Professional Documents
Culture Documents
GANDHARAN BUDDHISM
Archaeology, Art, Texts
I
u
www. a .ccesscopyright.ca.
15 14 13 12 II 10 09 08 07 06
543 2 I
by
ISBN-10: 0-7748-1080-7
1. Buddhism - Gandha.ra (Pakistan and Mgh anis ta n ) - H i t ory. 2. Gandhara ( Pakista n
and Afghani tan) - Antiquities. 3. Relics - Gandhara (Pakistan and Mghanistan). 4. Art.
Buddhist - G an dhara (Pa kistan and Afghanistan). 5. Art, Ga n dhar a . 1. B r an caccio, Pia,
1 9 66 - II. B ehre n dt , Kurt A, 1964- I I I . Tide. IV: Series.
294.3'0934
BQ286.G35 2006
C2006-900161-8
Program in Buddhist Studies at the
Canda
UB C Press gratefiilly acknowledges
dle
of the Canada C o un c il
for
the
Arts, and
Arts Council.
UBC Press
The University of British Columbia
2029 Wes t Mall
Vancouver. BCV6T IZZ
www.ubcpress.ca
Phot(l
Contents
Illustrations
/ IX
Acknowledgments
Introduction /
/ XII
/ 41
Mauriz'io Taddei
Behrendt
/ 107
Shoshin Kuwayama
6
135
Richard Salomon
Rhi
viii
Contents
The Nature and Use of the Bodily Relics of the Buddha in Gandhara / 183
Robert L. Braum
in
210
10
II
12
13
243
Fil(\tenzi
Contributors /
Index /
306
303
225
286
The existence of Mahayana in Gandharan B uddhism and its art has presented
intriguing questions both to Buddhit scholars and to art historiam . While
many scholars have presumed that Gandhara , or the northwestern part of the
Indian subcontinent, was a region that played a critical role in the emergence
of Mahayana in Indian Buddhism, they have been simultaneously faced with
the paucity of tangible evidence to support this understanding. The testi
mony by the Chinese pilgrim Faxian that most people in the region were
practicing Hinayana, not Mahayana, also provides a basis for skepticism among
some scholars. ' Those who are familiar with sculptural material, however,
have seen elements in Gandharan art that must be relatable to Mahayana , no
matter what form its followers took as a social entity. 3 All of the complex
aspects of this issue cannot be reviewed here ; instead, the focus will be on
how bodhisattvas were represented in Gandharan art.
The ideal of the bodhisattva is generally considered one of the central
features of Mahayana Buddhism . 4 Although the word bodhisattva originated
in the pre-Mahayana tradition, it gained particular importance with the rise
of Mahayanists and their scriptures, not only as a general ideal, but also as
referring to a class of venerated deities . Numerous bodhisattvas, besides those
in the pre-enlightenment stage of mal;t ui-bu ddhas who are also mentioned in
non-Mahayanic contexts, were mentioned in the Mahayana literature, and
some of them were worshipped and represented in visual form wherever
Mahayana Buddhism spread.
In Gandharan art, bodhiattvas commonly app ear as cult llnages or in the
relief, showing complex scenes, particularly in later p eriods, and they display
diverse iconographic forms that apparently signify the multiplicity of their
identity. Compared with their representations in other important regions of
early Buddhist art in I ndia, such as Mathura, Andhra, and the Western D eccan,
representatiom ofbodhisattvas in Gandharan art are much more conspicuous
in number and variety, and they constitute a prominent feature in Gandharan
I
1 52
Bu ddhism. For this reason, a se arch for bodhisattvas within Mahayana con
texts in Gandharan art could b e a useful starting p oint. l The questions ad
dressed in this chapter are straightforward, although answering them may not
be simple. They are : How many different types o f bodhisattvas were repre
s ented in Gandharan art? What kinds of p a tterns emerge in these typ e s ?
Wha t are their conne ctions with t h e presence of Mahayana i n Gandharan
Buddhism?6
Among a variety ofbo dhisattvas represented in Gandharan art, three promi
nent types emerge. A relief in the Taxila museum includes standing examples
of these three typ es along with Buddhas (figs .
7 . 1 , 7.2) . 7
B o dhisattva and
B u ddha figures alternate, and each bo dhisattva appears to form a pair with
the Bu ddha next to him . The one on the far left, with knotted hair and
holding a water vase, is a typical example of the bodhisattva Maitreya (who is
destined to attain Buddhahood in the future) ; his image appears consis tently
in sc enes of the seven Buddhas of the pas t . 8 Although the significance of the
chara cteristic hair and water vase, or
kammJ4alu,
brahmiil,la
family. 9 The Buddha next to him must, then, be one closely related to the
b odhisattva Maitreya , most probably Maitreya in fully enlightened form.
