The Brahmins who brought Indic culture and learning to the royal courts of mainland Southeast Asia during the early centuries of our era continued to play that role throughout the Classic period, and in the royal palaces of Cambodia and Thailand, right into modern times. These intellectuals acted as priests of the temples, teachers, royal chaplains, librarians, astrologers, and in all likelihood architects and calligraphers. It was they who cast the horoscopes for all important events, who interpreted the Vedas and the Hindu laws to the empire’s power brokers, who designed temples in which the great gods could reside, who conducted all ritual, who tended and carried the Sacred Fire, and who kept the calendars.
All this learning depended upon writing. There are over 1,200 inscriptions known for the ancient Khmer world, almost all from the Early Kingdoms and Classic periods. These were incised into polished stone, and most appear on the doorjambs of temples and on free-standing, four-sided stelae. They are read in horizontal lines from left to right, and from top to bottom, in a complex alphabet derived from the Nagari script of India. There are two kinds of inscriptions. The most prestigious were in Sanskrit, and almost always in the form of poem; as Claude Jacques comments:
These inscriptions were placed under the gaze of a particular god and seemingly were intended to attract that deity’s attention to the person who had had the sanctuary built in his honor or, more often, who was offering him gifts. Most of these donors were kings, the poems being composed upon their death. They were accompanied by a short eulogy (prashasti). Prose texts in Old Khmer comprised the other kind of inscription, frequently appearing on the same stone with the Sanskrit one; these had a very different, more prosaic, and far more informative subject matter. According to Jacques, the overwhelming majority are inventories listing the temple’s possessions – land, livestock, servants and furnishings. Some end with an imprecation formula, for example putting a curse upon any violator of the terms of the grant ‘as long as the moon and the sun shall last’.
These texts are generally fixed in time by the intricate calendar system of ancient Cambodia, itself partly dependent upon astronomical and astrological considerations. The solar year is given in terms of the Great Era (saka) that began on the Vernal Equinox of AD 79; thus, one is to add 78 to the saka date to reach a year in our system. The digits making up the saka numbers may be spelled out alphabetically, or they may be given by chronograms: for example, saka 1044 (AD 1122) might be given symbolically by ‘oceans [4], ‘oceans’ (again)[4], ‘sky’ [o], ‘moon’ [1]. There were 12 lunar months, each divided into a 15-day waxing period and a 15-day waning one. The astrologers were deeply interested in the current position of the moon against the band of stars that runs along the ecliptic in a kind of lunar zodiac; since the sidereal month is about 27 days, there were 27 of these ‘lunar mansions’ or nakshatras, each with an animal name (the moon generally traversing one mansion a day). Because the lunar calendar was always running ahead of the solar one, extra lunar months were occasionally intercalated in a complex system of Indian origin.
By Zhou Daguan’s day, solar years were also expressed in terms of a 12-year cycle, each year named for a specific animal, a system that they had borrowed from the Chinese – perhaps a reflection of Cambodian’s rapidly increasing trade with the Middle Kingdom.
Like their counterparts in peninsular India, the Cambodian astrologers were close observers and calculators of the positions of the five visible planets (Mercury, Venus, Mars, Jupiter and Saturn) as these moved across the solar zodiac, which was essentially the same one that is still in use in the Western world. One more calendrical statement appearing in the inscriptions is the day of the seven-day week, each day being linked to one of the planets or to the sun or to the moon, as it is with us. According to Zhou, the ordinary Khmer had no family or personal names, but were known by the day of the week on which they were born.
The Yugas – the huge cosmic cycle of successive creations and destructions – did not enter into their calendrical computations, but they certainly played a role in Cambodian cosmology, as Eleanor Morón Mannikka has shown in her study of the proportions and measurements of Angkor Wat.
