By Roy E. Jordaan
The first thing which should be reported is the reconstruction of the three temples, Candi A,
Candi Nandi, and Candi B, that stand across from the main temple [see ground-plan] and are
often mistakenly called candi wahana or v■hana temples. The
reconstruction of these temples was started in 1991, just before I left Indonesia, and was
completed in 1994. This is not the place to describe the exterior of the reconstructed temples,
the beauty of which can be admired and studied in detail by every visitor. Instead, I want to call
attention to the less well-known fact that, in the framework of the reconstruction, the Indonesian
Archaeological Service made a detailed study of the foundation of the Nandi temple taking this
temple apart to the last stone. The purpose of this exercise was to describe the composition and
structure of the foundation, before the temple was definitively reconstructed. In the course of
this activity several interesting matters came to light, such as some remains of so-called 'ritual
deposits'. It was further determined that the Nandi temple (just as probably all the other temples
in the central temple area) was solidly founded in the soil or 'anchored', so to speak (see below),
because the invisible foundation was constructed from a 3-metre thick layer of anthracite and
river-boulders. This lay on top of a thick layer of limestone blocks about 6 metres thick that had
been neatly placed on top of each other (personal communication Drs. Bambang Prasetya
Based on the results of the stratigraphic analysis of the soil around the Nandi temple and elsewhere, it can be noted that the soil profile of the temple area is structured in such a way that layers of different kinds of soil and other materials, such as sand, clay, gravel, and larger and smaller river-boulders, alternate. Contrary to Krom's (1923:451) suggestion, the local soil does not consist of sand, easily permeable by water. Rather, in view of the particular structure of the soil in the central temple area, permeability to water must have been practically zero. This is certainly true of the top layer, which, because it was intensively trodden upon, and contaminated by building materials, must have been so degraded since the beginning of the construction in the late 8th-early 9th century, that a so-called 'slaking' or 'sealing surface' came into being, through which water could hardly move at all (personal communication, Drs. W. Hoogmoed, soil tillage expert, University of Wageningen) This is supported by Soekmono's (1985:688) observation that 'the fine sand of which the soil seems to consist in reality is dried mud'.
These data adequately explain the poor drainage of the central temple area (a problem already mentioned in the Archaeological Reports of the colonial period) and are the reason for the rigorous measures taken by the Indonesian Archaeological Service to improve the drainage of the central temple area. These measures proved to be much more extensive than the few hand-dug gutters that I found in the western part of the temple area in 1991. No less than 20 concrete drainage pits were involved here, spread over the central temple area. From conversations with the official directly involved, and from the unpublished report Laporan pembenahan halaman pusat Candi Prambanan [Report on the improvement of the central courtyard of Candi Prambanan] (1993), it appeared that the drainage pits were connected to each other with an underground system of pvc drainage pipes. Furthermore, around each of the temples in the central temple area a ditch had been dug, filled with a layer of gravel. These were also connected to the above mentioned drainage pits.
Although understandable, and to some degree defensible, these measures are open to discussion and critique from a scientific point of view, because the question of the underlying reason for the drainage problems in the central temple area was never raised. Must we assume, as is usually done, that it concerns a short-coming or design error by the early Hindu-Javanese architects, or did these architects indeed intend the central temple area to be filled with water, based on considerations not fully understood, or not taken seriously enough by us. For some years now, I have been attempting to show that the latter is the case (Jordaan 1989; 1991).
Summarizing briefly, it is my opinion that in their concept of the temple complex, the architects wanted to give concrete form to the Hindu myth about the churning of the ocean, which, as is known, was the origin of amerta, holy water. On the basis of this myth, the temple complex was built in such a way, that the central temple area could be flooded with water on certain religious feast days, and function as a pool (tank) or a reservoir for the holy water that the priests made in a special temple ritual. This inundation of the compound brings me to the question of water supply, a point that up to now could only be guessed at. It seems, however, that a satisfactory answer has been found. In collecting the photographs and illustrations for the forthcoming book, I had the opportunity to consult the second part of IJzerman's Beschrijving der oudheden..., the so-called Atlas, which I was not, until recently, able to consult. On Plate XVII, fig. 67, there is a line [corresponding to the dotted line in the ground-plan], which indicates the course of an underground, stone water conduit: starting from the northern part of the third wall, it runs straight into and out of the temple complex, in the course of which it closely follows the contours of the central temple area, partially underneath the fourth row of subsidiary temples that have now disappeared. The course of the water conduit must certainly be based on technical considerations, in connection with the maintenance of the fall of the water, which had been diverted from the river at a higher point. This fall could not have been mauntained in the central temple area, which is constructed as a raised terrace. Old photographs of a stone water-course and an associated culvert, in the archives of the Archaeological Service (see OD photo no. 7760 and OD photo no. 11403-4), which, due to a lack of information in the accompanying report (Oudheidkundig Verslag 1931-1935), could not be pointed out before now, give a good impression of the size of the Prambanan water conduit. The culvert, in any case, was large enough for an adult man to stand up to waist-height in, which implies that it must have been fairly easy to raise the water from the river at these points, and transport it to the terrace. The new information concerning the course underground of the water conduit, as well as about the special structure of the soil of the central temple area, confirm the hypothesis that Prambanan was conceived of, and built as a holy water sanctuary. Architecturally, Candi Prambanan is as much of a wonder as Borobudur.
1993 Laporan pembenahan halaman pusat Candi Prambanan tanggal 15 Juni s/d 15 September 1993. [Bogem: Panitia Pemugaran Candi Wahana Candi Lorojonggrang Prambanan Daerah Istimewa Yogyakarta, Dinas Purbakala].
1891 Beschrijving der Oudheden Nabij de Grens der Residenties Soerakarta en Djogdjakarta. Batavia: Landsdrukkerij. [2 vols.]
Jordaan, Roy E.
1989 'A holy water sanctuary at Prambanan', Amerta: Berkala Arkeologi 11:17-41.
1991 'Text, temple, tirtha', in: Lokesh Chandra (ed.), The Art and Culture of South-East Asia, pp. 165-181. Delhi: Aditya Prakashan.
1923 Inleiding tot de Hindoe-Javaansche Kunst. 's-Gravenhage: Nijhoff.
1938 Uittreksel van de Oudheidkundige Verslagen van de Oudheidkundige Dienst in Nederlandsch-Indië, 1931-1935. Batavia: de Unie
1985 'Lumpur dalam konstruksi candi', in: Sulastin et al. (eds.) Bahasa-sastra-budaya: Ratna manikam untaian persembahan kepada Prof. Dr. P.J. Zoetmulder, pp. 684-696. Yogyakarta: UGM.
Dr Roy E. Jordaan obtained his PhD in Anthropology from the University of Leiden.