The b o dhisattva in the c entre, wearing a turban and with his left hand
pla c e d on his waist, is commonly identified as the b o dhisattva Siddhartha in
the appearance of a princ e . The similarity of this figure to some found in
narrative scenes from the life of S akyamuni Buddha has b een pointed out by
Miyaji . 'o Several epis odes, such as the Marriage and the Resolution for the
Great D ep arture, present the prince Siddhartha in an identical manner. n The
male figure in this form has precedents in earlier B uddhist art, such as the
p rince
in Gandha.ran narrative scenes . I3 In the Taxila relief, the turbaned male of this
type is p o sitioned at the centre to s how his primary importance . It is also
notewortllY that the B uddha in the c orresponding pair is the only one in an
active posture, with the right hand in
abhayamu dra;
c orresponding B u ddha in its pair make it obvious that this figure also repre
sents a bodhisattva . Who, then , is this third b o dhisattva? It woul d not be
entirely out of the question to consider that tIus figure, together with its
FIGURE
FIGURE 7 . 2 .
(after
D etail of Figure 7 . 1 ,
Ingholt, pI. I-I I ) .
paired Buddha, represents one o f the well-known Buddhas from the p ast,
such as Dipal;kara or Kasyapa, as a Buddha and a bodhi,attva. Dipal;kara and
Kasyapa are the only two Buddhas, other than S akyamuni and Maitreya, re
corded in extant inscriptional evidence on Gandharan art. '4 One has to note,
however, that this third bodhi,attva type, the wreath-bearer, is also seen in
larger statues and therefore must have been someone who had greater im
portance as an obj ect of veneration. It seems hardly likely that the bodhisattva
Ka syapa or D ipal;kara was worshipped using cult statues on s uch a scale, and
1 5 4 Juhyung Rhi
thus an alternative should be sought among those who would have been
more prominent in the contemporaneous c ult practice. The wreath-bearing
b o dhisattva has often been identified as AvalokiteS vara, unquestionably the
mos t popular bo dhiattva in the Mahayana world, who c o uld have been wor
shipped in Gandhara as well !l The wreath in the left hand has b een inter
preted as interchangeable with a lotus, the distinctive attribute ofAvalokiteSvara
in later Indian art and in early Chinese Buddhis t art . 1 6
If we accept this view and identify the b odhisattva in the third pair as
AvalokiteS vara , the identity of the Bu ddha ne}..t to him presents a question . It
could be Amitabha, who is traditionally asso ciated with AvalokiteS vara. The
relationship b e tween AvalokiteS vara and Arnitabha in this pair is obviously
different from the one in the o ther 1:\\'0 pairs , b e c ause Amitabha prior to
enlightenme nt was the bhiklt Dharmakara, not AvalokiteSva ra . However,
AvalokiteS vara , who would become a Buddha and preside over the Western
P ure Land after the final nirv31;la ofArnitabha, was commonly regarded as the
emanation and thus the spiritual son of Amitabha Buddha. He was already
referred to as buddhasu ta (son of the Buddha) in the
long b efore this relationship was codified elaborately in Vaj rayana B uddhist
iconography. S o it is possible that the inseparable tie b etween these two dei
ties was already re cognized in this p erio d. Amitabha is s aid to have appeared
in the world many kalpas b efore S akyamu ni ; I N therefore, he c ould theoreti
cally be ref,r arded as one of the past Buddhas , although still living, at an ines
timable age, in the Western Pure Land. These three p airs, then, can be read as
the B uddhas and bodhisattvas from the p a s t, present, and future who have
delivered or will deliver the dharma to sentient b eings at different moments
and in different modes.
The three b odhisattva typ es discussed above, Maitreya, Siddhartha, and the
wreath bearer, were predominantly employed for independent statues. Among
s tanding b odhisattva statues , we find numerous Maitreya images of the first
typ e, which make up by far the largest numb er of Gandharan bodhisattvas .
The maj ority of tllem have knotted hair arranged with a ringlet (fig. 7 . 3 ) . A
7 . 4) . The
anlples are invariably of inferior quality and have rather decadent features
(fig.
7.5).
BodhisatfIJas
if!
Gandharan Art 1 5 5
FIGURE
7 . 4.
Musee
These three basic types can be seen also in seated bodhisattva images , which
exhibit more variety, owing to modifications and insertion of new elements .
One of these types is a seated bodhisattva with a water vase, usually held in
the hands but occasionally carved on the pedesta1.21 Such figures could be
regarded as seated Maitreyas, corresponding to the first type of the standing
images. There are, however, some instances of a b o dhisattva having the same
characteristic hair and exhibiting the preaching gesture, but without a water
vase;" one wonders whether this figure should be regarded as an extension of
the seated Maitreya type or as a new iconographic invention for yet another
bodhisattva.2J
The statue of the prince in the Firs t Meditation in the Peshawar Museum
(fig. 7 . 6) is a representative example of the seated Siddhartha image. The
I S 8 Juh}'lll1g Rid
FI GURE
7 . 6 B odhisattva Siddhartha
(after
Kurita, vol.