Zhou Daguan makes no mention of inscriptions, but he does talk about writing on perishable materials: there are the manuscripts, which probably existed in quantity in the libraries, state archives and temples of Classic Angkor. Not one, however, has survived the vicissitudes of time, history and tropical climate, a tragedy for Khmer scholarship. The religious texts, whether Brahmanic of Buddhist, were contained in palm-leaf books or sastra; these consisted of fronds about 50 to 60 cm (20 to 24 in) long, bound together into a stack by loose cords. Each leaf was incised with a stylus, and the scratched lines filled with lampblack. In the Angkor Wat reliefs, Brahmin pandita or gurus accompanying the Sacred Fire carry such books in their hands or on their shoulders, while at Banteay Chhmar, a pandita reads one accompanied by a Khmer theorbo
It is likely that all secular books were paper screenfolds. Such accordionlike manuscripts were being produced in Cambodia until the middle of the twentieth century. The paper was manufactured from the inner bark of a member of the mulberry family; it was softened by soaking, then wrung out and finely shredded to separate the white from the brown fibres (these latter being used to produce black paper, the kind mentioned by Zhou). After this had been boiled with white lime and then washed and pounded, the resulting paste was spread onto cloth or screens and left to dry in sheets. White paper was treated with rice powder mixed with water and chalk, and black paper with soot or charcoal. The final stage was to polish the surfaces, and fold the paper into books. Manufactured paper was imported from China; while there is no mention of its use to make books, Zhou reports that the natives derived great amusement at seeing the Chinese use it as toilet paper.
Zhou describes the chalk pencils that were used to write black pages, and says that such pages could be easily erased; accordingly, official documents, such as revenue, corvée manpower and census tallies must have been kept in the white-paged screenfolds, which were written in black ink, probably with bamboo and/or metal pens. Assuredly some of these paper books contained astronomical tables, for Zhou assures us that their astronomers could calculate solar and lunar eclipses – an impossibility without the accurate accumulation of observational data over a very long period of time.
How literate were the Classic Khmer? Surely all the Brahmins of the empire could read and write, and so could the kings and princes, all of whom had been instructed by Brahmin teachers (in contrast to contemporary European rulers such as Charlemagne, who were often illiterate). The vast civil bureaucracy would have found it in their interest as revenue gatherers and beneficiaries to be literate, too. Both Mahayanist and Theravada Buddhist monks would by Sangha rules have to be able to read and recite the sacred texts of their faith. Add to this list the masters of works, the architects, and the master craftsmen who worked in stone and metal, and one can conclude that a substantial minority during Classic times was lettered. Nonetheless, the great majority of Khmer – the free peasants and the slaves – would have been unable to ‘read’ anything but the imagery of the reliefs and sculptures.
Angkor: City and State
With the exception of B.-P. Groslier, with his vision of an immense ‘hydraulic city’ containing almost two million souls, until recently few scholars had devoted much thought to what kind of city Angkor really was. As Roland Fletcher has said, ‘Angkor still needs to be reappraised as a place where people actually lived.’
There were many cities during pre-modern times in both mainland and insular Southeast Asia. Following a dichotomy first recognized for medieval France, John Miksic of the University of Singapore has proposed that they fell into two groups. Heterogenetic cities were found along coastlines and at the borders of ecological zones rather than at their centres; they had few public monuments, and were characterized by entrepreneurship and intensive trade, as well as by high population densities (pro-colonial Malacca would have been an excellent example of such a city).
Orthogenetic cities were located well inland, and were correlated with the production of a surplus staple crop – that is, rice – which could be commandeered by the authorities. Stability and ritual were the prevailing order, and there were impressive monuments of a religious nature. There was no money and little evidence for large markets and significant trade. ‘The permanent population of the orthogenetic city was composed of nobles, civil, religious and military bureaucrats, and their staff.’ In contrast with heterogenetic cities, overall population density was very low. From everything that we know about Angkor, it would appear to have been orthogenetic. Moving away from our area, so would have been the monumental Classic Maya cities such as Tikal, Copan and Palenque in Mesoamerica, with their royal courts and extremely dispersed patterns of settlement.
A clue as to what at least part of Angkor might have looked like comes from the old Siamese capital of Ayutthaya in Thailand, founded in 1351 and destroyed by the Burmese in 1767. It was a conscious clone of the Khmer capital, Angkor Thom, and covered about the same area; instead of being bounded by a huge moat, it was surrounded on all sides by rivers or by connecting canals, and by a wall. An account of Ayutthaya by a seventeenth-century Dutch traveler states:
The Streets of the walled Town are many of them large, straight and regular, with channels running through them, although the most part of small narrow Lanes, Ditches, and Creeks most confusedly placed; the Citizens have an incredible number of small boats…which come to their very doors, especially at floods and high water.
