Candi Singosari (Candi A)
Javanese temples are usually composed of three elements: a base, a temple body and a roof. The temple body is the only part of the building that can actually be entered. In Candi Singosari these divisions are not as clear. The temple has a low base and a roof, but the section that contains the temple chambers with the divine images has the form of a second base, while the structure on top of it is shaped like a temple body and contains a chamber-like space. It does not, however, contain an image and cannot be entered. Because of this unusual composition, the divine images seem to be placed within the base, while the part that most resembles the temple body is empty.
Many theories have been proposed to explain this architectural peculiarity. Some scholars think that Singosari is a double temple, with the lower part dedicated to Shiva and the upper part to Buddha, and that it demonstrates the precedence of Buddhism over Hinduism. However, no traces of any Buddha sculpture have been found. Other scholars, acknowledging the Hindu-Buddhist character of the temple, argue that the upper chamber was left empty on purpose: it symbolized, they suggest, the Buddhist concept of emptiness (shunyata), according to which all materiality in this world is impermanent and ‘empty’ of the identity attributed to it.
The temple itself does not provide any other evidence to confirm the hypothesis that Candi Singosari was both Hindu and Buddhist. Several Buddhist sculptures have been found in the village, but these may have belonged to one of the other temple structures at the site. So far, no sculpture or ornament from Candi Singosari itself has been found to have clear Buddhist features. On the other hand, the sculptures of Durga, Ganesha and Agastya in the temple chambers clearly confirm its Hindu character. One must conclude that, until there is real proof to the contrary, Candi Singosari was a Hindu sanctuary. It remains an enigma, however, as to why the divinities of the temple seem to have been placed within the base.
The temple and Mount Meru
Originally, each of the five rooms of Candi Singosari had an independent roof. Viewed from afar, the temple would have looked like Mount Meru, as it is described in the sacred texts of Hinduism and Buddhism: a central peak surrounded by four secondary peaks marking the directions of the compass.
During this period, it was common to compare temples with Mount Meru. The sacred mountain was not only the abode of the gods, it was also the centre of the universe and temples were perceived as replicas of the universe. The Nagarakertagama, for example, describes the sanctuary of Kagenengan, south of Singosari, as ‘imposing in appearance, holy, high, of the aspect of the mountain Meru’. According to Indian tradition Shiva resides on Mount Kailasha and this mountain is often, also in Java, identified with Mount Meru. Thus, the Tantu Panggelaran, a late, probably sixteenth-century, East-Javanese text, states that the Mahameru (the Great Meru) was also called Kailasha.
The same text tells the story of the transfer of the mountain from India to Java. In that story we find yet another name for Mount Meru, namely Mandara, the name of the mountain that was used by the gods as a churning stick to churn the ocean. According to the Old Javanese text, several pieces of this mountain, which was carried by the gods to Java, became various mountains in Java. Thus, the top became Mount Penanggungan (to the south of Surabaya), and the existing Bromo (Brahma) had to support the last remaining part of this mountain (now Semeru, the highest mountain on Java).
After the structural work had been completed, the stone carvers decorated the temple with ornamental relief work. Plants, flowers, garlands, birds and mythological animals were carved on the outside of the temple to enrich the metaphor of the cosmic mountain and turn the building into a god’s palace. For unknown reasons, the ornamentation of Candi Singosari was never finished. This gives us a unique opportunity to see the sculptors at work. Thus, we know that the decorations were carved starting at the top of the temple, and that the ornamentation was traced on the stone with black paint before being carved. This can still be seen in the scroll ornament around the niches of the temple body.
Apparently, the sculptors first cut the stone following the drawing, before adding relief and details to the carving. The larger sculptures, such as the protruding kala heads, were probably rough-hewn before being placed on the temple; the sculptors would then have finished them afterwards, but their work was interrupted and never resumed. The roof is the only part of the temple where the ornamentation was finished; the temple body was never completed.
Although very few remain today, numerous ornamented stones belonging to the superstructure were found in neighbouring villages during surveys carried out in the early twentieth century. J.L.A. Brandes and his team were able to gather enough pieces of the puzzle for the architect H.L. Leydie Melville to propose a satisfactory reconstruction of the temple roof.
According to Melville’s drawing, published in Brandes’s 1909 monograph, the superstructure was composed of three storeys of diminishing sizes and topped by a finial in the shape of a lotus bulb. Each storey rested on a plinth and was crowned by a cornice. Both the plinth and the cornice were interrupted by antefixes (at the centre of each side and at the corners). The three storeys were adorned by 16 turrets each (5 turrets were thus visible on each side), protruding from the back wall. The central and corner turrets were temple replicas with a small false door topped by a kala head. On the wall, in the space between the turrets, triangular banners with hanging leaves and bead pendants were carved.
