JIATS_03_2007

Journal of the
International Association
of Tibetan Studies
Issue 3 — December 2007
ISSN 1550-6363
An online journal published by the Tibetan and Himalayan Digital Library (THDL)
www.jiats.org
Editor: José Ignacio Cabezón
Book Review Editor: Kurtis Schaeffer
Assistant Editors: Alison Melnick, Zoran Lazovic, and Christopher Bell
Managing Director: Steven Weinberger
Technical Director: Nathaniel Grove
Contents
Articles
• A Look at the Diversity of the Gzhan stong Tradition (24 pages)

– Anne Burchardi
• Beyond Anonymity: Paleographic Analyses of the Dunhuang Manuscripts (23 pages)
– Jacob Dalton
• “Emperor” Mu rug btsan and the ’Phang thang ma Catalogue (25 pages)
– Brandon Dotson
• An Early Seventeenth-Century Tibeto-Mongolian Ceremonial Staff (24 pages)
– Johan Elverskog
• The Importance of the Underworlds: Asuras’ Caves in Buddhism, and Some Other
Themes in Early Buddhist Tantras Reminiscent of the Later Padmasambhava
Legends (31 pages)
– Robert Mayer
• Re-Assessing the Supine Demoness: Royal Buddhist Geomancy in the Srong btsan
sgam po Mythology (47 pages)
– Martin A. Mills
• Modernity, Power, and the Reconstruction of Dance in Post-1950s Tibet (42 pages)
– Anna Morcom
Book Reviews
• Review of Thundering Falcon: An Inquiry into the History and Cult of Khra ’brug,
Tibet’s First Buddhist Temple, by Per K. Sørensen et al (5 pages)
– Bryan Cuevas
• Review of Tibetan Songs of Realization: Echoes from a Seventeenth-Century Scholar
and Siddha in Amdo, by Victoria Sujata (6 pages)
– Lauran Hartley
• Review of Holy Madness: Portraits of Tantric Siddhas, ed. Rob Linrothe and Review
of The Flying Mystics of Tibetan Buddhism, by Glenn H. Mullin (8 pages)
– Serinity Young
ii
News
Abstracts
Contributors to this Issue
iii
Articles
A Look at the Diversity of the Gzhan stong Tradition1
Anne Burchardi
University of Copenhagen
Abstract: This article introduces two studies by classical Tibetan Buddhist scholars
that explain the range of meanings of the term gzhan stong. The two texts – one
by Padma bi dza (twentieth century), the other by Tāranātha (1575-1634) – are
analytical studies that summarize and compare the various views of previous
scholars who wrote on gzhan stong. Such interpretive studies are valuable in that
they present us with different ways of interpreting the heterogeneous material
classified under the rubric “gzhan stong.” They also suggest ways of contextualizing
the different levels of discourse found within this material.
Introduction
Descriptions of gzhan stong are frequently encountered in the context of polemical
discourse, where it stands in contradistinction to rang stong.2 Some scholarly
attention has been paid to the historical context of the controversies involving
prominent gzhan stong masters and their writings. But so far the attention given
to the actual differences of interpretation of the term gzhan stong in its various
hermeneutical and philosophical contexts has been quite limited in non-Tibetan
publications – limited, that is, when we consider the extent of primary sources
available in Tibetan.3 Fortunately, this situation is improving year by year, and it
1
This article is a revised and enlarged version of the paper presented at the Tenth Seminar of the
International Association for Tibetan Studies (6th-12th September 2003), Oxford. Thanks to Gene
Smith, David Seyfort Ruegg, José Cabezón, and Karl Brunnhölzl for advice.
2
Gzhan stong may be translated as “empty of other” and rang stong as “empty of self(-nature).”
There is general agreement about rang stong referring to relative or conventional phenomena. So-called
gzhan stong proponents will generally – but not necessarily – consider gzhan stong to refer exclusively
to ultimate phenomena. See Wangchuk Dorji, “The rÑiṅ-ma Interpretations of the Tathāgatagarbha
Theory,” Wiener Zeitschrift für die Kunde Südasiens 48 (2004): 171-213, and especially 171 n. 3, for
an account of various usages of gzhan stong.
3
The following are some significant studies in English relevant to the study of gzhan stong: David
Seyfort Ruegg, Buddha Nature, Mind and the Problem of Gradualism in a Comparative Perspective:
On the Transmission and Reception of Buddhism in India and Tibet (London: School of Oriental and
African Studies, University of London, 1989); Susan Hookham, The Buddha Within (Albany: SUNY
Journal of the International Association of Tibetan Studies, no. 3 (December 2007): 1-24.
www.thdl.org?id=T3128.
1550-6363/2007/3/T3128.
© 2007 by Anne Burchardi, Tibetan and Himalayan Digital Library, and International Association of Tibetan Studies.
Distributed under the THDL Digital Text License.
Burchardi: The Diversity of the Gzhan stong Tradition 2
is my hope that the present reflections will further our general knowledge of the
gzhan stong tradition in some small way.
There is a vast amount of material waiting to be explored that can inform us
about the gzhan stong views held by various Tibetan masters. Here I would like
to introduce two analytical studies that show various ways of viewing and defining
gzhan stong.
Padma bi dza
The first study is The Full Moon Dialogue (dri lan tshes pa’i zla ba),4 a relatively
recent text from early twentieth-century east Tibet, written by a mkhan po from
Zur mang named Padma bi dza (also known as Padma rnam rgyal)5 and
commissioned by the eleventh Si tu, Padma dbang mchog rgyal po (1886-1952).
It is a reply to twenty-five questions submitted to the monastic college at Dpal
spungs.6 The following is an excerpt from this text.
Press, 1991); Matthew Kapstein, The ’Dzam thang Edition of the Collected Works of Kun mkhyen Dol
po pa Shes rab rGyal mtshan: Introduction and Catalogue (New Delhi: Shedrup Books, 1992); Cyrus
Stearns, The Buddha from Dolpo: A Study of the Life and Thought of the Tibetan Master Dolpopa
Sherab Gyaltshen (Albany: SUNY Press, 1999); Matthew Kapstein, “We Are All Gzhan stong pas:
Reflections on The Reflexive Nature of Awareness: A Tibetan Madhyamaka Defence, by Paul Williams,”
Journal of Buddhist Ethics 7 (2000): 105-25; E. Gene Smith, Among Tibetan Texts: History and
Literature of the Himalayan Plateau, ed. Kurtis R. Schaeffer (Boston: Wisdom Publications, 2001).
4
Place and date of printing are unknown. This text was kindly given to me by the former director of
The National Library of Bhutan, Lopon Pemala. It is studied at the Nyima Lung Monastic College in
Bhutan.
5
See http://www.tbrc.org P5784.
6
The introduction informs us that a paper roll containing twenty-five questions concerning the
difference between Madhyamaka Rang stong and Gzhan stong in terms of the base, path, and fruit of
the essential Mahāyāna sūtra and mantra perspectives was brought to the great Bka’ brgyud seat of
Dpal spungs in Mdo khams called Thub bstan dar rgyas chos ’khor gling, the seat of ’Jam mgon si tu
rin po che. The paper roll stated that these questions were for the scholars based at ’Jam mgon gzhan
phan snang ba’s school, and that the questions came from someone from Rgyal rong called Karma
nges don. Upon thorough investigation of the words and the meaning, it was found that the wording
was probably that of a “realized one” (rtogs ldan), but from the point of view of the meaning, the
questions were generally considered to be indicative of the fact that the author had a certain level of
analytical realization (de la skabs ’dir rgyal rong ba ka rma nges don yin zer ba zhig gis/ mdo khams
bka’ brgyud kyi gdan sa chen po shar dpal spungs thub bstan dar rgyas chos ’khor gling du ’khod pa’i
byams mgon si tu rin po che’i chos sde/ ’jam mgon gzhan phan snang ba’i slob grwa pa gzhi byes kyi
mkhas pa rnams la dri ba yin zer ba dbu ma rang stong dang gzhan stong gnyis kyi khyad par dri bya’i
snying por gyur pa’i theg chen mdo sngag phyogs kyi gzhi lam ’bras bu’i skor ci rigs pa nas brtsams
te dri ba nyer lnga tsam zhig mdzad ’dug pa’i shog dril lag tu son te tshig don la legs par brtags pas/
tshig gi dag sdeb ni rtogs ldan phal cher gyi lugs su ’dug la/ don gyi cha rnams spyir rnam dpyod kyi
drod tshad nyul ba’i dri ba re yin/; Padma bi dza [Zur mang mkhan po padma rnam rgyal], Dri lan
tshes pa’i zla ba [n.p., n.d.], 2.4-3.4).
The questions are very pointed and include the consequences of both affirmative and non-affirmative
answers. For example:
Are concepts dharmakāya or are they delusion? If they are the true nature, then this ordinary mind
contradicts the uncreated dharmakāya, the special quality of mahāmudrā. If they are delusion, then
it follows that saṁsāra is dharmakāya (rnam rtog chos sku gnas lugs ’khrul lugs gang / gnas lugs
yin na tha mal shes pa ’di/ ma bcos chos sku phyag chen khyad chos ’gal/ ’khrul lugs yin na ’khor
ba chos skur thal/; Padma bi dza, Dri lan tshes pa’i zla ba, 14.1-14.2).
Journal of the International Association of Tibetan Studies 3 (December 2007) 3
Translation
In brief, the crucial exchange concerning the difference between rang stong and
gzhan stong [is as follows]:
1. A general presentation of the different opinions.

2. A specific description of the Jo nang intention.
3. A presentation of the validity of the two kinds of Madhyamaka.
A General Presentation of the Different Opinions

There are about seven main divisions concerning the meaning of the terms (rang
stong and gzhan stong):
1. The omniscient Jo nang, (Dol po pa shes rab rgyal mtshan [1292-1361])

father and sons, consider consciousness to be rang stong and pristine
awareness (ye shes) to be gzhan stong.
2. The supreme shākya, Gser mdog paṇ chen (1428-1507), considers the
appearance of phenomena to be rang stong and the luminous, true nature
of phenomena (dharmatā) to be gzhan stong.
3. Sa bzang ma ti paṇ chen (1294-1376) considers subject and object to be
rang stong and space and pristine awareness to be gzhan stong.
4. Karma pa bdud ’dul rdo rje (1733/4-1797/8) considers cyclic existence
(saṁsāra) to be rang stong and transcendence of suffering (nirvāṇa) to
be gzhan stong. This is a wholesome position.
Another question reads:

At the time of the ground, is the potential primordially pure or not? If it is, then what is the basis of
the temporary delusions? Similarly, the manner of delusion must be said to be a manner of purity.
If it is not primordially pure, then what are the natural qualities and the temporary impurities?” (gzhi
dus rigs pa ye dag yin nam min/ yin na glo bur ’khrul gzhi gang la bya/ ji ltar ’khrul tshul dang ni
dag tshul smros/ ye dag min na yon tan rang bzhin dang / dri ma glo bur ba de ci la ser/; Padma bi
dza, Dri lan tshes pa’i zla ba, 47.4-47.5).
The colophon states:
My master Byams mgon tā’i si tu rin po che held this paper roll of questions in his hand, and with
the command “give some satisfying answers to this,” he hit it on my head as a drum stick, like a
text worthy of respect. So nurtured by his kindness and the kindness of ’Jam mgon gzhan phan
snang ba and others, [I,] Padma bi dza from Zur mang, placed in the ranks of Dpal spungs thub
bstan dar rgyas chos ’khor gling, wrote this, which is what came to mind, when staying at the home
of Lto ru tsang, minister of the king of Sde dge, on my way to Rdzogs chen to have books printed
(rang re’i rigs kyi khyab bdag byams mgon tā’i si tu rin po ches ’di’i dri ba’i shog dril lag tu gnang
nas rnga la dbyug gus bskul ba’i dpe ltar lags pas/ ’di la lan yid tshim pa zhig thob cig ces bka’
stsal spyi bor phebs par brten nas/ rje de nyid dang ’jam mgon gzhan phan snang ba sogs kyi bka’
drin gyis rjes su ’tsho zhing / shar dpal spungs thub bstan dar rgyas chos ’khor gling gi gral mthar
’khod zur mang ba padma bi dzas mdo khams rdzogs chen phyogs dpe cha spar du ’gro ba'i lam
zhor/ sa skyong sde dge’i mdun mdzod lto ru tshang sar ’dug skabs blo thog nas shar byung du bris
ba dag par bshus nas bskur ba dge legs ’phel/; Padma bi dza, Dri lan tshes pa’i zla ba, 75.2-75.5).
The text consists of the answers to these questions.
5. The omniscient Mi bskyod (1507-54), father and sons, consider the pure
kāyas and pristine awareness to be rang stong in terms of their true nature
and to be gzhan stong in terms of the way they appear.
6. ’Jam mgon si tu paṇ chen (1700-74) considers the aspect that refutes to
be rang stong and the aspect that establishes to be gzhan stong. This is a
wholesome position.
7. Furthermore, Kaḥ thog dge rtse paṇ chen (1761-1829) considers properly
the context of mastery in equipoise to be rang stong and the context of
differentiation in post-meditation (subsequent attainment) to be gzhan
stong.
These seven positions have been arranged here in a summary in order to broaden
the intelligence of those with lucid minds.7
If you summarize the seven, they can be condensed into three: [1] the main Jo
nang pa [assertion that] pristine awareness is gzhan stong; [2] Shākya mchog
[ldan’s assertion that] the sphere [of reality] is gzhan stong; and [the assertions of]
the others [3-7] that both the sphere [of reality] and pristine [awareness] (dbyings
ye gnyis ka) are gzhan stong.8
They can also be condensed into two: the first five [1-5] are mainly presentations
of rang stong and gzhan stong as subjects to be determined,9 while the latter two
[6-7] are mainly presentations of rang stong and gzhan stong as methods of
ascertainment.10
So we can say that [1] consciousness [as rang stong] and pristine awareness [as
gzhan stong]; [2] phenomena [as rang stong] and dharmatā [as gzhan stong]; [3]
saṁsāra [as rang stong] and nirvāṇa [as gzhan stong]; [4] subject-object [as rang
stong] and space-pristine awareness [as gzhan stong];11 [5] true nature [as rang
7
mdor bsdus gnad kyi dris lan dbu ma rang stong dang gzhan stong gi khyad par la/ ’dod tshul gyi
dbye ba spyir bstan pa/ jo nang pa’i dgongs pa bye brag tu brjod pa/ dbu ma gnyis kyi ’thad tshul bstan
pa dang gsum las/ dang po la ming don ci rigs kyi dbye sgo gtso bor bdun tsam du ’dug ste/ kun mkhyen
jo nang yab sras ni/ rnam shes rang stong / ye shes gzhan stong du ’dod pa dang / shāka [read shākya]
mchog pa gser mdog paṇ chen ni/ chos can snang ba rang stong / chos nyid ’od gsal gzhan stong du
’dod pa dang / sa bzang ma ti paṇ chen ni/ yul dang yul can rang stong // dbyings dang ye shes gzhan
stong du ’dod pa dang / karma pa bdud ’dul rdo rje ni/ ’khor ba rang stong // myang ’das gzhan stong
tsam du ’dod na legs phyogs dang / kun mkhyen mi skyod yab sras ni/ dag pa’i sku dang ye shes kyang
gnas tshul rang stong / snang tshul gzhan stong du ’dod pa dang / ’jam mgon si tu paṇ chen ni/ dgag
phyogs rang stong / sgrub phyogs gzhan stong du ’dod na legs phyogs dang / kaḥ thog dge rtse paṇ
chen ni/ mnyam gzhag la zlo’i skabs rang stong / rjes thob shan ’byed skabs gzhan stong du ’dod pa
legs zhes pa ste/ ’dod tshul bdun po ’di dag ni blo gsal rnams kyi blo gros zor yangs pa’i ched du bsdus
te bkod pa yin no/ (Padma bi dza, Dri lan tshes pa’i zla ba, 60.3-61.4).
8
bdun pa de’ang bsdu na/ jo nang pa gtso bor ye shes gzhan stong / shāka [read shākya] mchog pa
dbyings gzhan stong / gzhan rnams dbyings ye gnyis ka’i gzhan stong ste gsum du ’du’o/ (Padma bi
dza, Dri lan tshes pa’i zla ba, 61.4-61.5).
9
gtan la dbab bya: that which is to be resolved or ascertained.
10
de’ang snga ma lnga gtan la dbab bya’i sgo nas dang / phyi ma gnyis ’bebs byed kyi sgo nas gtso
bor rang stong dang gzhan stong du bzhag pas gnyis su ’du’o/ (Padma bi dza, Dri lan tshes pa’i zla
ba, 61.5-61.6).
11
Padma bi dza has switched the order of view 3 and 4 here.
stong] and apparent nature [as gzhan stong]; [6] refutation [as rang stong] and
establishment [as gzhan stong]; [7] equipoise [as rang stong] and post[-meditation
as gzhan stong] are the seven [1-7], the three [1, 2, & 3-7], or two different types
[i.e., 1-5 as subjects to be determined and 6-7 as methods of determination] of rang
stong and gzhan stong.12
A Specific Description of the Jo nang Intention

Among the four general tenets, in Tibet there are several tenets of the Madhyamaka
(Central System), divided into the two of rang stong and gzhan stong. The first
was given the name rang stong, referring to the empty aspect mainly taught and
emphasized in the context of the intermediate turning.13
As for the second, in the context of covering the final [turning] with the intention
of Mantra[-yāna] (Path of Secret Syllables), it was appropriate to comment even
on the intention of most sūtras as Mantra[-yāna]. The ultimate meaning of the
sūtras of the final turning of the wheel was joined with the ultimate meaning of
the Highest Yoga Tantras (Anuttara-yoga-tantra). The illusory impurities to be
removed were taught to be rang stong. Although the qualities of full maturation
were not asserted to be present at the time of the ground, the basis of purification,
the aspect to be freed, space and pristine awareness, being naturally permanent,
stable, peaceful, and indestructible were ascertained to be gzhan stong. This is the
intention of the omniscient Jo nang, father and sons, and in Tibet this was given
the famous name of Gzhan stong Madhyamaka.14
A Presentation of the Validity of the Two Kinds of Madhyamaka
1. The differences between the two types of Madhyamaka.