Plans and watercolour drawings by Europeans show that it was crisscrossed by canals and streets, with the royal palace in the northwest sector (as in Angkor Thom); the only densely settled sector lay in the southeast. Comments by an early eighteenth-century observer are relevant here:
Considering the bigness of the City, it is not very populous…scarce the sixth part is inhabited, and that to the South-East only. The rest lies desart [sic] where the Temples only stand…there are abundance of empty space and large gardens behind the streets, wherein they let nature work, so that they are full of Grass, Herbs, Shrubs and Trees, that grow wild…
The houses of ordinary inhabitants were thatched, single-storey structures of bamboo and wood, built on piles, while foreign traders lived along the main north-south avenue in more substantial tile-roofed houses. Ayutthaya, whatever its Angkor-inspired beginnings, was slowly evolving from an orthogenetic to a partly heterogenetic city, due to the easy access that Chinese, European and Arab traders had from its waterways. Let us first consider Angkor Thom; in recent years its four quadrants have been surveyed in detail by Jacques Gaucher of the EFEO, using aerial photographs and ground ‘truthing’. The main axes of Angkor’s capital district were lined with canals, and, again like Ayutthaya, the Royal Palace was in the northwest quadrant; elsewhere, apart from the monumental constructions, there were numerous small water tanks, channels, and house mounds. Based on the results of this survey, Roland Fletcher suggests that while Angkor Thom could have held as many as 90,000 people (assuming a density of 100 persons per hectare), the population may have been only a quarter of that, given the amount of open space (as in Ayutthaya); the palace; the major temples; and the single-story dwellings.
Turning now to the city of Angkor as a whole, a survey carried out there from 1992 to 1998 by Christophe Pottier has shown that this landscape was dotted with low mounds that had once supported hamlets of about five to ten traditional, single-storey houses. These mounds were associated with hundreds of small, local shrines and medium-sized, rectangular water tanks, recalling Zhou Daguan’s statement that ‘every family has a pond – or, at times, several families own one in common’. Based on ground survey and upon radar imagery and aerial photography, Fletcher now estimates that the total area of Angkor’s urban complex is about 1,000 square km (386 square miles), within which the people were mainly living along linear features – canals and roads that extend out from central Angkor for about 20 to 30 km (12 to 18 miles) in all directions, probably less and less densely occupied as one moves towards the peripheries. Angkor Thom, then, was like a kind of spider sitting in the centre of a virtual web of settlement, with large open spaces, including even rice fields, between the ‘threads’. This web extended well north of Preah Khan into the foothills of Phnom Kulen; the lovely Banteay Srei was probably at its northern edge.
In Fletcher’s words, ‘Angkor was therefore a low density, dispersed urban complex with housing along linear features and scattered across the landscape in patches and on isolated mounds.’ Groslier’s estimate of 1.9 million persons is thus an impossibility. The true figure for Angkor at its apogee, during the reign of Jayavarman VII, was probably a fraction of this, but only a great deal of future research can give us an idea of the total population.
The system of government that made Classic Khmer civilization and the city of Angkor possible was a highly effective and powerful ‘top down’ one, supported by a command economy and by a massive and all-encompassing, revenue-generating bureaucracy operating on every level from palace to village. Historian David Chandler has summarized this kind of government in both Cambodia of the 1860s, and Cambodia under the rule of Angkor.
In both cases…government meant a network of status relationships whereby peasants paid in rice, forest products, or labour to support their officials. The officials, in turn, paid the king, using some of the rice, forest products, and peasant labour with which they had been paid. The number of peasants one could exploit in this way depended on the position one was granted by the throne; positions themselves were for sale, and this tended to limit the officeholders to members of the elite with enough money [absent, of course, in Classic times] or goods on hand to purchase their positions.
In an influential 1982 essay, Cornell University historian O. W. Wolters proposed that many of the early polities of Southeast Asia, including the Angkorian one, were mandalas; this is a technical Sanskrit term meaning ‘a particular and often unstable situation in a vaguely defined geographical area without fixed boundaries and where smaller centres tended to look in all directions for ssecurity.’ A mandala was a ‘circle of kings and brahmins’, in which one king would lord it over lesser ones, and those kings in vassal or tributary status would continuously try to repudiate this and build up their own network of vassals. This would certainly apply to what we know was going on among rival kings in mainland Southeast Asia prior to AD 802. But for most of the time during the next five centuries, in spite of sporadic revolts and foreign invasion that could have occurred in any empire, the Angkor state – that is, the king – had complete control over all Khmer territory. This was enforced out to Cambodia’s frontiers not only by his army and his judicial system, but by his roving inspectors (the tamrvach) and by periodic census-taking. ‘Supreme and extensive political dominion’ is how the Oxford English Dictionary defines ‘empire’. The city of Angkor may not have looked like imperial Rome, but the Classic Khmer Empire over which it ruled lasted as long as the Roman one.