The plinths and cornices of the three storeys were each decorated with a different set of motifs. On each side of the central antefix of the plinth on the first storey (the only one remaining today) are two lions with their backs to each other and facing a makara (a mythical aquatic creature with a crocodile’s mouth and an elephant’s trunk). On the same storey, the cornice shows griffon-like creatures facing one another. On the second storey, the plinth is decorated with geese and the cornice with peacocks, while on the third storey, a series of stylized antefixes replace the animals.
All these motifs are surrounded by lush vegetation, giving the roof the appearance of a garden of paradise. As in the real world, aquatic and terrestrial animals (makaras and lions) are depicted on the lower part of the roof, while birds occupy the upper parts.
Only the cornice, the upper frieze, the kala heads and the frames of the niches are decorated. Further ornamentation was probably planned, in a similar fashion to that found on Candi Kidal, a slightly earlier temple, also in the Malang region. There the band running at mid-height around the temple body is decorated, as are the antefixes marking its corners. Candi Kidal also has decorated plinths and the walls are decorated with medallions.
Today, the most striking elements of ornamentation are the four large kala heads crowning the niches. Such large monster heads are emblematic of ancient Javanese architecture. Although kala heads also occur in Indian and Khmer art, where they are known as kirtimukha, they were not as popular in these places as in Java, where monster heads occur above almost every entrance or niche. There is an Indian myth which explains the birth and function of the monster head.
The motif of the kala head is also depicted on the cornice that crowns the temple body. The kala heads are depicted en face, within an antefix-like frame, and flanked by a foliate spiral. The motif is unique to Candi Singosari and cannot be found on any other Javanese temple. At the centre of the frieze, one can see an antefix adorned with a winged conch shell. Below the cornice, at the top of the wall, runs a frieze. The frieze shows garlands hanging from foliate flower motifs and birds perching on the garlands.
When the sculptors were interrupted in their work, they were carving the niche frames. Only the southern niche gives us a full glimpse of what the ornamentation should have been like. At the base of the niche jambs one can see a lotus stem emerging. The stem climbs up along the jambs, looping, and ends at the centre of the lintel. A bird stands within the last loop. At regular intervals a small branch stems out from a node of the main stem and transforms into a lotus flower.
Art history comparisons
Both the architecture and the ornamentation are unique. Hence it is difficult to compare the monument with other temples and to place it clearly within the history of Javanese art. Various elements show links to earlier Central Javanese art (eighth to early tenth century), while other elements are characteristic of the East Javanese style of the thirteenth to fifteenth centuries. Because there are hardly any known temple remains from the intervening period (eleventh to twelfth centuries), it is difficult to interpret this ambiguity. Although architecturally Candi Singosari clearly belongs to the East Javanese tradition (thirteenth-fifteenth century), its style is atypical and bears certain similarities to the Central Javanese tradition (eighth to early tenth century).
The question remains why? There are actually two ways of interpreting the relationship between Candi Singosari and Central Javanese art: either the temple dates back to the early East Javanese period, of the eleventh or twelfth century, from which period little comparable architecture exists; or the sculptors who worked at Singosari were inspired by older monuments. The Central Javanese traits of Candi Singosari would then be the result of an archaistic style. Only further analysis of other data sources - sculptures, inscriptions, et cetera - will help us to decide which of these hypotheses is the most plausible.
The temple bears a number of characteristics that seem to have become fashionable from the thirteenth century onwards, such as the placement of the kala heads on the tympanum (rather than on the lintel), the band midway around the temple body, the fact that the turrets of the superstructure are in relief and thus part of the wall, and the fact that triangular ornaments decorate the spaces between the turrets.
Interestingly some of the ornaments seem to have been inspired by earlier Central Javanese traditions from the eighth to early tenth centuries, even though they are depicted in the low relief characteristic of the East Javanese style. The frieze of birds on the temple body is one clear example. Similar ornaments with birds perching on the garlands are, in a different style, common on Central Javanese temples.
The clearest resemblance is to a similar ornament on Candi Gunung Gangsir, an East Javanese brick temple that has Central Javanese ornaments carved in a Central Javanese style. The age of this temple is still an enigma. Its architecture corresponds to later East Javanese forms of architecture, but its ornaments have Central Javanese forms and correspond to the Central Javanese style. The decorated bricks and terracotta plaques were made using moulds, which could either have been from an earlier period or made from stone-carved Central Javanese motifs.
The treatment of the turrets is also quite unique. While, in other East Javanese temples, the five turrets are identical, this is not the case at Singosari. The turrets placed at the centre and at the corners of the roof are distinguished by the presence of false doors, following the Central Javanese tradition in which turrets are miniature temples, mimicking the tri-partite structure of the building.
The kala heads of the temple body also show similarities with some Central Javanese counterparts: they have no lower jaw and their paws have been transformed into a plant-like design.