Buddhist Art in Java

An essay on ‘Buddhist Art in Java’ by A. Foucher.

translated by L.A. Thomas and F.W. Thomas as Chapter VIII of
‘The Beginnings of Buddhist Art’

re-edited with photographs by
Ānandajoti Bhikkhu

A classic essay by Prof. Foucher identifying the main Avadāna reliefs at Borobudur, with photographs.



use j/k or left/right arrow
to navigate through the photos below

III: Buddhist Iconography in Java

1. Borobudur

We shall not undertake a detailed review of the bas-reliefs deployed along the upper galleries of the stūpa. We restrict ourselves to noticing that, as we mount, they assume a character more and more iconographic, less and less “narrative”, and that the edifying story finally [255] gives way to the image of piety. But see, supra, the identification of one of the bas-reliefs of the upper gallery, pp. 165-6 and pl. XXII (Great Miracle of Śrāvastī). Buddha, monks, nuns, Bodhisattvas of both sexes file past in twenties, at times seated under trees more or less stereotyped, most often installed under the open porches of temples, just as they are seen on the miniatures, or the clay seals, of India. Cf. Etude sur l’iconogr. bouddhique de l’lnde, I, 1900, pp, 45-6. The sculptors weary so much the less of all these repetitions as each one of them represents so much progress in covering the considerable surface which it was their task to decorate. There would be no advantage in noting here and there in passing a few specially characteristic figures, such as, in the second gallery, some Avalokiteśvaras with four or six arms, and a Mañjuśri carrying the Indian book (pustaka) on the blue lotus (utpala); or again, in the third gallery, a group composed of a Buddha between these same two Bodhisattvas, etc. The problem is much more vast, and demands a solution of very different amplitude. It would be necessary to make a census of all these images and each of their varieties, to draw up an exact and complete table of them, and to study attentively their graphic distribution; then only, after having allowed for the necessities of decoration and having among this crowd of idols discerned the really essential types, we might attempt the identification of what for the artists of Java constituted the Buddhist pantheon. We must hope that some Dutch archaeologist will find time to undertake this delicate and extensive task; it is unnecessary to say that it is forbidden to a simple visitor.

Neither shall we dwell upon the hundreds of statues which decorate this stupa of the “Many Buddhas” (for such would be the meaning of the word Borobudur): [This meaning is no longer accepted as correct] but [256] here the reason for our abstention is quite different. They were, in fact, classified long ago, and W. de Humboldt proposed to recognize among them, in accordance with Hodgson’s Nepalese drawings, the images of the five Dhyāni-Buddhas. The identification has since been generally admitted, and in principle we see no reason for contesting it: at the most it would need to be pressed further and completed. The arrangement of the groups must in any case be remade. Among these manifold replicas with heads generally well treated and expressive, but effeminate and bloated bodies, all seated in padmāsana and only differentiated by the gestures of the hands, we must, in fact, distinguish:

1. in the four first rows of niches (in the proportion of 92 to each facade), to the east, those in bhūmisparśamudrā; For the mudras, or gestures of the hands, cf. ibid., p. 68.
2. at the south, those in varamudrā;
3. at the west, those in dhyānimudrā;
4. at the north, those in abhayamudrā;
5. in the fifth row of niches, on the four facades (viz. 64 altogether) those in vitarkamudrā;
6. in the 72 little open cupolas of the three circular terraces, those in dharmacakramudrā;
7. the single image found under the great central cupola.

Whatever identification may be proposed, will, it is understood, have to take into account each of these varieties, without omission and without confusion. Therefore we cannot admit that of Humboldt, Cf. Leemans, loc. cit., p. 480. which confuses and mixes up nos. 4 and 5. If we must identify [257] 1 . Akṣobhya, by the gesture of calling the earth to witness, 2. Ratnasambhava, by the gesture of giving, 3. Amitabha, by the gesture of meditation, 4. Amoghasiddha, by the gesture of protection, it is clear that in the last row of niches we must recognize, 5. the fifth Dhyāni-Buddha, Vairocana, by the gesture of discussion, although the gesture of teaching is more usually reserved for him and although, on the other hand, the vitarkamudrā is scarcely distinguished from the abhayamudrā by the fact that in it the index-finger is joined to the thumb. It follows likewise that with the five rows of niches belonging to the polygonal galleries we have, as was natural, exhausted the list of the five Dhyāni-Buddhas.

