Kingdom of Janggala

It matters because it is the clearest case in Javanese history of a state erased by its sources. Janggala left seven inscriptions, five of them issued by one king; Kediri, its southern twin, left almost nothing until 1117. Yet the standard histories of Java describe a "Kediri period" and treat Janggala as a footnote — a distortion that a 2026 epigraphic survey has begun to correct.

At a glance

What was Janggala?

The northern of the two kingdoms into which Airlangga divided Kahuripan in 1052/53. Its territory was the lower Brantas and the coast — the ports and rice plains that had made the old kingdom rich — and its capital was Jiwana, which the charters equate with Kahuripan, Airlangga's own seat.

That gives the division an odd logic: the northern kingdom kept the capital and the trade, and the southern kingdom, Kediri, kept the interior and the rice. The arrangement looks less like a partition between equals than a division of functions, and it did not last: the brothers were at war within a year.

Why was it created?

Because Airlangga divided his kingdom in 1052/53, and nobody knows why. The charters show a single state until the early 1050s; then two kings, two capitals and two sets of charters. The Javanese tradition, written centuries later, says he divided the realm to prevent his sons from fighting over it.

The evidence suggests the division did not prevent the war. The Garamān charter of 1053 — Śaka 975, one year after the partition — already names the king of Pañjalu, Kediri's ruler, as Garasakan's enemy. The brothers were fighting within a year of the division, which is the best argument that the division was a settlement of an existing conflict rather than a preventive measure.

Where was it?

In the north, not the highlands. The older scholarship — Coedès, and Hall as late as 2011 — placed Janggala in the Malang highlands of eastern Java, on the assumption that the two kingdoms divided the country between coast and interior in a simple way. The findspots of Janggala's inscriptions contradict that: they come from Tuban, Lamongan, Mojokerto and southern Gresik, all in the northern delta.

This is a good example of how Javanese geography is reconstructed. A state's territory is inferred from where its charters were carved and where the villages they name can be located, and when the findspots change, the map changes. The 2026 survey's redrawing of the two kingdoms is the current state of the question.

How do we know about it?

From seven inscriptions. Janggala's documentary life is astonishingly short: the charters run from 1052 to 1059, and five of the seven were issued by Garasakan. Then silence — not because the kingdom ended, but because its kings stopped issuing charters that survive, or because the charters were carved on perishable materials, or because the political centre moved.

The inscriptions are land grants, like all Javanese charters of the period: sima grants to temples and villages, with boundaries, witnesses and curses.

They give us the king's name, the officials, the places and the religious foundations — and almost nothing else. Janggala's history is a list of dated grants, and the narrative has to be inferred from them.

What did its charters grant?

Land, tax exemption and religious merit. Each of Garasakan's charters frees a piece of land from royal dues and gives its income to a religious foundation or a village, and records the boundaries, the witnesses, the offerings and the penalties for anyone who disturbs the grant. The formula is identical to Airlangga's, which is the point: Janggala presented itself as the continuation of Kahuripan, not as a new state.

The grants also name the officials who administered the kingdom — the rakryan, the samget, the tanda — and the local lords whose consent made a grant effective. They are the only surviving record of how a small Javanese state worked, and they show a government of negotiation: the king could not simply command a village, he had to buy its loyalty with tax relief.

What was the Garuda-mukha?

A seal, not a coin. Janggala's charters describe documents "sealed with the Garuḍa-face" — the garuḍa-mukha — and a carved emblem of the same design appears on the stelae. The device was a royal authentication mark, the equivalent of a great seal, and it is the continuity proof that links Garasakan's charters to Airlangga's: the same emblem, the same chancery practice, the same state.

The confusion with a coin or amulet comes from the seventeenth-century Javanese chronicles, which describe a garuda emblem as a royal regalia. The charters are unambiguous: it is a seal and a carved device, and it tells us that the Javanese state had a documentary culture as well as a temple culture.

Why did historians doubt it existed?

Because the sources were read through a lens that had no place for it. Nineteenth- and early twentieth-century scholarship organised Javanese history into periods named after capitals, and the period from the eleventh to the thirteenth century was labelled "Kediri" — so Janggala appeared as a rebel province, an interruption, or a mistake. Krom's standard history fixed the label, and later writers repeated it.

The doubt was not unreasonable: a kingdom that issues seven surviving charters in seven years and then falls silent looks like a phantom. What changed the picture was epigraphy — the systematic re-reading of the charters and the mapping of findspots — which showed a state with a capital, a chancery, a seal and a territory, and forced the 2026 survey to redraw the map of eastern Java.

How did it end?

Gradually, and without a battle. Kediri, the southern kingdom, grew stronger through the twelfth century; by the 1180s the king of Janggala was a subordinate of Kameswara of Kediri, and by the early thirteenth century the northern kingdom had ceased to have its own line of kings. The formal union is credited to Singhasari after its victory at Ganter in 1222 — but Ganter was a battle between Singhasari and Kediri, not between Janggala and Kediri, and no source records Janggala's end.

That is why the entry's dates run to the 1130s on the map and to the twelfth century in the text: the border data gives a drawn range, the sources give a process, and the process has no date. The honest statement is that Janggala was absorbed gradually between the 1130s and the 1220s, and that nobody recorded the moment.

What survived of the name?

An appanage title and a formula. Under Majapahit, Janggala reappears as a princely domain: Hayam Wuruk, the great king of Majapahit, was installed as a prince "in the realm of Jiwana", and the paired formula jaṅgala-kaḍiri — Janggala and Kediri — appears in charters from 1330 to 1486 as a way of naming the eastern provinces together.

So the name outlived the state by two centuries, as a title and a geographical expression. That is a common fate for the loser in a dynastic division: its territory is remembered, its kings are not.

Common myths about Janggala

Timeline of Janggala

What the map shows next

At 1092 on the map Janggala and Kediri are two kingdoms, and the Mataram Kingdom is a century and a half in the past.

At 1138 on the map the border data ends for Janggala, and Singhasari is still eighty years away.