Medang

Two monuments

Drafted from Cowing, The Indianized States of Southeast Asia (1968)formation

Borobudur and Prambanan stand about twenty kilometres apart, on the same volcanic plain in central Java, built within roughly a century of each other. Borobudur is a Mahayana Buddhist stupa-mountain of some two million blocks with five kilometres of narrative relief around its galleries. Prambanan is a Shaivite complex whose central tower stands forty-seven metres and whose reliefs carry the Ramayana. Between and around them are Kalasan, Sewu, Plaosan, Mendut, Sambisari and forty or so others.

One small kingdom on one plain built the lot, in about a hundred and fifty years.

Buddhist and Hindu, twenty kilometres apart

The obvious question about that plain is how one small kingdom produced the two monuments at once, and the older answer was that it did not: that there were two dynasties, the Buddhist Sailendras and the Shaivite line of Sanjaya, ruling in rivalry or in alternation.

That reading has been argued down over the last half-century toward something messier: a single ruling group with both patronages inside it, intermarrying, in which the labels describe what a particular king endowed rather than which house he belonged to. Nobody has settled it, and a page like this is not the place to try.

What is worth saying is that the kingdom carried neither religion anywhere. Shaivism and Mahayana Buddhism arrived in Java by the same channels as everything else Indic, from India by sea, and Medang's contribution was not to spread them but to build more of them, in stone, than anywhere outside India. Pagan did the same thing in Burma three centuries later with ten thousand temples and equally little missionary effect.

Enormous devotion and no propagation at all is a combination that turns up more often than the usual story about religions suggests.

Balaputra

Drafted from Cowing, The Indianized States of Southeast Asia (1968)peak

One man connects the two halves of the archipelago, and he did it by losing.

Around 856 Rakai Pikatan consecrated the Shiva complex at Prambanan and, on one reading of the Shivagrha inscription, defeated Balaputradeva — the last Sailendra claimant in Java. Balaputra left.

A few years later a copper-plate charter from Nalanda in Bengal records a monastery endowed there for the benefit of a king of Suvarnadvipa named Balaputradeva, who gives his descent from the Sailendras of Java.

Suvarnadvipa is Sumatra. He had become king of Srivijaya.

What that is and is not

It is not a conquest, an inheritance or a claim. Balaputra did not assert Java from Sumatra, and Srivijaya did not become Medang; the two kingdoms were contemporaries and stayed contemporaries.

What it is, is one man walking away from a defeat on one island, arriving on another as its king, carrying a dynastic name with him, and endowing a monastery in a third country that happened to write it down. Java and Sumatra are otherwise two separate worlds in this period — different rivers, different crops, different sources, almost no direct traffic in the record — and this is the one place where a single individual is visible crossing between them.

The Shivagrha reading is argued over, so the whole chain is provisional. But if it holds, the greatest maritime power in the region acquired its most famous king because a Javanese succession dispute went the other way.

Putting the kingdom's height here rather than at Borobudur is a deliberate choice. Rakai Pikatan's reign is simply the one the inscriptions let you see most clearly: a succession settled, a monument finished, and a defeated rival who can be traced afterwards to a throne four hundred miles away.

The plain they left

Drafted from Ricklefs, A History of Modern Indonesia since c.1200 (2008)contraction

Around 929 Mpu Sindok moved the court from central Java to the Brantas valley in the east. Central Java was never again the seat of a Javanese kingdom.

Nothing explains it. There is no defeat, no invasion, no succession crisis; the inscriptions simply stop being issued on one plain and start being issued four hundred kilometres away. Merapi erupting, an epidemic, a Srivijayan raid, and a deliberate move toward the rice surplus and the trade of the Brantas have all been proposed, and not one of them is attested.

What the move left behind is the point. The largest concentration of eighth- and ninth-century monumental architecture in Southeast Asia stands on a plain that a state walked away from, and it stood there through everything that followed — Kediri, Singhasari, Majapahit, Demak, the sultanates, the Dutch — without anybody ruling from it again. Borobudur was under volcanic ash and vegetation when Raffles's men were shown it in 1814.

Where it went instead

The move east is where the rest of Javanese history starts.

Every kingdom that follows sits on the Brantas or its plains: Kediri from 1042, Singhasari from 1222, Majapahit from 1293, and the sultanates after them. Seven hundred years of Javanese states, all of them downstream of a decision taken in 929 for reasons nobody wrote down.