6. The 72 images of the circular terraces would then all be consecrated to the historic Buddha, Śākyamuni, and would exhibit him teaching.

7. As for the purposely unfinished statue which was discovered under the great central cupola, it has been the subject of many hypotheses. Dr. Pleyte regards it as the last enigma of Borobudur: “The great Dāgaba”, he says, C. M. Pleyte, Die Buddhalegende in den Skulpturen des Tempels von Boro-Budur, Amsterdam, 1901-2, p. ix. For the bibliography see ibid., notes on pages i-iii. “was formerly without any opening; but at present one can have access right into the interior, part of the wall having been removed. The removal brought to light a hidden image of Buddha, which represents him seated in bhūmisparśamudrā. This image of Buddha is thus the centre of the sanctuary. By reason of its incomplete form it is considered by Groeneveldt to be a representation of the Ādi-Buddha. This would be a manner of symbolizing the abstract essence of this supreme divinity of Mahāyānism. Kern, on the contrary, recognizes in this unfinished [258] figure an embryo Buddha: this would be an allusion to the Bodhisattva in the womb of his mother...” If these diverse interpretations fail to satisfy us any more than they did Dr. Pleyte, the short resume which he gives of them is at least sufficient for our purpose. We do not indeed pretend to discuss here the greater or less degree of probability in these theories. Still less shall we stop to criticize that of Wilsen, who saw in this same statue a rough model of a future Buddha, prepared for subsequent completion by the cunning priests. Cf. Leemans, op. cit., pp. 486-7. In truth, speculations of this kind are scarcely more susceptible of refutation than of proof; and it is this which makes us suspicious of them. If we in our turn venture a new hypothesis, it is because we should prefer to seek the solution of this problem of archaeology elsewhere than in the messianic, symbolical or theistic conceptions more or less familiar in such and such forms of Indian Buddhism.

Let us, then, make a tabula rasa of all this metaphysics and consider again, as briefly as possible, the essential elements of the question. Under the central dome of the stūpa of Borobudur, at the spot where we should expect to find the usual deposit of relics, or at least the upper deposit – for it happens sometimes that there are along the perpendicular which joins the summit to the base several of them, one above the other– was discovered an image of Buddha, whose emplacement sufficed to prove its specially sacred character. Now this statue was intentionally left unfinished; “The hair, the ears, the hands and the feet are not completed”, says Leemans; and further on he adds: “One is forced to admit that the artist who made the plan of the whole really had a premeditated intention [259] of leaving the statue of the central sanctuary in the state in which we possess it”. See the discussion, loc. cit., pp. 484-6. On the other hand, this image shows us Buddha seated, his legs crossed in the Indian manner, the left hand resting in his lap, his right hand hanging down, the palm turned inwards and the fingers stretched toward the ground. Before committing ourselves to any apocalyptical explanation of this figure it is well, in point of method, to ask ourselves, first of all, whether the iconography of India, the recognized model for that of Java, does not comprise any type of Buddha composed in the same attitude and presenting the same peculiarity of incompletion.