2. How these two ultimately do not contradict each other.
The Differences Between the Two Types of Madhyamaka

This differentiation between Rang stong and Gzhan stong Madhyamaka should be
known in an unmistaken way to [correspond to the difference between] the meaning
12
smras pa rnam shes ye shes chos can chos nyid dang / ’khor ’das gzung ’dzin dbyings ye gnas
snang tshul/ dgag sgrub mnyam rjes sgo nas rang stong dang / gzhan stong dbu ma’i khyad par bdun
gsum gnyis/ shes par skabs kyi tshigs su bcad pa’o/ (Padma bi dza, Dri lan tshes pa’i zla ba, 61.6-62.1).
13
gnyis pa jo nang pa’i dgongs pa bye brag tu brjod pa ni/ spyir grub mtha’ bzhi las/ bod kyi phyogs
su dbu ma’i grub mtha’ ’dod tsul [read tshul] ’ga’ zhig la rang stong dang gzhan stong gnyis su dbye
rigs pa’i dang po ni/ bka’ ’khor lo bar pa gtso bor gyur pa’i stong phyogs gtso bor ston skabs dbu ma
rang stong du ming chag pa dang / (Padma bi dza, Dri lan tshes pa’i zla ba, 62.1-62.3).
14
gnyis pa ni gong ma sngags kyi dgongs pas gung ’gebs skabs/ mdo phal cher gyi dgongs pa yang
sngags ltar ’grel rung bas/ ’khor lo tha ma’i mdo don mthar thug dang / bla med sngags kyi rgyud don
mthar thug gnyis lto sbyar te sbyang bya ’khrul pa’i dri ma rang stong du bstan nas/ rnam par smin
pa’i yon tan gzhi dus su mi ’dod par gsungs kyang bral ba’i phyogs kyi sbyang gzhi dbyings dang / ye
shes rtag brtan zhi ba g.yung drung gi rang bzhin gzhan stong yin ces gtan la phab pas/ bod du kun
mkhyen jo nang yab sras kyi dgongs pa dbu ma gzhan stong zhes yongs su grags pa’i ming chags pa
yin no/ (Padma bi dza, Dri lan tshes pa’i zla ba, 62.3-62.6).
and intention of Sūtra[-yāna] (Path of the Buddha’s Words) and Mantra[-yāna,

respectively] through a presentation of
1. Difference in terms of subject matter.

2. Difference in ascertaining.
3. Difference in terms of manner of ascertainment.15
Difference in Terms of Subject Matter

It is appropriate to describe the difference between the two kinds of Madhyamaka
in terms of subject matter as follows: “Faults are rang stong and qualities are gzhan
stong.”16
The empty aspect emphasized in the context of the sūtras of the second turning
of the wheel and the texts commenting on their intention constitute that which is
to be purified, the faults and defects which are temporary and illusory. These are
considered rang stong. That which is emphasized in the treatises of the final turning
and the Mantra[-yāna], tantras, and so forth constitutes the basis of purification,
the qualities which are space and pristine awareness. These are considered gzhan
stong.17
Difference in Ascertaining
Concerning the differences between the two types of Madhyamaka [in terms of
method], it is appropriate to say: “The logic that refutes is rang stong and the logic
that affirms is gzhan stong.”18
As Nāgārjuna (second century) says: “No object whatsoever has ever come into
existence either from itself, from something else, from something other than these
two or without a cause.” This manner of teaching, through the logic of refutation
emphasized in the intermediate turning of the wheel and so forth is the rang stong
method of teaching.
As the regent Maitreya says: “Sentient beings possess tathāgatagarbha, since
perfect buddhakāya radiates, since suchness cannot be differentiated, and since
15
gsum pa dbu ma gnyis kyi ’thad tshul bstan pa la/ dbu ma gnyis kyi khyad par mdzad tshul dang
/ de gnyis mthar thug ’gal med tshul gnyis las/ dang po ni/ de ltar dbu ma rang stong dang gzhan stong
gi khyad par dbab bya/ ’bebs byed/ ’bebs tshul gsum gyi sgo nas mdo sngags kyi dgongs don phyin ci
ma log pa ltar shes dgos pa las (Padma bi dza, Dri lan tshes pa’i zla ba, 62.6-63.2).
16
dang po gang zhig gtan la dbab bya skyon rang stong yon tan gzhan stong gi sgo nas dbu ma gnyis
kyi khyad par bstan ces brjod kyang rung ste/ (Padma bi dza, Dri lan tshes pa’i zla ba, 63.2-63.3).
17
gtso cher ’khor lo bar pa’i mdo dang dgongs ’grel sogs las stong cha bstan pa’i skabs sbyangs
bya glo bur ’khrul pa’i nyes skyon rang stong dang / gtso cher ’khor lo tha ma’i gzhung dang sngags
rgyud sogs las sbyangs gzhi dbyings dang ye shes kyi yon tan gzhan stong du bstan par ’dod pa ltar
ro/ (Padma bi dza, Dri lan tshes pa’i zla ba, 63.5-64.1).
18
gnyis pa gang gis gtan la ’bebs byed dgag phyogs kyi rigs pas rang stong dang / sgrub phyogs kyi
rigs pas gzhan stong gi sgo nas dbu ma gnyis kyi khyad par yang rung ste/ (Padma bi dza, Dri lan tshes
pa’i zla ba, 64.1-64.2).
they have the potential.” It is appropriate to say that teaching through the logic of
affirmation in the final turning of the wheel is the gzhan stong method.19
Difference in Terms of Manner of Ascertainment

It is appropriate to say that the manner of teaching the sphere of reality
(dharmadhātu) as a generality/universal is rang stong, and the manner of teaching
it in terms of its actuality/particularity is gzhan stong, just as master Asaṅga (ca.
310-90) differentiated between dharmadhātu as a generality and as an actuality.20
It is also appropriate to say that the manner of ascertaining the conceptual image
of dharmadhātu by way of [the teachings on] the sixteen kinds of emptiness and
so forth in the intermediate turning of the wheel is rang stong, and the manner of
determining dharmadhātu in terms of its actuality, sugatagarbha (heart of bliss),
by way of [the teachings on] the luminosity of mind itself and so forth in the final
turning of the wheel and Mantra[-yāna] is gzhan stong.21
So, it is appropriate to understand the postulations of the two types of
Madhyamaka through proper investigation of the sources of the teachings.
Furthermore, as explained just above, one should gain certainty in three ways.
To summarize in verse form:
Faults and qualities, refutation and affirmation, conceptual and actual dhātu
[Are] the differences between rang stong and gzhan stong
In terms of what they are, how they [are ascertained], and in what manner.22
How These Two Ultimately Do Not Contradict Each Other

Through a difference in the tradition of expression on these points, the names of
the two types of Madhyamaka came about. Therefore, the aspect that expresses
space (dbyings) and illusory nature (sgyu ma’i rang bzhin) of the intermediate
turning of the wheel was given the name Rang stong Madhyamaka. The aspect
19
mgon po klus/ bdag las ma yin gzhan las min/ gnyis las ma yin rgyu med min/ dngos po gang dag
gang na yang / skye ba nam yang yod ma yin/ ces sogs gsung pa ltar gtso bor ’khor lo bar pa la sogs
pa’i dgag phyogs kyi rigs pas rang stong du bstan tshul dang / rgyal tshab byams pas/ rdzogs sangs
sku ni ’pho phyir dang / de bzhin nyid dbyer med phyir dang / rigs yod phyir na lus can kun/ rtag tu
sangs rgyas snying po can/ zhes sogs gsungs pa ltar ’khor lo tha ma’i sgrub phyogs kyi rigs pas gzhan
stong du bstan na yang rung ba ltar ro/ (Padma bi dza, Dri lan tshes pa’i zla ba, 64.2-64.5).
20
gsum pa ji ltar gtan la ’bebs tshul chos dbyings spyi mtshan bstan pa’i tshul gyis rang stong dang
/ rang mtshan bstan pa’i tshul gyis gzhan stong du bstan kyang rung ste/ slob dpon thogs med kyis chos
dbyings la spyi mtshan dang rang mtshan gnyis su phyes pa ltar ro/ (Padma bi dza, Dri lan tshes pa’i
zla ba, 64.5-64.6).
21
’khor lo bar pas stong nyid bcu drug sogs kyi sgo nas chos dbyings spyi mtshan gtan la ’bebs pa’i
tshul gyis rang stong dang / ’khor lo tha ma sngags dang bcas pa las sems nyid ’od gsal sogs kyi sgo
nas chos dbyings rang mtshan bde gshegs snying po gtan la ’bebs pa’i tshul gyis gzhan stong du bstan
na yang rung ba’i phyir/ (Padma bi dza, Dri lan tshes pa’i zla ba, 64.6-65.2).
22
gsungs pa’i khungs rnams la legs par brtags pas/ dbu ma gnyis kyi ’jog mtshams go yang rung
zhing / de’ang bshad ma thag pa ltar tshul gsum gyi sgo nas nges par bya’o/ ’dir smras pa’i sdom/
skyon dang yon tan dgag dang sgrub/ phyi dang rang gi mtshan nyid dbyings/ gang zhig gang gis tshul
ji ltar/ rang gzhan stong pa’i kyad par gsum/ zhes par skabs kyi tshigs su bcad pa’o/ (Padma bi dza,
Dri lan tshes pa’i zla ba, 66.2-66.4).
that expresses space and pristine awareness of the final turning of the wheel as the
base of Mantra[-yāna] was given the name Gzhan stong Madhyamaka.23
In fact, even the rang stong pas must assert that, while dharmatā is free of all
elaborations of eternalism and nihilism, yet it appears as the ground, the nature of
the kāyas (bodies), and pristine awareness. Even the gzhan stong pas must accept
that the appearing yet empty space is free of all elaborations. Therefore, it would
be good to establish the intended meaning of the two Madhyamakas as one and
the same: a great union ultimately transcending all the elaborations and terms of
eternalism and nihilism.24
Analysis
It is well known that the views held by proponents of the gzhan stong tradition
have varied considerably from the time of Dol po pa shes rab rgyal mtshan up to
the present day. Padma bi dza identifies seven different positions held by seven
different masters spanning a period of more than six centuries. He defines a number
of different categories for his comparison. In the first context, rang stong and gzhan
stong are shown to refer to phenomena belonging to two different levels of reality.
In the second context, rang stong and gzhan stong are presented as different
strategies of ascertaining a given subject, and in the third context he shows rang
stong and gzhan stong as different methods of gaining realization.
The first five masters cited – Dol po pa shes rab rgyal mtshan, Shākya mchog
ldan, Sa bzang ma ti paṇ chen, Karma pa bdud ’dul rdo rje, and Karma pa mi
bskyod rdo rje – are presented as viewing rang stong and gzhan stong as differing
in terms of subjects to be determined. The first four of these masters hold that
various relative phenomena are rang stong while ultimate phenomena are gzhan
stong.25 However, although the fifth master cited, the Eighth Karma pa, is portrayed
as going against this pattern, he is nevertheless included in the list perhaps on
account of those of his writings that advocate gzhan stong.26 The sixth master cited,
23
gnyis pa dbu ma gnyis po mthar thug ’gal ba med tshul ni/ de ltar khyad par gsum po tsam zhig
rtsal du ’don lugs kyis dbu ma gnyis kyi mtshan du chags pa des na ’khor lo bar pa’i dbyings dang
sgyu ma’i rang bzhin rtsal du bton pa’i cha nas dbu ma rang stong du ming chags pa dang / ’khor lo
tha ma sngags dang bcas pa’i gzhi dbyings dang ye shes kyi rang bzhin rtsal du bton pa’i cha nas dbu
ma gzhan stong gi ming du chags pa tsam gyi khad par las/ (Padma bi dza, Dri lan tshes pa’i zla ba,
66.5-67.1).
24
don la dbu ma rang stong pas kyang chos nyid rtag chad kyi spros pa thams cad dang bral bzhin
du snang ba’i gzhi sku dang ye shes kyi rang bzhin ’dod dgos pa dang / gzhan stong pas kyang snang
bzhin du stong pa’i dbyings spros pa thams cad dang bral ba ji bzhin du khas len dgos pa’i phyir na/
dbu ma gnyis po mthar thug spros mtshan rtag chad thams cad las ’das pa’i zung ’jug chen por dgongs
don gcig tu grub na legs te/ (Padma bi dza, Dri lan tshes pa’i zla ba, 67.1-67.3).
25
This conforms to Dol po pa’s general definition of relative phenomena being rang stong (chos can
rang stong) and their ultimate nature or dharmatā being gzhan stong (chos nyid gzhan stong).
26
Mi bskyod rdo rje commented upon the Abhisamayālaṃkāra in accordance with the gzhan stong
philosophy in his Shes rab kyi pha rol tu phyin pa’i lung chos mtha’ dag gi bdud rtsi’i snying por gyur
pa gang la ldan pa’i gzhi rje btsun mchog tu dgyes par ngal gso’i yongs ’dus brtol gyi ljon pa rgyas
pa (Sikkim: Karma Shri Nalanda Institute, n.d.). See Karl Brunnhölzl, A Commentary on the Perfection
of Knowledge: The Noble One Resting at Ease (Sackville: Nitartha Institute, 2001) for a partial translation
Si tu paṇ chen (1700-74), is presented as regarding rang stong and gzhan stong as
different methods of ascertaining a given subject, while the seventh master, Kaḥ
thog dge rtse paṇ chen,27 is portrayed as holding rang stong and gzhan stong as
different ways of gaining realization.
The seven positions are also summarized into three positions with Dol po pa
representing the first position that emphasizes the aspect of pristine awareness (ye
shes) as gzhan stong, Shākya mchog ldan representing the second position that
emphasizes the aspect of the sphere of reality (dbyings) as gzhan stong, and the
others representing a third group of positions that asserts a combination of space
and awareness (dbyings ye gnyis ka) as being gzhan stong.
Padma bi dza concludes on a conciliatory note, showing how rang stong and
gzhan stong do not contradict each other. They can thus be considered to belong
to different levels of discourse, and therefore lack the basic criteria for meaningful
comparison. Ruegg has aptly observed: “It may be possible to think in terms of
complementarity (or incommensurability) between two theories belonging to
distinct universes of religious-philosophical discourse rather than in terms of
contradiction between theories competing on the same level.”28
As we have seen, gzhan stong is used to refer to a number of different but related
fields of inquiry. Let us explore this issue a bit further.29 Certain Tibetan Buddhist
masters use the term gzhan stong to refer to a philosophical tenet system (grub
mtha’, siddhānta); others use it to refer to a philosophical point of view theory (lta
ba). At other times it is used to refer to a combination of theory and practice (lta
sgom) or to a practice tradition (sgom lugs). Finally, many – such as Padma bi dza
– argue that rang stong and gzhan stong represent Sūtrayāna and Mantrayāna,
respectively.
of this text. Mi bskyod rdo rje also endorsed gzhan stong in his Dbu ma gzhan stong smra ba’i srol
legs par phye ba’i sgron me, published in Dbu ma gzhan stong skor bstan bcos phyogs bsdus deb dang
po (Sikkim: Karma Shri Nalanda Institute, 1990), 12-48. However, his commentary on
Madhyamakāvatāra, the Dbu ma la ’jug pa’i rnam bshad dpal ldan dus gsum khyen pa’i zhal lung
dwags brgyud grub pa’i shing rta (Seattle: Nitartha International Publications, 1996) is in accordance
with the rang stong theory, while at the same time taking issue with the emptiness propounded by the
Dge lugs pa school. For studies of this last text see Paul Williams, “A Note on Some Aspects of Mi
Bskyod Rdo Rje’s Critique of Dge Lugs Pa Madhyamaka,” Journal of Indian Philosophy 11, no. 2
(1983): 125-146, and David Seyfort Ruegg, “A Kar ma bka’ brgyud Work on the Lineages and Traditions
of the Indo-Tibetan dBu ma (Madhyamaka),” Orientalia Iosephi Tucci Memoriae Dicata, vol. 3 (Roma:
Istituto Italiano per il Medio ed Estremo Oriente, 1988), 1249-80. See also Karl Brunnhölzl, The Center
of the Sunlit Sky: Madhyamaka in the Kagyü Tradition (Ithaca: Snow Lion, 2004), 445-526, where Mi
bskyod rdo rje’s views on gzhan stong are elaborated in the fourth chapter, entitled “Is There Such a
Thing as Shentong-Madhyamaka?”
27
His association of rang stong with equipoise and gzhan stong with post-meditation differs from
the views of, for example, Shākya mchog ldan and from that of many gzhan stong proponents’ positions,
but this is a subject for future study.
28
Ruegg, Buddha Nature, 7-8, 11; and also David Seyfort Ruegg, Three Studies in the History of
Indian and Tibetan Madhyamaka Philosophy, Studies in Indian and Tibetan Madhyamaka Thought,
part 1 (Vienna: Arbeitskreis für tibetische und buddhistische Studien, Universität Wien, 2000), 81.
29
It is important to remember that although gzhan stong, in the course of centuries of discussion
across the categories mentioned above, has become a somewhat “loaded” term, there are no grounds
for the ascription of a substantive ontology to the term itself.
Broido has pointed out that Dol po pa never intended his gzhan stong to fulfill
the requirements of a philosophical tenet system.30 Instead, Dol po pa “calls his
darśana ‘Dbu ma chen po,’ and intends it to be connected with experience and to
be contrasted with dbu ma as a siddhānta.”31 However, later holders of the gzhan
stong lineage seem to have given gzhan stong the status of a philosophical tenet
system.32 It was on this level of discourse that polemical comparisons between
rang stong and gzhan stong began to proliferate. Both Dol po pa and later gzhan
stong proponents assert that gzhan stong incorporates and bases itself on rang
stong as a philosophical tenet system. They then go on to say that gzhan stong
supersedes this level of discourse. Nevertheless, it seems that some later gzhan
stong proponents argue for the supremacy of gzhan stong even on the philosophical
tenet system level. Although this may be nothing but a strategy for defending the
tradition against criticism from its opponents, it can become, as Broido points out,
“a source of serious confusion,”33 at least for those attempting to study the tradition.
Tāranātha
Tāranātha (1575-1634) is an example of a gzhan stong proponent who used the
term gzhan stong in a way that encompasses both the sense of philosophical tenet
system (grub mtha’) and of practice tradition (sgom lugs).34 He is said to have had
30
Michael Broido, “The Jo-nang-Pas on Madhyamaka: A Sketch,” The Tibet Journal 45, no. 1 (1989):
86-90.
31
Broido, “The Jo-nang-Pas,” 87. He further says: “In S’s [Dol po pa shes rab rgyal mtshan’s] usage
a siddhānta (grub-mtha’) is a fixed philosophical position based on axioms and set rules of argument;
a darśana (lta-ba) is a point of view in a broad sense, including what derives directly from experience.”
32
This is defined by the Tibetan tradition in general as “Limit of Establishment,” i.e., the final
conclusion reached by a system based on logic and scripture, referring to tenet systems.
33
Broido, “The Jo-nang-Pas,” 89.
34
Tāranātha was a prolific writer on many subjects. In a short text called The Central Causal and
Resultant Drop (Rgyu ’bras dbu ma’i thig le), Tāranātha describes tathāgatagarbha in its sūtra and
tantra context; see Tāranātha, The Collected Works of Jo-naṅ Rje-btsun Tāranātha, vol. 13 (Leh,
Ladakh: C. Namgyal & Tsewang Taru, 1982-87), 463-68. In The Heart of Zhentong (Gzhan stong
snying po; Tāranātha, Collected Works of Jo-naṅ Rje-btsun Tāranātha, vol. 4, 491-514), he outlines
the position of gzhan stong in relation to various topics. For a study of this work see Klaus-Dieter
Mathes, “Tāranātha’s Presentation of trisvabhāva in the gŹan stoṅ sñiṅ po,” Journal of the International
Association of Buddhist Studies 23, no. 2 (2000): 195-223. Finally, The Ornament of Gzhan stong
Madhyamaka (Gzhan stong dbu ma rgyan; Tāranātha, Collected Works of Jo-naṅ Rje-btsun Tāranātha,
vol. 4, 797-824) is, according to Cyrus Stearns in The Buddha from Dol po: A Study of the Life and
Thought of the Tibetan Master Dolpopa Sherab Gyaltsen (Albany: SUNY, 1999), 69, one of Tāranātha’s
most important works, devoted solely to the explication of the gzhan stong view. Lastly, we should
mention materials that have only recently come to light, namely texts in the ’Dzam thang edition of
Tāranātha’s Collected Works: Tāranātha, Collected Works, ’Dzam thang Edition, TBRC W22276. Here
we find his extremely interesting interpretations of The Heart Sūtra from a gzhan stong point of view;
see Matthew Kapstein, Reason’s Traces (Boston: Wisdom Publications, 2001), 304-6. The texts are
the Shes rab kyi pha rol tu phyin pa’i snying po’i don rnam par bshad pa sngon med legs bshad
(Tāranātha, Collected Works, ’Dzam thang Edition, vol. 17, 571-759) and the Sher snying gi tshig ’brel
(Tāranātha, Collected Works, ’Dzam thang Edition, vol. 17, 759-83). Here Tāranātha identifies three
different passages (762, 765, and 771) that he maintains clearly show that The Heart Sūtra teaches
gzhan stong. While gzhan stong presentations of tathāgatagarbha are quite common, it is rare to see
gzhan stong elucidations of the second turning of the wheel. Although, for Yogic Conduct System
(Yogācāra) precursors to these, see Donald S. Lopez, Jr., The Heart Sūtra Explained: Indian and Tibetan
visions of Dol po pa during the composition of several of his texts. His writings
served in turn as a significant inspiration for ’Jam mgon kong sprul blo gros mtha’
Yas’s writings on gzhan stong35 and other subjects. ’Jam mgon kong sprul is an
example of a later gzhan stong proponent who presents gzhan stong as a separate
philosophical tenet system.36
We will now use Tāranātha’s text, The Twenty-One Profound Points (zab don
nyer cig pa),37 as a lens into the positions of the first two masters cited in Padma
bi dza’s list, Dol po pa shes rab rgyal mtshan and Shākya mchog ldan. Their
positions are compared in detail by Tāranātha with respect to twenty-one issues.
The fact that Tāranātha introduces the discussion using the term gzhan stong dbu
ma’i lta sgom (theory and practice of the Empty of Other Central System) indicates
that his text seeks to encompass the discourse of two fields of inquiry: philosophical
point of view theory (lta ba) and meditation (sgom). He says:
I think that it is impossible for any contradiction to exist in the intention behind
the explanation of those who see the profound [reality]. However, seeing what is
needed for training various people, differences were [explained] conventionally.
Here, the Lord of Dharma, the great omniscient Brton pa bzhi ldan [Dol po pa],
and the great paṇ ḍita, the victorious Shākya mchog ldan, agree on the essential
point of the view and meditation of Gzhan stong Madhyamaka. However, when
Commentaries (Albany: SUNY, 1988), 60, where the statement “form is emptiness, emptiness is form,
emptiness is not other than form; form is not other than emptiness” is interpreted as an affirming
negative.
35
Such as ’Jam mgon kong sprul blo gros mtha’ yas, Gzhan stong dbu ma chen po’i lta khrid rdo rje
zla ba dri ma med pa’i ’od zer, in Rgya chen bka’ mdzod (Paro: Ngodrug, 1975/6), vol. nya (8), 581-621;
Nges don dbu ma chen po la ’khrul rtog nyer gsum gyi ’bur ’joms pa legs bshad gser gyi tho ba, in
Rgya chen bka’ mdzod, vol. ta (9), 211-18; Lta ba gtan la bebs pa las phro ba’i gtam skabs lnga pa
lung dang rig pa’i me tog rab tu dgod pa, in Rgya chen bka’ mdzod, vol. ta (9), 29-68; and Ris med
chos kyi ’byung gnas mdo tsam smos pa blo gsal mgrin pa’i mdzes rgyan, in Rgya chen bka’ mdzod,
vol. ta (9), 69-100.
36
See his Shes bya kun la khyab pa’i gzhung lugs nyung ngu’i tshig gis rnam par ’grol ba legs bshad
yongs ’du shes bya mtha’ yas pa’i rgya mtsho (Beijing: Mi rigs dpe skrun khang, 1985), 35.1-41.1:
glegs bam gsum pa/ skabs bdun pa/ lhag pa shes rab kyi bslab pa rim par gye ba/ ’khor lo gsum gyi
drang nges dang bden gnyis rten ’brel rnam par nges pa’i skabs/ lnga pa/ grub mtha’ so so’i ’dod
tshul. Translated into English by Anne Burchardi and Ari Goldfield, Gaining Certainty about the
Provisional and Definitive Meanings in the Three Turnings of the Wheel of Dharma, the Two Truths,
and Dependent Arising: The Root Text and Commentary Section Two of Chapter Seven from The
Treasury of Knowledge by Jamgön Kongtrül Lodrö Thaye (Kathmandu: Marpa Institute, 1997). Here
he presents the Vaibhāṣika, the Sautrāntika, the Cittamātra, the Svātantrika madhyamaka, the Prāsaṅgika
madhyamaka, and the Gzhan stong Madhyamaka view of the two truths as an ascending scale. The
presentation of gzhan stong as a form of Madhyamaka contrasts with, for example, Mi bskyod rdo rje’s
position that gzhan stong is more correctly a form of Yogācāra. See Brunnhölzl, Center of the Sunlit
Sky, 445.
37
Tāranātha, Collected Works of Jo-naṅ Rje-btsun Tāranātha, vol. 4, 781-95. For an excellent study
and translation of this text see Klaus-Dieter Mathes, “Tāranātha’s ‘Twenty-one Differences with Regard
to the Profound Meaning’ – Comparing the Views of the Two gŹan stoṅ Masters Dol po pa and Shakya
mchog ldan,” Journal of the International Association of Buddhist Studies 27, no. 2 (2004): 285-328.
ascertaining their view provisionally there are many minor differences between
their philosophical tenet systems.38
Tāranātha presents Shākya mchog ldan’s position first, referring to him as

