Gupta Empire

The daughter's son

Drafted from Fleet, Inscriptions of the Early Gupta Kings and Their Successors (1888)formation

The Gupta dynasty's own inscriptions give its founder's grandson a title that no other imperial house in India advertises: Licchavi-dauhitra, the daughter's son of a Licchavi.

It appears in the genealogy every time. On the Allahabad pillar, on the copper plates, on the coinage — a king describing himself, in his own formal titulature, by reference to his mother's family.

Why it is there

The consensus reading is that the Guptas started small and married up.

The first two names in the genealogy, Gupta and Ghatotkacha, carry only the title maharaja — a substantial lord, not a paramount sovereign. The third, Chandragupta I, is maharajadhiraja, supreme king of great kings, and he is the one who married Kumaradevi of the Licchavis.

The Licchavis were an old and prestigious lineage of the Ganges basin, a gana-sangha — one of the oligarchic clan-states that had existed since before the Mauryas, and one the Buddha himself had known. Marrying into that gave a provincial house something it could not manufacture.

Chandragupta struck coins with his own portrait on one face and his queen's on the other, named, with the legend Licchavayah — the Licchavis — on the reverse. A dynasty putting its wife's clan on the money is telling you where it thinks its claim comes from.

The other reading

Fleet noticed something else. The name-ending gupta, on the authority of the Vishnu Purana, was the ending prescribed for a Vaishya — the merchant varna, third of the four.

He is careful, and points out counterexamples: the astronomer Brahmagupta can hardly have been anything but a Brahmin. But the suggestion he records, and which has been argued ever since, is that the Guptas were not of high birth, and that this is exactly why the Licchavi connection is repeated so insistently.

If that is right, then the greatest patrons of Sanskrit court culture in Indian history — the dynasty under which the classical literary and legal canon takes its shape — were a merchant family who had married into an old clan and never stopped mentioning it.

That is not a discreditable origin and the inscriptions do not treat it as one. It is simply the ordinary way a new house becomes an old one, and the Guptas left the receipt in their own titles.

What they started with

Magadha and the middle Ganges: the same ground the Mauryas had started from six hundred years earlier, with the same capital at Pataliputra.

The era they used — the Gupta era, running from 319 or 320 — is conventionally taken to date from Chandragupta I's accession, which means the dynasty considered itself to have begun with the marriage.

Whatever the family was before, that is the year they decided to count from.

The pillar

Drafted from Fleet, Inscriptions of the Early Gupta Kings and Their Successors (1888)expansion

Samudragupta's account of his conquests is cut on a sandstone column at Allahabad that already carried the edicts of Ashoka.

That is the first thing to notice about it. A fourth-century emperor recorded his wars on a pillar a third-century-BC emperor had raised to announce that he had given up war, and did not remove the earlier text. Both are still there, and so is an inscription Jahangir added in the seventeenth century.

What the inscription does

It is a prashasti, a formal eulogy, composed by a named court officer, Harishena, who describes himself in it. Fleet's translation preserves the gaps where the stone is damaged, and the surviving passages set out something more precise than a list of victories: a taxonomy of ways of being subdued.

Exterminated. Of the kings of Aryavarta, the northern plain, the inscription says he increased his majesty by violently exterminating them, and names nine: Rudradeva, Matila, Nagadatta, Chandravarman, Ganapatinaga, Nagasena, Achyuta, Nandin, Balavarman, and many other kings.

Made servants. The kings of the forest countries — the tribal territories between the settled regions — he made to become his servants.

Obeisance. The frontier kings of Samatata, Davaka, Kamarupa, Nepala and Kartripura, and the tribal peoples — Malavas, Arjunayanas, Yaudheyas, Madrakas, Abhiras, Prarjunas, Sanakanikas, Kakas, Kharaparikas — gratified his commands by giving all kinds of taxes and obeying his orders and coming to perform obeisance.

