Evidence page · independent description

Three Chinese monks spent fifty years in India and wrote it all down.

Everything the Sanskrit tradition says about itself is said by people with a stake in the answer. The most substantial independent description of that society comes from Buddhist pilgrims who travelled there to study, wrote in Chinese, and had no position in any Indian dispute about caste, priority or precedence.

YEARS SPENT IN OR NEAR SOUTH ASIA approximate; Bardaisan never travelled — he interviewed an embassy Xuanzang · 629–645 CE · overland 17 yrs Yijing · 671–695 CE · by sea 25 yrs Faxian · 399–414 CE · overland out, sea back 15 yrs Megasthenes · c. 300 BCE 10 yrs Bardaisan · c. 218 CE · by interview 0 yrs

Time spent in the region. The Chinese pilgrims are outliers on this axis by a wide margin.

Eleven centuries of contact

CHINESE CONTACT WITH SOUTH ASIA — 1,100 YEARS ■ pilgrim ■ translator ■ envoy ■ military ■ trade 138 BCE Zhang Qian's mission west; reports of Shendu (India) reaching Bactria 2 BCE Earliest Chinese notice of Buddhist teaching 67 CE Traditional date for Buddhism's arrival at the Han court 148 CE An Shigao translating at Luoyang 178 CE Lokakṣema translating Mahāyāna sūtras into Chinese 266 CE Dharmarakṣa — large-scale translation programme 399 CE FAXIAN departs; 15 years — Gandhāra, Magadha, Sri Lanka 401 CE Kumārajīva reaches Chang'an; the translations that shape Chinese Buddhism 518 CE Song Yun and Huisheng sent to Gandhāra by the Northern Wei 629 CE XUANZANG departs; 16 years, a decade in India, years at Nālandā 643 CE Wang Xuance's embassies begin — four missions over two decades 648 CE Wang Xuance raises Tibetan and Nepalese troops and defeats a usurper in Magadha 671 CE YIJING departs by sea via Śrīvijaya; 25 years 720 CE Vajrabodhi and Amoghavajra carry tantric Buddhism to the Tang court 751 CE Battle of Talas; Tang power in Central Asia checked 842 CE Tang persecution of Buddhism; the overland exchange contracts 1000 CE Sea route dominant; Chinese ceramics appear in South Indian and Sri Lankan strata translation into Chinese begins two and a half centuries before the first pilgrim travels the other way

Selected events. Dates for the pilgrims are departure years.

What the chronology shows that the three accounts alone do not

Translation runs in the other direction first, and by a long way. An Shigao is working at Luoyang in 148 CE and Lokakṣema by 178 — two and a half centuries before Faxian sets out. Texts moved east long before travellers moved west.

The exchange was never only religious. Zhang Qian's report of Shendu in the second century BCE is an intelligence brief. Wang Xuance's four Tang embassies are diplomacy — and in 648 he raised Tibetan and Nepalese troops and defeated a usurper in Magadha, which is a Chinese envoy conducting a military action in India.

And the route changes. Faxian and Xuanzang went overland through Central Asia. Yijing went by sea, and stopped for years at Śrīvijaya in Sumatra — because by the late seventh century the maritime route was viable and Sumatra was a Buddhist centre worth studying in. The shift from land to sea is visible inside one century of these accounts.

A pilgrim's route is a map of which corridors were open, and Yijing's tells you the sea lane had won.

What contracted the exchange was internal to China. The Tang persecution of Buddhism in 842 and the loss of Central Asian position after Talas in 751 closed the institutional pipeline from the eastern end, before anything decisive had happened at the Indian one.