“former” (snga ma), and Dol po pa’s position last, referring to him as “latter” (phyi
ma).39 They are introduced in this order in the text as a way of expressing a
hierarchical order of profundity according to the author. My impression is that
Tāranātha wishes to restate Dol po pa’s original gzhan stong by showing how it
differs from Shākya mchog ldan’s version of gzhan stong. But does Tāranātha’s
admiration for and identification with Dol po pa allow for an accurate portrayal
of Dol po pa’s position? Tāranātha seems in many cases to be speaking on Dol po
pa’s behalf, and so we can to a certain extent read what is portrayed as Dol po pa’s
position as being actually Tāranātha’s own position.
Another issue is whether Tāranātha uses a given point in time as definitive for
portraying Shākya mchog ldan’s position in regard to the twenty-one points,40 since
Shākya mchog ldan’s position vis-à-vis the rang stong/gzhan stong problematic
is known to have changed during his lifetime. He has been described by Thu’u
bkwan blo bzang chos kyi nyi ma (1737-1802) as starting out as a (rang stong)
mādhyamika, then becoming a Cittamātra (Mind Only System) proponent, and
then ending up as a Jo nang pa (gzhan stong mādhyamika). According to Dreyfus,41
the second period, which started when Shākya mchog ldan was around the age of
forty-nine, is more accurately described as Central Yogic Conduct System
(Yogācāra-madhyamaka), and in particular as False Aspectarian Yogic Conduct
System, which surpasses the Ṃind Only ṣystem.42 During the last phase, starting
when he was around fifty-six, he claimed the Jo nang pa position to be “a more
appropriate way to account for the experiential side of the realization of
emptiness.”43
Among the points covered in Tāranātha’s text are the issues of whether the
second turning is of literal or non-literal meaning and whether gzhan stong might
38
om swa sti/ zab mo gzigs pa rnams la ni/ bzhed dgongs ’gal ba mi srid snyam/ on kyang gdul bya
tha dad dang / dgos pa’i dbang gzigs tha dad bsnyad/ ’dir chos rje kun mkhyen chen po brton pa bzhi
ldan dang / paṇḍi ta chen po rgyal ba shākya mchog ldan gnyis/ gzhan stong dbu ma’i lta sgom gyi
gnad gcig kyang / gnas skabs lta ba de gtan la ’bebs pa’i skabs/ grub mtha’ mi ’dra ba than thun mang
dag yod pa rnams ’dir ngos bzungs bar bya ste (Tāranātha, Zab don nyer cig pa, 782.1-782.3).
39
Even though Shākya mchog ldan was born sixty-seven years after Dol po pa’s death.
40
See Mathes, “Tāranātha’s ‘Twenty-one Differences,’” 315 for an analysis of this.
41
Georges B. Dreyfus, Recognizing Reality: Dharmakirti’s Philosophy and Its Tibetan Interpretations
(Delhi: Sri Satguru Publications, 1997), 28-29.
42
See also Iaroslav Komarovski, Three Texts on Madhyamaka by Shakya Chokden (Dharamsala:
Library of Tibetan Works & Archives, 2000), 59 n. 27.
43
Dreyfus, Recognizing Reality, 29. Shākya mchog ldan’s brilliant expositions relevant to gzhan
stong research include, among others: Byams chos lnga’i nges don rab tu gsal ba zhes bya ba’i bstan
bcos, in Collected Works, vol. 11 (Thimphu: Kunzang Tobgye, 1975), 1-38; Byams chos lnga’i lam
gyi rim pa gsal bar byed pa’i bstan bcos rin chen sgron gyi sgo ’byed, in Collected Works, vol. 11,
39-156; and Rgyud bla’i rnam bzhad sngon med nyi ma, in Collected Works, vol. 13, 113-58. For a
translation of the latter see Anne Burchardi, Shakya mchog ldan’s Literary Heritage in Bhutan (The
National Library of Bhutan, forthcoming).
also be taught in the second turning and not exclusively in the third turning of the
wheel of dharma. Tāranātha remarks in this connection that there is a difference
between the rang stong intention of The Victorious One and his sons, and that
which is known as rang stong “these days”:44
The former [viz., Shākya mchog ldan] says, “On the level of philosophy (lta ba),
rang stong is profound for cutting through elaborations. On the level of meditation
(sgom pa), gzhan stong is profound for bringing experience (nyams su len pa).”45
Here rang stong is identified as the tradition of Prāsaṅgika, of the Svātantrika, as
well as that of the model treatises.46 The latter [viz., Dol po pa] says that the rang
stong philosophy explained by the Victorious One and his sons is excellent for
cutting through elaborations, but the philosophical tenet (lta grub)47 comprising
gzhan stong is not different in this respect. That which is known as rang stong
these days takes the view of Prāsaṅgika, Svātantrika, as well as the model treatises
to mean that the ultimate is unreal. That is a mistake. It is not a wholesome way
of cutting elaborations through philosophy since it is a denigration.48
Shākya mchog ldan’s position is depicted as sympathetic to gzhan stong while

remaining within the generally accepted framework of logical discourse when it
comes to theory. He acknowledges the role that gzhan stong plays on the level of
meditation. He is shown to hold rang stong as the highest view since he argues
that because non-dual pristine awareness (gnyis med ye shes) cannot withstand
logical scrutiny it cannot have an ultimate place in the context of logical discourse.
Dol po pa argues that non-dual pristine awareness can withstand logical scrutiny
since it is beyond the reach of the logical mind.49 In this context the two masters
are obviously speaking from two different perspectives. While Shākya mchog ldan
is speaking of view, Dol po pa is speaking of that which transcends view.
The two masters are shown to differ on the value they give to trisvabhāva
(three-fold nature) vis-à-vis the two truths, on their positions concerning rang rig,
44
deng sang rang stong gi lta bar grags pa (Tāranātha, Zab don nyer cig pa, 783.5). I assume that
deng sang refers to Tāranātha’s time and not to Dol po pa’s. Tāranātha has the following comment on
what he considers the original meaning of rang stong in his Rgyu ’bras dbu ma’i thig le. He says:
“When it was taught that the ultimate is empty of own essence, the intention was that, when observing
the ultimate, there is nothing for the intellectual mind to hold on to” (don dam rang gi ngo bo stong pa
ltar/ gsung pa’i dgongs pa don dam la dmigs nas/ blo yis bzung ltar ma grub pa la dgongs; Tāranātha,
Rgyu ’bras dbu ma’i thig le, in Collected Works of Jo-naṅ Rje-btsun Tāranātha, vol. 13, 466.7-467.1).
45
Nyams su len is often translated as “practice,” meaning “bringing into experience.”
46
The Madhyamaka treatises of Nāgārjuna and Āryadeva.
47
Combined abbreviation of lta ba and grub mtha’.
48
yang snga mas/ lta bas spros pa gcod pa la rang stong zab/ sgom pas nyams su len pa la gzhan
stong zab ces te/ de’i rang stong yang thal rang gzhung phyi gsum gyi lugs la ngos ’dzi[n]/ phyi mas
ni/ rgyal ba sras bcas bzhed pa’i rang stong gi lta ba de spros pa gcod byed mchog yin yang / gzhan
stong du ’dus pas lta grub logs logs pa min la/ deng sang grags pa’i rang stong thal rang gzhung phyi
gsum gyi lta ba don dam bden med du ’dod pa ni nor pa yin pas/ lta bas spros pa gcod pa la bzang ba
min te/ skur ’debs su ’gyur ba’i phyir yin zhes gsung / (Tāranātha, Zab don nyer cig pa, 784.4-784.7).
49
Literally: “Since non-dual wisdom withstands logical analysis, when analyzing it, it is said that it
is one’s own confusion [that one ends up analyzing]” (gnyis med ye shes rigs pas dpyad bzod yin pas/
de la dpyod pa de rang gi ’khrul gsung; Tāranātha, Zab don nyer cig pa, 785.4).
on the status of tathāgatagarbha, and on their understanding of Mantra[-yāna],