Devotion. And then, furthest out, the Daivaputras, Shahis, Shahanushahis, Sakas and Murundas — the Kushan and Saka rulers of the north-west and beyond — who were won over simply by devotion and obeisance, offering their daughters in marriage, giving Garuda-tokens, and asking for charters confirming them in their own territories.

The graded empire

Four rings, four relationships, one sentence apiece. Annihilate the neighbours, enslave the forest, tax the frontier, accept homage from the far edge.

That is a much more honest description of what a large Indian empire was than the word "empire" supplies. The Guptas ruled a core directly and everything else by arrangement, and the arrangements were different at every distance. The inscription is not hiding that; it is boasting about it, because a king who can extract four different kinds of submission from four different kinds of people is demonstrating range rather than admitting limits.

And the south

The southern campaign gets its own formula, and it is the most interesting clause in the document.

Samudragupta went down the eastern coast, defeating a series of kings whom the inscription names one after another — it begins with Mahendra of Kosala. And then he gave them back their kingdoms. The formula the eulogist uses is that his glory was produced by the favour shewn in capturing and then liberating them.

A raid, in other words, conducted for tribute and prestige rather than territory, at the end of which the defeated kings resumed their thrones as clients. Nothing in the Deccan or the far south was annexed.

Set that beside the ninth name in the extermination list and the empire's actual shape appears. Close to home, kings are removed. Far away, kings are collected.

The pillar at Meharauli

Drafted from Fleet, Inscriptions of the Early Gupta Kings and Their Successors (1888)aside?What aside means

Nine miles south of Delhi stands an iron column twenty-three feet eight inches high, tapering from sixteen inches across at the base to twelve at the top, cut with an inscription that names no dynasty, gives no date, and was already unreadable when a Muslim dynasty in Delhi took an interest in it centuries before any Englishman saw it.

What the inscription says

It is a Vaishnava dedication, recording that a king named Chandra raised the pillar — called, in its own text, a dhvaja, a standard — to the god Vishnu on a hill named for Vishnu's footprints. Before it gets to the dedication, it spends four lines on why the king deserved to make it.

He is the one "on whose arm fame was inscribed by the sword" when, fighting in the Vanga countries, he beat back with his own chest a confederacy of enemies who had united against him. He is the one who crossed the seven mouths of the river Sindhu in battle and conquered the Vahlikas. The breezes of the southern ocean, the inscription says, are perfumed even still by his prowess. And having won sole supreme sovereignty by his own arm, and held it a very long time, he set up the standard bearing his own beautiful, full-moon face.

Whose pillar

Fleet, editing the text in 1888, records that nobody before him had managed to read the king's own name correctly — early transcriptions of a badly weathered lithograph made the inscription "quite unintelligible from beginning to end," and it took two further re-readings, forty years apart, before Bhau Daji established in 1875 that the name was Chandra.

Establishing the name did not establish which Chandra. Fleet lays out the candidates without settling on one: Fergusson, struck by the Persian influence visible in the pillar's capital, argued for one of the early Gupta Chandraguptas and dated it to 375 or 400 AD. Fleet's own inclination, on separate grounds, was to allot it to Chandragupta I, the first Gupta ruler to take the title Maharajadhiraja — though he flags the one real objection to his own guess: the inscription says nothing about the Indo-Scythian rulers the early Guptas are known to have had to overthrow to establish themselves at all. He does not rule out Chandragupta II either, and neither did most of what was written after him.

An object that outlasted the argument about it

Whichever Chandra commissioned it, the pillar itself has survived in a condition the inscription's own readers kept remarking on: "the writing... is in a state of excellent preservation throughout, owing, of course, to the nature of the substance on which it is engraved." A column of wrought iron, more than fifteen hundred years old by any of the proposed datings, standing in the open air the whole time, with its dedication still legible when nobody could agree whose reign it belonged to.

