The Sigiriya Ladies: Who were they, why were they painted?

Sigiriya maidens
Tara is a goddess who was worshipped by Mahayana Buddhists in Sri Lanka. The divine females at Sigiriya are splendidly adorned with jewellary, bracelets, necklaces, tiaras, diadems, chaplets of flowers and are of three complexions red, yellow and green. The goddess Tara too has numerous manifestations in regal splendor and may be of red, yellow, green, blue or white complexion.
Sigiriya maidens

By former Archaeological Commissioner Dr. Raja De Silva

The theories and views on the subject of the paintings may be divided into two broad categories: firstly they portray human beings, i.e., queens, princesses and maids of the court of Kassapa I (AC 479-497) in a secular or religious milieu (Bell 1897, 14; Wijesekera ND 1943, Raghavan MD 1948; Chutiwongs, Prematilleke, Roland Silva 1990). Secondly, they depict celestial beings – either semi-divine nymphs known as apsaras, or a class of divine beings, i.e. goddesses (Coomaraswamy 1927; Hocart 1929; Paranavitana 1947; Deraniyagala PEP 1948, Ratnasuriya MD 1950; Mirando AH 1955; de Silva Raja 1990, 2002; Bandaranayake S. 1999)., The majority of interpretations of the second category too are based on the attribution of the paintings to Kassapa.

Human Beings

Bell's interpretation ended with the following words

"Each figures is imbued with divergent traits in face, form, pose and dress, which would seem to stamp it as an individual likeness."

Though Bell did not suggest any reason for the paintings, Paranavitana (1947 264 – 269) put it succinctly when he stated:

 "If we accept Mr. Bell's opinion, we are to believe that King Kassapa used the rock cavity as a portrait gallery of the queens and princesses of his court."

Further on, Paranavitana stated:

 "The figures in the Sigiriya paintings are thus seen to have served an essentially decorative purpose."

Because various individual likenesses were recognized, Bell took it that Kassapa commissioned the painting of members of his palace. It will be shown below that this is not necessarily the logical conclusion (also reached by Wijesekera and Raghavan) regarding the subject of the Sigiriya paintings.

The master artist was commissioned by his patron, the king, to paint innumerable females of a specified class in a certain coherent understandable theme, and would have set about his work accordingly. When the stage was set for the actual painting process, the master painter and his atelier would have commenced work high up on scaffoldings erected by the prepared surface of the great rock. They would have had a free hand to contemplate on the features of females familiar to them and to paint in their facial and bodily likenesses, encouraged by the knowledge that their work would not be subject to scrutiny, be it from the sangha or from the court.

It should be remembered that the paintings on the high surface of the rock, including those in the fresco pockets, were to be seen from a distance at a lower level, and what would have mattered from the viewpoint of imparting their message was not the details of their facial features (except possibly their youth or maturity), which in any case cannot be recognized from a distance, but their environment or milieu, regal bearing, complexion and posture, gesture of the hands or mudra, and the attributes they carry. That is why each painter was able without let or hindrance to freely indulge his whims in painting several "portraits" of women familiar to him. This is well exemplified in the figures Pocket A, No. 5, Pocket B, No. 12, No. 13. The figures are depicted with a naturalism that brings to mind the naturalism of great painters like Caravaggio (ca. 1600), who about a thousand years after our period under consideration at Sigiriya, was well known for painting sacred themes which included the physical features of characters known to him from the streets of Rome. Thus the fact that the features of peculiarly situated female figures were naturalistic and realistic, does not entitle us to conclude summarily that the patron wanted specific human beings portrayed, particularly ladies of the royal court.

Graffiti and Mathmaluwe

Incidentally, Mathmaluwe (The Island 5 March) in arguing against my documentation of a monastic complex at Sigiriya wants us to remember that "the earliest graffiti found on the mirror-wall are as early as the 6th century AC" and wonders how it is that "there is no reference to religious structures when the graffiti were so close to the reign of Kassapa." May I remind this writer that Bell or Paranavitana did not publish any graffiti that were dated to the 6th century, and that the graffiti published by Paranavitana were dated largely to the period 8th to 10th century AC, which were by no means close to the period of Kassapa. Besides, because the graffiti hitherto published do not mention any monastery at Sigiriya, to state that there was no monastery there is not straight thinking - it was not necessary for any visitor to state the obvious fact on the gallery wall that the site was dotted with monastic remains. We should, however, remember that some graffiti mention gods (No. 300, 601), goddesses (No.909, 551), and divine beings (No. 144). The subject of the paintings could well have been forgotten many centuries after they were done, when the general public wrote their comments on the gallery walls.

