Tibetan paintings Table of Contents:
Front matter 0 Preface 1 Introduction 2 Visions and Visualizations 3 Kadampa Style 4 Sakyapa Style 5 Thankas from Western Tibet 6 Landscape Tradition 7 Age of the Dalai Lama 8 Appendix Color Plates Monochrome Figures Bibliography
There seems sufficient evidence now to postulate that a uniform style of painting developed in Tibet in
the second half of the eleventh century mostly in association with the Kadampa monasteries. The religious
order of the Kadampas was established by Dromton, the chief disciple of Atīśa Dipankara Srījñāna who
arrived in western Tibet in 1042. After spending two years in the region, Atīśa moved on to central Tibet
where he died in 1054 in the Sña-than (Nethang) monastery. Atīśa’s biography makes no mention of any
artists accompanying him to Tibet, but it is almost certain that he must have carried with him
manuscripts, some of which may have been illustrated, as well as paintings and bronzes.
It is interesting that among the gifts sent by Tsongkhapa to Karmapa Dezin Shegpa
was a bronze image of Maitreya belonging to Atisa. Thinley, p.75
But whether he did so or not, there was vigorous traffic at this period between the monks and teachers of
the two countries. In particular, Tibetan monks were a constant presence in the well-known monasteries of
ancient Magadha (modern Bihar), known generally as Vangala to the Tibetans. These included Vajrasana
(Bodhgaya), Nalanda, and Vikramaśīlā in Bihar and Jagaddala and Somapuri in Bangladesh.
Pl. 5 Four Illuminations from a Buddhist manuscript India (Bihar or Bengal), 12th century Palm leaf Average illumination: 6 x 5.5 cm R.H. Ellsworth. Ltd.
That there was an active centre for writing manuscripts and for painting in the ancient port of
Tāmaralipti in West Bengal is evident from the Chinese pilgrim Faxien (Fa-hsien) who visited the city
early in the fifth century. Prima facie evidence for a flourishing school of painting during the rule of
the Pāla dynasty (9th through the 11th century) is provided by Buddhist manuscript illuminations which
are now well known to scholars (Pl. 5). However, that there were also religious paintings on cloth which
were the forebears of the Tibetan thankas is clear from the Mañjuśrīmūlatantra where an extensive
description of painting such a pata is given.
See Lalou 1930.
Moreover, a Chinese source of the Song dynasty gives some interesting information about paintings on
cloth in Nalanda that seems to have been generally overlooked by Indian scholars. The passage runs: ‘In
India, at the temple of Nalanda, the priests paint many Buddhas, Bodhisattvas, and Lohans, using the
linen of the West.’
H. A. Giles, An Introduction to the History of Chinese Pictorial Art (Shanghai:
Kelly and Walsh, 1918), p.157.
Although neither murals nor cloth paintings of this school have survived, a good idea of how they must
have looked can be formed today with the vestiges of murals in early Kadampa shrines in Tibet and from
the group of thankas which is the subject matter of this chapter.
Tucci was the first scholar to publish both murals and thankas in this style and considered them to be
‘Nepalese.’ The next publication (in Chinese) which illustrates some murals in this style has nothing to
say about the paintings except to suggest dates which we will presently discuss.
Liu I-se, Figs. 17-19, 23.
In 1969, while discussing the largest thanka in this style (Pl. 7), now in the Los Angeles County Museum
of Art, I suggested that the style was distinct from the Nepali style and pointed out, I believe for the
first time, that the same style is visible at both Dunhuang and in thankas recovered from Kharakhoto
Pal 1969, p.131.
(Figs. 8-9). This is now generally accepted by most scholars who have written on the subject.
See Karmay 1975 and Hatt.
There also seems no doubt that the style derived directly from the Pāla style of eastern India and owed
very little to Nepal, at least at the early stages. However, by the early fourteenth century, as we will
discuss shortly, the style was modified considerably by Newari artists, especially in Sakyapa monasteries.
Fig. 8
Goddess Tara
China (Kharakhoto), Before 1227
Embroidered Silk, 98 x 45 cm.
The Hermitage Museum, Leningrad
Fig. 9
Dancing Vajravarahi
China (Kharakhoto), Before 1227
Watercolours on hemp cloth; 111 x67 cm.
The Hermitage Museum, Leningrad
Fig. 10
The Life of Sakyagri (1127-1225)
Central Tibet (Sakyapa monastery)
ca.1300 85.8 x 66.4 cm. R. H. Ellsworth, Ltd.
Pl 7
The primary reason why I feel it would be perfectly appropriate to designate this style as Kadampa is its
consistent association with early establishments of that order. The vestiges of the style may be observed
in the murals of Iwang, Samada, Nethang, Nenying, Chasa and in the Jokhang in Lhasa. All these
establishments were closely associated with the Kadampa order during the period, and hence, it would not
be inappropriate to refer to the style by the designation of the order. The Chinese publication referred
to earlier has reproduced a number of murals executed in this style from a monastery whose Chinese name
referred to in the book is difficult to transcribe into Tibetan. Most scholars I have consulted have read
the name as Ladong. In all probability this is the Landang (Glang-Thang) monastery founded in 1093 by
Dorje Senge, a disciple of Potopa. Landang is situated in ’Phan-yul in central Tibet near the monastery
of Nalendra, which however was a later foundation.