That is, in miniature, the shape of most Gupta history. The court that produced the age later ages would call classical left behind pillars, plates and coins bearing names, titles and boasts - and left the work of sorting out which king did which conquest to everyone who came after, arguing from the same handful of durable objects the dynasty happened to leave standing.

What Faxian saw

Drafted from Faxian, A Record of Buddhistic Kingdoms (1886)peak

A Chinese monk walked across northern India at the height of the Gupta empire, spent about a decade there, and wrote down what he saw. He was not interested in the government, which is what makes his account of it worth having.

The state that did not register anybody

Faxian's description of the Middle Kingdom — the Gangetic heartland — is a few sentences long and every clause in it is a claim about administration.

The people are numerous and happy; they have not to register their households, or attend to any magistrates and their rules; only those who cultivate the royal land have to pay a portion of the gain from it. If they want to go, they go; if they want to stay on, they stay.

No household registration. No compulsory attendance on officials. Revenue drawn from crown land rather than from a general assessment. Freedom of movement.

Then the criminal law:

The king governs without decapitation or other corporal punishments. Criminals are simply fined, lightly or heavily, according to the circumstances of each case. Even in cases of repeated attempts at wicked rebellion, they only have their right hands cut off.

And a detail that is easy to skip and is the whole fiscal argument: the king's body-guards and attendants all have salaries.

Why this is remarkable and what to be careful about

Set it beside the Mauryan state Megasthenes described seven hundred years earlier — the committees registering every birth and death, the land measured and the sluices inspected, the quarter of the produce taken — and it is the opposite kind of government.

That is broadly what the Gupta inscriptions suggest too: an empire that held its core directly and everything else through subordinate kings, with a lighter and less intrusive apparatus than the Mauryas had run.

The cautions are real. Faxian was a monk moving between monasteries on pilgrimage routes, describing the country as it looked to a traveller who was fed and housed by religious institutions. He is comparing everything silently to China, where household registration and corporal punishment were pervasive, so their absence is what strikes him. And he says nothing about how the subordinate kings taxed their own people.

But he had no reason to flatter anybody. He was not at court, he was not petitioning, and his book was written for Chinese monks who wanted to know where the relics were.

The other half of the same paragraph

He follows all of that with this:

Throughout the whole country the people do not kill any living creature, nor drink intoxicating liquor, nor eat onions or garlic. The only exception is that of the Chandalas. That is the name for those who are held to be wicked men, and live apart from others. When they enter the gate of a city or a market-place, they strike a piece of wood to make themselves known, so that men know and avoid them, and do not come into contact with them.

In the markets there are no butchers' shops and no dealers in drink. Only the Chandalas fish, hunt and sell meat. Money is cowries.

So the same society that does not register its households and does not mutilate its criminals requires a category of people to announce themselves with a clapper so that everybody else can get out of the way.

Both sentences are in the same paragraph, by the same observer, on the same day. A state can be light on its subjects and a social order can be merciless, and the classical age of Indian civilisation is the place to notice that these are different questions.

The endowments

One more mechanism, and it is how the whole religious economy worked.

After the Buddha's parinirvana, Faxian says, the kings and the heads of the Vaishyas built viharas for the priests and endowed them with fields, houses, gardens and orchards, along with the resident populations and their cattle — and the grants were engraved on plates of metal, so that they were handed down from king to king without any one daring to annul them.

Copper-plate grants of villages, in perpetuity, transferable across dynasties. That is a real institution and the plates survive in quantity. It is also the beginning of something the Guptas' successors would live with: land alienated from the revenue, permanently, by an instrument no later king felt able to revoke.

A classical reference point

Drafted from Thapar, Early India: From the Origins to AD 1300 (2002)peak

The Gupta empire held the Gangetic plain from around 320 and never covered as much ground as the Mauryas had, governing much of its territory through subject kings rather than directly. Its reputation rests on what its court sustained rather than on what it conquered.

