CHAPTER THREE

THE ARCHITECTURAL PROJECT

The initiator of a building indicated to the master craftsman the general parameters of the proposed construction, and the project represented the sum of the constraints imposed. Excepting the eco• nomic and technical constraints, which could not be varied, the architectural particulars could be altered. Each structure is the ex• pression of a different project; even if sometimes two constructions seem follow the same building programme, they varied in their composition because of an evolution in techniques, the choice of materials, and above all the economic means placed at the disposal of each master craftsman. But as time went by the master craftsman liberated himself from architectural models, whilst still retaining in his memory an impor• tant fund of monumental knowledge. Given this, we have selected six complexes to reconstitute, in so far as it is possible, the original project: a palace, and five temples, , , the Western Mebon, Thomanon and the . While the structural programme of the palace is easy to unravel, those of the temples are extremely complex, especially in their sym• bolism. Each is, above all, the dwelling of the god, and as B. Dagens showed, is also the body of the divinity1• This in architecture is al• ready difficult to express, but each temple also comprises an icon• ographic programme which imposes its far from negligible shape and dimensions to a building. The temple thus comes to express the relationship among the gods in a genuine theogony. In 1952, dur• ing the excavations of the temple of Bat Chum, flagstones showing a yantra were discovered2 in the foundations of the north tower and in the interior of the central tower, in front of the base of its en• trance stairway. This yantra, which G. Coedes was able to reconsti• tute, is in the form of a chessboard with 49 squares containing the

1 B. Dagens, 'Le temple comme corps du dieu', Reuue de l'lnstitut des eludes indiennes, 1997. 2 G. Coedes, 'Un yantra recemment decouvert a ',JoumalAsiatique, 1952, pp. 465-477. THE ARCHITECTURAL PROJECT 19 letters of the Sanskrit alphabet3• In spite of the trouble which Coedes went to in order to link this yantra to the Buddhist divinities of the temple of Bat Chum, such as they are described in the inscriptions of the door jambs of the shrines, we believe that these flagstones formed part of the original function of the temple, which was to be Hindu; in effect, in the first instance, the yantra was linked to Brah• minism4. The design of this yantra is certainly linked to the layout of the towers of Bat Chum, for which a square had sometimes to be constructed with the following dimensions: 5, 5, and 7 for the hy• potenuse. It is obvious that the square is incorrect since 25 and 25 equal 50, not 49. This detail shows the degree to which the archi• tectural programme of a temple expresses not only religious sym• bolism but also acquired knowledge up to that point, a kind of sum• mum of structural intelligence. The significance of architecture everywhere evolves over time and sometimes a project is no longer clearly evident, or is at the very least profoundly altered. For example, the Bapuon (cf. part two, chapter five), which was conceived as a Sivaite temple, became Buddhist at the end of the fifteenth century. The initial programme was very often forgotten as beliefs changed, without the architecture being profoundly modified. The forms the master craftsmen thought had self-evident meaning had over time, lost their original significa• tion and acquired a new one, sometimes completely different. So our attempt to discover the original building programme which led to the laying out of these six complexes may seem to be a hopeless wager, but it is not completely impossible since the architectural history is often clearly discernible. Wooden buildings certainly comprised the essential Khmer home during the period under consideration. The forms which appear on the bas-reliefs of major temples have a very simple plan and the load• bearing structure is clear. Most of the buildings consist of one storey, which allows us to sketch a restoration (in fig. 98; part two, chapter eight) without being able to supply details concerning the space enclosed. However, there is in the large relief on the south side of

3 G. Coedes' reply to the article ofJ. Filliozat on this subject of the interpre• tation of a yantra in the first volume of L'Inde Classique, pp. 568-569. Yantra are an immediate projection of the divine image; the yantra is of greater value than the real image because it permits interpretations which are both comprehensive and more abstract, and which allow a real theogony to be established. 4 Cf. J. Filliozat, op. cit., p. 568.