Wylie 1962, p.162 regarding Langdang.
In any event, it seems clear that the earliest diffusion of the Kadampa style occurred in central Tibet
where Atīśa spent most of his twelve-year residence in the country and where his principal disciples
established the earliest temples and monasteries of the order, mostly in the second half of the eleventh
century. It is conceivable that they brought artists from Magadha to decorate some of the monuments.
In western Tibet the extension of the style can be observed principally in the murals of the Lhakhang
Soma at Alchi (Pls. 10 and also at Saspol, which is not far from Alchi).
Murals in this style may also be seen in the Guru Lakhang and Gyung drung gompa,
though these may be slightly later than the Lhakhang Soma.
The style probably was also prevalent in the Kailash region, and although originally all these monuments
were associated with the Kadampa order, I have elsewhere shown that the Lhakhang Soma murals may have
been rendered under Kagyupa influence in the thirteenth century.
Pal 1982B.
The Kagyupas may also have been responsible for the spread of the style at Kharakhota in northwestern
China, as has been suggested by Heather Karmay.
Karmay 1975, pp.41-42.
As to the murals in this style in some of the Dunhuang caves, probably painted in the thirteenth century
during the reign of Kublai Khan, the most likely order which would have undertaken the task would have
been the Sakyapas who were closely associated with the emperor. Unfortunately, most pre-sixteenth century
murals at Sakya itself have been destroyed, but it may be pointed out that some of the paintings at
Shalu, executed in the first half of the fourteenth century under the supervision of the renowned Buton
Rinpoche (1290-1364), reflect a modified version of the Kadampa style.
I have a slide in my possession given to me by Michael Henss which illustrates a
thanka in this style.
After all, the new temple was built on a site consecrated by Atīśa himself, and it is very likely that
the older murals of both Shalu and Sakya were painted in the Kadampa style.
That the early Sakyapa murals were painted in the Kadampa style is evident from a thanka depicting the life
of Śākyaśrī (1127-1225) known as Khache Panchen or the Pandita of Kashmir (Fig. 10).
At the invitation of Tröphu Lotsava (b. 1173), Śākyaśrī came to Tibet in 1204 and returned to Kashmir in
1213. The painting is based on the biography of Śākyaśrī written by Tröphu Lotsava, and hence the
painting may have been rendered in the second half of the thirteenth century at the Tröphu monastery.
Like the early fourteenth century Shalu murals, it shows an admixture of both the Kadampa and the Nepali
manners which is not surprising since Tröphu Lotsava was closely associated with Nepal.
Thus, it is clear that as was always the case in Tibet, a particular style was not confined to a
particular order, and certainly the Kadampa style was the first pan-Tibetan style which spread beyond
the confines of the country. Having reviewed the spatial diffusion of the style, let us now discuss its
broad temporal parameter.
The mid-eleventh century can easily be determined as the terminus a quo of the style for most of the
monasteries and temples of central Tibet where its remnants are visible were built during the second half
of the eleventh century. It has already been mentioned that by the early fourteenth century when the
murals of the Serkhang at Shalu were painted, the style had become considerably altered by the inclusion
of both Nepali and Chinese elements. The early thirteenth century thankas of Kharakhota (Figs. 8-9) are
rendered in a more cursive version of the style, as is also the case with the murals of Saspol and Alchi.
The Dunhuang murals, however, probably executed around 1275 appear to be much closer in style to some of
the thankas published here which we consider to be typical examples of the pure Kadampa style.
Among the murals in this style, none can be dated with certainty. The fragmentary remains at the Jokhang
and those at Iwang and Samada may be the earliest and may belong to the late eleventh or early twelfth
century.
Slides of the Jokhang were also supplied to me by Michael Henss. For others see
Tucci 1949 and 1956.
Such a date would be a likely period for the first renovation of the Jokhang after the revival of
Buddhism. Iwang, Samada, Nenying and Chasa are all early temples and there is no reason why the murals
should not belong to the twelfth century.
In publishing the murals in the Ladong monastery, the Chinese author has suggested a 1039 date for a
composition with two Lotsava and generally twelfth century for the others.
See note 4.
It is of course possible that 1039 is a misprint for 1093 when the monastery was founded. It is more
likely, however, that this painting too is of the early twelfth century and the two monks engaged in
conversation may represent the founder and his guru. Among the many thankas painted in this style, only
one bears an inscription which provides some circumstantial evidence for suggesting a late twelfth
century date. This is the now well-known example at Los Angeles (Pl. 7) and I have discussed the evidence
at length elsewhere.
Pal 1983B.
Otherwise, the Kharakhota thankas and the Xixia prints remain the principal
datable material of the Kadampa style.
Karmay 1975.
One may generally state that the Kadampa style flourished unmodified from the second half of the eleventh
through the end of the thirteenth century.