It also put the Javanese court on top of the granary. Melaka, with no farmland of its own, bought most of its rice from Java; Demak on the north coast sold it. The surplus that fed the trade of the whole strait came off the plains Mpu Sindok moved onto.

What Sindok built

Drafted from Hall, A History of Early Southeast Asia: Maritime Trade and Societal Development, 100-1500 (2011)contraction

Mpu Sindok did not simply relocate a court. He spent his reign turning a displacement into a functioning kingdom, and within a decade of the move east he had to prove it could survive a direct attempt to undo it.

An attack that was answered rather than absorbed

An inscription dated 937 records that around 928-929 an army from Srivijaya's Malayu base landed on Java's coast and pushed inland as far as Nganjuk before Sindok's forces defeated it. The stone that records the victory credits Sindok personally with saving the kingdom, which is worth noticing: the same Srivijayan reach that would later be blamed for helping end the dynasty in 1016 had already tried once, within a few years of the capital's move, and had been beaten back rather than accommodated.

A kingdom built on water rather than on a battlefield

What actually filled the ninety years between the move and the pralaya was not warfare but plumbing. Between 900 and 1060, royal intervention founded more than forty new communities in the developing Brantas River delta, and the surviving inscriptions of this period are overwhelmingly about irrigation rather than conquest: dams, dykes and canals built to bring a turbulent river under control, each one recording a king's own involvement in commissioning it and instructing local elders on maintaining it afterward. The Brantas flooded its banks often enough that unmanaged farmland could not be trusted to feed a court; a managed one could, and every east Javanese king after Sindok kept building the systems that made that true.

What that investment was worth

The scale of the payoff shows up precisely when the infrastructure failed. When the river burst its dikes in 1033 under Airlangga - by then ruling what had been rebuilt after 1016 - the resulting flood was remembered as a genuine crisis: trading vessels could no longer reach the royal port, and the maharaja's own income, along with his subjects', collapsed until two new dams restored the old course. That a single breach could be recorded as a catastrophe is itself the measure of how much an intact system had been worth. Sindok's own successors had spent the ninth century's end and the tenth building exactly the kind of centralised water management whose sudden absence, decades later, was disaster enough to write down.

Pralaya

Drafted from Cowing, The Indianized States of Southeast Asia (1968)end

In 1016 a vassal named in the inscriptions as Wurawari attacked Dharmawangsa's court during the wedding of the king's daughter and killed him. The Javanese word for what followed is pralaya, the dissolution at the end of a world age.

Airlangga, the king's son-in-law and then sixteen years old, took refuge among hermits on Mount Vanagiri.

Twenty years to get it back

He stayed on the mountain three years. Then notables and brahmans came to plead with him to take the royal power as his father-in-law's successor, and in 1019 he was crowned — over a territory that did not reach beyond a strip of the north coast between Surabaya and Pasuruhan.

He waited another decade before trying to recover the rest, and the delay was probably prudence: Srivijaya, which may have had a hand in 1016, was itself wrecked by the Chola raid of 1025.

The reconquest ran from 1028 to 1035, against a country divided among competitors, and the inscriptions record it in their own register. A residence burned in 1031. In 1032 he rid the country of a woman "endowed with a formidable power, similar to a rakshasi", and devastated the southern region, "which he burned with his tongue like a fiery serpent". The prince of Wengker, beaten and fleeing "abandoning his wife, children, treasures, and royal vehicles", was seized by his own troops and died in 1035.

In 1037, "having placed his feet on the head of his enemies", Airlangga took his place on the throne of lions. He moved his residence east, to Kahuripan, whose site has still not been identified.

His court produced the Arjunawiwaha, one of the great Old Javanese poems, in which Arjuna withdraws as an ascetic, resists temptation, and is given a divine weapon — written for a king who had spent four years on a mountain and twenty getting his kingdom back.

And then he divided it

In 1042 Airlangga split Kahuripan between his two sons.

The Javanese account gives the reason as a father's even-handedness. What it produced was two kingdoms, Janggala and Panjalu, which fought each other; Panjalu won, took the name of its capital's district, and became Kediri.

So the line from Medang to Kediri is not a line. It runs through a kingdom assembled from nothing by a teenager, held for twenty years, and then deliberately cut in half — and only one of the halves survived to matter. Dharmawangsa's kingdom ended in 1016 with a wedding and a massacre, and what followed was not its continuation but its reconstruction by somebody else, under another name, before being taken apart again on purpose.