1 , pI. 1 3 1 ) .
many questiom. Should the wreath and lotus be indeed equated as icono
graphic signs for the same deity? If so, what is the reason for the replacement
of a wreath with a lotus in seated images? Or does the seated type with a lotus
represent yet another, possibly a fourth, bodhisattva ?26
The bodhisattvas of this category are usually seated in paryal'ikasa/w or the
ordinary cross-legged pose. But there are quite a few examples ofbodhisattvas
seated on a stool in the pensive pose, one foot resting on the opposite leg,
FlGURE 7 . 8 .
which h an gs o ff the stool (fig. 7 . 8) . The essential similar ity i n icon o grap hi c
attributes with AvalokiteSvara im age s among the Kashmir bronzes is quite
remarkable!7 We also recall that this p ose was used for the prince Siddhartha
in a narrative relief panel, now in th e Lahore Museum, depicting th e First
Meditation.28 If the o ccurrence of this pose in narrative context prece de d
that of the independent statues, we may suppose that this position, initially
r,'
'
1 62 jtlhyllllJ; Rhi
FIGURE 7 . 1 0 .
King As oka, and the one on the left is the Display of the Snake B owl to
Uruvilva Kasyapa. If these scenes and the bodhisattva in the centre are con
nected, the bodhisattva should be someone closely related to S akyamuni Bud
dha, such as one of his previous incarnations , the prince Siddhartha or the
bodhisattva in Tuita H eaven . In any case, it is clear that neither the crossed-
ankle nor the p ensive pose was used exclusively for any p articular bodhisattva
in Gandharan B uddhist iconography.
Turbaned bodhisattva statues occasionally have small figurines on the me
dallion of the headdress; these might have been used as a dditional icono
graphic signs, and they often include a small Buddha or Garuc.la c arrying a
niiga or niigini. There are several examples in which such figurines remain on
the headdress and several more where the medallions have become detached.
Some bodhisattva statues aho have a small proj ection in front of the headdress
for attaching small medallions, so" such figurines could have been quite com
mon elements in turbaned bodhisattva s tatues .
The motif of Garuc.la carrying a 11iiga or niigini (fig. 7 . n ) is generally be
lieved to be related to the Gre ek myth of the abduction of Ganymede by
Zeus in the form of a gigantic eagle. Al though this th eme appears wi d ely in
ancient art throughout west and soutll Asia and h a s been interpreted in vari
ous ways , its significance in the context of the bodhisattva's headdress remains
obscureY Alexander Soper suggested that it signified AvalokiteS vara's special
position as the bearer of souls to Sukhavati, but this interpretation is no more
dlan a speculative suggestion for justifYing a preconceived identification, and
there seems no particular reason to connect it to Aval okit eSvara . 3 3 We should
note that this motif is shown on the headdress of the famous "Foucher
bodhisattva" (fig. 7 . 4) , which we identify on the basis of typology as the
prince Siddhartha. So it may well have been the iconographic sign ofSiddhartha
- or it may have been used there with a generic religious meaning, not as an
iconographic signifier.
A small Buddha in the headdress is a well-known symbol for AvalokiteS vara
in Indian Buddhist art, at least from the Gupta period onward . It is also
widely adopted for AvalokiteS vara images in east Asia, and many scholars have
presumed that the tradition had its origin in early Indian Buddhist ar t . A
bodhisattva figure in the Peshawar Museum of the third standing type, that is,
holding a wreath, bears a small Buddha in the preaching gesture on its head
dress (fig. 7. 12) , and this may justify using the small Buddha as a sign for
Avalo ki t eS vara . One should note, however, that a bodhisa ttva standing on the
right side of a Buddha triad in the Claude de Marteau collection that would
be classified as the second standing'tYEe (Siddhartha) by the hand gesture also
has a s mall Buddha on it headdress . J4 Furthermore, the presence of a Buddha
in the headdress as a distinctive iconographic indicator for AvalokiteSvara was
not an establi.,hed phenomenon even in early Chinese Buddhist art, as is conven
tionally presumed. During the fifth and sixth centuries CE in China, it was
Maitreya seated in the crossed-ankle pose (possibly in the Tuita Heaven) , not
AvalokiteSvara, that commonly ha d a Buddha in the headdress . J5 Therefore,
1 64 JUhYlllIg Rhi
FIGURE 7 . I I .
pi. 1-4) .
there seems little reason to suppose that the tradition already existed in
Gandhara. The B uddha in the headdress may well have served as a general
si gn indicating that the bodhisatrr;df ft th e stage of ekajatipratibuddha, or
destined to b ecome a Buddha very soon, and could have been used for any
bodhisattva at this s tage, including Siddhartha or AvalokiteSvara.
B esides a small Buddha image or Garu<;la c arryin g a naga or niig ini, several
o ther motifs appear on the headdress , such as a figure holding a garl an d one
in the pose of anjalimu drii, a lion head, and even S Urya riding in a horse
chariot.36 It is unlikely that all these functioned as iconographic indicators for
particular bodhisattvas .