The Los Angeles thanka of Amitayus (Pl. 7) is typical of a group of Kadampa style paintings in which the
dominant figure in the centre is invariably a bejeweled and diademed Buddha, obviously of the
sambhogakāya. Two other closely similar thankas are also illustrated here (Pls. 8-9). In all three, the
eminent Buddha is flanked by two standing bodhisattvas and a group of bodhisattvas and Buddhas on either
side of the central figure’s head. The number of figures in this group differs from one thanka to
another, but they are always shown seated in the same manner. Invariably also, a group of Buddhas,
bodhisattvas, or other deities are represented in different combinations along the bottom of the thanka,
below the lotus on which the central Buddha sits. In some thankas a monk’s figure is added in the lower
left-hand corner. Generally the bodhisattvas, including the two who are standing beside the principal
Buddha, are twelve in number, and only in the Amitayus painting are there two additional Buddhas among
the group. It must also be pointed out that Vairocana in the Benares thanka makes the bodhyanagimudrā
(chi ken-in in Japanese) which is his prescribed gesture in his sambhogakāya image.
Pal 1983A.
Pl. 6 Portrait of a Sakyapa hierarch Central Tibet (Sakyapa monastery?). 12 century 86x75 cm. Ernst Jucker Collection, Basle. Ettingen
Pl. 7 Buddha Amitayus and Acolytes Central Tibet (Kadampa monastery). 1145-1189 (?) 245.1 at 149.8 cm. The Nash and Alice Heeramaneck Collection, Los Angeles County Museum
Pl. 8 An Buddha Mahavairocana Central Tibet (Kadampa monastery). Late 12th century 137 x78 cm. Bharat Kala Bhavan, Banares
Pl. 9 Buddha Ratnasambhava Central Tibet (Kadampa monastery). Late 12th century 81 x 64.5 cm. Ernst Jucker Collection. Basle-Ettingen
This combination of a central Buddha surrounded by bodhisattvas appears to have constituted a principal
iconographic programme in most of the early Kadampa temples visited by Tucci.
Tucci 1956.
It is found both in sculptured and painted representations in Iwang, Samada, Chasa and most other early
existing temples. Although generally the number of bodhisattvas is eight in these shrines, following the
Mandala of the Eight Bodhisattvas (which was especially popular in the Nalanda region),
Pal 1972-73.
the presence of ten or twelve bodhisattvas is also in perfect harmony with Mahāyāna philosophy. For
example, the ten bodhisattvas may represent the ten spiritual stages (daśabhūmi) of Mahāyāna and the
number of bodhisattvas, as we know from other sources, can also be as many as sixteen. In any event, the
iconographic scheme of this group of thankas representing sambhogakāya Buddhas related to the Yogatantra
can clearly be associated with Kadampa beliefs.
Among the other principal subjects represented in the surviving thankas of the style are Tārā,
One of the earliest and finest thankas of the Kadampa style now in the possession
of Mrs. Alice Heeramaneck represents the goddess Tara. Indeed, in my opinion this thanka may well be
of Indian origin, or was perhaps painted by an Indian artist in Tibet. If the former, then it may be
the only example of a Pala painting on cloth known to date. A later and more cursive version of this
thanka is illustrated in a mural in the Guru Lakhang in Ladakh
cycles of Samvara and Vajravārāhī, Mahākāla and portraits of eminent lamas. It may be pointed out that
these subjects also predominate in the murals at Alchi and Saspol as well as among the thankas of
Kharakhota. All these themes were expounded by Atīśa and played an important part in Kadampa ritual.
Tārā was the tutelary deity of Atīśa, and, according to his biography, he was in constant communication
with the goddess. Whenever he encountered difficulties, the goddess appeared to him in a dream or a
vision and instructed him to make the right decision. The cult of Mahākāla was brought to Tibet from
Bodhgaya by the great Rinchensangpo, and it was popular with the Kadampas and the Kagyupas as well as the
Sakyapas. However, the surviving thankas represent only a fraction of the paintings that must have been
dedicated during the period, and it is certainly not our intention to suggest that these were the only
iconographic themes depicted. In terms of the portraits, only the representation of Śākyaśrī can be
identified with any certainty.
Inscriptions in the Iwang temple inform us that the murals there were rendered in the style of India and
Li yul (Khotan).
Tucci 1973, p.94.
It is not clear, however, whether these two styles were represented separately or whether they were
synthesized in the murals of Iwang. Apart from the fact that it is difficult to determine what exactly
was the style of Khotan, a comparison with the art of Khotan, as recovered by Sir Aurel Stein, offers
very little that can be specifically related either to the Iwang murals or to the Kadampa style in
general. The paintings that have survived at such sites at Rawak, Dandan-uliq or Endere in Khotan, where
Tibetan inscriptions abound, are considerably earlier than the twelfth century Kadampa paintings and
reflect a style related much more directly with the art of Kashmir than with that of Magadha. Certainly
some Khotanese monks and artists may have sought refuge in Tibetan monasteries after the conquest of
Khotan by the Muslim Turkish dynasty of Kashgar around the turn of the first millennium. It is rather
curious that although artists and styles of Li yul are frequently mentioned in Tibetan textual tradition,
Buddhist monks of Khotan are alluded to only once in the Blue Annals and that too in the context of the
eighth century.