The relevance of the writings on the mirror wall as a basis for interpreting the subject of the paintings would need to be considered. Many of the graffiti belonging to a period several centuries after Kassapa refer to the paintings as ladies grieving for the loss of their lord (Wijesekera 1945, 275). This centuries-late literary evidence is relied upon to interpret the subject of the paintings.

The reasonable conclusion to be drawn from the relevant graffiti is that these late scribblers thought that the paintings depicted ladies of the court of Kassapa, the king who was known to have been a patron associated with the monastery of Sigiriya, where he had donated a vihara commemorating his two daughters, Bodhi and Uppalavanna. The graffiti are of not much relevance for the elucidation of the subject of the paintings.

Bell and those who have followed him have based their interpretations on the supposition that Kassapa lived in a palace on Sigiriya rock and that he was responsible for the paintings which were connected in some way with him. It was shown earlier (The Island of 3, 6, 10 March) that the story of a palace of Kassapa on Sigiriya rock was most likely to be untrue. Therefore the human being-interpretation has no basis as there was no reason to have hundreds of royal ladies painted anywhere at Sigiriya.

Siri Gunasinghe (2008) has a similar theme (beautiful nondescript women) serving a decorative purpose, due to Kassapa, the aesthete. (favoured by my friend Asanga Tilakaratne (AT), Professor and specialist in Pali and Buddhist studies, The Island 17 March) This interpretation too was shown to be unsound (The Island 27 March).


Bell also thought that the paintings were "possibly goddesses on account of the fact that the demi-figures were made to emerge from clouds". He soon abandoned this theory in favour of taking the clouds to have been a device to dock the ladies of their lower bodies while being painted on a concave surface within the fresco pocket. This idea becomes untenable when it is considered that there are clouds painted in the fresco pocket not only between figures but across the chest of a painting (A2) as well. Furthermore, in Deraniyagala's cave (B7) on the escarpment, clouds are depicted even where the surface of the plaster on the soffit is more or less flat.

The clouds serve another purpose: to show the spectator that the milieu of the paintings was the sky, air, finite space, i.e., space occupied by bodies, which cannot be indicated per se in a painting. Appreciation of this function of the clouds is of assistance in explaining the subject of the paintings – which is the depiction of beings that live in the heavens.

Celestial Beings (Apsaras, Goddesses)

The females associated with clouds in the heavens where divinities tend to gather are either apsaras or goddesses, and it is to be concluded that the subject of the paintings is the artist's conception of a celestial being.

1. Semi-divine Beings (Apsaras)

Ananda K Coomaraswamy (AKC) [1913], the well known art historian, would have known of Bell's entertainment of the likelihood of the paintings representing goddesses, and he accepted Bell's interpretation taking Sigiriya as the fortress of Kassapa. The crucial question is; why did AKC not accept the goddess-interpretation but demote the goddesses to the lower status of apsaras? He interpreted the paintings as semi-divine beings disguised as mortals.

We can deduce the quandary AKC was placed in. He had gone beyond the point of no return in following Bell on the path of the Kassapa-fortress theory. He could not reconcile this with the proposition that the paintings were of goddesses for it would entail his accepting them as of a religious theme, and therefore, the site itself as an important religious one, to wit, a monastery. So AKC patched up his craft in mid-air (so to speak), forgot the idea that the paintings represented goddesses, and invoked instead semi-divine females to inhabit the clouds. All this in his knowledge of the iconographical difference between goddesses with their tiaras, coronets, diadems and spots of hair on the foreheads (urna lomas), and apsaras who are not entitled to these distinguishing attributes. No mention was made as to why "apsaras" were painted at various places in Sigiriya, but even if that were so, it had to follow that the paintings were of a religious, not of a secular nature.

Thirty three years later, Senerat Paranavitana [SP] (1947, 1961), Archaeological Commissioner and scholar, who on retirement from the public service, was appointed the first university Professor of Archaeology, was not satisfied with AKC's interpretation; he restored the fallen fair females to their original status of goddesses (see below).