,
FIGURE 7. 1 2 .
mound
B o dhisattvas were also commonly incorp orated in the Buddha triad for
mat, of which two triads from S ahri-Bahlol are the best-known examples
(fig. 7 . 1 3 ) . 37 Two standing figures on each side of the c entral Buddha in this
triad are identified as b odhisattvas, not devas, because the two smaller inter
mediate figures b etween the Buddha and the two bodhisattvas are easily
identified as Brahma and lndra . One of the two bodhisattva figures is the
1 66 juhYII/lg Rhi
Museum (after
1 6 8 Juhytlug Rhi
in Buddha triads generally consisted of either the Maitreya type and the
Siddhartha type or, more commonly, the Maitreya type and the AvalokiteSvara
typ e .
In the late p eriod of Gandharan art, there seem to have been considerable
modifications in this format. In a stele from Sahri-Bahlol mound D (fig. 7. 14) ,
a Buddha triad ap p ears in an elaborate architectural settin g , and the turbaned
bodhisattva on the left has his left hand on his waist just like the Siddhartha
type. However, in the same hand, he holds a lo n g lotus stem that makes him
identifiable with AvalokiteSvara. The ' b ddhis attva on the opposite side, with
jatamukuta, holds a water vase in the left hand like an ordina ry Maitreya, but
displ ays t'aradamudra with the right hand . D espite minor modifications, this
triad still seems to represent S akyamuni with Maitreya and AvalokiteS vara.
In the triad format with seated bodhisattvas as attendants, they are gener
y
all seated in the pensive po s e, but there are very few examples with both
figures preserved. A triad fi'om Lor iyan Tmgai is among these rare e xamples
(fig. 7. IS) . In this triad, the bodhisattva on the Buddha 's left we ars a turban
and hold a lotus, features that can usually be associated with AvalokiteSvara .
Interestingly, the b odhisattva on the opposite side holds a book, not a water
vase, and he does not wear a turban like Maitreya . It must be that the iconog
raphy of Maitreya was modified or M aitreya was replaced by a different
bodhisattva, perh aps from a Mahayan a co ntext . Since the ico n ography of
Mai treya was firmly established in Gandharan art, it seems doubtful that the
former was the case. We recall that the b o dhis attva in the pensive pose with a
book exists as an independent statue (fig. 7 . 9) . B e c ause the head of the statue
is b ro ken, it is impossible to confirm which bodhisattva' type was represented
in this ins tance . In fact, bodhisattvas holding a book are relatively common in
Gandharan art , especially among obj ects datable to the late phase.40 A stele in
the Peshawar Mmeum also has such a turbaned, crossed-ankle bodhisattva
inside a shrine, though it is no t clear th 4 a specific bo dhisattva is meant.4'
However, if one focuses on the 60k liio e, one is tempted to identify this
type of b odhisattva as Mai1jus ri, who was quite important in early Mahayana
Buddhism as the embodimen t of wisdom, as stated in the Prqjfiaparamita sutras
or the Avatall1saka st1tra. Because the iconograp hy of M a ilj us ri characterized
by a scrip ture is generally known from relatively late textual sources of eso
teric Buddhism and such visual representations b ecame common in later
esoteric contexts,42 their presence in Gandh ara may be questioned . However,
a Buddhist text preserved in Chinese translations datable to as early as the
third century CE , Wensllllshirib annieiJal1iing (Maillsriu irva1Ja s Citra) tells us that
Mai1jusri hold a Mahayana scripture in the left hand,4J and thus the possibil
ity of i de ntifYi n g our figure with Maiijusri c ann ot be ruled out.44
FIGURE 7. 1 5 .
AilS.
1 70 JUhYlll1g Rhi
, FIGURE 7 . I 6 . Part of B uddha triad. John and Mable Ringling Museum ofArt, the
State Museum of Florida (after Davidson, pI. 23 ) .
1 72 JuhYli ng Rhi
left and right the meditation and preaching gestures . The figure on one side
of the Buddhas is in the p ensive pose; the one on the other side is in the
crossed-ankle pose. If we read these seven figures as symb olically converging
toward the central Buddha, not in mere juxtaposition, we can naturally con
sider that some sort of hierarchy is present in their ordering; i.e. , the preach
ing bodhisattvas in the crossed-ankle pose would appear above the pensive
bodhisattva. . 52
This is all the more intriguing whe we recall that a similar hierarchical
scheme is present in early Chinese B uddhist art. In the fifth-century-CE
caves at Yungang and Dunhuang We find numerous figures of bodhis at tvas in
the crossed-ankle pose. They are generally identified as Maitreya in the Tuita
Heaven, although I am skeptical whether all of them should be so regarded, as
some could also be a previous incarnation of S akyamuni in the same heaven.