The other suggested stylistic source for the Kadampa style is Nepal. While it is undeniable that Nepal is
a constant presence in the art of Tibet, and hence certain Nepali traits may always be discernible in a
given Tibetan painting, by and large the Kadampa style is related much more directly to the Pāla rather
than to the Nepali tradition. The figural type employed in these Kadampa style thankas is far closer to
those seen in Pāla than in Nepali manuscript illuminations. The svelte, slim figures of bodhisattvas with
their long faces and pointed chins are derived directly from Pāla illustrations (Pl. 5). There too the
figures sway with exaggerated stances and languorous grace. The postures of the seated figures in the
upper section are identical to those seen in Pāla pictures. The garment design of the principal Buddhas
with its striped patterns is very similar to that seen in Pāla rather than Nepali illuminations. One of
the peculiarities of the garment worn by the standing bodhisattvas is that it is densely colored around
the hips, appearing like underpants, and then becomes completely diaphanous down to the ankle, thereby
clearly revealing the smooth contours of the thighs and legs. This is a characteristic of Pāla rather
than of Nepali illuminations, where the male figures generally wear shorter dhotis which are painted with
uniform density. It may be pointed out though that in a series of Ashtabodhisattva Mandala paintings from
Nepal, which I have dated to the thirteenth century but which may in fact belong to the twelfth century,
Pal 1978, Figs. 68-70.
the standing bodhisattvas are given diaphanous dhotis but their designs and fashions are completely
different.
There are many other features in these Kadampa style thankas that are derived from Pāla sources. Perhaps
the most prominent is the shape and form of the rocks against which the figures are seated in the
beautiful lineage thanka (Pl. 6).
Not only is this distinctly different from the Nepali formula for rocks, which is derived from Ajanta
murals, but it is obviously borrowed from Pāla manuscript illuminations (Pl. 5). Similar rocks also occur
in the Kharakhoto woven Tārā (Fig. 8) and in the paintings of the Ananda Temple in Burma.
G. H. Luce, Old Burma-Early Pagan 3 vols. (New York: 1970).
There can be little doubt that both stem from a common source which must be Pāla art. Another obvious
borrowing is the form of architecture employed in the thanka of Śākyaśrī (Fig. 10) and the beautiful Tārā
in the Cleveland Museum (Pl. 18). Although this painting may actually have been painted by a Nepali
artist, for reasons we will discuss later, the shrine in which the goddess sits so gracefully copies
faithfully the kind of temple architecture that was prevalent in Bihar and Bengal and which is best
preserved in the temples of Pagan in Burma. Other details that relate to Pāla rather than Nepali manners
will be pointed out in the discussion of the individual thankas that follow, but it should be clear to
the reader by now that just as the early Khache-Tibetan paintings help us to glean the lost style of
painting that was once prevalent in Kashmir, these Kadampa style thankas are of equal significance for
determining what the religious paintings created by the artists of ancient Magadha looked like.
Pl. 10 Buddha Vairocana Ladakh (Alchi, Lakhang Soma). Early 13th century Mural
Pl. 14 Mahakala Ladakh (Alchi, Lakhang Soma). Early 13th Century Mural
Pl. 18 The Goddess Tara Central Tibet (Sakyapa monastery?). ca.1300 52.1x43 cm. The Cleveland Museum of Art. The J.H. Wade Fund by exchange
What is perhaps most striking about the Kadampa style thankas is the smooth but vivid colours - with
tonalities that are quite different from those encountered in the early Alchi murals (Pls. 1-2) or in the
Lhakhang Soma (Pls. 10 and 14) and Kharakhoto thankas. They seem to confirm the Chinese historians’ claim
that the artists of Nalanda had developed a very distinctive sense of colouring. Although red always
predominates in paintings of the Indo-Nepali tradition, it is counterbalanced in these pictures by
glowing yellows, greens and blues. Indeed, altogether the palette is much richer than that which one
encounters in either Pāla or Nepali manuscript illuminations, though the comparison perhaps is not
altogether fair because of the diminutive size of the illuminations. The difference in the media as well
as the talent of the individual artists certainly contributed to the stylistic differences between the
various representations of the style. For instance, the figures in the Pāla illuminations are far more
perfunctorily and freely drawn than those in the Nepali style miniatures. Sometimes the features, such as
the eyes and the hair, are almost impressionistic dabs in Pāla illustrations, but the Nepali artist drew
their figures with greater care and with a steadier hand. As a result the Pāla illuminations appear more
lively and spontaneous, whereas the Nepali manner is elegantly dignified and gently sensuous. Similarly,
the figures in the Lhakhang Soma murals and in the Kharakhota thankas (Figs. 8-9) are more summarily
drawn, the standing figures often being more awkward and disproportionate than those in the central
Tibetan Kadampa style thankas. The colours in the former group are not quite as smooth or vibrant, and
even though some attempt is made at shading, the modelling is far more perfunctory, so that the figures
appear flat.