If it were permissible to judge by the facility of the solution, the question would in this case be well put: at least, in order to answer, it is not necessary to push far our interrogation of the Indian tradition. The two most celebrated prototypes of the pretended portrait images of Buddha are that of Kauśāmbī (or Śrāvastī) and that of Mahābodhi, near to Gayā. The former is in this case out of the question. Concerning the second we possess two versions of an identical legend, the one reported by Hiuan-tsang, the other by Tāranātha. See, for the first, the translation of Stan, Julien, II, pp. 465 sqq., or of S. Beal, II, pp. 120 sqq.; and, for the second, the translation of Schiefner, p. 20. Anxious before all to guarantee the authentic resemblance of the image, they naturally attribute its execution to a supernatural artist: on two points they seem none the less in harmony with historic truth. First of all, we learn from the texts, in the most formal manner, that the original work was regarded rightly or wrongly, matters not as not being finished, an accident which people were unanimous in explaining as due to an unfortunate [260] interruption in the mysterious work of the divine sculptor. Among the unfinished parts Tāranātha cites especially the toe of the right foot and the locks of hair. Whilst these material details would he less easy of verification than might be thought in the obscurity in which, as Hiuan-tsang tells, the majesty of the idol was hidden, there is some appearance that this general belief in its state of incompleteness was in one way or another well-founded. In the second place, and in any case, it is a fact attested by the monuments, as also by the descriptions of the texts, that it represented Buddha seated, the left hand at rest, and the right hanging, at the moment when, disturbed from his meditations by the assaults of Māra, he touched the earth with his fingers, in order to invoke it as witness. The references are to be found in our Etude sur l’Iconographie bouddhique de l’Inde, I, pp. 90-94. In short, the image of Vajrāsana of Mahābodhi, to use the term under which it was known, made the gesture of bhūmisparśa, and was, or which for us comes to the same thing, passed for being, incomplete.

We leave to experts the task of concluding. To us this double rapprochement appears sufficiently precise to allow of our putting forward the idea that the central Buddha of Borobudur, incomplete and in bhūmisparśamudra, is, or at least intends to be, nothing but a replica of the statue of Bodhgayā. In addition to its simplicity, the hypothesis has also this great advantage, that it frees us from the necessity of attributing exceptionally to the artists of Java, always so respectful towards Indian tradition, the creation of a new model which India would not have known. Finally, if it does away with one difficulty, in our opinion a considerable one, we do not think that it raises another in its [261] place. It is a fact historically established by Chinese evidence that from the VIIth to the XIth century of our era, that is, during the period covering the construction of Borobudur, which is attributed to the second half of the IXth century the “True Visage of the Throne of Diamond”, or “of Intelligence”, was the most venerated Buddhist idol in India, and even the model most in request of exportation, Cf. Ed. Chavannes, Les Inscriptions chinoises de Bodh-Gayā in Revue de I’Histoire des Religions, vol. XXXIV, I, 1896. whilst the temple of Mahābodhi had become the greatest centre of pilgrimage. This would explain without effort why a more or less faithful copy of this miraculous image should have been able to assume a character sufficiently sacred to merit being placed by the Javanese architects in the hollow of the great stūpa of the Indian Archipelago, just as the original reposed under the niches of the famous sanctuary of Magadha.

Such, at least, is the hypothesis which we could not help long ago submitting to Indianists, with all the respect inspired by the experience of our predecessors and the reservations imposed by the necessity, in which we still were, of trusting to the descriptions of others. At the time of our visit to Borobudur we found nothing to add concerning this statue, inasmuch as it was still in the same state in which Dr. Pleyte had seen it, once again covered up to the neck and left in a state of abandonment very unworthy of all the ink which it had caused to flow. Thus we were obliged to restrict ourselves to reiterating the wish that it might once again be cleared and more closely studied. If we have returned in some detail to this subject, it is because in the interval this wish has been fulfilled, and because the kindness of Major Van Erp allows [262] us to produce at last a photograph of this famous idol. [A more recent photograph is substituted above, courtesy of Okkisafire.] Perhaps this latter will be for the reader a disillusionment: in fact it merely sketches in a rather rough fashion the ordinary type of Buddhas of Borobudur, and it is quite clear that, if a replica of the image of Vajrāsana is really intended, it was executed freely and not from a moulding. But upon a moment’s reflection it will be seen that this was exactly what was to be expected: and, in any case, it is well once for all to place before the eyes of the public the decisive piece of evidence in a dispute which otherwise would run the risk of being endless.