Senake Bandaranayake [SB] (1993) revived the apsara-interpretation of AKC and sought to make good the omission of raison-d'etre for their ubiquitous presence by suggesting that they were painted for possibly two reasons;

  1. To betoken or express the grandeur and status of Kassapa
  2. To artistically evoke, i.e., call up, life at Kassapa's court, with "aesthetic and erotic dimensions (emphasis added).

Both these interpretations are tantamount to explaining that the paintings were luxurious decorative motifs calculated to extol Kassapa, who is believed to have lived in a palace on Sigiriya rock.


There being no evidence for a palace on Sigiriya, no reason existed for enhancing Kassapa's status at that site. In any event, SB has departed from allowing apsaras their traditional role of attending on higher deities as gods and goddesses, Buddha and bodhisattvas, and appointed them to a lower duty of extolling the royal grandeur of a king in a secular setting, a king who was not even a great warrior like Dutthagamani or Vijayabahu I. They have not been known to glorify lesser living beings, mortal humans, not even kings as Kassapa, a Buddhist patron. This is an unusual departure initiated by SB, which does not hold water, even though apsaras started life as water-sprites.

 2. Divine Beings. Goddesses

 Lightning princesses, cloud damsels

Paranavitana interpreted the paintings as a type of goddesses, which was stated to have been a well known motif in ancient Sinhala art: lightning princesses (vijju-kumari) and cloud damsels (megha-lata). They were personifications of the natural phenomena of lightning and clouds which can be seen at the higher levels of Sigiriya. They were painted, he said, for the purpose of glorifying the image of Kassapa as a god-king, done on his instructions high up on Sigiriya, his abode.

Click to visit
Pabbata-raja made the mountain or rock his centre of ritual worship. The mountain at Mihintale, known in pre-Buddhist times as Missaka pabbata, and the rock at Sigiriya (above) were two of the main centres where pabbata-raja kings held their festivals, involving both rain-making and fertility.

N D Wijesekera (1974) examined SP's literary allusions, found them wanting, and concluded that there were no lightning princesses and cloud damsels in Sinhala paintings.

There was no reply by any art critic, and this cogent criticism has been ignored by SB too. This interpretation was further deconstructed by Raja de Silva (2002), and SP's god-king theory was refuted by de Silva (2005); there has been no comment by any art-critic on that score too. Whereas we need to accept that the paintings represent some type of divine being, their identification as lightning princesses and cloud damsels is untenable.


My predecessors and other art critics have had their conjoint vision of the subject of the paintings blurred by the cataract of unquestioning belief that Sigiriya and the paintings were done by Kassapa for Kassapa. Once this illusion is removed, the way ahead is clear: given that Sigiriya was a long-standing Buddhist monastery, and the paintings are those of goddesses, the identification of the subject as an important Buddhist goddess is "elementary" (as Sherlock Holmes said to Watson).

In the mid-part of the first millennium AC Sigiriya had a considerable presence of dhammarucis with their Mahayanist leanings within the framework of Theravada Buddhism. In this environment (material and social), the painting of numerous Taras is understandable, as was intuited by the German Swamy Gauribala in 1967, whose idea of Sigiriya as a Mahayana centre for the initiation of aspiring bodhisattvas was published by Roloff Beny in his beautiful photographic documentary titled Island Ceylon (Thames and Hudson 1971).

Why Tara?

Two matters required explanation: on what basis are the paintings recognized as those of Tara, and why were they painted there? Ample literature is available on Tara from her earliest forms of representational existence (nascent) to more elaborate manifestations (M Ghosh 1990; the earliest Tara statue found in India is dated to the 6th century AC.). A number of devotional hymns to the goddess give information on her iconography, which includes complexion, posture, gesture and attributes, which are generally flowers that she holds, largely of the lotus family. An early Tara from India carrying a flower other than a lotus is known (M Ghosh). So a Sigiriya goddess that does not hold a lotus flower but has other features of Tara should not be precluded from being a Tara and considered to be some other deity. This is another example of Mathmaluwe's erring ways: he says that the local paintings are not of Tara because flowers other than the lotus are also carried (The Island 3 March 2010). In fact, it is not essential that a Tara should carry a flower, it is the mudra that is more relevant. A comparison of the factors in the Sigiriya paintings with the known iconography of Tara leaves no doubt about the identity of the local goddess. They are various early, ca 6th century AC, manifestations of Tara.