Fig. 1
Bodhisattva Avalokitegvara
China (Dunhuang)
9th century, 50 x 14 cm.
The National Museum, New Delhi
Fig.2 Mahasiddha Sengepa
Central Dr Eastern Tibet
ca.1650, 20 x 20.3 cm.
Los Angeles County Museum of Art
Gift of Dr. and Mrs. Pal in memory of A. Quincy Jones
The central Tibetan version of the Kadampa style, on the other hand, although strongly linear, shows much
more sophisticated modelling of the figures which seem to stand out from their background, as do those in
the Pāla miniatures. The outlines in the central Tibetan Kadampa style thankas are drawn much more
carefully than in the Pāla illustrations; consequently, the loss of a sense of liveliness is compensated
by greater elegance. This may well be a modifying influence of the Nepali manner or the Newar artists who
may have been responsible for some of the thankas. The lithe, weightless figures sit or move gracefully
and buoyantly on the surface of the thankas and all the various elements, the figures, the lotus flowers,
the rocks, trees and streams, the thrones (whether simple or elaborate), the architectural designs or the
leaping flames, are rendered with meticulous care and with mellifluous rhythm. Whether the figures stand
or sit, kneel or prance, dance or fly, there is an underlying rhythmic vitality that make the forms
fluid and animated. Technical certitude is not allowed to stifle the unstinted fluency and seductive
grace of the overall design. Although the composition is invariably the same, with a large central figure
surrounded by Buddhas, bodhisattvas, acolytes and human teachers, these figures are not confined within
elaborate arches and pilasters, as they are in Nepali paintings. Rather, they are separated by simple
arches or are represented in medallions formed with meandering vines that enhance the rhythmic quality of
the composition.
The beautiful lineage thanka (Pl. 6) is probably the earliest of a type of hieratic portraiture which is
characteristic of Tibetan art. Whether such lineage paintings were known in India cannot be determined
today. Certainly the vast body of surviving Pāla and Nepali manuscript illuminations contain no
portraitures of any Buddhist luminary, whether realistic or idealized. In Tibetan lineage paintings,
however, although a stock formula is repeated, sometimes the faces are remarkably realistic, as in this
example, and at others they are highly conventionalized. It appears to have been much more common in
Tibet to create an image of a high dignitary soon after his death. Whether this is characteristic of
Tibetan culture or whether it was borrowed from China, where realistic portraiture was an ancient custom,
is a question worth looking into. Be that as it may, this sensitively rendered and resplendent thanka of
a high religious dignitary, as well as the two other early Kadampa style portraits known,
See note 4.
represent the highest achievements of early Tibetan portraiture and anticipate the later Sakyapa
tradition of such paintings. Unfortunately we do not know the identity of the hierarch who is the
principal subject of this thanka. A clue to the order to which he belonged may be provided by his
luxuriant garments and some of the deities represented. In later lineage thankas such garments are
usually worn by Sakyapa hierarchs.
Moreover, the deities represented above are Mañjuśrī, Vajradhara,
Vajrasattva and Vajrayogini; the blue figure in the middle of the bottom row is either Mahākāla or
Jambhala. The central position given to Mañjuśrī would tend to associate the thanka with the Sakyapa
order. In that case, the principal figure may well represent Konchok Gyelpo (1034-1102) of the noble
family of Khon who founded the Sakya monastery in 1073, or it may depict his son and successor Kunga
Nyingpo (1092-1158), who was the real founder of the Sakyapa order. Although I am not quite sure of the
very interesting and unusual investiture (abhisheka) scene included at the bottom right (the enthroned
figure must be a royal person), there seems little doubt that this is an early twelfth century thanka.
Therefore, if it belongs to the Sakyapa order, then the central figure must be Konchok Gyelpo and it may
have been painted soon after his death.
Pl. 11 Mandala of Samvara Central Tibet. 12th century J.P. Goenka Collection, India.
Pl. 12 Samvara and Nairatmya in Yab-yum Central Tibet (Kadampa monastery). 12th century 43 x 27 cm Private Collection
Equally colourful and joyously vibrant is a mandala of Samvara which is probably the earliest known
mandala in the Kadampa style (Pl. 11). In terms of colours this is clearly a symphony in blue highlighted
with yellows and reds which enhance the brilliant intensity of the basic hue. Otherwise the colours are
of the same basic tonality as the lineage or the other Samvara paintings (Pl. 12) and differ strongly
from the Nepali manner of colouring where the deep red is a purer vermillion. However, some Nepali
elements are present, especially in the seated bodhisattva figures in the cemeteries, and in the forms of
the trees and the rocks in two of the vignettes along the top. The convention of painting such rocks,
although derived ultimately from Ajanta, was a favourite device of Nepali artists. Especially imaginative
are the trees with sinuous trunks and bent and extended branches capped with leaf domes, which are even
more fanciful than anything encountered in either Pāla or Nepali illuminations and anticipate some of the
delightfully decorative trees seen in fifteenth century Indian paintings.
