigenous poetry thus converge on the same picture — about as strong a corroboration as ancient Indian social history ever gets.
- Hiuen Tsang’s account of Harsha’s court finds support in Banabhatta’s Harshacharita and in Harsha’s own Madhuban and Banskhera copper-plate inscriptions, and the scale of Nalanda is confirmed both by excavation at the site and by donative epigraphy recording later royal endowments to the monastery.
- But Hiuen Tsang’s claims of negligible taxation and an absence of forced labour sit awkwardly against contemporary evidence for vishti (forced labour) and multiple tax categories in Gupta and post-Gupta inscriptions and Smriti texts — a reminder that a pilgrim grateful to his royal patron had reason to idealise the reign he was describing.
As Romila Thapar has cautioned, a foreign account has to be read against its author’s vantage point, language competence and purpose before its content can be treated as social fact rather than impression.
“It is only when foreign notices can be checked against indigenous sources — inscriptional, literary or archaeological — that they acquire genuine evidentiary value for the historian.”
(A two-column table setting each foreign claim against its corroborating or contradicting indigenous source — Megasthenes/slavery vs Arthashastra, Periplus/trade vs Arikamedu and Sangam verse, Hiuen Tsang/taxation vs Gupta inscriptions — would make this argument visually sharper than prose alone.)
The corroboration exercise, in the end, does not discredit Graeco-Roman and Chinese accounts; it disciplines their use. Where independent strands of evidence converge, as with Indo-Roman trade or Gupta-period untouchability, these outsider testimonies become some of the most reliable social history available for periods that indigenous sources describe only obliquely. Where they stand alone or conflict with indigenous evidence, as with the claims about slavery or Harsha’s supposedly tax-light rule, they end up revealing at least as much about the observer’s own assumptions as about India itself — which is precisely why no foreign account can be used as a stand-alone source, only as one voice in a wider triangulation.