AT, however, while not refuting my iconographic identification of the paintings as those of Tara, but having favoured Siri Gunasinghe's theory of beautiful women, is reluctant to accept that the paintings are indeed of the beautiful Tara (The Island 17 March). As silence on my part could well be taken as tacit acceptance of all that he has wondered about Sigiriya, it is desirable that I make a short reply to his references to my interpretation. His main points of reasoning are as follows: Tara became popular in Vajrayana in India around the 7th century AC, and it was a remote possibility to have a huge monastery dedicated to Tara outside India, particularly in a Theravada stronghold like Sri Lanka, even if Kassapa were alive in the 7th century AC.,


Even if Tara became popular in India in the 7th century, that cannot be taken to preclude or make improbable the worship of Tara in Sigiriya and the nearby Kandalama in its early times from the 5th /6th century, particularly when it is well known that this country had close contact before ca 500 AC with Nagarjunikonda which was associated with the teachings of the great Mahayana scholar Nagarjuna. Furthermore, the date of Kassapa is not relevant to the identification and dating of the Sigiriya paintings, though both Siri Gunasinghe and AT have cast their lot with Kassapa, and think that the paintings must therefore be dated to his reign, i.e., the last quarter of the 5th century AC.. I have dated the local paintings, after detailed comparison with dated Ajanta paintings (e.g., Cave 2) [showing their closest similarity] to "sometime after the beginning of the 6th century AC, probably towards the middle ". Thus, it is unlikely on this count alone that Kassapa had the paintings done, no matter what date (either 5th or 7th century AC) is assigned to him.

With reference to the Theravada stronghold of Sri Lanka that is referred to by AT, what I mentioned about Tara worship was that it was resorted to in the dhammaruci stronghold that was Sigiriya.

Regarding Siri Gunasinghe's acceptance of the Mahavamsa authors' account of Kassapa and Sigiriya, AT states "to accuse them of telling untruth is not reasonable". What I had to say is different; "All this is not to impute that the compiler-monk of the second part of the Mhv. consciously and purposefully recorded untruths concerning secular matters, but he might have acted like Herodotus long before his time, who wrote ‘My business is to record what people say, but I am by no means bound to believe it".

Why paintings of Tara?

 Devotees on pilgrimage to Sigiriya would have seen hundreds of images of their favoured saviouress-goddess Tara. This "reduplication" in the background of the sky, space or the heavens would mean that the deity is omnipresent and can be seen from whichever direction a devotee comes up the escarpment to the main rock. They would have contemplated on the goddess in a disinterested way. The intention of the artist was to induce the beholders to piously believe in and venerate the true beauty of their adored goddess, whose abode was the Remembrance Rock (sihi-giri).

At a higher level (pun intended) the extended reason for painting Taras was to show devotees that they are offering flowers to the Buddha (varnagandho gunopethang, etc.), who was symbolically represented by the dagoba at the highest level of the summit.

Parthian shot

The Central Cultural Fund plus its former kingpins and the Director of National Museums have "muscled in' and assumed the role of de facto guardians of our material cultural heritage exhibited at the Sigiriya museum. The Director General of Archaeology has, alas, not been allowed to do duty in planning this archaeological site museum, but has become only the hapless de jure authority at Sigiriya and all that has been found there since 1894.

The former parties have propagated (in the new museum they have organized) the myth that Kassapa lived at Sigiriya and had the paintings done for his own glorification. All this, in the face of undeniable published evidence (unrefuted since 2002) of such an extreme improbability, nay impossibility.

I say, "Forgive them O Lord Buddha, for they know not what they do."

-Raja de Silva

"Sigiriya paintings and graffiti: What KNO Dharmadasa should know" by Raja de Silva
"Sigiriya: Who are these maidens and what is their message?"
"Sigiriya: Palace in the sky or Mahayana Mandala?" review of the author's book Sigiriya and its Significance
"Archaeological Department, Central Cultural Fund and the Law" by Raja de Silva
Overview of the author's book Sigiriya and Its Significance home