Many features in this mandala appear to be unique and are not encountered in more typically Nepali
mandalas or those portrayed in the early murals at Alchi. The broad shape of the petals of the central
lotus as well as the peculiar treatment of the petals of the peripheral lotus - which are given unusual
volume - are noteworthy. The design of the portals is also unusual, although the shape of the red and
white bell-shaped towers is somewhat similar to those seen in the dhoti of Avalokiteśvara in the Sumtsek
temple at Alchi.
Pal 1982B, P1.54.
As a matter of fact, the easy, relaxed disposition of the figures in the cemeteries
and the rich tonality of the blue are other elements that the mandala share with the early Alchi murals,
but the styles are distinctly different. Another interesting detail is the manner in which the prongs of
the thunderbolt on either side of the portal rise directly from the wall of the citadel. Both in Alchi
and in Nepali mandalas the prongs invariably emerge from the mouth of a makara and appear as its tongue.
Finally, the lively ring of fire as well as the mode of representing the flames behind some of the
deities in the cemeteries have exact parallels in the Pāla miniatures. In most Nepali mandalas the
tongues of the flames were far more stylized and increasingly assumed ornamental shapes.
Like some of the others already discussed, this painting, too, was closely modelled after a Pāla
original. In fact, if someone were to cut off the lower register, one could easily suggest a Maghadhan
origin for this painting. Except for the Indian pandit wearing a yellow cap at the lower left, all the
others definitely represent Tibetan monks. One of them (at the far right) wears the type of broad-
brimmed hat that is commonly seen in western Tibetan murals. This, together with the other minor
similarities, may tempt us to suggest a western Tibetan origin for the thanka. But it is possible that
the monk alone may have been from western Tibet; several western Tibetan monks, such as the Zanskar
lotsava, were equally well-known in central Tibet. Moreover, this type of hat was worn generally by
Tibetan monks during travel. What does deter us from suggesting a western Tibetan provenance is the
notable difference between this thanka and the Kadampa style murals in that region (Pls. 10 and 14). In
its accomplished drawing, delicacy of its ornamental features, polished elegance of its figural forms and
the sensitivity of its vividly evocative colors, this thanka, along with the Sakyapa portrait is a
classic example of the high Kadampa style that I believe flourished in the temples and monasteries of
south and central Tibet in the twelfth century. Notwithstanding its borrowed elements, it demonstrates
how adept the Tibetan artists were in assimilating various manners to create an original and beautiful
work of art.
But for one figure wearing a pink garment of floral print (third figure from the right in the topmost
row), the extraordinarily beautiful thanka of Samvara has nothing that is typically Tibetan (Pl. 12).
This is true also of a small but certainly early Mahākāla (Pl. 13), which was probably rendered for a
Kagyupa. The same Mahākāla sits at the extreme end of the bottom register in the Samvara thanka, and this
form of Mahākāla is closely associated with Kagyupas. He is prominent among the Lhakhang Soma murals (Pl.
14), where too he is surrounded by bird- and animal-headed figures, as in the thanka illustrated here.
Moreover, the iconography of this Mahākāla thanka differs notably from that of the more typical and later
Sakyapa representations (Pls. 27-28). Among the most striking visual elements of the Samvara thanka are
the two golden footprints prominently placed on either side of the central deities, each footprint being
supported by a lotus. They almost certainly represent the footprints of the Buddha.
Pl. 12 Samvara and Nairatmya in Yab-yum Central Tibet (Kadampa monastery). 12th century 43 x 27 cm Private Collection
Pl. 13 Mahakala Central Tibet (Kagyupa monastery?). ca.1200 19 x 15 cm. Private Collation
Pl. 14 Mahakala Ladakh (Alchi, Lakhang Soma). Early 13th Century Mural
Pl. 27 Mahakala and companions Central Tibet (Sakyapa monastery), 15th century 53 x 45 cm. Private Collection
In any event, except for the details mentioned, nothing in these paintings betrays their Tibetan
character. On the contrary, the figural forms, especially the two kneeling figures with pots in the
Mahākāla thanka and the two nāgas kneeling at the corners of the throne in the other, have exact
parallels in Pāla bronzes. Similarly, the meandering vine on either side of the throne contains figures
that may just as well have been rendered by the same artists who did the manuscript illuminations at
Nalanda and Bodhgaya. Just as the Samvara mandala in the Kadampa style may actually have been executed by
Newari artists, these two thankas could have been painted by Indian artists from Magadha.
The close relation of both these paintings, though their colouring differs considerably, to other Kadampa
style thankas is quite evident. For instance, Mahākāla’s elaborate necklace with fringes is quite
different from the type of necklace he wears in later paintings, but is very similar to those seen in
some of the Kadampa thankas (Pls. 8-9). The design of the lotus as well as the figural forms of these
Kadampa thankas and the Mahākāla painting are also very similar. The Mahākāla could hardly have been
painted much later than the Lhakhang Soma mural of the same subject.