In any case, the Maitreya-type figures seated with crossed ankles are com
mon, along with bodhiattvas in the pensive pose as attendants (fig. 7. 17), thus
suggesting a hierarchical relationship between these bodhisattva groups. 53 I
believe that in early Chinese Buddhist art these two types were used with
hierarchical implications, p erhaps to indicate multiple stages within the
bodhisattva practice, rather than for iconographic identification of particular
bodhisattvas . Although we do not know exacdy where this convention origi
nated, we have to seriously consider the possibility that it started in Gandhara.
1 74 Juhytmg Rhi
sattva's ability to reveal himself in infinite forms to convert all sentient beings .
According to this passage,AvalokiteSvara preaches dharma with versatile skill
fulness, taking the shape of various bJi.1l,gs,including a B uddha, bodhisattva,
pratyekabuddha, Brahma, Indra, IS vara ; and Vajrapa1;li. 56 Taddei rightly pointed
out, however, that this ability is not unique to AvalokiteSvara and c ould be
extended to numerous other bodhisattv;L. 57
A similar account is found in a more generic form in o ther Mahayana
scriptures, such as the Da. abhumika siUra and the Bodhisattvabhami. The former,
which eventually constituted an essential p art of the Avatarf1saka sa tra, ex
plains how a bodhiattva, reaching the s eventh stage of the bodhisattva prac
tice, acala, attains the capability to transform himself into innumerable forms :
Now for th e B odhisattva who has reached this stage th e power of B o dhi
sattva- c o n duc t is atta i n e d thro ugh the divers i ty of i n n umerable bodies , the
pro duction of in numerable voices , inmleasurable b o di es , the immeasur
able numb er of bi rth s . . . He adapts and s ustain s hi s own body in accor
d anc e with the birth and the attainment of b o di es o fliving b eings in o rde r
to mature them . . . In c o nfo rmity with the variety of bodies (i. e. colors ,
mark, form, length and width) and the vari ety of intention and mental
d is p o sitio m of living beings, so he shows in vario us places hi own b o dy
resp e c tively in
a dap ta
tion to
chical s tages in the bodhisattva pra ctice may have underlain such imagery in
the Northwest.
In Gan dh aran art, not only were numerous bo dhisa ttva figures rep res e nt ed ,
but also many different kinds existed among them. Their number and variety
obviously exc eed what would have been possible within the boundary of
no n-M a h ayani c traditions . Extant b odhisattva figures seem to include those
most p rob ably w or shipp e d exclusively by Mahayanists, such as AvalokiteS vara
and Maiijusri. Further mo re , there are images, such as the bodhisa ttva in multi
plicatio n , that could have been based on the s ophisticated idea of the multiple
manifestations of a Mah ay a na bodhisattva in a h igh stage of practic e . This
tradition, which is not conspicuous in the early p er i od of Gandharan art,
seems to have become quite proInin ent in it later stage, particularly in b o dhi
sattva statues, tri ads, and steles of c omplex scenes . It appears that in the later
years demand for such i mage s ro s e s i gni fic a ntly in Buddhist monasteries.
Several monastic s i tes , s u c h as Mohammed Nari,Takht-i-bahi, Sahri-Bahlol,
and Lo riya n Tanf,rai, s eem to be p a r tic ul arly important in tIus connection.
Some of the most spectacular steles depic ting complex scenes, which incor
p orate various types of b o dhisattva s , c ome from Mohammed Nari, Takht-i
b ahi and Sahri-Bahlol, p e rh ap s the most important centres of B uddhist votive
carvings in stone in Gandhara, arid whkh yielde d numerous b odhisattvas,
Bu ddha triad , and steles of complex scenes. Loriyan Tarigai is notable for
several bo dhisa ttva statues h oldin g a wreath or a lotus, and Buddha triads
with seated b o dh is a ttva s . 62 One is tem p t e d to vi s u ali z e vigorous activities of
M ah ayanis ts at these monasteries. It might b e hasty, however, to c onclude
that they were Mahayana monasteries, for e p i grap h ic al evidence from Takht
i-bam indicates that they were possibly affiliated to the Kasyapiya , one of the
most traditional Buddhist s ch ools generally p o sit e d against Mahayana.63
This p erplexing situation may be best understood by s up p o sin g that most
Mahayanists in Gandhara still exi s t ed within monastic establishments of tra
di t ion al schools, p ossibly as their progressive wings . Sinill ar opinions have
been exp ress e d by B u ddhi s t scholars . E tienne Lamotte, for instance, has po inte d
out that Mahaprqjnnpnramita-upadda, tra di ti on ally attributed to N agarjuna, was
ac tually composed by on e or more monks trained in the Sarvastivada from
the northwestern region of ancient India.('4 Heinz B echert remarked that
monk. of two branches, Ma hay ana and Hinayana (or non-Mahayana) , lived
to gether peace fully in the same monastic c ong rega ti ons . 6 5 These opinions
seem to coincide with o ur supposition, and this may explain why the Chi
nese p ilgrim Faxi an reported in his travelogue that the maj ority of B u d dhis ts
in Gandhara were practicing Hin ayana .66
Such mon a st e r ies as Takh t- i -b ah i , Sahri-Bahlol, Mohammed Nari, and
Lo iiy an Tfuigai, where b odhisattvas i emerge as important cult obj e cts within
1 76 JuhYli ng Rhi
the body of imagery, may not have been declared Mahayana monasteries .
