This is just a little post about the Austroasiatic language family, sometimes known as Mon-Khmer. I have been reading quite a bit about it, and I thought I'd use this to summarise the topic, both for my benefit and, possibly, for yours.
The Austroasiatic language family is probably the oldest in mainland southeast Asia. Its members include Cambodia (Khmer) and Vietnamese, as well as a number of languages in eastern India (the Munda languages). The orang asli ('aborigines' or 'original inhabitants') of the Malay Peninsula also speak Austroasiatic languages, so the range of people who spoke Austroasiatic languages in ancient times includes blowgun-shooting horticulturalists, adivasis in South Asia, and monument-building Hindu-Buddhists in southeast Asia, including the Khmers and the Vietnamese.
The Khmers were, of course, responsible for a lot of lovely things, including Angkor Wat and other sites associated with the Angkorian state, which flourished in mainland southeast Asia between 800 and 1431 CE. Viet attacks on Angkor were one of the factors in its decline (along with environmental factors and, more importantly, the invasion of Tai-Kadai speakers - Thai and Lao people). The Vietnamese also ousted an Austronesian-speaking kingdom in the south, known as the Cham, in the second millennium CE, replacing most of the Austronesian languages of the eastern mainland with Austroasiatic. Austroasiatic speakers in India are separated from those in Cambodia and Vietnam by Thailand, which is Tai-Kadai-speaking, and Burma, which is primarily Tibeto-Burman-speaking (a sub-branch of Sino-Tibetan). There are also plenty of Indo-Aryan languages in the region (probably) formerly dominated by Austroasiatic, including Assamese and Bengali.
So the area is quite a messy one, linguistically-speaking. There are still a few Austronesian languages spoken in the highlands of mainland southeast Asia, including Jarai, but the largest concentration of Austronesian speakers is in the Malay Peninsula, which is usually considered under the maritime, rather than mainland, aspect of southeast Asia, due to its cultural connections with Java, Sumatera, and the rest of the Malay world (although in reality, Malay states often paid homage to the King of Siam before the advent of British control in the nineteenth century). Austroasiatic languages were probably spoken throughout mainland southeast Asia before the arrival of these Austronesian speakers in the south and east, and before the invasions of southeast Asia by Tai-speakers from what is now southwest China in the early second millennium CE. There may have been a simple and unambiguous spread of these languages from eastern India through Burma and out to Vietnam and peninsula Malaysia before the arrival of these other groups, and the Mundas would have been a little less mysterious in origin than they seem today.
That is the view most consistent with the linguistics of Austroasiatic. It seems like proto-Austroasiatic was a language spoken in a riverine environment, probably in mainland southeast Asia, around 4000 BP (BP = 'before present', so 2000 BCE). That's a late origin - a thousand years after the claimed Austronesian dispersal, and somewhat closer to the propagation of Indo-Aryan - but according to Paul Sidwell of the Australian National University, this is the most reasonable interpretation of the data. Earlier studies had claimed that it was a lot older, but this seems unlikely. It seems that proto-Austroasiatic was a dialect continuum that gradually spread out, instead of different branches splitting off. This means that instead of there being any nested sub-branches like 'Mon-Khmer' and 'Vietic', or even 'Munda' for that matter, there was actually heavy borrowing between all branches as they were forming. It was just one big language, proto-Austroasiatic, that, as it spread out, maintained some form of continuity between the speakers before they were cut off, resulting in borrowings that 'flatten' the family. Speakers of proto-Austroasiatic in Thailand may not have been able to understand speakers from Vietnam, but they would each be able to understand some other Austroasiatic speaker - a Cambodian, perhaps. It was this continuum of overlapping dialects that allowed new vocabulary to spread even as the speakers spread further apart from one another, distorting the branching of the family. This means that the immense age usually ascribed is not justified.
The reconstructed language has a lot of terminology for boats and river life, and living along rivers would have been a logical step in an environment of tangled vegetation such as is found in mainland southeast Asia. The river would provide a medium for transport as well as a source of food in the form of fish. It would also have provided abundant land for growing rice, which was and is the staple crop of mainland southeast Asia (and which came in many forms, including so-called 'floating' or deepwater rice). Rice agriculture was introduced to mainland southeast Asia from the
north between 2300 and 2000 BCE. It is a perfect crop to grow in the
riverine environment of the region, as it is a rather thirsty cereal.
The regular flooding of the rivers provided perfect land - land that would not have required buffalo to till the soil and removing the need to transplant the plants after sowing, and thereby reducing the amount of labour required for a high yield. It's quite easy to see how a river-based culture would spread far and wide, and why it might be more successful than societies without access to riverine resources given the surpluses provided by flood-retreat agriculture. That is probably how Austroasiatic became the dominant language family of the region. Its subsistence base provided a large amount of food in an environment that allowed for ease of transport, and it is easy to see how the language of the rice-farmers would have accumulated prestige in the eyes of other people in the area.
Rivers remained, and remain, important in mainland southeast Asia for agricultural purposes. Even at the height of the Angkor kingdom centred on Cambodia and Thailand, the rivers were the fertilisers of the soil. It was once thought that the large baray of Angkor - enormous reservoirs purposefully constructed by the citizens and slaves of Angkor under the direction of the king - were designed to provide water for irrigation, but there are no references to this in inscriptions and, indeed, at the height of baray construction, flood retreat agriculture was clearly still being practised. It seems that the baray represented the waters in the Hindu conception of the world, above which rose Mount Meru, the home of the gods, represented in Angkor by the great temples found between the baray. The Khmers were clearly obsessed with water, but they weren't obsessed with its agricultural properties. It is a common notion among archaeologists that other people must be as interested in subsistence as they are, but usually they aren't, and religious belief and state power seem to have acted as greater motivators for the building of these gigantic reservoirs than the desire to increase yield (which was probably already high, and supplemented by other methods anyway).
So Austroasiatic developed along the rivers of mainland southeast Asia around 2000 BCE, when rice agriculture became popular in the area. It is one of the few times in which the first-farmer hypothesis proves to be true: the first farmers in the area really did spread a language family. And it was a language family to have the most profound impact on southeast Asia, even allowing for the enormity of the Indian influence on the region. By way of Khmer art, architecture, and religion, ways of life with origins in Austroasiatic-speaking communities have retained considerable power even since the arrival of the Thais and Laos.