As to the Samvara it is difficult to understand why it has recently been suggested that this is a
‘Western Trans-Himalayan’ painting of the late thirteenth century.
Klimburg-Salter: p.123.
The accomplished and sophisticated rendering of this thanka makes a western Tibetan origin most unlikely.
As it has been noted already, the western Tibetan version of this style is considerably more crude and
provincial. Moreover, its similarities with the other examples of the Kadampa style paintings from
south-central Tibet are clear. Compare the various representations of Mañjuśrī in the second row from the
top with those at the bottom of the Banaras Vairocana (Pl. 8), or the pedestal below the lotus with their
grinning faces with that in the Lineage thanka (Pl. 6). The figures in the Lhakhang Soma generally lack
the suave elegance of those in the Samvara or the Lineage thankas, while the grinning faces in the
western Tibetan representations are even more cursive and impressionistic. Thus, if the Lhakhang Soma
murals and the Kharakhoto thankas were rendered in the first quarter of the thirteenth century, it would
be difficult to date the Samvara thanka later than the end of the twelfth, and certainly it must be
regarded as a fine example of the high Kadampa style of south-central Tibet.
mahāsiddhas. Moreover, the dancing figures both at the top and the bottom appear to be more Chinese.
Two more outstanding paintings of the Kadampa style represent the Goddess Vajravārāhī or the Adamantine
Sow dancing with her companions (Pls. 15-16). Both thankas may also be compared with a third from the
Kharakhoto group (Fig.9) which was painted before 1227 A.D. The two thankas must at least be contemporary
with the Kharakhoto painting, if not slightly earlier.
mahāsiddhas. Moreover, the dancing figures both at the top and the bottom appear to be more Chinese.
Pl. 15 Dancing Vajravarahi Central Tibet (Kadampa monastery). ca. 1200 82.9 x 60 cm. Private Collection
Pl. 16 Dancing Vajravarahi Central Tibet (Kadampa monastery). ca.1200 J.P. Goenka Collection, India
Fig. 9
Dancing Vajravarahi
China (Kharakhoto), Before 1227
Watercolours on hemp cloth; 111 x67 cm.
The Hermitage Museum, Leningrad
What is interesting is that although painted in the same basic style, the three paintings differ
considerably not only iconographically but also in their visual elements. All eight cemeteries are
represented with greater detail in one of the thankas (Pl. 15), while in the other two (Pl. 16 and Fig.
9) they are more perfunctorily added within the central zone, almost as an afterthought. Both in Pl. 15
and Fig. 9, the goddess, accompanied by six (female) companions, dances on the orb of the sun placed on a
prostrate body or corpse which lies fully stretched on a large lotus. But in Pl. 16 she is accompanied by
a larger number of female companions and her left foot rests directly on a crouching figure which is
placed on a small lotus. The figure of Vajravārāhī herself differs considerably in the three
representations in their proportions as well as their physiognomic features. The representations in Pl.
15 and Fig. 9 are considerably more linear, but the outline of the figure in Pl. 16 has been reinforced
with broader brushstrokes to create more modelled forms. In fact, most of the dancing figures in this
thanka have been similarly treated. Monks and mahasiddhas are included in all three paintings but in
their iconography as well as forms, those in the two Tibetan thankas are more alike. The Kharakhoto monks
and mahasiddhas are not as well proportioned but are certainly livelier. The Kharakhoto painting was very
likely meant for a Chinese patron as is indicated by the empty cartouches besides the monks and the
As has already been noted, the Kharakhoto paintings may have been associated with the Kagyupa order. It
is interesting to note that the dark monk in the vignette immediately to the left of Vajravārāhī’s lotus
also figures among the monks in the lower register of Pl. 16 (second figure from the right). However, the
mahāsiddhas appear to be different in the two paintings. As a matter of fact, the mahāsiddhas in the
Kharakhoto thanka can be identified more easily than they can be in the other two. The figurative forms
of the monks and mahasiddhas in both the Tibetan Vajravārāhīs are much closer to those in the early
Lineage or Samvara thankas (Pls. 6 and 12), and there can be little doubt that despite their differences,
they belong to the same basic style.
Pl. 16 Dancing Vajravarahi Central Tibet (Kadampa monastery). ca.1200 J.P. Goenka Collection, India
Pl. 17 Vajrasattva and Prajna Western Tibet. late 13th century 36.6 x 32.5 cm. Zimmerman Family Collection
Pl. 18 The Goddess Tara Central Tibet (Sakyapa monastery?). ca.1300 52.1x43 cm. The Cleveland Museum of Art. The J.H. Wade Fund by exchange
These variations within the same given style clearly demonstrate that the artists not only enjoyed
considerable freedom but that the iconographic constrictions placed on them did not always suppress their
creative flair. How excitingly different the three paintings are in their compositional arrangements,
visual passages, application of colours as well as in their overall aesthetic effect!