However, Mahayana activities existed at these centres on a considerable scale
and may have grown significantly as time went by. In light of this conclusion,
the bodhisattva. Maitreya and Siddhartha as they appear in Gandharan art,
although not distinctly Mahayana figures , may also need to be reappraised in
relationship to the growing Mahayana faith present in Gandharan Buddhism .
N OTES
I am gra t eful for valuable c omm e nts on thi chapter by Ronald Davids on , Gregory Sc h op en ,
M.au rizio Taddei, and John Rosenfield d uring the . c o nfere n c e at McMa. ter Univers i ty.
geneous groups,
s ep ara te s ocia l
lo o s ely
existe d at diverse levels: those wh o "vrot e sc riptures , those who rea d them , those who
venerated deities discuss e d in th e
as pre
scrib e d by the s cripture s o r k nowle dg eab l e c l eri c s . M ahayan is ts at the la.t two levels may
not nec ess ar i ly have had a clear u nd e rs t a n din g of the d o c tri na l implications of th e deities
th ey ve n era te d i n v is u al images , as is the case with the maj o rit y o f l ay B u ddhists in the
M a h ayan i t
more profound
understanding initiated such practices, an d that many fo llowers p arti cip a ted , rega rdle ss of
their awareness of do c tr in al i mp li c a ti o ns .
Dutt, Mahi:iyi:illa
daijii Imkkya
flO
ketlkyli, 3 -9 .
To b e more pre c is e , this attempt may b e defined as a search for b odhisattvas that c o ul d
o nly have been worshipped by Mahayanists . One might
c o n s id e r
s u ch a. Am i tab ha and Alqobhya that appear only in the context of Mahayana. H owever, in
G andh aran art B u ddha im a ges are hardly disti ng u ishable in ide ntity, and it i do ub tfu l that
there was originally any intention to make them disc ernible even if they were meant to be
6 Alfre d
Foucher's disc ussion of G andh aran b odhis a tt va s in his L'Art grew-b ouddh ique dll
assessment rega rdin g the pres e n ce of M a h aya na bo dhisattvas in th e art is g en er ally nega
th e b asis of the ic o n ograp hy of the advanced M ahayana B uddhist p anth e o Il of the Eso
teric stage, with which h e b e c ame a c qu ai nte d thro ugh his p r evio u s research on Pala
manu sc ripts and s culpture ( s ee Fou cher, Etl lde my "iconographic b(Juddhiqll e) . A recent study
on the i c o nogra phy of bo dhi sa ttvas in Gandharan Buddha triads by the J a panese scholar
Akira Miyaji ( " G andara no sanzon keishiki no ry o kyoji b osatsu no zlIZ6") has many
Carolyn W.
Sclmlidt's 1990 PhD di se rta ti on "Bodhiattva Headdresses and H airstyles " provides a
10
II
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
3I
32
Cataloglle cif Candhara SClilpwres in the S ta t e MuseHm, LlJcknow (L uck n ow : S tate Museum,
1 9 69) , fig. D. Fou cher regarded all snch bodhisattvas as Siddhartha (L'Art greco-hollddhiqllc
tiu Candhara, vol. 2, 228-3 1 ) .
Also see W Z walf, A Catalogue !if the Gandhara SCflipt1lYe ill the British Museum (Londo n:
British M u s eu m , 1996), pis. 3 8 , 39; Archaeological Slirvey of iHtiia, A Hnuai R.eport 1907-1908,
Calcutta : Office of the Su p erinte n dent of Government Printing, 1910, pI. XLVlb.
Osamu Tak a ta ( " Ga n dara bijuts u ni okeru daj oteki chosho: miroku z o to bnnonzo,"
24-2 5 ) q u estio n e d wh e ther the wreath sho uld b e regarde d as a u n iq u e sign for
AvalokiteSvara , altho ugh the lotus may be so re garded.
Pal, Bronzes of Kas h m i r, pI. 45. This figure is ge n erally dated to the seventh or eighth
century CE and understoo d as being relate d to similar rock-cut b o dhisattva images in the
Swat valley.
I ngholt, Can dharall Art ill Pakistal1 , pl. 161 (D ) .
Cf. Miyaj i , " Gandara no hanbshiyu no zuzo," 3 2 1 -53 .