2. The Candi Mendut

It would be a task more within our reach to identify, by way of a specimen, the images which decorate the Candi Mendut. This edifice, placed in the axis of the oriental gate of, and at three kilometres from, Borobudur, consists, in fact, of a cella only, with a vestibule in front. The whole is, according to the Javanese custom, perched on a terrace in the same manner as are the Brahmanic temples of Prambanam. In Buddhist terminology it is what is properly called a vihāra. We know that the meaning of this term (temple of divinity or monk’s cell) has been unduly extended by European archaeologists to the whole of the monastery (Cf. Art gréco-houddhique du Gandhâra, p. 99). – We deliberately leave aside the other Buddhist edifices which we likewise visited in the neighbourhood of Jogyakarta under the guidance of Dr. J. Groneman, and on which we may consult his guide, entitled Boeddhistische Tempelen Klooster-Bouwvallen de Parambanan-Vlakte, Soerabaia, 1907. Naturally it shelters statues, and the walls of its entrance vestibule, like the exterior faces of the building, are decorated with figures whose purely Buddhist character may be recognized at once by anyone who is a little familiar with the Indian iconography of this religion. The building, fairly well preserved, except in the upper parts, has been the object of a restoration the architectural [263] details of which we shall not undertake to discuss. The three enormous statues of the cella have been replaced on their pedestals. Cf. B.E.F.E.O., IX, 1909, p. 831. They are characterized by a curious detail. Whereas at Borobudur, and even on the walls of the Candi Mendut, the nimbuses of the divine personages retain, as in Southern India, the simply oval form, those of the three figures rise to a point, like the leaf of the Bodhi-tree, in the Sino-Japanese fashion. It would be interesting to date as exactly as possible the appearance of this form in Java. It would, in fact, mark with sufficient certainty the moment when the two great currents of artistic influence, which, diverging from their common Indian source, had followed respectively the land routes through Central Asia and the sea route south-eastwards, met again in the island and there, so to speak, closed their circuit. Cf. B.E.F.E.O., IX, 1909, p. 831.

The central statue, about m. 2,50 high, cut out of an enormous block of andesite, represents a Buddha seated in the European manner, the hands joined in the gesture of teaching. Not only the āsana and the mudrā, but even the details of the hair, the lotus-stool, the throne with a back, etc., recall in a striking manner the images found at Sārnāth, in the northern suburb of Benares, on the traditional site of the master’s first preaching. Besides, to cut short all discussion, the lower band of the pedestal is still stamped with a wheel of the law, accompanied by the two characteristic antelopes of the Mṛgadāva.

On each side of the teaching Śākyamuni, on a throne having a back likewise adorned with superposed animals, a Bodhisattva is seated in lalitākṣepa, the left leg bent back, the right foot hanging down and resting on a lotus. At the [264] right of Buddha Avalokiteśvara may at once be recognized, thanks to the effigy of Amitābha which he bears in his headdress. As usual, his right hand makes the gesture of charity; his left is folded back in the position of discussion, but without at the same time holding a lotus.

His counterpart, with the palm of his left hand leaning on the ground and the right hand turned back in front of his chest, does not present any particular mark allowing us to determine his identity. It is solely the traditional force of custom which compels us to attribute to him the name of Mañjuśri: the more so as, after having despoiled these two acolytes of every characteristic attribute, the sculptor must for a means of recognition have relied upon their simple presence by the side of Buddha [this statue has later been identified as the Bodhisattva Vajrapāṇi].

The walls of the vestibule bear on the right and left, in panels of about m. 1,90 x m. 1, figures of the genius of wealth [Kuvera] and his wife Hariti [above], which have already been published by Dr. J. Ph. Vogel. B.E.F.E.O., IV (1904), pp. 727-730: cf. above, p. 141.

Of the principal facade of the temple– exceptionally oriented towards the north-west instead of to the east – only the wall to the left of the entrance is preserved; it bears a standing Bodhisattva, holding a lotus surmounted by a stūpa: it seems that we must by this sign recognize Maitreya.