It has already been noted that the murals in the Kadampa style may be seen as far west as Saspol and
Alchi in Ladakh (Pls. 10 and 14). It would appear that the style may have been carried to these regions
from south-central Tibet by the Drigungpas, a subsect of the Kagyupas, perhaps early in the thirteenth
century. Earlier murals in the monasteries and temples in the region are decorated primarily in the
Khache-Tibetan style derived from Kashmir. Curiously, very few thankas in the Kadampa style have emerged
from western Tibet or Ladakh. Some were recovered apparently from the Kailash region and have been
published elsewhere.
Olschak Wangyal, pp.50-52.
One of the finest thankas painted, probably in western Tibet, but in the Kadampa style, depicts
Vajrasattva and his consort (Pl. 17). Very likely, the thanka formed the frontispiece of a series from
which another example depicting a monk is now in the Los Angeles County Museum of Art. The two figures of
Vajrasattva and his consort stand out boldly against the overwhelming red which is the predominant
colour. The feeling of tenderness and conjugal harmony between the two is well expressed not only by
their intimate postures but also by their smiling countenances, as one sees in Pāla images of
Umā-Maheśvara. The jewelries they wear and the striped garments are especially attractive. The figures
surrounding the pair are either Buddhas or bodhisattvas. At the bottom, however, are the seven auspicious
symbols consisting of the king, the queen, the horse carrying the jewel, the elephant carrying the wheel
and the general. The monk worshipping in the corner wears the red robe typical of Indian monks in early
Tibetan paintings.
The attire of the king and queen provides one clue that the thanka was painted in western rather than
south-central Tibet. This may be further confirmed by a comparison with the Lhakhang Soma murals (Pls. 10
and 14), though the artist responsible for the thanka wielded a finer brush. In spite of the fact that
the figures in the sambhogakāya thankas (Pls. 7-9) are not as suave and carefully drawn as the figures in
this thanka, they are certainly more graceful than those in the Lhakhang Soma murals. This is also the
case with the lions and the elephants that decorate both the sides and the base of the thrones in the
mural and the thanka.
By the fourteenth century the Kadampa style appears to have run its course. The monasteries in Magadha
and Bengal ceased to function by the thirteenth century, and thereafter, Nepal became a more important
source for Tibetan Buddhism. Newari artists, of course, were present in Tibet from a much earlier period,
but one can well imagine that as long as the Maghadhan and Kashmiri monasteries were flourishing the
Tibetans continued to maintain direct contact with them. After their dissolution, however, it was natural
for them to turn to Nepal both for spiritual and artistic nourishment.
This is clearly evident from a painting of the Goddess Tārā now in the Cleveland museum (Pl. 18). We do
not know exactly where the painting was done, but it reflects a combination of both the Kadampa and the
Nepali styles, as does the Śākyaśrī thanka (Fig. 10). Although the red predominates as in Nepali
paintings, its tone is somewhat different from the Nepali red. The goddess is elaborately enthroned
within a shrine whose tiered roofs are beautifully fringed with flowering foliage of various shades of
green. While the shrine itself is copied from a Pāla miniature and is obviously meant to portray a famous
sanctuary of the goddess in India, the elaborate throne with its intricately designed torana is clearly
Nepali. Such thrones are not encountered in Pāla illuminations and constituted a stock motif of eleventh
century Nepali miniatures. That a Nepali artist was responsible for this vivid and sparkling painting is
clear and the fact that it contains no Tibetan figures led us earlier to consider this as a Nepali painting.
Pl. 45 Cosmic Form of Avalokitesvara Central Tibet (Gyantse style), 15th century 50x40.5 cm. Professor M. Driesch Collection, Cologne
It is doubtful if one will ever be able to determine absolutely whether this was rendered in Nepal or in
Tibet. However, one or two features that indicate a Tibetan relationship must be pointed out. The
grinning faces of lions and elephants are not encountered in any Nepali paintings but, as we have already
seen, occur frequently in Kadampa style thankas in Tibet. The shape and features of the face are drawn
with the same kind of sharpness and icy precision that one finds in the thirteenth century murals of
Ladong and the later wall paintings of Shalu and Gyantse. Indeed, the facial features and their precise
delineation are comparable to the face of Avalokiteśvara in a later painting which may have been painted
in Gyantse (Pl. 45). Finally, the treatment of the lotus in which the goddess sits and the sense of
colouring in general are closer to such Kadampa style thankas as the Samvara mandala (Pl. 11) than they
are to Nepali paintings of the period.
While archaistic survivals of the Kadampa style may be traced in many of the western Tibetan and Ladakhi
monuments, by and large, the Kadampa style was supplemented in the fourteenth century in central Tibet by
a new style that drew its inspiration from Nepal. The monastic order that was primarily responsible for
the emergence of the new style was that of the Sakyapas in the province of Tsang.
Pl. 5
Fig. 8
Fig. 9
Fig. 10
Pl. 6
Pl. 7
Pl. 8
Pl. 9
Pl. 10
Pl. 14
Pl. 18
Fig. 1
Fig.2 Mahasiddha Sengepa
Pl. 11
Pl. 12
Pl. 13
Pl. 27
Pl. 15
Pl. 16
Pl. 17
Pl. 45