The Sanskrit literary canon takes its classical shape here — Kalidasa is placed in this period — along with the codification of Hindu legal and philosophical texts, temple architecture in stone, and the mathematics. The decimal place-value system with a symbol for zero is attested in India in this period, passed to the Islamic world where al-Khwarizmi wrote on it in ninth-century Baghdad, and reached Europe from there as Arabic numerals.

That is not a writing system, a religion or a dynasty, and it is therefore invisible to almost every way anybody has of comparing empires. It is also, by some distance, the most widely used thing any state in this list produced.

The empire itself fragmented in the sixth century under pressure from the Hunas and from two centuries of granting away its own revenue.

The Hunas, and the plates

Drafted from Thapar, Early India: From the Origins to AD 1300 (2002)end

The Gupta empire was pushed over from outside and had already been weakened from inside by a mechanism nobody at the time could have described as a weakness, because it looked like piety.

The invasions

The Hunas — the Hephthalites, or White Huns, a steppe people of the same broad movement that produced the Huns in Europe — came through the north-west passes from the later fifth century.

Skandagupta beat them, and said so. The Bhitari pillar inscription describes him restoring the fallen fortunes of his family, and Junagadh records him repairing a dam. The successful defence is real and it is also the last thing the dynasty does that looks imperial.

Within a generation the Hunas were back and holding the Punjab and much of western India; Toramana and his son Mihirakula ruled from Sakala, and Mihirakula appears in the Buddhist sources as a persecutor and in the Gwalior inscription as a great king. He was eventually beaten by a coalition of Indian rulers in the 530s, and the Guptas were only one member of it.

Beating an invader in coalition is how a paramount power discovers it is no longer paramount.

The plates

Underneath that, something slower.

The Gupta and post-Gupta period is when land grants to Brahmins and to religious foundations become a normal instrument of royal patronage, recorded on copper plates. Faxian saw them and described the terms exactly: the fields, the houses, the orchards, along with the resident populations and their cattle, granted in perpetuity and handed down from king to king without any one daring to annul them.

Read that as a fiscal document rather than as an act of merit. Every grant removes a village from the revenue permanently, and transfers with it the right to collect from the people on it. The grantee gets the taxes; increasingly, he gets the administrative and judicial authority as well.

Do that for two hundred years and the state has given away its own base. The king still reigns; the villages belong to somebody else; and the intermediate holders have their own resources, their own dependants and no particular reason to be governed.

Some historians have called the result Indian feudalism and the label has been argued about for fifty years. The mechanism does not depend on the label. A government that alienates revenue in perpetuity, on tamper-proof instruments, to a class it needs the approval of, has a diminishing amount of country left to run.

The dissolving

There is no date and no battle. The Gupta line continues in Magadha into the sixth century in a reduced form; regional dynasties — the Maukharis, the Vakatakas' successors, the Maitrakas of Valabhi — appear as independent powers in territory that had been Gupta or Gupta-client.

By the middle of the sixth century there is no paramount authority in northern India. Harsha assembles one again in the 600s and it lasts as long as he does.

What the end was not

It was not a catastrophe, and this matters for how the period afterwards is usually described.

The temples go on being built — the stone temple architecture that begins under the Guptas develops without interruption into the regional styles of the seventh century. The Sanskrit literary and scientific tradition continues: Varahamihira and Brahmagupta both work in the century after the empire ends, and Brahmagupta is the man who writes down the rules for calculating with zero.

Nalanda, the monastic university that becomes the intellectual centre of the Buddhist world for six hundred years, is founded under the Guptas and does its greatest work after them, sustained by exactly the kind of land grants that were dissolving the empire.

The state came apart. Almost nothing the state had been paying for did.

Zero

Drafted from Thapar, Early India: From the Origins to AD 1300 (2002)aside?What aside means

The decimal place-value system with a symbol for zero, treated as a number rather than a blank, is attested in India in this period. Aryabhata's astronomical work of 499 uses place-value; the Bakhshali manuscript and the inscriptional record carry the notation; Brahmagupta in the seventh century writes the rules for operating with zero and with negative numbers.