but in particular they are shown to differ on whether the nature of non-dual pristine
awareness is permanent or impermanent, whether it is an entity or a non-entity,
and whether it is compounded or uncompounded. Here Dol po pa’s arguments
transcend (or disregard) the logical categories by stating that while non-dual pristine
awareness is permanent in the sense that it is beyond the three times and is
uncompounded in the sense that it is beyond compounded and uncompounded, it
is neither an entity nor a non-entity.
Dol po pa believes – according to Tāranātha – that epistemology is insufficient
when it comes to establishing the ultimate. He does accept the logic of non-dual
pristine awareness as a cognition, but refutes its inclusion in the logical category
of entities (dngos po), where cognition normally belongs, at least according to a
good deal of Buddhist epistemological literature. He says, “To assert a non-entity
as existing ultimately is not a logical teaching, but to assert an entity as being
ultimate is the tradition of proponents of substantialism.”50 Whether this last remark
in fact is Tāranātha’s way of defending Dol po pa against his later critics can only
be asserted by studying Dol po pa’s own writings in detail. However, this exchange
serves to illustrate the complexities involved when an already ahistorical and fictive
discussion covers more than one level of discourse.
Conclusion
These two comparative studies – the texts of Padma bi dza and of Tāranātha – prod
us to look more closely at the writings of the individual masters cited in order to
investigate whether these were indeed their positions. At the same time, they show
some of the complex issues faced by authors when portraying and comparing the
philosophical positions of important Buddhist scholars who preceded them.
Such studies are valuable because they show us how Tibetans themselves crafted
hermeneutical strategies for bringing order to the heterogeneous complex of views
known as gzhan stong, allowing us to explore the semantic ranges of the term and
its usage. Such strategies are also useful because they provide us with a starting
point for contextualizing the statements found in the actual writings of the scholars
quoted.
The literary output of several of these Tibetan Buddhist scholars is immense.
In some cases their positions may have changed during their lifetime. Furthermore,
an author may defend a philosophical position for a rhetorical purpose without
holding this position himself, and as Tāranātha has mentioned, their presentations
may be colored by pedagogical concerns.
For those of us who study this tradition it is therefore, in some cases, premature
to announce any definitive conclusion as to the position of individual authors before
their entire literary output has been studied in detail. It will require patience before
50
dngos med don dam du ’dod pa chos mi rigs la/ don dam dngos por ’dod pa dngos smra ba’i lugs
so (Tāranātha, Zab don nyer cig pa, 788.1-788.2).
a clear picture emerges of the exact positions of even the most influential Tibetan
Buddhist scholars on gzhan stong.
Glossary
Note: glossary entries are organized in Tibetan alphabetical order. All entries list
the following information in this order: THDL Extended Wylie transliteration of
the term, THDL Phonetic rendering of the term, English translation, equivalents
in other languages, dates when applicable, and type.
Ka
Wylie Phonetics English Other Dates Type
kaḥ thog dge rtse paṇ Katok Getsé Penchen 1761-1829 Person
chen
karma nges don Karma Ngedön Person
karma pa Karmapa Person
karma pa bdud ’dul Karmapa Düdül Dorjé 1733/4-97/8 Person
rdo rje
karma pa mi bskyod Karmapa Mikyö 1507-1554 Person
rdo rje Dorjé
bka’ brgyud Kagyü Organization
Kha
mkhan po khenpo Term
Ga
grub mtha’ drupta philosophical tenet San. siddhānta Term
system
dge lugs pa Gelukpa Organization
rgyal rong Gyelrong Place
rgyu ’bras dbu ma’i Gyundré Umé Tiklé The Central Causal Text
thig le and Resultant Drop
rgyud bla’i rnam Gyülé Namshé Text
bzhad sngon med nyi Ngönmé Nyima
ma
sgom gom meditation Term
sgom pa gompa meditation Term
sgom lugs gomluk practice tradition Term
sgyu ma’i rang bzhin gyumé rangzhin illusory nature Term
Nga
dngos po ngöpo entity Term
snga ma ngama former Term
Cha
chos can rang stong chöchen rangtong relative phenomena Term
[being] rangtong
chos nyid gzhan stong chönyi zhentong ultimate nature Term
[being] zhentong
Ja
jo nang Jonang 1292-1361 Person
jo nang pa Jonangpa Organization
’jam mgon kong sprul Jamgön Kongtrül 1813-1899 Person
blo gros mtha’ yas Lodrö Tayé
’jam mgon gzhan Jamgön Zhenpen 1871-1927 Person
phan snang ba Nangwa
’jam mgon si tu paṇ Jamgön Situ Penchen 1700-1774 Person
chen
’jam mgon si tu rin po Jamgön Situ Rinpoché Person
che
Nya
nyams su len pa nyamsu lenpa bringing experience Term
nyams su len nyamsu len practice Term
gnyis med ye shes nyimé yeshé non-dual pristine Term
awareness
Ta
gtan la dbab bya tenla wapja that which is to be Term
resolved or
ascertained
rtogs ldan tokden a “realized one” Term
lta grub tadrup philosophical tenet Term
lta sgom tagom combination of Term
theory and practice
lta ba tawa philosophical point Term
of view theory;
philosophy
lto ru tsang Toru Tsang Person
brton pa bzhi ldan Tönpa Zhiden 1292-1361 Person
Tha
thub bstan dar rgyas Tupten Dargyé Monastery
chos ’khor gling Chökhorling
thu’u bkwan blo Tuken Lozang Chökyi 1737-1802 Person
bzang chos kyi nyi ma Nyima
Da
dol po pa Dölpopa 1292-1361 Person
dol po pa shes rab Dölpopa Sherap 1292-1361 Person
rgyal mtshan Gyeltsen
dri lan tshes pa’i zla Drilen Tsepé Dawa The Full Moon Text
ba Dialogue
mdo khams Dokham Place
sde dge Degé Place
Pa
padma rnam rgyal Pema Namgyel twentieth Person
century
padma bi dza Pema Bidza twentieth Person
century
padma dbang mchog Pema Wangchok 1886-1952 Person
rgyal po Gyelpo
paṇ ḍita pendita Term
dpal spungs Pelpung Monastery
dpal spungs thub Pelpung Tupten Monastery
bstan dar rgyas chos Dargyé Chökhorling
’khor gling
spyi mtshan chitsen generality; Term
universal
Pha
phyi ma chima latter Term
Ba
byams mgon tā’i si tu Jamgön Tai Situ Person
rin po che Rinpoché
byams mgon si tu rin Jamgön Situ Rinpoché 1886-1952 Person
po che
byams chos lnga’i Jamchö Ngé Ngedön Text
nges don rab tu gsal Raptu Selwa Zhejawé
ba zhes bya ba’i bstan Tenchö
bcos
byams chos lnga’i lam Jamchö Ngé Lamgyi Text
gyi rim pa gsal bar Rimpa Selwar Jepé
byed pa’i bstan bcos Tenchö Rinchen
rin chen sgron gyi sgo Dröngyi Gojé
’byed
dbu ma Uma San. Madhyamaka Doxographical
Category
dbu ma chen po Uma Chenpo Doxographical
Category
dbu ma gzhan stong Uma Zhentongkor Text
skor bstan bcos Tenchö Chokdü Dep
phyogs bsdus deb Dangpo
dang po
dbu ma gzhan stong Uma Zhentong Mawé Text
smra ba’i srol legs Söl Lekpar Chewé
par phye ba’i sgron Drönmé
me
dbu ma la ’jug pa’i Umala Jukpé Namshé Text
rnam bshad dpal ldan Penden Düsum
dus gsum mkhyen pa’i Khyenpé Zhellung
zhal lung dwags Dakgyü Druppé
brgyud grub pa’i Shingta
shing rta
dbyings ying sphere of reality; San. dhātu Term
space
dbyings ye gnyis ka ying yé nyika both the sphere of Term

reality and pristine
awareness
Ma
mi bskyod Mikyö 1507-1554 Person
Dza
’dzam thang Dzamtang Place
rdzogs chen Dzokchen Monastery
Zha
gzhan stong zhentong empty of other Term
gzhan stong snying po Zhentong Nyingpo Text
gzhan stong pa zhentongpa Term
gzhan stong dbu ma Zhentong Uma Gyen The Ornament of Text
rgyan Gzhan stong
Madhyamaka
gzhan stong dbu ma’i zhentong umé tagom theory Term
lta sgom and practice of the
Empty of Other
Central System
Za
zab don nyer cig pa Zapdön Nyerchikpa Text
zur mang Zurmang Monastery
Ya
ye shes yeshé pristine awareness Term
Ra
rang stong rangtong empty of Term
self[-nature]
rang stong pa rangtongpa Term
rang mtshan rangtsen particularity Term
rang rig rangrik Term
Sha
shākya Shakya Term
shākya mchog Shakya Chok Person
shākya mchog ldan Shakya Chokden 1428-1507 Person
sher snying gi tshig Shernyinggi Tsikdrel Text
’brel
shes rab kyi pha rol tu Sherapkyi Paröltu Text

phyin pa’i snyin po’i Chinpé Nyingpö Dön
don rnam par bshad Nampar Shepa
pa ngon med legs Ngönmé Lekshé
bshad
shes rab kyi pha rol tu Sherapkyi Paröltu Text
phyin pa’i lung chos Chinpé Lungchö
mtha’ dag gi bdud Tadakgi Dütsi
rtsi’i snying por gyur Nyingpor Gyurpa
pa gang la ldan pa’i Gangla Denpé Zhi
gzhi rje btsun mchog Jetsün Choktu Gyepar
tu dgyes par ngal Ngelsö Yongdü Tölgyi
gso’i yongs ’dus brtol Jönpa Gyepa
gyi ljon pa rgyas pa
Sa
sa bzang ma ti paṇ Sazang Mati Penchen 1294-1376 Person
chen
si tu Situ Person
si tu padma dbang Situ Pema Wangchok 1886-1952 Person
mchog rgyal po Gyelpo
si tu paṇ chen Situ Penchen 1700-1774 Person
gser mdog paṇ chen Serdok Penchen 1428-1507 Person
Sanskrit
Wylie Phonetics English Sanskrit Dates Type
Abhisamayālaṃkāra Text
Highest Yoga Anuttara-yoga- Doxographical
Tantra tantra Category
Āryadeva second/third Person
century
Asaṅga ca. Person
310-390
buddhakāya Term
Mind Only System Cittamātra Doxographical
Category
darśana Term
dharma Term
sphere of reality dharmadhātu Term
dharmakāya Term
true dharmatā Term
nature of
phenomena;
ultimate
nature
body; state kāya Term
Madhyamakāvatāra Text
mādhyamika Term
mahāmudrā Term
Mahāyāna Doxographical
Category
Maitreya Person
secret syllables mantra Term

Path of Secret Mantrayāna Doxographical
Syllables Category
Nāgārjuna second Person
century
transcendence of nirvāṇa Term
suffering
Prāsaṅgika Doxographical
Madhyamaka Category
cyclic existence saṁsāra Term
Sautrāntika Doxographical
Category
heart of bliss sugatagarbha Term
Buddha’s Word sūtra Term
Path of the Sūtrayāna Doxographical
Buddha’s Words Category
Svātantrika Doxographical
Madhyamaka Category
tantra Term
Tāranātha 1575-1634 Person
tathāgatagarbha Term
three-fold nature trisvabhāva Term
Vaibhāṣika Doxographical
Category
Yogic Conduct Yogācāra Doxographical
System Category
Central Yogic Yogācāra- Doxographical
Conduct System Madhyamaka Category
Bibliography
Broido, Michael. “The Jo-nang-Pas on Madhyamaka: A Sketch.” The Tibet Journal
45, no. 1 (1989): 86-90.
Brunnhölzl, Karl. A Commentary on the Perfection of Knowledge: The Noble One
Resting at Ease. Sackville: Nitartha Institute, 2001.
———. The Center of the Sunlit Sky: Madhyamaka in the Kagyü Tradition. Ithaca:
Snow Lion, 2004.
Burchardi, Anne. Shakya mchog ldan’s Literary Heritage in Bhutan. Thimphu:
The National Library of Bhutan, forthcoming.
Burchardi, Anne, and Ari Goldfield. Gaining Certainty about the Provisional and
Definitive Meanings in the Three Turnings of the Wheel of Dharma, the Two
Truths, and Dependent Arising: The Root Text and Commentary Section Two
of Chapter Seven from The Treasury of Knowledge by Jamgön Kongtrül Lodrö
Thaye. Kathmandu: Marpa Institute, 1997.
Dreyfus, Georges B. Recognizing Reality: Dharmakīrti’s Philosophy and Its
Tibetan Interpretations. Delhi: Sri Satguru Publications, 1997.
Hookham, Susan. The Buddha Within. Albany: SUNY Press, 1991.
’Jam mgon kong sprul blo gros mtha’ yas. Rgya chen bka’ mdzod. Paro: Ngodrub,
1975/6.
———. Shes bya kun la khyab pa’i gzhung lugs nyung ngu’i tshig gis rnam par
’grol ba legs bshad yongs ’du shes bya mtha’ yas pa’i rgya mtsho. Beijing:
Mi rigs dpe skrun khang, 1985.
Kapstein, Matthew. The ’Dzam thang Edition of the Collected Works of Kun
mkhyen Dol po pa Shes rab rGyal mtshan: Introduction and Catalogue. New
Delhi: Shedrup Books, 1992.
———. Reason’s Traces. Boston: Wisdom Publications, 2001.
———. “We Are All Gzhan stong pas: Reflections on The Reflexive Nature of
Awareness: A Tibetan Madhyamaka Defence, by Paul Williams.” Journal of
Buddhist Ethics 7 (2000): 105-25.
Komarovski, Iaroslav. Three Texts on Madhyamaka by Shakya Chokden.
Dharamsala: Library of Tibetan Works & Archives, 2000.
Lopez, Donald S., Jr. The Heart Sūtra Explained: Indian and Tibetan
Commentaries. Albany: SUNY, 1988.
Mathes, Klaus-Dieter. “Tāranātha’s Presentation of trisvabhāva in the gŹan stoṅ
sñiṅ po.” Journal of the International Association of Buddhist Studies 23, no.
2 (2000): 195-223.
———. “Tāranātha’s ‘Twenty-one Differences with Regard to the Profound

Meaning’ – Comparing the Views of the Two gŹan stoṅ Masters Dol po pa
and Shakya mchog ldan.” Journal of the International Association of Buddhist
Studies 27, no. 2 (2004): 285-328.
Mi bskyod rdo rje. Shes rab kyi pha rol tu phyin pa’i lung chos mtha’ dag gi bdud
rtsi’i snying por gyur pa gang la ldan pa’i gzhi rje btsun mchog tu dgyes par
ngal gso’i yongs ’dus brtol gyi ljon pa rgyas pa. Sikkim: Karma Shri Nalanda
Institute, n.d.
———. Dbu ma gzhan stong smra ba’i srol legs par phye ba’i sgron me. In Dbu
ma gzhan stong skor bstan bcos phyogs bsdus deb dang po, 13-48. Sikkim:
Karma Shri Nalanda Institute, 1990.
———. Dbu ma la ’jug pa’i rnam bshad dpal ldan dus gsum khyen pa’i zhal lung
dvags brgyud grub pa’i shing rta. Seattle: Nitartha International Publications,
1996.
Padma bi dza [Zur mang mkhan po padma rnam rgyal]. Dri lan tshes pa’i zla ba.
N.p., n.d.
Ruegg, David S. “A Kar ma bka’ brgyud Work on the Lineages and Traditions
of the Indo-Tibetan dBu ma (Madhyamaka).” Orientalia Iosephi Tucci
Memoriae Dicata. Vol. 3, 1249-80. Roma: Istituto Italiano per il Medio ed
Estremo Oriente, 1988.
———. Buddha Nature, Mind and the Problem of Gradualism in a Comparative
Perspective: On the Transmission and Reception of Buddhism in India and
Tibet. London: School of Oriental and African Studies, University of London,
1989.
———. Three Studies in the History of Indian and Tibetan Madhyamaka
Philosophy. Studies in Indian and Tibetan Madhyamaka Thought, Part 1.
Wiener Studien zur Tibetologie und Buddhismuskunde 50. Vienna: Arbeitskreis
für tibetische und buddhistische Studien, Universität Wien, 2000.
Shākya mchog ldan. Collected Works. Thimphu: Kunzang Tobgye, 1975.
Smith, E. Gene. Among Tibetan Texts: History and Literature of the Himalayan
Plateau. Edited by Kurtis R. Schaeffer. Boston: Wisdom Publications, 2001.
Stearns, Cyrus. The Buddha from Dol po: A Study of the Life and Thought of the
Tibetan Master Dolpopa Sherab Gyaltsen. Albany: SUNY, 1999.
Tāranātha. The Collected Works of Jo-naṅ Rje-btsun Tāranātha. 17 vols. Leh,
Ladakh: C. Namgyal & Tsewang Taru, 1982-87.
———. Collected Works. ’Dzam thang Edition, TBRC W22276.
———. Zab don nyer cig pa. In The Collected Works of Jo-naṅ Rje-btsun
Tāranātha. Vol. 4, 781-95. Leh, Ladakh: C. Namgyal & Tsewang Taru,
1982-87.
Wangchuk Dorji. “The rÑiṅ-ma Interpretations of the Tathāgatagarbha Theory.”

Wiener Zeitschrift für die Kunde Südasiens 48 (2004): 171-213.
Williams, Paul. “A Note on Some Aspects of Mi Bskyod Rdo Rje’s Critique of
Dge Lugs Pa Madhyamaka.” Journal of Indian Philosophy 11, no. 2 (1983):
125-46.
Beyond Anonymity: Paleographic Analyses of the
Dunhuang Manuscripts
Jacob Dalton
Yale University
Tom Davis
University of Birmingham
Sam van Schaik

The British Library
Abstract: This article presents a new paleographic approach to the Tibetan

manuscripts from Dunhuang. By adapting the techniques of forensic handwriting
analysis to the Tibetan alphabet, we can identify groups of manuscripts written in
the same hand. After introducing this new approach, the present paper applies it
to the works of a single scribe, taken as an initial example. Once this particular
group of manuscripts has been identified, a range of further insights into this
person emerge – his many connections to the kingdom of Khotan, his unique writing
style, and his interest in the external ritual practices relating to water and fire
offerings, stūpas, rosaries, and the like. This new approach promises to alter
significantly our understanding of the Tibetan Dunhuang documents. No longer
are we confronted with a mass of undigested material; now we can begin to date
and ascribe names to whole swathes of the collection.1
An Introduction to Forensic Handwriting Techniques

Working with manuscripts is certainly a laborious affair, but one that can have
almost magical moments. After years of careful analysis, one may begin to feel
an almost personal bond with the scribes of the distant past. The present article
1
The authors thank the AHRC, the International Dunhuang Project, the American Philosophical
Society, and the American Academy of Religion for their support in funding this collaborative project.
The original research began during a three-year AHRC-funded project on the tantric manuscripts from
Dunhuang, published in Jacob Dalton and Sam van Schaik, Tibetan Tantric Manuscripts from Dunhuang:
A Descriptive Catalogue of the Stein Collection at the British Library (Leiden: Brill, 2006).
1550-6363/2007/3/T3106.
© 2007 by Jacob Dalton, Tom Davis, Sam van Schaik, Tibetan and Himalayan Digital Library, and International
Association of Tibetan Studies. Distributed under the THDL Digital Text License.
Dalton, Davis, and van Schaik: Beyond Anonymity 2
grew out of just such a strange and ill-defined series of experiences. Its goal,
however, is to justify these intuitions in as much detail as possible, to shed light
on these murky insights. For three years, from 2002 to 2005, Jacob Dalton and
Sam van Schaik worked together on a project to catalogue the Tibetan tantric
manuscripts in the Stein collection, held at the British Library. After about a year
of working every day with these manuscripts, they began to recognize the individual
handwritings specific to certain scribes. Over the following months they became
increasingly convinced of their identifications, until many seemed quite obvious.
When these theories were presented to other scholars they were met with interest,
but also with skepticism. More proof was required, and while their initial, largely
intuitive recognitions of different scribes’ handwritings had been relatively easy,
it would be far more difficult to explain precisely what lay behind these
identifications.2
In 2004 van Schaik and Dalton contacted Tom Davis, the third author of this
article and an expert in forensic handwriting analysis. Over the following year the
three authors met numerous times to develop a forensic-style approach to the
paleography found in the Dunhuang manuscripts. Developing a firm basis for their
theories proved difficult, and some of the early identifications had to be abandoned,
but a reliable method did finally emerge. What follows is an introduction to this
new paleographic approach to the Tibetan Dunhuang collection. The article includes
four sections:
1. An introduction to the basic practice of forensic handwriting analysis and