If we now commence on the terrace the pradakṣiṇā of the monument, we come first to the north-eastern facade. In the middle of the central panel, framed by pilasters bearing atlantes in their capitals, we see, seated on a throne covered with a lotus and under a stereotyped tree, a feminine divinity with eight arms. Unfortunately the head is broken; but it seems, in fact, that it had only one face; and this [265] suffices to put aside the identification with the Vajra-Tāra with four faces in favour of Cunda. Her right arms do hold the shell, the thunderbolt, the disc, and the rosary. Of her left arms, the first from the top is broken; the three others carry an elephant’s hook (aṅkuśa), an arrow, and some object which we could not distinguish. On either side stands a Bodhisattva holding a flyflap: the one on the right has further the pink lotus of Avalokiteśvara, the one on the left the blue lotus of Manjuśri. Finally, on the two lateral panels, the same standing Bodhisattva, his right hand in the varamudrā, bears a flower quite analogous to the nāgapuṣpa of Maitreya.

On the next facade the central figure is an Avalokiteśvara with four arms. One of its right arms, which is broken, must have been lowered in the gesture of giving, whilst the other holds up a rosary. A pink lotus and a book adorn the left hands; the flagon of ambrosia rests upon another lotus on the same side. Two feminine attendants, doubtless forms of Tara, worship him. In the Bodhisattvas figured on the two lateral panels the thunderbolt with which both are armed proclaims Vajrapāṇi.

The principal figure of the south-western, and last facade is again feminine (pi. XLIV). She is seated in the Indian manner upon a lotus supported by two nāgas. The two attributes of the upper pair of hands, on the right the rosary and on the left the book, should indicate the Prajñāpāramitā with four arms. But in that case the normal hands should make the gesture of teaching, instead of that of meditation. Similarly, if she were a four-armed Tara, the first right hand should make the gesture of charity. The symbols and the attitudes [266] combine, therefore, to indicate a second representation of the goddess Cunda, the form with four arms. The two Bodhisattvas, her attendants, reproduce exactly those of her counterpart on the opposite facade. As regards those of the lateral panels, they carry on blue lotuses a sword and a book respectively: we must, therefore, see in them two replicas of the same Mañjuśrī, of whom these are the two traditional emblems.

To sum up: in the personages who decorate the exterior of the three unpierced faces of the temple of Mendut we propose at first sight to recognize, in the middle, two images of Cunda with four and eight arms, and one of Avalokiteśvara with four arms; on the sides, two replicas each of Maitreya, Vajrapāṇi and Mañjuśri: all being important figures of the Buddhist pantheon. But, naturally, this preliminary review would have to be severely tested. It would be necessary, in particular, to examine these bas-reliefs more closely with the help of ladders or a hanging stage, so that no detail could escape; and, this minute labour accomplished, it would still be necessary to verify by comparison with other Buddhist statues of Javanese origin whether there is not occasion to modify in some measure, for local reasons, the Indian attribution of these images. At that cost only could these too rapid identifications become reasonably certain.

3. The Museum of Batavia Batavia is the old colonial name for what is now Jakarta, most of the collection seems to have been eventually incorporated into the National Museum of Indonesia.

We have just spoken of a kind of general confrontation of the Buddhist statues of Java. The material would not be lacking, in spite of the relatively restricted number of Buddhist monuments in the island. Many of them have already been brought together, both in a building near to the residency of Jogyakarta [267] and in the museum of the Asiatic Society of Batavia. Of the first collection a catalogue has been published by Dr. Groneman. The most interesting objects to be mentioned in the second are some inscribed images of the Dhyāni-Buddhas Akṣobhya (no. 224) and Ratnasambhava (no. 225), of the śakti Locanā (no. 248), of Tārā in the form of Bhṛkuṭī (no. 112), of Hayagrīva (no. 76), etc. Everyone will appreciate the interest of these names, Several of these statues have already been published by the late J.L.A. Brandes, Beschrifving van de ruine… Tjandi Djago, The Hague and Batavia, 1904. taken at hazard from our notes on the lapidary museum.