An example of this is in the Musemn fii r Indische Kunst in Berlin ( G r u nwedel, Buddhist
A rt in Iudia, fig. 1 39) ; a n other piece is in the Tokyo N ational Museu m (Kur ita , Candara
bijutsH, vol . 2, pI. 9) .
See, for example, Fo ucher, L'Art greco-bollddhiqlle dll Calldhara, vol . 2 , fig. 42 6.
Grunwedel, Bu ddh is t A rt ill India, 1 09- 1 0 ; Foucher, L'A rt greco-bouddhique du Candhara,
vol. 2, 3 2-40; C oomaraswamy, "Rape of a N agi," 3 3 3 -40 ; Zimmer, A rt '?f Indian Asi a, vol.
1 , 48-67.
Fussman , " Numismatic and Epigraphical Evidence," 73 -74; Iwamatsu , " Gandara chokoku
to amidabutsu " ; Fussman, "La place des SlIkhalJati Ilyliha dam Ie bou ddhisme indien," 5 5 0 .
47 Salomon and Schopen, "On an All e ged R eference to Amitabha " ; ( Grego ry S chopen,
"Ins cription on the Kuan Image ofAmitabha," 1 3 0 , n. 5 0 . Akira Miyaj i initially suggested
that the inscription may have been added later ("Gandara sanzon keishiki no ryo k.-yoj i
bosatsuzo n i tsuite," i n Indo Pa k is ta n 11 0 bukkyii zuzii [Tokyo: Yoshikawa kobunkan, 1985 1 ,
20-21 ) , and then revised hi skep tic al opinion (" Gandara no sanzon keishiki no ryo kyoji
bosatsu no zuzo," 2 7 2) . Salomon cautiously ruled out the p ossibility of later insertion of
the inscription (personal col1llTlunication) .
48 Fouch-er, " Great MiraC le at S ravasti " ; Rhi, " Gandharan Images of the S ravast i Miracle."
49 Rosenfield, DYl1astic A rts f the Kl Is h a ns, 2 3 6- 3 8 .
5 0 John H untin gto n once identified, with elaborate. arb'1.lments, the Mohammed Nari stele as
a representation of Amitabha's pure land, Sukhavati ("A Gandharan Image ofAmitayus'
Sukhavati") . Japanese scholars (for exam ple Toyomune Minamoto, "Shaeij o no shinben"
and ''Jodohen no keishiki") also noted its resemblance to the depictions of Sukha:vati in
East Asian Buddhist art. D espite the apparent similarity, however, there does not seem to
be enough evidence for connecting such a sc ene to Alnitabha's S ukhavati.
51 F ouc h er, " Great Miracle at S ravasti," pI. XXVII.
52 Miyaj i , who also notic ed the significant pairing of bodhisattvas in the p emive and crossed
ankle poses in Gandharan art, suggests that the pensive pose may have implied reflection
on worldly concerns and the release from them, and heading toward the sacred realm of
liberation . See Miy;tii, " G andara no hankashiyu no zuzo," 3 4 8 .
5 3 This i s particularl y notable at the Yungang caves. S e e Zhonggllo shikll y llngan;S! shiku, vol. I ,
pI . 4 3 ; vol. 2 , p is. 1 9 , 5 4, 77, 8 9 , 2 1 0 , 2 1 8 .
5 4 Taddei, "Non-Buddhist Deities i n Gandharan Art."
55 Taddei, ibid. , cited three examples, including one at the N aprstek Museum in Prague that
displays only emanati n g Buddhas . I add here one more piece, at Ch andigarh Museum (no .
23 5 5 , unpublished) .
56 Taddei cited Mallmann's French tramlation ( in tro d uction a 1'6(utie d'Avalokitewara, 3 1 -3 2)
based on the Sanskrit version e dited by Kern and Nanjio. The details of the manifestations
ofAvalokiteSvara vary depending on the versions . C Kern , SaddharmapUl.1qarika, 4 1 O- I 2 ;
T. 2 6 2 : 57a-b; Hu rvitz , Scripture <!f the Lotus Blossom, 3 14- 1 5 .
5 7 Taddei , "Non-Buddhist Deities i n Gandharan Art ."
5 8 Honda, "Annotated Translation of the Dasabllll ll1ika s U lra , " 2 24-2 5 .
5 9 Trans lation based o n Wogihara, Bodhisaltwbhti llli, 64.
6 0 E xta nt examples of se a t ed M aitreya bodhisattvas in dhyanamudra invariably hold a ,vater
vas e . S ee, for example, Maj umdar, Gllide 10 the Sw/ptl lres in the Indialt l\,fusellm, part 2, pI .
IVb; Fou cher, L'A rt greco-brl lddhiqllc du C;diifhili(/} vol. 2, fig. 42 2 ; I ngholt, G a u dha ran A r t in
Pak istan, pI. 3 0 2 .