how it may be applied outside the courtroom.
2. A discussion of how these forensic techniques were adapted to the Tibetan
script and how the handwriting of one scribe can be described by a few
simple rules.
3. A brief overview of the writing practices and the social milieu of tenth
century Dunhuang.
4. A review of further internal evidence indicating that the manuscripts
written in the hand identified in part two are all the work of a single person.
Terminology
First, a note on the terminology. When someone looks at handwriting, what they
see on the page is a series of graphs – in an alphabetic script, the letters as they
actually appear on that particular page. Each graph is an individual, necessarily
unique, representation of a grapheme, which in our alphabetic writing means a
2
Cristina Scherrer-Schaub has discussed the paleography of early Tibetan manuscripts in Cristina
A. Scherrer-Schaub, “Towards a Methodology for the Study of Old Tibetan Manuscripts: Dunhuang
and Tabo,” in Tabo Studies 2: Manuscripts, Texts, Inscriptions and the Arts, ed. Cristina A.
Scherrer-Schaub and Ernst Steinkellner, 3-36 (Rome: Istituto Italiano per l’Africa e l’Oriente, 1999);
Cristina A. Scherrer-Schaub and George Bonani, “Establishing a Typology of the Old Tibetan
Manuscripts: A Multidisciplinary Approach,” in Dunhuang Manuscript Forgeries, ed. Susan Whitfield
(London: The British Library, 2002), 184-215. However, these discussions have concerned script
typologies rather than individual handwritings.
letter considered abstractly. A reader sees a graph (a unique mark on a piece of

paper), recognizes a grapheme (“that’s an /a/”), and is thus able to read.
If a given writer produces a graphic form of the grapheme /n/ that resembles
his (graphic) version of the grapheme /u/, so that the two cannot be distinguished,
a reader may wonder “which letter is that?” What he or she is actually asking is,
which grapheme am I seeing here? This kind of graphic habit, if it occurs regularly
in a particular hand, is idiographic. It is a variation in graphic form that is
characteristic of that particular writer and thus provides evidence of individuality.
The most valuable idiographic items for the purpose of identification are those that
are not entirely under his or her conscious control, as this makes them difficult
both to forge and to disguise. Also characteristic of individual writers is allographic
variation. Some writers, for instance, use a flourished cursive form of the capital
/T/, others a plain block capital form. This difference is common and normally
conscious, and so only weakly idiographic; it is a licensed and recognized variation
in the representation of the grapheme and therefore allographic.
We have, then, a hierarchy that can be listed from most general to most specific:
grapheme, allograph, idiograph, graph. The grapheme /a/ is the letter considered
independently of any particular realization of it. An allograph is an accepted version
of that grapheme. An idiograph is the way (or one of the ways) in which a given
writer habitually writes /a/. A graph is a unique instance of /a/, as it appears on a
particular page.
Forensic handwriting examination concerns itself with questioned documents.
There will be a dispute, usually as to the authorship of a particular document. In
order to decide that dispute, sample writing is obtained, writing that indisputably
was produced by the alleged author of the questioned document. Since in forensic
cases there is always the possibility of deliberate deception, the sample writing
will be of two kinds: writing that was produced for the purpose of the examination
(request writing), and writing that was produced without the knowledge that it
would be examined by an expert (naturally occurring writing). The function of
the latter is to act as a control sample, to test the validity of the request writing.
Request samples must be treated with caution, since they may be disguised in order
to hinder identification of the questioned writing. Why then have request samples
at all? Because their content can be controlled. If the questioned writing is an
extended text, the analyst can ask for the request writing to contain the same content
as the questioned writing. Handwriting identification depends on letter-by-letter
comparison, and if the same letters occur in the same place in each document, such
comparisons are much easier.
Methodology
Step 1: Analysis
The normal practice of forensic handwriting examination is a three-stage process:
first the handwriting under examination is analyzed, then compared, and finally a
conclusion is derived from the comparison. In the first stage, the usual procedure
is to prepare analytic charts. These are tables containing cells for instances of each
of the graphemes that the writing contains, one cell for each of the lower case
letters, each of the upper case, punctuation, numerals, and common abbreviations
such as the ampersand. Each of the cells will be filled in (as far as possible, as
some letters may not be represented in the documents examined) with a description
of each of the significantly different forms of each letter in the hand under
examination. Verbal descriptions are not very useful for analysis of graphic shapes;
the normal procedure is to draw the shape, with added arrows indicating line
direction and other significant features, sometimes with a brief verbal comment.
A certain skill – ironically, a forger’s skill – is required in order to produce a
satisfactory analytic chart.
The analytic chart is essentially a private document; it is part of a laboratory
notebook and will normally only be seen by the examiner and his colleagues in
the laboratory, though on occasion the defence will ask to see a prosecution expert’s
notebook. The ability to create such a chart is very much based on experience. The
examiner must know what is likely to constitute significant variation for the
purposes of determining authorship. Supposing, for instance, the document being
analyzed contains twenty instances of a lower-case /a/. Each of those graphs as
they appear on the document will be unique, but it will also be (usually)
recognizable as representations of the same grapheme, because otherwise the hand
would be illegible. The examiner’s experience enables him or her to assess the
significance of these differences and similarities in each case and how they may
be of use in determining authorship. For this purpose there may be only one form
of /a/ in the document, represented twenty times with sufficient fidelity to the
writer’s internal model as to make them, for purposes of identification, the same.
Or there may be two, three, or even more idiographically distinct forms of /a/ in
the document, each one represented one or more times.
Once the sample writing has been analyzed in this way, the questioned writing
is similarly analyzed, in a separate chart or charts. These analyses are kept separate
in order to ensure a disciplined approach to the business of identification. It is easy
to fall into the trap of hypothesizing a theory of authorship early in the procedure,
and then looking for evidence that confirms the hypothesis and ignoring evidence
that refutes it. If the analysis takes place before the comparison, the analyst is
prevented from doing that.
Step 2: Comparison
The second step is to compare the two analyses of the questioned and sample
writings, with constant reference to the original documents. Each version of each
grapheme in the chart of sample writing is compared meticulously with the versions
of the corresponding grapheme in the chart of questioned writing. Again, this is
experience-based. The examiner must know which graphic forms are likely to be
idiographic, and which allographic; in other words, which variations of the
grapheme are sufficiently unusual as to be useful for identification. This requires
a familiarity with the way in which that particular language is written in a particular
time, place, and by a particular kind of writer. It involves a knowledge of all of

the factors that can influence writing, particularly at the idiographic level. One
must consider the physiology of the writing movement, the characteristics of the
writing implement used, the writing surface on which the inscription has taken
place, and the way in which all three of these interact.
However, in spite of the level of expert knowledge behind a given opinion,
forensic experts are very conscious of the fact that at some point they may have
to convince a jury of their findings. The comparison that is the basis of these
findings must contain a preponderance of judgments that are based on clear and
evident similarities or differences. Although forensic experts are permitted to give
opinion evidence, they are not allowed to take refuge in a private, incommunicable
expertise – an intuitive mystery, what lawyers call a “black box.” The basis of the
opinion must be expressible and comprehensible to non-experts.
Step 3: Opinion
Finally, as a result of this comparison, an opinion is formed. The first thing to say
about this opinion is that it will probably be qualified. Forensic expertise in
handwriting is as objective as possible, but complete objectivity simply cannot be
attained. All the information on which a particular opinion is based can never be
fully available to consciousness: one cannot recall every instance of a particular
kind of /o/ that one has seen, though that submerged knowledge provides the basis
for one’s opinion. Moreover, since the opinion concerns the characterization of
aspects of a graphic shape, its basis cannot be made entirely available for inspection.
An expert will say: “this /o/ is not significantly the same as that /o/ because it
differs in certain respects.” Yet the two /o/s are necessarily the same in other
respects (they must be, or they would not be recognizable as instances of /o/). Thus
in the case of these two different /o/s, the forensic opinion essentially decides that
the differences are significant while the similarities are not, simply because in
one’s experience, the similarities are not idiographic. Ultimately, therefore, the
deciding factor remains: experience.
However, this is not a black box. All human beings (indeed, all sentient beings)
are skilled recognizers and do it constantly. Handwriting experts can communicate
their findings on the basis of this shared ability. A jury can be shown what the
expert sees, usually in the form of enlarged photographs, and be persuaded to agree
on the basis of this universal ability to recognize.
And experts can be wrong. In order to minimize this possibility two strategies
are adopted by handwriting experts: caution and overkill. Handwriting experts are
notoriously cautious: the police frequently complain that it is hard to get a usable
opinion out of a handwriting expert. When they get one, they are happy to have it,
because it will stand up well in court. This is because of the other strategy:
handwriting analyses are exhaustive. Every letter in the document is examined,
described, and compared with every other letter in all of the documents under
consideration. The effect of the evidence, when well presented, is intended to be
overwhelming. Thus the opinion will be supported by a second chart, for public
consumption, that shows the evidence on which this opinion is based. This consists
of a photographic display, in which the actual idiographs, enlarged and in context,
are shown. The examiner will use the display chart as the basis of his or her
evidence, adding a verbal commentary that points out the idiographic similarities
or differences.
A Forensic Approach to Paleography

When a document examiner is given a non-forensic case to look at, a number of
differences between the forensic and non-forensic practice are apparent, some of
them rather disconcerting. Foremost, in non-forensic work the writers are usually
dead, and can have been dead for a long time; the Tibetan manuscripts that are the
focus of the present study are at least a thousand years old. Their writers are thus
not available for questioning or for the provision of request samples. This and a
range of other differences make paleography a very different affair from that of
forensic science. After years of learning the extreme and unforgiving rigor required
in forensics, it is not easy to make this adaptation to paleographic work.
There are compensations. Forensic handwriting analysis is used to put people
in prison (or get them out). Huge amounts of money can be at stake. The utmost
caution is essential. Paleography, on the other hand, would rarely address issues
of anything like that importance, and normally investigates questions that are only
of interest to relatively few scholars. The standard of proof can afford to be
considerably lower: one can, however reluctantly, relax a little.
Another reason for relaxing is that the forensic handwriting expert is always
aware of, and more often than not directly concerned with, the possibility of forgery
and deliberate disguise. Thus the examination must be conducted on an extremely
subtle level. Forged or disguised writing is by definition unnaturally written, and
this unnaturalness manifests in a lack of fluency of line quality (among other
things), which can sometimes only be seen under magnification. Forgery of old
documents does of course occur, but the paleographer normally faces a much
simpler problem: basic identification. Were those two documents written by the
same person? At this level of examination the need to examine the original
manuscripts is much less imperative.
Also helpful have been recent advances in technology. Digital cameras, scanners,
entry-level computers, and high-level image editing programs are now easily
available, and can produce digital photographs of documents that can even, in
some ways, be better than the original for identification purposes. Once the images
are in the computer they can be manipulated in remarkable ways. Different
occurrences of the same letter can be lifted from different documents and easily
juxtaposed, for instance. The routine procedure developed by Tom Davis in his
paleographic work is to create (for alphabetic writing) three PowerPoint
“slideshows,” for capitals, lower case, and numerals/punctuation. These are used
for the initial step of analysis (described above). In creating these slideshows, each
document is first scanned or photographed to produce a Photoshop file. Each
idiographically significant form of, say, the letter /a/ can then quickly and easily
be lifted in Photoshop from the image, together with the word in which it occurs,
and pasted into the corresponding /a/ slide in the PowerPoint slideshow. This
process creates a complete picture of the handwriting as it is found in the document
under examination, and the PowerPoint slides can be annotated as required.
The next step is to use the images of individual words and letters, to enable the
comparison. For this, HTML framesets are ideal; it is easy to create web pages
that enable side-by-side comparison of any letter as it occurs in different documents
under examination. HTML is easy to use and universally readable; moreover,
whereas in forensic work one’s examination is normally carefully kept confidential,
the scholarly tradition in paleographic work is to offer one’s evidence for peer-group
consideration and discussion, and, of course, the World Wide Web makes this
almost absurdly easy.
The Adaptation of Forensic Techniques to the Tibetan Script
Building a Tool for Forensic Examination

Over several meetings the authors discussed the possible application of the forensic
methods outlined above to the Tibetan script. The advantage was that we were
dealing with another alphabetic script. Before conducting our analysis, however,
we first had to decide how many graphemes the Tibetan alphabet contained. This
was not as simple as going through the standard alphabet of ka, kha, ga, nga, and
so on, mostly due to the Tibetan practice of stacking letters. The shape of the letter
ka, for example, is significantly altered by the addition of the subscribed ya (ya
brtags), so that kya becomes a distinct grapheme. In some cases, vowel signs may
also change the form of a letter so that the combination may be considered a separate
grapheme. Thus, ku is often more than a mere combination of ka with the “u”
vowel (zhabs kyu). We also added non-letter forms, including the punctuation
marks shad (phrase-marker) and tsheg (dot). In all, we counted around 110
individual graphemes. We also had to distinguish between the headed (dbu can)
and the headless (dbu med) forms of the letters, which, like our own roman upper
and lower case letters, are actually different graphemes. Thus a manuscript in
which the scribe has written both headed and headless letters may contain over
two hundred graphemes.
Analysis
Having identified our objects of analysis, we were able to begin. Our work here
would be at a disadvantage, for we could never be as familiar with the scribal
conventions of tenth-century Dunhuang as we are with those of our own alphabet.
Despite having worked closely with the original manuscripts for a number of years,
our relative lack of experience would inevitably make it difficult to distinguish
between which scribal characteristics were conscious stylistic features (allographic
variations) and which unconscious quirks (idiographic variations).
For help in this regard we utilized contemporary Tibetan handwriting guides,

though with caution, as their relationship to tenth-century handwriting practice has
yet to be established.3 Modern Tibetan calligraphy is a precisely prescribed practice,
but the scripts seen in the Tibetan Dunhuang documents are far less regular. In our
analytic discussions (and in the present article), we adopted many of the standard
modern names for the individual strokes in a graph. Thus the top horizontal line
is designated the “head” (mgo), the main vertical line the “leg” (rkang ba), and so
on.
With these modern terms and practices in mind, we began our analysis of the
graphs found in a particular manuscript that we suspected belonged to a wider
group of manuscripts in the same hand. For our starting point we chose ITJ425.
In our analysis we employed the usual tools of the forensic handwriting analyst,
examining the order in which strokes were written, the length and angle of a
particular stroke, and superfluous strokes known as “ticks.”4
We began by analyzing all of the graphs present in ITJ425. For the purposes of
the intertextual comparisons to come, however, we soon realized we would have
to limit our scope. Given the number of graphemes in the Tibetan alphabet
(sometimes over two hundred) and the size of the documents involved (here sixteen
folio sides, but often more), a comparison of every grapheme in both texts under
comparison would not be possible. For this reason, we used our analysis to reduce
the scope of our comparisons to only those graphemes in which we perceived
significant idiographic variation. Our analysis of ITJ425 thus resulted in a set of
idiographic benchmarks that were unique to that hand. These benchmarks could
then be used to identify other manuscripts in the same hand.
Comparison
Next we selected a group of manuscripts that we suspected contained the works
of no more than two to three different scribes. Working with high-resolution digital
images of the manuscripts in question, we extracted several examples of every
grapheme we could find in each manuscript (some were not represented) by “cutting
them out” using Photoshop. The forensic method discussed in the previous section
requires one to enter each and every graph into the analytic chart, but this was not
possible in the case of our manuscripts, which can comprise over a hundred folios,
much longer than the usual notes and letters analyzed by forensic experts. Instead
in constructing our charts, we tried to select a representative sampling of all the
graphs present. For each manuscript we cut out hundreds of thumbnail images,
labeling them with their graph name and location, and linking them into a website.
For the website we adopted the method of HTML framesets described above, the
3
See Rev. G. Tharchin, The Tibetan Primer of Current Hand Writing (Kalimpong: Tibet Mirror
Press, 1970); and Dpa’ ris sangs rgyas, Bod yig ’bri tshul mthong ba kun smon (Beijing: Mi rigs dpe
skrun khang, 1997).
4
By stroke we mean a mark made by the pen between one pen-lift and the next. The Latin term ductus
or duct is often used in paleographic writings rather than stroke.
result of which can be found at the following URL: http://idp.bl.uk/handwritings/

index.html.5
The website is divided into three columns, each with three sections. The upper
section contains a list of manuscripts to choose from. Selecting one of these brings
up a list of the available graphs from that manuscript in the middle section. Selecting
one of these graphs brings up one or more examples of that graph, extracted from
the selected manuscript, in the lower section. Using this website, graphs from three
different manuscripts can thus be displayed next to each other, allowing the kind
of comparisons discussed in the previous section.
Let us now look at the list of benchmarks that grew out of our analysis of the
manuscript ITJ425. In this particular case, we have established a list of four firm
benchmarks (other less firm benchmarks were proposed, but because they are less
reliable, we have not included them in the present article). Each benchmark is
described below, with a sample graph shown alongside a graph from another
manuscript (in what we believe is another hand) for comparison. For further
examination beyond these basic examples, the reader is directed to consult the
website mentioned above.
Small Heads and Long Legs

First, in almost all of the graphs in this hand, the “head” (mgo) of the graph is
proportionately smaller and the “leg” (rkang ba) longer than in other handwritings.
This can be demonstrated by comparing two sample graphemes, na and zha:
Images 1 & 2: A na graph from ITJ425 and

ITJ594.
Images 3 & 4: A zha graph from ITJ425 and

ITJ321.
5
The website is hosted by the International Dunhuang Project (see http://idp.bl.uk), under the auspices
of which the cataloguing work which resulted in the present paper took place.
Ticks
Our second distinguishing benchmark is that the leg of the da graphs are often
finished with a small “tick” at the bottom. This unusual feature occurs frequently
(in about fifty percent of cases) and appears to be an unconscious idiographic
element of the handwriting.
Images 5 & 6: A da graph from ITJ425 (with

tick) and from ITJ321.
The Grapheme Ga
Our third benchmark is the unusual form of the ga grapheme. As can be seen here,
the “leg” of the graph rises some way above the “head” before descending below
it, an unusual feature.
Images 7 & 8: A ga graph from ITJ425 and

ITJ594.
The Grapheme Nya

Fourth, we have the grapheme nya, which is usually written in a single pen stroke.
In most examples of handwriting in the Dunhuang manuscripts (as indeed in later
Tibetan model handwriting), the pen is lifted after the first curve of the graph, and
then placed down again for the second curve, as in the example to the right from
ITJ647.
Images 9 & 10: A nya graph from ITJ425 and

ITJ647.
Opinion
Individually, none of the above features may be unique to this handwriting, but
the occurrence of all of them in a manuscript is persuasive evidence, if not outright
proof, that the manuscript is from the same hand as ITJ425. Having developed this
set of benchmarks, we were able to apply them to a number of additional
manuscripts. These comparisons resulted in a group of manuscripts that we judged
all to have been written by the same scribe who wrote ITJ425. Having identified
this manuscript group, we found a range of further convincing reasons for seeing
all these manuscripts as the work of a single person, and it is to this supporting
evidence that we now turn.
Writing Practices around Dunhuang