It travelled. Indian astronomical texts reached Baghdad in the eighth century, al-Khwarizmi wrote on Indian calculation in the ninth, and his work reached Latin Europe through translations in the twelfth — which is why the digits are called Arabic in English and Indian in Arabic, and why algorithm is his name.

Every number anyone writes down today, on any monument, in any ledger, in any scientific measurement, is written in it.

It is not a script, and it is not a religion, and no later state claimed to inherit it — so it falls outside every category by which the influence of an empire is normally reckoned. A notation is not a thing a dynasty can be credited with in the way a conquest or a conversion is; it belongs to the mathematicians who worked it out, several of whom lived under governments that no longer existed by the time they wrote.

Which is roughly the point. The Guptas held less ground than the Mauryas and ruled much of it through other people's kings. Judged on the things empires are usually judged on, they are middling. What came out of the century they presided over is the arithmetic of the entire world, and there is no scale on which those two facts can be weighed against each other.

The word "classical"

Drafted from Thapar, Early India: From the Origins to AD 1300 (2002)afterlife

The Gupta period is called the classical age of India, and it is worth knowing who called it that and when, because the label is doing work.

Where the term came from

It is a nineteenth- and early-twentieth-century coinage, applied by scholars working with a Greek and Roman analogy in mind: a period of political unity, formal perfection in literature and art, and a canon that later ages refer back to. Vincent Smith and others used it, and Indian nationalist historiography took it up gladly, because it supplied a golden age that was indigenous, Hindu, and demonstrably not a gift from anybody.

There is a real thing underneath the label. The Sanskrit literary canon does take its classical shape here — Kalidasa is conventionally placed at the Gupta court, and the Shakuntala is the play that Goethe wrote a quatrain about and that made Sanskrit literature a European subject in the 1790s. Temple architecture in dressed stone begins here. The Ajanta paintings are Vakataka work of the same decades. Aryabhata writes in 499.

And what it hides

Three things, and they are the reasons to be careful with the word.

It flattens the century after. The scientific tradition peaks after the empire: Varahamihira, Brahmagupta, and the mathematics that mattered most all belong to the seventh century, when there was no Gupta state at all. So does Nalanda's greatest period, and most of what Xuanzang came to India to study.

It quietly assumes that unity produces culture. Some of the most productive periods in Indian intellectual history had no paramount power in them, and some of the empires had very little to show.

And it has been used politically. A golden age that is Hindu, Sanskritic and ends when invaders arrive is an argument as much as a chronology, and it has been deployed as one — which is not an argument against the period's achievements but a reason to be exact about their dates.

The thing that actually travelled

Not the literature, which stayed inside its language. Not the dynasty, which nobody claimed to succeed.

The decimal place-value notation, with a symbol for zero treated as a number rather than a blank. It is attested in India in this period; Brahmagupta wrote the rules for operating with it, including with negative numbers, in 628. Indian astronomical texts reached Baghdad in the eighth century; al-Khwarizmi wrote on Indian calculation in the ninth; his work reached Latin Europe through translations in the twelfth.

Which is why the digits are called Arabic in English and Indian in Arabic, and why algorithm is a Persian mathematician's name.

Every date written down anywhere today is written in it.

What that says about measuring empires

The Guptas held less ground than the Mauryas, ruled much of it through other people's kings, and were dismantled by a combination of steppe invasion and their own charitable endowments.

Judged on territory, duration or armies, they are a middling empire. What came out of the century they presided over is the notation the entire world does its arithmetic in — and there is no way to weigh those two things against each other, because they are not the same kind of quantity.

That is not a defect in this particular case. It is what happens whenever anybody tries to rank states: the things that can be counted and the things that turn out to matter are only loosely related, and the second list is only visible afterwards.