6 1 According to the Biography qf Vasuba ndhll ( T. 2049) trans lated into Chinese by Paramartha
during the sixth century CE , Asailga was born in Puruapura (Pes h awar) . Xuanzang gave
a sirnilar accOlU1t regarding Asallga and his younger brother Vasubandhu, who was another
eminent exponent of Yogacara thought. S ee B eal, Si- Ytl-Ki, vol. 1 , 9 8 .
6 2 Regarding the dominant importance o f these sites as the provenance o f Buddha triads
and steles of complex scenes, see table 2 ("Provenance Sites of the So-called ' Sravasti
Miracle' Images"), in Rhi, " Gandharan Images of the Sravasti Miracle ," 1 9 2 .
63 Konow, Kh a ros h th i Inscriptions, no. 2 2 . A fragmentary inscription on a p otsherd fromTakht
i-bahi reads "sal11ghe cadl/diSe ka . . . (in the four quarters) ." There is no question about the
reading of the first two words, but "ka . . . " is more problematic. Konow interpreted "ka . . . "
as " KasyaJ'iyalla" (Kisyapiyanlim) . In Kharohi inscriptions, the phrase sa l! lgh e cadudiSe was
46
1 79
u sually followed by the wo rds i ndi cating dle plac e and the school. On the o the r hand, in
3 4)
no s . 33 and
the equivale nt parts read, res p e c tive ly, " sa lJtgh e cadudise Utararamae IakaiiTac Kasaviana
tance o f the K a s yapiyas) and "sa/!lghe cadudiSe Urafaraje acarya {ne]lIa Kal'aviyana" (in the
SaIllgha of the fo u r qu arters , in the Ura s a kingdom, of the Ka syapiya t each e rs) . Therefore,
" ka . . . " c o ul d j u s t as well have been a p l a ce name, and hence Konow's s u gg e s ti on may well
b e questioned. However, su c h a fo r mula app ears m ost freque n tly in the inscriptions o f lli e
K a syapiya , and the only exa mpl e used for o th er schooh is a lion capita l fi:om Mathura
dedicated to the S arvas tiva da (Konow, Kharoshthi IHscriptiollS, no. I S ) . This sect was domi
naI1t in Ga n dh ara , as c o nfir me d by the fact that five out of ten ins c r ip ti o ns from the
region state tha t s ec ta r ia n a ffil i ations b e l o n ge d to tins school. I n l ight of the above. it
se ems p ossible that the in s c ri p ti o n from the Takh t-i-ba hi p o ts h erd mentions the Ka syapiya
B I B LI O G R A P H Y
Ackermann, H . C . Narrative Stonc RcliifsfrolH Galldhara in the Victo ria and A lbert Musellm in
Londoll . Rome: Is titu to I taliano p er il Me di o e d E strem o Or i e nt e ,
Barger, E . md P.
W right .
1975 .
AJha nistatl . M e mo irs of the Archae ological S urvey of India . vol. 64. N ew D elhi: Ar
chaeological Survey of India,
1 94 1 .
Reco rds
In German Sch o la rs Oil india, v o l . I , Cultural D e p art m e nt of the E mb assy of the Fed
eral Republic of G er ma ny, 6 - 1 8 .V.lra ;lasi, India : Chowkhamba Sanskrit S eries Office,
1973
TClUrincIIsia 10 ( 1 9 8 2 ) : 65 -70.
C o o maraswamy, A.K. "Th e R ap e of a N agi: An Indian Gupta S ea l." In Selected Pap ers:
IlrulitiollalArtl aud Sym bolism, 3 3 1 -40. Princeton, N] , : 'Princeton University P re ss,
1977.
Czuma , S.]. Ku shan SClllptllre: Im agesfrom Early In dia. Cleveland, O hi o : Cleveland Museum
1985 .
1976.
rh i .E?egimdngs
'.'
' .
011
the Development
'!f Early B lIddhis t atld Hindu Iconography Held at the A1uscum of In dia u A rt, Bcrlin it! lvla)'
1 9 86, e dit e d by M . Yaldiz and W. Lobo, 67- 8 8 . B erlin : M u s eum fur Indische Kunst, 1 9 8 7 .
5 2 2- 8 6 .
Ie b o u d dhismc
indien."Journal Asia ti q u e
190 1 .
287
( 1 999) :
(Transl a t ed b y A. C.
Gibson and revised md enlarged by J . Burgess trom tile o r i ginal in German. Buddhistische
Ku ns t in Ind i en , 1 89 3 . )
lSI
1 82 Juhyttng Rid
Muller, M. "The Larger SukhavativyHha ." In B u ddhist Malwyalta Tb.-Is, 1 - 8 6 . Sacred Books
of th e East, vol. 49. Oxford: Clarendo!1, 1.8 94.. :
N akamura, H. Indian Budd his m : A SI /IVcy with Bibliographical Notes. H ira ka ta City, J ap an:
KUFS, 1 9 8 0 .
25 , 1 -2 (2002 ) : 3 - 3 2 .
S c hmidt ,
Pre ss , 1960.