The three

WhoWhen and whereWhat the account contains
Faxian
法顯
399–414 CE. Overland through Central Asia and Gandhāra, down through Mathurā to Magadha, then Sri Lanka, returning by sea A Record of Buddhistic Kingdoms. Monastic populations, the state of sites, and social observation — including the earliest outsider description of untouchability as a practice
Xuanzang
玄奘
629–645 CE. Overland via the Tarim and Bactria; more than a decade inside the subcontinent, including years at Nālandā; travelled south as far as the Deccan and Kāñcī Great Tang Records on the Western Regions. Over a hundred kingdoms described: rulers, revenue, climate, crops, script, monastic and brahmanical institutions, and the four castes as seen from outside
Yijing
義淨
671–695 CE. By sea, via Śrīvijaya in Sumatra, where he stopped for years to work Record of Buddhist Practices Sent Home from the Southern Sea. Monastic daily practice in detail — food, medicine, hygiene, ordination procedure — and the maritime route
Why these accounts carry unusual weight

They travelled to the source. Every other outside account of South Asian material reaches the writer through intermediaries; the Greeks got their India through Alexandria and Latin Europe got it through Arabic. The pilgrims had no chain behind them.

They stayed. Xuanzang was in the subcontinent for over a decade and studied Sanskrit at Nālandā. He was not a traveller collecting impressions; he was a student inside the institutions he describes.

They had motive to be accurate and no motive to flatter. They came for texts and teachers, and wrote for readers in China who wanted to know what was actually there. They had no position in Indian arguments about precedence.

And they are preserved in an entirely independent manuscript tradition, in Chinese, copied in China — immune to every editorial pressure that operated on Sanskrit transmission.

Two literate traditions describing the same society, with no shared scribes and no shared interest. That is as close to an independent check as this field gets.

Faxian, and untouchability at c. 400 CE

An external, dated attestation of a practice, not of a doctrine

Faxian records that people of the lowest classification lived apart, and that on entering a town or market they struck a piece of wood to announce themselves so that others could avoid contact.

This matters because of what it is and is not.

It is not a text prescribing a rule. This platform's corpus work shows when the vocabulary appears — caṇḍāla absent from the Rigveda, the Atharvaveda, the Taittirīya Saṃhitā, four Brāhmaṇas and the Bṛhadāraṇyaka, arriving in the Chāndogya and reaching 3.4 per ten thousand words in Manu. But a word in a law code is a rule someone wrote. It is not evidence that anyone obeyed it.

Faxian's observation is the other kind of evidence entirely: a foreigner walking through a market and recording what he saw people doing.

Together the two lines converge. The prescriptive vocabulary appears in the Dharmasūtra window; an independent observer describes the practice operating several centuries later. Neither would be sufficient alone — a rule without evidence of practice, or a practice without evidence of when it began.

Xuanzang on the four classes

Described from outside, by someone with no place in the scheme

Xuanzang sets out the four categories — brahmins, kṣatriyas, vaiśyas, śūdras — as an observed feature of the society, with the occupations attached to each, and notes that marriage does not cross them.

He also records the groups placed outside the scheme entirely — butchers, fishermen, executioners, scavengers — as living apart from the settlements, with their dwellings marked so they could be avoided.

The value of this is that it is not an insider's account of an ideal. The Dharmaśāstra describes a system it is arguing for. Xuanzang describes a system he found in operation, and he had no reason to make it look either better or worse than it was.

And he describes Buddhism as widespread and institutionally powerful in the seventh century — thousands of monasteries, tens of thousands of monks, Nālandā functioning at scale. Which dates the decline: whatever happened to Buddhist institutions in India happened after 645 CE, not before.

What they should be used for, and what they should not

Reliable forUnreliable for
What the observer saw and did — routes, distances, institutions, practices witnessedNumbers. Monastic populations and city sizes are reported figures, often round, sometimes second-hand
The state of Buddhist institutions, which they came to inspect and understoodBrahmanical doctrine, which they understood as outsiders and sometimes as rivals
Dating. Their travels are fixed within a year or twoLegend and local history, which they recorded as told to them
Geography, crops, climate, script and language distributionAnything they were told about the deep past
And the general rule they illustrate

An outside account credits the place the writer got the thing from, and the number of intermediaries determines how wrong that is.

Greek writers credit Egypt, because Alexandria is where they met the material. Latin Europe credits Arabs. Syriac translators credit the Greek they translated from. Each is accurate about the immediate source and silent about what lay behind it.

The pilgrims had no intermediaries at all — and they credit India, directly and by name. The reliability of an outside account is inversely proportional to the length of the chain behind it, and that is a testable property rather than a matter of opinion.