Angkor did not appear from nowhere. For about six centuries before its first kings, the lower Mekong and the plains around the great lake were home to rice farmers, traders and rulers who built in brick, worshipped Indian gods and wrote in both Sanskrit and Khmer. Historians call this the pre-Angkorian period, roughly the first century to the eighth century CE. It laid down the language, religion and habits of kingship that Angkor would inherit.
From the delta to the lake
The earliest known center of power lay in the Mekong delta, in a realm that Chinese writers called Funan, which flourished from about the first to the sixth century on the sea trade between India and China. Its port towns were linked by canals, and its rulers adopted Indian titles and gods. As sea routes shifted south and trade passed it by, the center of gravity moved inland, toward the middle Mekong and the Tonle Sap, where rice rather than commerce supported the rulers.
Chenla and the brick towers
Chinese histories of the Sui and Tang dynasties name a new power in the region, and Chenla is a name Chinese records gave to the inland Khmer realm that rose in the sixth century. The Sui history says that Chenla had once been a vassal of Funan and later conquered it.
Inscriptions, however, give us local names. A king called Isanavarman I, who ruled in the early seventh century, made his capital at Isanapura, a city whose ruins lie in central Cambodia at a site now called Sambor Prei Kuk. Sambor Prei Kuk, the ruins of Isanapura, holds well over a hundred brick temples, some with unusual eight-sided towers, and it was listed as a World Heritage site in 2017.
The oldest Khmer words
The Khmer language belongs to the Austroasiatic family, which includes Vietnamese and Mon, and it is unrelated to Sanskrit. Yet for its writing it borrowed an Indian script. The earliest dated inscription in the Khmer language, from Angkor Borei in the south, was carved in 611 CE. Like many later stones, it uses Khmer for practical matters, recording gifts of servants and goods to a sanctuary.
Pre-Angkorian inscriptions also show who held power locally. The title poñ appears in seventh century Khmer inscriptions for local notables who controlled land and temples, and it disappears in the Angkorian period, as kings drew more authority to themselves. Its fading is one of the clearest signs of a political change that came before the great monuments.
One kingdom or many?
Tang sources report that early in the eighth century Chenla split in two, into a "Land Chenla" in the north and a "Water Chenla" near the coast. Older histories took this as the break up of a single state. The historian Michael Vickery argued that Chenla was never one unified state but a cluster of smaller rival polities, which Chinese observers lumped under one name. Inscriptions from this time name many rulers in many places, often with no sign that they answered to anyone else. On this view, the real achievement of the founders of Angkor was to pull these rival centers into one realm for the first time.
Indianized or localized?
Why did Khmer rulers take Sanskrit names, worship Shiva and Vishnu and build temples on Indian plans? The French scholar George Coedès, writing in the mid twentieth century, spoke of "Indianized states", and Coedès's term suggested that Indian culture was carried to Southeast Asia and largely shaped its early states. Later historians, Wolters among them, stressed localization: local people chose the Indian ideas that suited them and reshaped them for their own purposes. There is no evidence of Indian conquest or settlement on a large scale; the borrowing was done by Khmer elites who found in Indian religion a powerful language for kingship.
A favorite pre-Angkorian image illustrates the point. Harihara, a god who is half Shiva and half Vishnu, was especially popular in pre-Angkorian Cambodia, far more than in India itself. The choice suggests a local taste for joining two great cults rather than ranking them.
Worked example: the trouble with "Chenla"
Imagine two textbook sentences. The first says, "In the eighth century the kingdom of Chenla collapsed into civil war and broke into two states." The second says, "Chinese records of the eighth century describe two Chenlas, a northern and a southern one, while local inscriptions name many separate rulers."
The first sentence sounds confident, but it hides three choices. It assumes the Chinese name matched a real, unified state. It assumes the split was a collapse rather than a description of a world that was always divided. And it uses the word "civil war", which no source states. The second sentence reports what the sources say and leaves room for doubt.
This habit matters throughout Khmer history, because so many key labels came from outsiders: Chinese envoys named the early kingdoms, and French scholars later named the periods, styles and temples. Asking who chose a word, and what they could see, is the first step to reading the past on its own terms.