Before we look in detail at the work of our scribe, however, we should perhaps
briefly review his scribal setting. Most of his works are written on Tibetan-style
Indian-style book (po ti, pustaka) manuscripts, with a few written on Chinese-style
scroll sheets. He probably wrote with a wooden pen (of which some examples
have been found in Central Asia), which he would dip regularly into an ink
receptacle.6
Although our scribe wrote in Tibetan, it is not clear that he was of Tibetan
origin. The colophons of many Tibetan sūtra manuscripts show that there were
certainly Chinese scribes at Dunhuang who were able to write Tibetan, and as we
will see below, there are indications that he may even have been Khotanese. One
of the mysteries of the library cave manuscripts is why manuscripts in so many
different languages (Chinese, Tibetan, Khotanese, Uighur, and more) were stored
together. If they mostly represent the small library of Sanjie Monastery (sanjie si),
as Rong Xinjiang argues, then we are left with the question: Did people of different
language groups live together in this monastery, or were the manuscripts collected
from elsewhere?7 Such questions remain to be answered.8
A related issue is the location of Sanjie Monastery. Rong Xinjiang has argued
for a location directly in front of the cave site, perhaps in front of Cave 16 where
some unidentified wooden ruins remain.9 Against this conclusion, Robert Sharf
has recently suggested that the monastery where the library originated, be it Sanjie
Monastery or otherwise, was likely not in the immediate vicinity of the Mogao
6
On the increased use of wooden and reed pens, rather than brushes, during and after the Tibetan
occupation of Dunhuang, see Fujieda Akira, “Chronological Classification of Dunhuang Buddhist
Manuscripts,” in Whitfield, Dunhuang Manuscript Forgeries, 111–12.
7
Rong Xinjiang, “The Nature of the Dunhuang Library Cave and the Reasons for its Sealing,”
Cahiers d’Extrême-Asie 11 (1999-2000): 247-75.
8
See Tokio Takata, “Multilingualism in Tun-huang,” Acta Asiatica: Bulletin of the Institute of Eastern
Culture, no. 78 (2000), Tun-huang and Turfan Studies: 49–70, for a discussion of the interaction of
different cultures at Shazhou/Dunhuang.
9
Rong Xinjiang, “The Nature of the Dunhuang Library Cave,” 264.
caves.10 According to Sharf, the caves themselves were less meditation caves than
mortuary shrines or temples under the jurisdiction of a number of powerful families
and clans. There is also ample evidence that the site served as a place of pilgrimage
where large festivals took place. However, Sharf argues, we must be careful to
distinguish the site’s monumental functions from what we normally think of as
“monastic” practices, such as ascetic, repentance, and meditative practices, scriptural
study, or scribal activities relating to the library discovered there. Such monastic
activities, and perhaps Sanjie Monastery itself, were more likely located somewhere
off-site, probably nearer to the town of Shazhou, a few miles away. In short, we
are left with lamentably little clear evidence for the monastic setting within which
our manuscripts were penned.
Another question is the role of the writer: Was he a professional, writing out
manuscripts for others, or did he write manuscripts for his own use and that of his
immediate circle? A clear thematic consistency to the writings in the hand of our
scribe, and what appear to be revisions of the same work, suggest that he was not
a jobbing scribe writing for a variety of individuals, but that he wrote either for
his own benefit or at least for another person with surprisingly specific interests.
An Example Scribe: Supporting Internal Evidence
The Identified Manuscript Group

We have shown how the method of forensic handwriting analysis can be adapted
to the Tibetan Dunhuang manuscripts, using digital technology to facilitate both
the “laboratory work” and the dissemination of this technique. Now we will look
at the case of the particular scribe whose handwriting was identified using this
technique. By examining the interrelationships and thematic consistencies within
the group of manuscripts that we have attributed to this scribe, we can offer
substantial supporting evidence for our forensic handwriting identification.
The scribe here is one whom we have identified as responsible for the sample
manuscript ITJ425. By comparing a number of actual manuscripts against the
isolated graphs in our table, we were able to expand significantly the group of
manuscripts attributable to this scribe. For the purposes of this article, the group
comprises ITJ318, 338, 340, 341, 343, (344), 377, 407, 422, 423, 424, 425, and
688.11 Within this group, we can discern a number of thematic links. There is also
clear evidence that some manuscripts represent revisions of the same work. To
begin with, let us look at the thematic links.
10
Robert Sharf, “The Enigma of the Dunhuang Caves” (paper presented at the annual meeting of the
American Academy of Religion, San Antonio, TX, November 22, 2004). In forming his argument,
Sharf points to evidence from other cave sites, such as Longmen and Yungang, which were clearly not
sites for “monastic practice,” and to the lack of any hard evidence on the location of Sanjie Monastery.
11
We believe the group could easily be expanded much further, but these are the manuscripts discussed
herein.
Thematic Links between the Manuscripts

The first theme linking some of the manuscripts is a Khotanese connection. The
kingdom of Khotan was located on the southern Silk Route, to the west of
Dunhuang. It was still active during the tenth century, though increasingly under
threat from Turks and Arabs, whose Islamic religion was to supplant and eventually
eliminate Khotanese Buddhism in the eleventh century. During the tenth century,
Dunhuang and Khotan enjoyed strong diplomatic ties, Khotanese Buddhists visited
famous teachers in Dunhuang, and there also seems to have been a Khotanese
settlement in the area.12
Several of the manuscripts in our group contain Khotanese page numbers:
ITJ338, 340, 423, 424, and 425. The numbers are either written in Tibetan
transcription or in Khotanese numerals. The use of Khotanese numbers is unusual
in the Dunhuang Tibetan manuscripts and suggests that our scribe had a particular
connection to Khotan, perhaps even that he was of Khotanese origin.13
Another Khotanese connection can be seen in ITJ688, a treatise on the rosary
(’phreng ba, māla).14 This short treatise discusses the different kinds of rosaries
appropriate for practitioners of the different buddha families. At the outset the
author divides his discussion into seven topics, though he ends up addressing only
four of them and appends an additional one. The seven topics listed initially include:
(1) the teacher responsible for it, (2) the tantra from which it is taken, (3) the
original source, (4) the class to which it belongs, (5) the method for counting,
taught correctly for each buddha family, (6) the correct meditation, and (7) how
it is said to surpass.15 These topics roughly correspond to those of a similar
Khotanese treatise on rosaries found in the Dunhuang manuscript IOL Khot 55,
fols. 1r.4-1v.1.16 Furthermore, the topics that are listed but not discussed in the
Tibetan treatise are addressed in the Khotanese treatise, and in a similar order.17
When the two treatises are compared in more detail, several differences emerge.
The Tibetan text ascribes the teaching on rosaries to the well-known Indian scholar
Vimalamitra (eighth century), whereas the Khotanese text ascribes it to one
12
See Takata, “Multilingualism in Tun-huang,” 52–53. On the Khotanese manuscripts from Dunhuang,
see Prods Oktor Skjœrvø, Khotanese Manuscripts from Chinese Turkestan in the British Library: A
Complete Catalogue with Texts and Translations (London: The British Library, 2002).
13
These Khotanese numbers have been identified and discussed in Tsuguhito Takeuchi, “Old Tibetan
Buddhist Texts from the Post-Tibetan Imperial Period (mid-9 C. to late 10 C.),” in Proceedings of the
Tenth Seminar of the International Association for Tibetan Studies (Leiden: Brill, forthcoming).
14
A discussion and French translation of this text can be found in Rolf A. Stein, “Un genre particulier
d’exposés du tantrisme ancien tibétain et khotanais,” Journal Asiatique, no. 275 (1987): 265-82.
15
ITJ688, fol. 1.1-1.3: slobs dpon gang gis mdzad pa dang / rgyud gang las ’dus pa dang / ’khungs
gang las byung ba dang / mtshams gang du gtogs pa dang / so so’i rigs ma nor bar bstan pa bgrang
ba’i thabs dang / bsam rgyud ma nor ba dang / don las mtshan du gsol ba’o.
16
A transcription and translation of IOL Khot 55 can be found in Skjærvø, Khotanese Manuscripts,
292-96.
17
Stein claims that “the Khotanese text is without doubt based on a Tibetan model” (Stein, “Un genre
particulier,” 269). Unfortunately, he gives no reasons for this conclusion, and in this instance, the more
complete Khotanese treatise seems more likely to have been the model for the Tibetan version.
Baśrīmajsu.18 Also, where the Khotanese treatise simply classifies the teaching as
Vajrayāna, the Tibetan text classifies it as Mahāyoga and goes on to say that it was
gathered from a tantra bearing the title Descent of the Sage [Tantra] (thub pa a ba
da ra, *Muni-avatāra[-tantra]). This title remains unidentified, though it also
appears in Pelliot tibétain 849.19 Despite these specific differences, the overall
similarity between the Tibetan and Khotanese treatises indicates a close relationship,
perhaps that the somewhat garbled Tibetan version is a reworking of the Khotanese
text.
Even more interesting is the fact that several other manuscripts in our group
are also treatises divided into numbered topical divisions, a feature that is not at
all common in the Tibetan Dunhuang manuscripts in general. This feature can be
seen in the treatise on stūpas entitled Narrative Setting and Benefits of Stūpas
(mchod rten gyi gleng gzhi dang phan yon) in the manuscript ITJ338, a manuscript
that also bears Khotanese page numbers. The same topical structure appears in the
’Chu gtord gyi gzhung, which appears in two very similar versions in the
manuscripts ITJ340 and 341/1. Again, both manuscripts contain Khotanese
numbering. Finally, we have a treatise on the fire ritual (sbyin sreg, homa) called
Abridged Instructions on the Aspects of the Peaceful Fire Offering (zhi ba’i sbyin
sreg lag len man ngag ’khol bur phyung ba). This also exists in two manuscript
versions, ITJ422 and 423.
A final thematic coherence to this group of manuscripts may be their shared
concern with the practice of external rituals: activities such as making and using
rosaries, making and worshipping at stūpas, water offerings, and fire rituals. This
is not common elsewhere in the Tibetan Dunhuang texts, and thus is yet another
feature suggesting that this group of manuscripts reflects the interest of a single
scribe.
Multiples and Revisions

Our group of manuscripts is lent still further coherence by the existence of multiple
revisions of the same work, seen in a number of the relevant manuscripts. As
already mentioned, ITJ340 and ITJ341/1 are both versions of the Exposition on
Water Offerings (’chu gtor gyi gzhung).20 ITJ341 itself includes two quite different
18
Stein suggests that this person may be Rba Manydzu (also Dba’ Manydzu, Sba Manydzu).
19
Pelliot tibétain 849, fol. 20. See also Joseph Hackin, Formulaire Sanskrit-Tibétain du Xe siècle
(Paris: Librairie Orientaliste Paul Geuthener, 1924), 7.
20
Still another version of the same text is found in the Peking Bka’ ’gyur (Q. 4593), where it is titled
the Ritual Manual for Water Offerings (chu gtor gyi cho ga). A comparison of ITJ340 to the canonical
edition shows that they are largely identical, apart from a few word changes and a couple of passages
that are added or missing from each. More work needs to be done comparing these two versions to the
one found in ITJ341. The colophon to the Peking edition attributes the authorship of the work to
Jayasena and its translation to Rin chen bzang po (958-1055). If we are to accept this attribution, we
must conclude that the Jayasena who authored this work is not the same person who translated a number
of Tibetan works contained in the Bstan ’gyur. The latter translator worked alongside Sum pa lo tsā ba
Dar ma yon tan and was active in the second half of the twelfth century (see Dan Martin, “Tibskrit
Philology” [Kurt Keutzer, 2006], 906-8, http://www.eecs.berkeley.edu/~keutzer/martin/TibskritUni.pdf
variations on the same text (referred to here as ITJ341/1 and ITJ341/2), so that the
two items begin with identical opening lines, but part ways on the second line. A
further connection may be made between ITJ341 and ITJ407, yet another manuscript
from our handwriting group. The first folio of ITJ341/2 (10v.2) goes on to direct
the reader to mix the offering ingredients in a bowl, “in accordance with the
instructions in the Eight Jewels.”21 ITJ407 may contain precisely that work, as it
bears the title An Offering of the Eight Jewels.22 The latter seems to be a liturgy to
be recited in conjunction with an offering rite, possibly that being prepared in
ITJ341/2. In these ways the three manuscripts, ITJ340, 341, and 407, are closely
linked.
Two additional pairs of matching manuscripts are ITJ422 & 423 and ITJ424 &
425. All four of these manuscripts are on the same topic – the peaceful fire
offering.23 The first two, ITJ422 and 423, are variations on the same text, titled the
Abridged Instructions on the Aspects of the Peaceful Fire Offering.24 Similarly,
ITJ424 and 425 seem to be two versions of one text; both are arranged into the
same seven topics. Moreover there is clearly a relationship between the two pairs
of manuscripts, as they share much of the same terminology.
A final, though less certain, example of multiple copies of the same text may
be that of the Narrative Setting and Benefits of Water Offerings (chu gtor kyi gleng
gzhi dang phan yon). This text is found in three manuscripts in the Stein collection:
ITJ343, ITJ344, and ITJ377. The latter, ITJ377, is not written in our identified
hand. Significantly however, the final folio side of this manuscript, which follows
the close of the water offerings text itself, does contain four lines of writing by our
scribe.25 This is a particularly interesting case of our handwriting, as it includes
examples both of our scribe’s headless script common to the manuscripts we have
studied so far, and of his headed script.
A similar headed script also appears in ITJ343, another of our three copies of
the Narrative Setting and Benefits of Water Offerings. Unfortunately ITJ344 is
presently missing from the collection, so our identification of this text is based
solely on the relevant entry in de la Vallée Poussin’s catalogue.26 While we cannot
[PDF]). Given the presence of the Ritual Manual for Water Offerings among the Dunhuang manuscripts,
the work was more likely authored by another Jayasena who lived in the eighth century. Note too that,
given the ritual forms described in the text, we can safely date the work as post-seventh century, so
that its author must also be distinguished from still another Jayasena who is said to have taught Xuanzang
(602-44) during the latter’s visit to India (Martin, “Tibskrit Philology,” 907).
21
rin po che sna brgyad.
22
Rin po che sna brgyad kyi mchod pa.
23
zhi ba’i sbyin sreg, śānti-homa.
24
Zhi ba’i sbyin sreg lag len man ngag ’khol bur phyung ba.
25
In our forthcoming book-length study of the paleographic groups in the Tibetan Dunhuang
collections, we discuss the possible identity of the scribe responsible for ITJ344. We find our present
scribe’s work on a number of manuscripts ascribed to this person, and it seems quite possible that these
two figures knew one another.
26
Louis de la Vallée Poussin, Catalogue of the Tibetan Manuscripts from Tun-Huang in the India
Office Library (Oxford: Oxford University Press, 1962).
be sure ITJ344 is another example of our scribe, the paper size is identical to that
used by ITJ343 (17.5 cm x 6.2 cm), and both have four lines to a page. Moreover,
as we have already seen in ITJ338 (Narrative Setting and Benefits of Stūpas), the
same genre of “narrative setting and benefits” (gleng gzhi dang phan yon) is used
elsewhere by our scribe.
Possible Headed Manuscripts