We must likewise mention as belonging to the museum of the capital a considerable collection of small figures of more or less precious metals (gold, silver, or bronze), which are for the most part already classed. For access to this collection we are indebted to the kindness of Dr. C. M. Pleyte, who was so good as to take the trouble of opening the glass-cases for us. Let us cite among others some very artistic statuettes of Avalokiteśvara, Vajrasattva, Kuvera, Tārā, Mārīcī, etc. All have this in common, that they are remarkably faithful to their Indian models.

There is one at which it is perhaps worthwhile to stop for a moment, because of the rarety of the type in India and the success which it has had in the Far East. We have already had to occupy ourselves with the sole example preserved by chance at Bodhgaya. Now Dr. Pleyte – and we apologize for not having known this reference at the time – had for his part published three Javanese replicas, Cf. Bijdragen lot de Taal-, Land- en Volkenkunde van Ned.-Ind., Zesde Volgreeks, Tiende Deel, afl. 1 and 2, pp. 195-202, and our Etude sur l’Iconographie bouddhique de l’Inde, II, 1905, fig. 4. one of which is now in London, another at Leiden, [268] and the third at Batavia. He had likewise the merit of discovering in Schiefner A. Schiefner, Eine tibetische Lebensbeschreibung Śākyamuni’s, p. 244. a legend which explained the bellicose pose of this divinity, whose left foot treads upon the face of a man, and his right upon the bosom of a woman. This would be a mode of deciding, with no possible equivocation, the question of the supremacy of a simple Buddhist guardian of the law over the great god of the Brahmans. Śiva had the imprudence to refuse obedience to Vajrapāṇi under the pretence that the latter was only a yakṣa: contemplate for your own edification the punishment of his crime. We in our turn may note that on this point the descriptions of sādhanas, or magic charms, confirm the Tibetan tradition by likewise giving to the persons overthrown the names of Maheśvara and his wife Gauri: while for the genius, instead of making of him simply a furious transformation of Vajrapāṇi, they use the more precise appellation of Trailokyavijaya. Let us add that this last reappears among the divinities of the Japanese pantheon under the vulgar designation of Gosanze. His pose has not changed, nor his double, living pedestal; and, if he has no longer more than one pair of arms, his hands, at least, continue to execute the vajrabhūmikāramudrā characteristic of his anger and common to all his representations. Cf. J. Hoffman, Pantheon von Nippon (vol. V, of the Beschreibung von Japan of von Siebold), p. 75 and pi. XIX, fig. 164; and Si-do-in-dzou (Ann. du Musée Guimet, Bibl. d’études, vol. VIII, Paris, 1899), pp. 100-101 and pl. XII. On the Javanese statuette we find again the four visages which the Sanskrit manuscripts and the stele of Magadha ascribe to him, and even the eight attributes (sword, disc, arrow and bell, [269] thunderbolt, elephant’s goad, lasso and bow) which they agree in placing in his eight hands.

Any special inquiry would lead us, we believe, to this double conclusion: on the one hand, the close filiation of the Javanese Buddhist images in relation to their Indian prototypes, and, on the other hand, their more or less distant kinship with the Tibetan, Chinese, or Japanese idols, derived from the same origin. If no profound divergence from the composition or style of the common models seems to guarantee to this province of Buddhist iconography any great originality, its interest, on the other hand, promises to extend far beyond the local horizon. It is important for the general advancement of Asiatic studies that it should at last form as a whole the subject of some publication. Not only would the harvest be abundant, but we have carried away the impression that it is ripe and ready to be gathered. It is much to be desired that the enlightened government of the rich colony should provide some Dutch savant with the necessary leisure.


Buddhist Iconography


Photographs by Anandajoti Bhikkhu

About this Website

Creative Commons License
Creative Commons Attribution-ShareAlike 3.0 Unported License