We have left until now a further speculation regarding the work of our identified
scribe. It seems quite possible that a number of manuscripts written in the headed
script can also be attributed to this same person. For the purposes of this study, we
would like to limit our suggestion to the manuscript ITJ377 (just mentioned) as
well as ITJ457, 584, and Pelliot tibétain 103.
ITJ457 is an unusually small scroll, mostly written in the same style of headed
script as ITJ584. It is an important piece of evidence for identifying the headed
with the headless script of our model scribe, as the handwriting transforms from
one script to the other and back over the course of the manuscript. The clearly
headless sections closely resemble the style of handwriting we have analyzed above
in ITJ425, ITJ340, and so forth, and in this way, ITJ457 provides a kind of bridge
between our group of manuscripts written in headless script and the additional
headed manuscripts being suggested here.
Still other thematic links can be made between ITJ457 and our other manuscripts.
The prayer found in ITJ457’s first item deals with the many sufferings of death
and rebirth and how to avoid these. The same verses are also found written in
Tibetan on the verso of another Khotanese manuscript from Dunhuang, IOL Khot
140. The Khotanese text is a record of a monastic shopping trip to Shazhou. Its
relationship to the Tibetan text is unclear, but is further suggestion of close
relationships between Khotanese and Tibetan Buddhists at Dunhuang.
Like ITJ457, ITJ584 is also a confessional prayer. Matthew Kapstein has
demonstrated that most of its verses are to be found in the Stainless Confession
Tantra (dri med bshags rgyud), a tantra of the Na rag dong sprugs cycle.27 Kapstein
suggests that the prayer in ITJ584 may have been one of the source texts used by
the compilers of the Stainless Confession Tantra. He goes on to suggest a possible
connection between this headed manuscript (ITJ584) and another headless
manuscript, ITJ318. The latter is, once again, an example of our scribe’s hand, but
Kapstein links the two manuscripts on the basis of their contents. ITJ318 describes
a maṇḍala in the form of a 108-petalled lotus, a maṇḍala, Kapstein shows, that
can also be linked with the Na rag dong sprugs cycle. In this way ITJ584 (in headed)
27
See Matthew T. Kapstein, “La formation du Bouddhism tibétain à travers les documents de
Dunhuang” [in English; summary of the January 2001 EPHE Vème Section Lecture Series], IDP News,
no. 17 (Winter, 2001), article 3, http://idp.bl.uk/archives/news17/idpnews_17.a4d#3; and Matthew T.
Kapstein, “Between Na-rak and a Hard Place: Evil Rebirth and the Violation of Vows in Early
Rnying-ma-pa Sources and their Dunhuang Antecedents,” in Aspects of Esoteric Buddhism at Dunhuang,
ed. Matthew T. Kapstein and Sam van Schaik (Leiden: Brill, forthcoming).
and ITJ318 (in headless) both contain materials that were later to be incorporated
into the Na rag dong sprugs.
Site Numbers
In addition to the thematic and stylistic links discussed so far, there is still another
way in which our identification of this manuscript group is reinforced. When Stein
first discovered the Dunhuang manuscripts, he assigned site numbers to them,
marking each according to where he found it. Questions remain about how precisely
we should understand these numbers.28 For now, however, we can point to a striking
fact about the site numbers borne by the Stein manuscripts discussed in the present
article: all of them come from the same bundle, labeled Ch.73.iii. This would seem
to indicate that all of them were placed in the cave together, further evidence that
they are the work of a single author.
Speculative Conclusions
With this article we have introduced a new paleographic approach to the Tibetan
manuscripts from Dunhuang. We have attempted to show how techniques from
the field of forensic handwriting analysis can be adapted for the ancient Tibetan
handwriting seen in the Dunhuang manuscripts. This approach can be applied to
any number of manuscripts, but for the purposes of this introduction we have
focused on a single, relatively tightly defined group of manuscripts that we believe
were penned by the same hand. A more extensive, book-length study of several
other handwriting groups that we have identified within the Dunhuang collections
is forthcoming. For now, however, we close by outlining some of the wider
consequences that result from this new approach, consequences for our
understanding of the collection as a whole.
The identification of groups of manuscripts sharing the same hand significantly
alters the shape of the Tibetan Dunhuang collections. No longer are we confronted
with an overwhelming mass of manuscripts. Rather, we can begin to make sense
of large swathes of the collection by dividing it into a relatively small number of
manuscript groups. This, in turn, allows us to assign scribal names to many of the
Dunhuang manuscripts; the Tibetan Dunhuang manuscripts are only rarely signed,
but we need only a single signature on one manuscript to apply it to all manuscripts
in the same hand. From the names, we can often identify the cultural affiliation of
the writer, which reveals much about the shape of the Buddhist communities around
Dunhuang during the relevant period.
Similarly, few Tibetan items are dateable, but now once a single item has been
dated a range of manuscripts can be dated to roughly the same period. This helps
to correct a long-standing misconception regarding the dating of the Dunhuang
collection. For the past century scholars have commonly suggested that most of
28
For a preliminary analysis of Stein’s Tibetan site numbers, see Takeuchi, “Old Tibetan Buddhist
Texts.”
the Tibetan manuscripts probably date from around the time of the Tibetan
occupation of Dunhuang (c. 781-848 CE).29 Takeuchi has recently started to
question this assumption by compiling a still-growing list of Tibetan Dunhuang
manuscripts that can be firmly dated to the tenth century.30 Our paleographic
analysis of the manuscripts strengthens the case for dating large portions of the
collection to the tenth century, as the dateable manuscripts can be linked to many
more manuscripts in the same hands. We may further speculate that a large number
of the manuscripts are the work of a relatively small number of scribes, many of
whom may have known and worked alongside one another during the same
historical period.31
Reading the manuscripts in this light also allows us to distinguish the interests
particular to each scribe; each manuscript group reveals a surprisingly distinct set
of concerns. The sample scribe discussed here, for example, specialized in external,
small-scale ritual texts relating to rosaries, stūpas, water and fire offerings, and so
forth. Elsewhere we have shown how reading multiple works by a single scribe
can reveal links between texts and topics that would otherwise be considered
distinct and unrelated.32 The present article is offered in the hope that at least some
of the ideas contained here might help other scholars to make similar discoveries
about the religious and social milieu behind these ancient treasures from Dunhuang.
29
The date of the Tibetan conquest of Dunhuang remains disputed. See Bianca Horleman, “A
Re-evaluation of the Tibetan Conquest of Eighth-Century Shazhou/Dunhuang,” in Tibet, Past and
Present: Tibetan Studies 1, ed. Hank Blezer (Leiden: Brill, 2002), 49-65, for a summary of previous
arguments and an argument that the conquest took place earlier than has previously been suggested,
in the 760s.
30
Takeuchi, “Old Tibetan Buddhist Texts,” builds on the earlier work of Géza Uray, “L’emploi du
Tibétain dans les Chancelleries des États du Kan-sou et de Khotan postérieurs à la Domination
Tibétaine,” Journal Asiatique, no. 269 (1981): 81-90. See also Géza Uray, “New Contributions to
Tibetan Documents from the post-Tibetan Tun-huang,” in Tibetan Studies: Proceedings of the Fourth
Seminar of the International Association for Tibetan Studies, ed. Helga Uebach and Jampa L. Panglung,
515-28 (Munich: Kommission für Zentralasiatische Studien, Bayerische Akademie der Wissenschaften,
1988).
31
Further evidence for these claims will be elaborated in our forthcoming book-length paleographic
study of the Tibetan Dunhuang manuscripts.
32
See Sam van Schaik and Jacob Dalton, “Where Chan and Tantra Meet: Buddhist Syncretism in
Dunhuang,” in The Silk Road: Trade, Travel, War and Faith, ed. Susan Whitfield (London: The British
Library, 2004), where we demonstrate a clear case of one scribe’s synthesis of Mahāyoga and Chinese
Chan, two Buddhist traditions normally held never to have met.
Glossary
Ka
bka’ ’gyur Kangyur Textual
Collection
rkang ba kangwa leg Term
Ga
gleng gzhi dang phan lengzhidang penyön narrative setting Term
yon and benefits
mgo go head Term
Ca
chu gtor kyi gleng Chutorkyi Narrative Setting Text
gzhi dang phan yon Lengzhidang Penyön and Benefits of
Water Offerings
chu gtor gyi cho ga Chutorgyi Choga Ritual Manual for Text
Water Offerings
mchod rten gyi gleng Chötengyi Narrative Setting Text
gzhi dang phan yon Lengzhidang Penyön and Benefits of
Stūpas
’chu gtord gyi gzhung Chutorgyi Zhung Exposition on Text
Water Offerings
Tha
thub pa a ba da ra Tuppa Abadara Descent of the Sage San. *Muni- Text
[Tantra] avatāra[-tantra]
Da
dri med bshags rgyud Drimé Shakgyü Stainless Text
Confession Tantra
Na
na rag dong sprugs Narak Dongtruk Doxographical
Category
Pa
po ti poti Indian-style book San. pustaka Term
Pha
’phreng ba trengwa rosary San. māla Term
Ba
dba’ manydzu Wa Mandzu Person
dbu can uchen headed Term
dbu med umé headless Term
rba manydzu Ba Mandzu Person
sba manydzu Ba Mandzu Person
sbyin sreg jinsek fire ritual; fire San. homa Term
offering
Tsha
tsheg tsek dot Term
Zha
zhabs kyu zhapkyu “u” vowel Term
zhi ba’i sbyin sreg zhiwé jinsek peaceful fire San. śānti-homa Term
offering
zhi ba’i sbyin sreg lag Zhiwé Jinsek Laklen Abridged Text
len man ngag ’khol Menngak Khölbur Instructions on the
bur phyung ba Chungwa Aspects of the
Peaceful Fire
Offering; Pith
Instructions on the
Practice of the
Peaceful Fire
Offering Arranged
into Sections
Ya
ya brtags yatak subscribed ya Term
Ra
rin chen bzang po Rinchen Zangpo 958-1055 Person
rin po che sna brgyad rinpoché nagyé eight jewels Term
rin po che sna brgyad Rinpoché Nagyekyi An Offering of the Text
kyi mchod pa Chöpa Eight Jewels
Sha
sha cu Shachu Chi. Shazhou Place
shad shé phrase-marker Term
Sa
sum pa lo tsā ba dar Sumpa Lotsāwa Person
ma yon tan Darma Yönten
Sanskrit
Baśrīmajsu Person
Jayasena Person
maṇḍala Term
Great Yoga Mahāyoga Doxographical
Category
stūpa Term
sūtra Term
Diamond Vehicle Vajrayāna Doxographical
Category
Vimalamitra eighth Person
century
Chinese
Wylie Phonetics English Chinese Dates Type
Chan Doxographical
Category
Dunhuang Place
Longmen Cave
Mogao Cave
Sanjie Monastery Sanjie si Monastery
Yungang Cave
Xuanzang 602-644 Person
Bibliography
Dunhuang Manuscripts Discussed
IOL Tib J 318, 321, 338, 340, 341, 343, 344, 377, 407, 422, 423, 424, 425, 457,
584, 594, 647, 688. (Note: in the body of the article ITJ is an abbreviation of
the full pressmark IOL Tib J.)
Pelliot tibétain 103, 849.
IOL Khot 55, 140.
Canonical Primary Sources

Asc. Jayasena. Chu gtor gyi cho ga zhes bya ba. Q. 4593.
Secondary Sources
Akira, Fujieda. “Chronological Classification of Dunhuang Buddhist Manuscripts.”
In Whitfield, Dunhuang Manuscript Forgeries, 103-14.
Dalton, Jacob, and Sam van Schaik. Tibetan Tantric Manuscripts from Dunhuang:
A Descriptive Catalogue of the Stein Collection at the British Library. Leiden:
Brill, 2006.
de la Vallée Poussin, Louis. Catalogue of the Tibetan Manuscripts From
Tun-Huang in the India Office Library. Oxford: Oxford University Press, 1962.
Dpa’ ris sangs rgyas. Bod yig ’bri tshul mthong ba kun smon. Beijing: Mi rigs
dpe skrun khang, 1997.
Hackin, Joseph. Formulaire Sanskrit-Tibétain du Xe siècle. Paris: Librairie
Orientaliste Paul Geuthener, 1924.
Horleman, Bianca. “A Re-evaluation of the Tibetan Conquest of Eighth-Century
Shazhou/Dunhuang.” In Tibet, Past and Present: Tibetan Studies 1, edited by
Hank Blezer, 49-65. Proceedings of the Ninth Seminar of the International
Association for Tibetan Studies, Leiden, 2000. Leiden: Brill, 2002.
Kapstein, Matthew T. “La formation du Bouddhism tibétain à travers les
documents de Dunhuang” [in English; summary of the January 2001 EPHE
Vème Section Lecture Series]. IDP News, no. 17 (Winter, 2001), article 3,
http://idp.bl.uk/archives/news17/idpnews_17.a4d#3.
———. “Between Na-rak and a Hard Place: Evil Rebirth and the Violation of
Vows in Early Rnying-ma-pa Sources and their Dunhuang Antecedents.” In
Aspects of Esoteric Buddhism at Dunhuang, edited by Matthew T. Kapstein
and Sam van Schaik. Leiden: Brill (forthcoming).
Martin, Dan. “Tibskrit Philology.” Kurt Keutzer, 2006.
http://www.eecs.berkeley.edu/~keutzer/martin/TibskritUni.pdf (PDF).
Rong Xinjiang. “The Nature of the Dunhuang Library Cave and the Reasons for
its Sealing.” Cahiers d’Extrême-Asie 11 (1999-2000): 247-75.
Sharf, Robert. “The Enigma of the Dunhuang Caves.” Paper presented at the
annual meeting of the American Academy of Religion, San Antonio, TX,
November 22, 2004.
Skjærvø, Prods Oktor. Khotanese Manuscripts from Chinese Turkestan in the
British Library: A Complete Catalogue with Texts and Translations. London:
The British Library, 2002.
Scherrer-Schaub, Cristina A. “Towards a Methodology for the Study of Old
Tibetan Manuscripts: Dunhuang and Tabo.” In Tabo Studies 2: Manuscripts,
Texts, Inscriptions and the Arts, edited by Cristina A. Scherrer-Schaub and
Ernst Steinkellner, 3–36. Rome: Istituto Italiano per l’Africa e l’Oriente, 1999.
Scherrer-Schaub, Cristina A., and George Bonani. “Establishing a Typology of
the Old Tibetan Manuscripts: A Multidisciplinary Approach.” In Whitfield,
Dunhuang Manuscript Forgeries, 184-215.
Stein, Rolf A. “Un genre particulier d’exposés du tantrisme ancien tibétain et
khotanais.” Journal Asiatique, no. 275 (1987): 265-82.
Takata, Tokio. “Multilingualism in Tun-huang.” Acta Asiatica: Bulletin of the
Institute of Eastern Culture, no. 78 (2000), Tun-huang and Turfan Studies:
49–70.
Takeuchi, Tsuguhito. “Old Tibetan Buddhist Texts from the Post-Tibetan Imperial
Period (mid-9 C. to late 10 C.).” In Proceedings of the Tenth Seminar of the
International Association for Tibetan Studies. Leiden: Brill (forthcoming).
Tharchin, Rev. G. The Tibetan Primer of Current Hand Writing. Kalimpong:
Tibet Mirror Press, 1970.
Uray, Géza. “L’emploi du Tibétain dans les Chancelleries des États du Kan-sou
et de Khotan postérieurs à la Domination Tibétaine.” Journal Asiatique, no.
269 (1981): 81-90.
———. “New Contributions to Tibetan Documents from the post-Tibetan
Tun-huang.” In Tibetan Studies: Proceedings of the Fourth Seminar of the
International Association for Tibetan Studies, edited by Helga Uebach and
Jampa L. Panglung, 515-28. Munich: Kommission für Zentralasiatische Studien,
Bayerische Akademie der Wissenschaften, 1988.
van Schaik, Sam, and Jacob P. Dalton. “Where Chan and Tantra Meet: Buddhist
Syncretism in Dunhuang.” In The Silk Road: Trade, Travel, War and Faith,
edited by Susan Whitfield. London: The British Library, 2004.
Whitfield, Susan, ed. Dunhuang Manuscript Forgeries. London: The British
Library, 2002.
“Emperor” Mu rug btsan and the ’Phang thang ma
Catalogue
Brandon Dotson
School of Oriental and African Studies
Abstract: Confusion shrouds the events surrounding the death of Emperor Khri
Srong lde btsan (742-c. 800) and the succession of his sons at the turn of the ninth
century. Tibetan religious histories, Old Tibetan sources, and Chinese sources
offer conflicting pictures of the order of events and the identities of those involved.
Fortunately, a newly published source, the ’Phang thang ma Catalogue, throws
new light on the royal succession by referring to Mu rug btsan, the elder brother
of Khri Lde srong btsan, as nothing less than an emperor (btsan po). Considering
this new information alongside Old Tibetan inscriptional evidence, this article
attempts to establish the order of events around Khri Srong lde btsan’s abdication
and death, and to locate “Emperor Mu rug btsan” within them.
Introduction
The royal succession of the Tibetan emperors (btsan po) sets out the signposts for
early Tibetan history and its periodization.1 The order of succession is given in the
Royal Genealogy (PT 1286), a document that properly forms a part of Tibet’s first
royalist epic history, the Old Tibetan Chronicle. The regnal dates are clear from
the entries in the Old Tibetan Annals, which records the births, deaths, and
enthronements of the Tibetan emperors with considerable precision. The yearly
entries, however, only begin in earnest in 650, after the death of Khri Srong btsan,
alias Srong btsan sgam po (d. 649), and end in the year 764, with a few entries
missing in between. Therefore the regnal dates for those who ruled during the
second half of the Tibetan Empire must be established with recourse to pillar edicts,
other Old Tibetan documents, Chinese sources such as the Tang Annals, and
post-dynastic Tibetan religious histories. Despite this relative abundance of sources
1
Part of the research for this article was undertaken with the support of a Fulbright Grant to conduct
research in China and Tibet. I here express my appreciation to the Institute of International Education
for its support. I would also like to express my gratitude to Prof. Cristina Scherrer-Schaub, whose
comments on an earlier draft of this article significantly improved its form.
1550-6363/2007/3/T3105.
© 2007 by Brandon Dotson, Tibetan and Himalayan Digital Library, and International Association of Tibetan Studies.
Dotson: “Emperor” Mu rug btsan and the ’Phang thang ma Catalogue 2
through which to triangulate regnal dates, some tenacious problems persist

unresolved. Notable among these are the dates for the birth and enthronement of
Srong btsan sgam po and the names and dates of the sons and successors of Khri
Srong lde btsan (742-c. 800). Here I will attempt to unravel the latter problem in
light of a newly published source, the ’Phang thang ma Catalogue.2
The ’Phang thang ma Catalogue

Before moving on to an examination of the problems surrounding the succession
to Khri Srong lde btsan (742-c. 800), it will be useful to examine the nature of our
newly published source, the ’Phang thang ma Catalogue. The ’Phang thang ma
Catalogue is one of the three catalogues of Buddhist texts compiled during the
latter part of the Tibetan Empire. When Khri Srong lde btsan officially adopted
Buddhism in c. 779, this emperor established official sponsorship for the support
of such monasteries and temples as the Ra sa ’phrul snang (the Jo khang), the Ra
mo che, Bsam yas Monastery, and Khra brug Monastery. The emperor also traveled
with his own personal saṃgha at the Tibetan court.3 Further, this emperor
established at the court a Buddhist council (literally, council of the representative(s)
of the bhagavat at the court) and a committee for the translation of the dharma.4
This is evident from the Tabo fragments of the Sgra sbyor bam po gnyis pa, a
document that formed part of a project to regulate and standardize the translation
of Buddhist texts.5 In conjunction with this project, the Tibetan emperors sponsored
2
Rta rdo, ed., Dkar chag ’phang thang ma/ sgra sbyor bam po gnyis pa (Beijing: Mi rigs dpe skrun
khang, 2003).
3
pho brang ’khor gyi dge ’dun. These circumstances are evident from Khri Srong lde btsan’s Bsam
yas Edict, preserved in the Mkhas pa’i dga’ ston of Dpa’ bo gtsug lag phreng ba (Dpa’ bo gtsug lag
phreng ba, Dam pa’i chos kyi ’khor lo bsgyur ba rnams kyi byung ba gsal bar byed pa mkhas pa’i dga’
ston [Beijing: Mi rigs dpe skrun khang, 1985], 372; mKhas-paḥi-dgaḥ-ston by
Dpaḥ-bo-gtsug-lag-phreng-ba, ed. Lokesh Chandra, Śatapiṭaka Series 9 [4] [New Delhi: International
Academy of Indian Culture, 1965], 109b.3). For ease of reference, citations of this text employ both
the Beijing edition (abbreviated as Mkhas pa’i dga’ ston B) and the Delhi version (abbreviated as
Mkhas pa’i dga’ ston D).
4
pho brang du bcom ldan ’das kyi ring lugs [kyi mdun sa] dang / dar ma bsgyur ba’i lo cha pa’i
sgra/. This quotation is taken from the Tabo fragments of the Sgra sbyor bam po gnyis pa, which date
to the reign of Khri Srong lde btsan (756-c. 797, c. 798-c. 800). The bracketed text, kyi mdun sa, is
from a later version of the Sgra sbyor bam po gnyis pa assembled under Khri Lde srong btsan in 814.
For a lexical comparison of this passage in both versions, see Jampa Panglung, “New Fragments of
the sGra-sbyor bam-po gnyis-pa,” East and West 44, no. 1 (1994): 170 and Cristina A. Scherrer-Schaub,
“Enacting Words: A Diplomatic Analysis of the Imperial Decrees (bkas bcad) and Their Application
in the sGra sbyor bam po gnyis pa Tradition,” Journal of the International Association of Buddhist
Studies 25, nos. 1-2 (2002): 288. The Tabo version dates to either 783 or 795, and the merit of these
respective dates is discussed in Scherrer-Schaub, “Enacting Words,” 289-92. The Ldan dkar ma
Catalogue also attributes its compilation to the “council of the representative(s) of the bhagavat at the
court” (bcom ldan ’das kyi ring lugs kyi mdun sa; Marcelle Lalou, “Les Textes Bouddhiques au temps
du Roi khri-sroṅ-lde-bcan,” Journal Asiatique 241 [1953]: 319). The Old Tibetan text is only lightly
edited here in order to retain the original orthographies, and corrections are offered only where the
intended meaning is not readily apparent. In accordance with THDL and JIATS protocols, the reverse
gi gu, which is commonly transliterated as a capital “I,” is rendered as a lower case “i” preceded by a
dash, i.e., “-i.”
5
For an excellent overview of this project, see Scherrer-Schaub, “Enacting Words.”
three catalogues of Buddhist texts, namely the Lhan dkar ma Catalogue/Ldan dkar
ma Catalogue, the Mchims phu ma Catalogue, and the ’Phang thang ma Catalogue.
The Ldan dkar ma Catalogue is found in the Bstan gyur, and has been studied by
Lalou, among others.6 The other two catalogues, the Mchims phu ma Catalogue
and the ’Phang thang ma Catalogue, were long thought to have disappeared.
Fortunately, however, the latter has recently come to light and was published in
Beijing in 2003 together with a copy of the Sgra sbyor bam po gnyis pa from the
Fifth Dalai Lama’s library at ’Bras spungs Monastery.
In his recent treatment of the ’Phang thang ma Catalogue, Halkias demonstrates
that the ’Phang thang ma Catalogue was compiled after the compilation of the
Ldan dkar ma Catalogue.7 Specifically, Halkias claims that the compilation of the
’Phang thang ma Catalogue commenced sometime during the reign of Khri ’U’i
dum brtan (r. 841-42)8 or that of Khri ’Od srung (842/43-c. 890), but that it was
“an open register which, being the last of its kind, was susceptible to scriptural
augmentation yielding several variants from the second transmission of Buddhism
(phyi dar) onwards.”9 The late date of the ’Phang thang ma Catalogue’s initial
compilation is evident, Halkias notes, from the text’s mention of Emperor ’U’i
6
Lalou, “Les Textes Bouddhiques.”
7
Georgios Halkias, “Tibetan Buddhism Registered: A Catalogue from the Imperial Court of ’phang
thang,” The Eastern Buddhist 36, nos. 1-2 (2004): 54-55. Concerning the date of the Ldan dkar ma
Catalogue, it was composed in a dragon year, when the court resided at Lhan dkar, which many scholars
identify as 812. It is clear, however, that a later date cannot be ruled out, and indeed Yamaguchi dates
its compilation to 824 (Zuiho Yamaguchi, “The Fiction of King Dar Ma’s Persecution of Buddhism,”
in De Dunhuang au Japon: Études Chinoises et Bouddhiques Offertes à Michel Soymié, ed. Jean Pierre
Drège [Geneva: Librairie Droz, 1996], 243 n. 15). The east inscription at Zhwa’i lha khang offers the
strongest reason for abandoning 812 as the date of this catalogue, since this inscription is dated to “the
benevolent later dragon year, when the court resided at ’On cang do” (bka’ drin ’brug gi lo phyi ma
la/ pho brang ’on cang do na bzhugs pa’-i tshe/ [lines 22-23]; infra n. 39; Fang Kuei Li and W. South
Coblin, A Study of the Old Tibetan Inscriptions [Taipei: Institute of History and Philology, Academica
Sinica, 1987]). This corresponds to 812, and according to the imperial dating formula, which names
the site of the court or council to distinguish one year in the twelve year cycle from another, this year
would differ from the dragon year when the court resided at Lhan dkar whence the Lhan dkar ma
Catalogue was compiled. This does not completely rule out 812 as the year of the Lhan dkar ma
Catalogue’s compilation, however, as the royal court was stationed at different sites in summer and
winter.
The Dunhuang document PT 1085 may in fact be the key for dating the dragon year of the Lhan
dkar ma Catalogue’s compilation. PT 1085 is a response to the town prefect (rtse rje) of Sha cu,
dispatched from the court at Lhan dkar in the middle winter month of the dragon year (pho brang lhan
kar nas// ’brug-i lo ’-i dgun sla ’bring po ’-i ngo la bkye ’i phyag rgya phogs te; PT 1085, line 1). This
is, no doubt, the same dragon year mentioned in the preamble to the Ldan dkar ma Catalogue. The
reply from the court is “signed” by Minister Khrom bzher and Lha bzang gsas mthong (blon khrom
bzher dang / blon lha bzang gsas mthong g-i sug rgya ’og nas gthad; PT 1085, lines 8-9). While the
former name is too common to be of any use, the latter, once identified, may provide a clue concerning
the date of this dragon year. For a further discussion of the date of the Lhan dkar ma Catalogue, see
Halkias, “Tibetan Buddhism Registered,” 48 n. 4, 51-54.
8
On the validity of these dates, see Yamaguchi, “The Fiction of King Dar Ma’s Persecution,” 250
and Yoshiro Imaeda, “Chinese Texts on the Tibetan King Glang Dar ma,” Bukkyô-gaku Seminâ (Buddhist
Seminar) 74 (2001): 31.
9
Halkias, “Tibetan Buddhism Registered,” 77.
Dum brtan, alias Glang dar ma.10 Under the misleading heading “commentaries
on sūtras and śāstras by Khri Srong lde btsan,”11 the ’Phang thang ma Catalogue
lists texts attributed to emperors, queens, and monks.12 Among these is “Emperor
Dba’ dun brtan’s edict on Madhyamaka with accompanying commentary.”13
“Emperor Dba’ Dun brtan” is an obvious error for Dpal dum brtan. Remarkably,
Yamaguchi, in reading a citation of the ’Phang thang ma Catalogue by Bu ston,
already noted the importance of the ’Phang thang ma Catalogue’s attribution of
this Madhyamaka commentary to ’U’i dum brtan: it indicates that “the ’Phang
thang ma Catalogue must have been compiled after his reign.”14 Indeed, unless
we take this particular entry in the catalogue as a later addition – and there seems
little or no reason to do so – this places the date of the catalogue’s compilation in
the reign of Khri ’U’i dum brtan or that of his successor, Khri ’Od srung, at the
earliest. Assuming this to be the case, the earliest possible date for the ’Phang
thang ma Catalogue, compiled in a dog year, is 842. This does in fact harmonize
quite well with what we know of ’Od srung, who resided in ’Phang thang for much
of his reign, and eventually died there. Further, the ’Phang thang ma Catalogue’s
attribution of a Madhyamaka commentary to Glang dar ma ’U’i dum brtan adds
weight to the growing body of work that calls into question Glang dar ma’s
supposed opposition to Buddhism.15
Near the end of the ’Phang thang ma Catalogue, what might be called an
appendix reveals the methodology by which the catalogue was assembled. This
has already been discussed by Halkias, so I will only summarize briefly before
moving onto a finer point.16 The heading of this section reads:
Sūtras and teachings whose titles do not quite correspond after comparison with
three sources, but correspond in two sources, are mostly added to the ad hoc
10
Halkias, “Tibetan Buddhism Registered,” 57-58.
11
mdo sde dang bstan bcos kyi ti ka/ btsan po khri srong lde btsan gyis mdzad pa/.
12
Rta rdo, Dkar chag ’phang thang ma, 54.
13
btsan po dba’ dun brtan gyi dbu ma’i bka’ bcad bshad pa dang bcas pa (Rta rdo, Dkar chag ’phang
thang ma, 57).
14
Yamaguchi, “The Fiction of King Dar Ma’s Persecution,” 243.
15
See Yamaguchi, “The Fiction of King Dar Ma’s Persecution,” Cristina A. Scherrer-Schaub, “Prières
pour un apostat: fragments d’histoire Tibétaine,” Cahiers d’Extrême-Asie 11 (1999-2000): 217-46,
Samten G. Karmay, Btsan po lha sras dar ma dang de’i rjes su byung ba’i rgyal rabs mdor bsdus
(Dharamsala: Library of Tibetan Works and Archives, 1986), and Samten G. Karmay, “King Lang
Darma and His Rule,” in Tibet and Her Neighbours: A History, ed. Alex McKay (London: Hansjörg
Mayer, 2003), 57-66.
16
Halkias, “Tibetan Buddhism Registered,” 75-76.
(dkyus)17 [catalogue]. Those that appear in one source, but not in two, are recorded
elsewhere.18
The above passage reveals that the compilers of the ’Phang thang ma Catalogue
triangulated titles by comparing three main sources (dpe). The identity of one of
these sources is revealed in the next sub-heading: “[titles] found in Bandhe Gzhon
nu snying po’s source (dpe), but not found in the two old sources.”19 The other
two sources are referred to as the “old sources,” and, given the late date of the
’Phang thang ma Catalogue, it is tempting to follow Halkias and understand these
“sources” as none other than the Ldan dkar ma Catalogue and the Mchims phu ma
Catalogue.20 Such a conclusion may, however, be premature at present.
Gzhon nu snying po’s “source” is most likely a catalogue held by this Tibetan
master. Fortunately, Gzhon nu snying po is mentioned in other Old Tibetan sources.
IOL Tib J 689 (2), a Dunhuang document translated by Karmay, lists, by region,
various lineages of kalyāṇamitras in Tibet.21 The first of these enumerates the
“kalyāṇamitras of Bsam yas Monastery and ’Phrul snang upwards.”22 Famously,
Lha lung Dpal gyi rdo rje, the supposed assassin of ’U’i dum brtan, fills the ninth
place in this list, and this has prompted numerous scholars to regard him,
17
The term dkyus in the phrase dkyus su stsal is somewhat puzzling here, but the above reading can
be confirmed with recourse to the Sgra sbyor bam po gnyis pa. The term usually means “race” or
“crooked,” but the Sgra sbyor bam po gnyis pa states that when a new term is submitted for approval,
once it is accepted by the council of the representative of the bhagavat at the court and the committee
for the translation of the dharma, it is added to the dkyus/skyus of the catalogue (dkar chag gi dkyus
su bsnan do). Panglung (“New Fragments,” 165), presumably reading dkyus as dkyus ma, meaning
“common” or “everyday,” translates dkar chag gi dkyus as the “general register.” This is plausible, but
the grammar suggests that dkyus is a part of the catalogue, and not a description of it. While it is tempting
to read this term as “appendix” or “addenda,” it is perhaps more prudent simply to understand dkyus
as indicating an ad hoc catalogue, in this case of the holdings of the Tibetan court’s library of Buddhist
texts, to which titles could be added at any time. The relationship between this ad hoc catalogue (dkyus)
and the formal catalogues is uncertain. The former could be a sort of limbo for texts awaiting ratification.
Another possibility is that the formal catalogues – the Ldan dkar ma Catalogue, Mchims phu ma
Catalogue, and ’Phang thang ma Catalogue – were the updated hand-lists of the Tibetan court’s
holdings, named after the court sites at which the cataloguing was undertaken. The Ldan dkar ma
Catalogue, Mchims phu ma Catalogue, and ’Phang thang ma Catalogue would then be successive
catalogues of the court library, or updated versions of essentially the same document. This would
account for the fact that the ’Phang thang ma Catalogue contains nearly all of the texts listed in the
Ldan dkar ma Catalogue, and is supported by the method of the catalogue’s compilation as described
above. Skilling, it should be noted, already advanced the germ of such an interpretation (Peter Skilling,
“From bKa’ bstan bcos to bKa’ ’gyur and bsTan ’gyur,” in Transmission of the Tibetan Canon: Papers
Presented at a Panel of the 7th Seminar of the International Association for Tibetan Studies, Graz
1995, ed. Helmut Eimer [Wien: Verlag der Österreichischen Akademie der Wissenschaften, 1997],
92).
18
gsung rab mdo sde rnams kyi mtshan byang dpe gsum la gtugs nas nyung shas mi mthun pa dpe
gnyis las mthun par ’byung ba rnams ni mang brtsan du byas nas dkyus su stsal/ dpe gcig las ’byung
ste gnyis las ma byung ba rnams gud du bris pa/ (Rta rdo, Dkar chag ’phang thang ma, 49-50).
19
dpe rnying rnams gnyis las ni ma byung / bandhe gzhon nu snying po’i dpe las byung ba/ (Rta
rdo, Dkar chag ’phang thang ma, 50).
20
Halkias, “Tibetan Buddhism Registered,” 75.
21
Samten Gyaltsen Karmay, The Great Perfection: A Philosophical and Meditative Teaching of
Tibetan Buddhism (Leiden: Brill, 1988), 78-80.
22
bsam yas dang ’phrul snang yan cad kyi dge ba’i shes nyen/.
erroneously, as “the ninth abbot of Bsam yas Monastery.” Less famously, Gle’u
Gzhon nu snying po occupies the place immediately preceding as the eighth in the
lineage of “kalyāṇamitras of Bsam yas Monastery and ’Phrul snang upwards.”23
As Karmay pointed out in his analysis of this list, it is a temporal lineage (rgyud),
so it necessarily covers several decades. This fact, coupled with Gzhon nu snying
po’s proximity to Lha lung Dpal gyi rdo rje, places his tenure in the late part of
the reign of Khri Gtsug lde brtsan (r. 815-41), which is to say approximately the
mid-830s. Assuming that we are dealing with the same Gzhon nu snying po whose
“source” was used for the compilation of the ’Phang thang ma Catalogue, we can
locate this source temporally, and perhaps doctrinally as well, since Gle’u Gzhon
nu snying po is also found in Gnubs Sangs rgyas ye shes’s Bsam gtan mig sgron
as the author of a meditation manual (sgom lung), or manual for “simultaneist”
(cig car ba) meditation.24
23
The “office” of kalyāṇamitra is explained by a passage in Khri Lde srong btsan’s Skar chung Edict,
preserved in the Mkhas pa’i dga’ ston, and in a nearly identical passage on the Skar chung Pillar itself
(lines 33-43; Li and Coblin, A Study of the Old Tibetan Inscriptions, 319, 327). It is evident from this
passage that the kalyāṇamitra is a privileged teacher of Buddhism, appointed as a personal “spiritual
friend” to the emperor, and is above the rank of “representatives of the bhagavat” (bcom ldan ’das kyi
ring lugs). Their geographical jurisdiction over Bsam yas Monastery and the Jo khang in IOL Tib J
689 (2) also indicates that those in this list had some authority at these places, but does not necessarily
mean that they served as abbots. The passage from the Skar chung Edict in the Mkhas pa’i dga’ ston
is as follows:
The emperor’s successors will appoint from among the monks kalyāṇamitras [to serve them] from
their youth until they act as rulers. Whatever dharma one puts in his head, he should learn it, and
all of Tibet should open the door to the study and practice of the dharma. Never will the door to
liberation be blocked to the Tibetan subjects, from the upper classes on down, but the faithful will
be established in liberation. Those among them who are able will be permanently appointed as
representatives of the bhagavat. Never will those who act as representatives of the bhagavat and
accord with the word of the teachings in all actions, and accord with the work and responsibilities
of the teachings not be appointed to act as kalyāṇamitras (btsan pho/ dbon sras sku chung ngur
bzhugs pa yan chad/ chab srid kyi mnga’ bdag mdzad pa man chad kyang dge slong las dge ba’i
bshes gnyen bsgos te/ chos thugs su ci chud chud du bslab jing / bod yongs kyis kyang chos slob
cing spyad pa’i sgo bcad ste/ nam du yang bod ya rabs mand cad bod ’bangs las thard bar gzung
ba’i sgo mi bgag par/ dad pa’i rnams las thar par btsud ste/ de’i nang nas nus pa las/ bcom ldan
’das kyi ring lugs rtag tu bsko zhing / bcom ldan ’das kyi ring lugs byed pa’i rnams/ chos ’khor nas
bya’o cog gi bka’ la yang gtogs te/ chos ’khor gyi las dang dbang byed cing / dge ba’i bshes gnyen
byed par myi bsko re/; Dpa’ bo gtsug lag phreng ba, Mkhas pa’i dga’ ston B, 410-11; Mkhas pa’i
dga’ ston D, 129a.7-129b.3).
See also the discussion of this term in Michael Walter, “The Significance of the Term ring lugs:
Religion, Administration, and the Sacral Presence of the btsan-po,” Acta Orientalia Scientiarum
Hungaricae 51, no. 3 (1998): 312-13 n. 3.
24
Karmay, The Great Perfection, 103. On the accuracy of “simultaneist” as a translation of cig car
ba, as opposed to “instantaneist” or “subitist,” see Rolf A. Stein, “Sudden Illumination or Simultaneous
Comprehension: Remarks on Chinese and Tibetan Terminology,” in Sudden and Gradual: Approaches
to Enlightenment in Chinese Thought, ed. Peter N. Gregory (Honolulu: University of Hawaii Press,
1987), 41-66. That such strains of Buddhism persisted in Tibet long after the “Council of Tibet” is well
known, and the ’Phang thang ma Catalogue also provides indirect iconographic evidence of this. On
the first page of the ’Phang thang ma Catalogue manuscript, the unknown redactor reveals that he
made his copy from an illustrated manuscript. Among the captioned illustrations in the original, he
states, were those of Nāgārjuna, Asaṅga, Vasubandhu, Candrakīrti, Śāntarakṣita, Kamalaśīla,
Vimalamitra, Pad ma ’byung gnas, and “the great Chan teacher Hwa shang ma hwa ya na” (bsam gtan
gyi mkhan po chen po hwa shang ma hwa ya na; Rta rdo, Dkar chag ’phang thang ma, 2), who was,
of course, the famous representative of the “simultaneist” (cig car ba) tradition at the “Council of
Having discussed the nature of the ’Phang thang ma Catalogue and demonstrated
its source value, it is now possible to consider its contribution to the history of the
royal lineage and the chronology of the succession after Khri Srong lde btsan.
Contributions of the ’Phang thang ma Catalogue to the History

of the Royal Lineage
The chronology of events during the reign of Khri Srong lde btsan poses some
very important problems concerning the history of the Tibetan Empire. Chief
among these are the dates for the foundation of Bsam yas Monastery, and the dates
for the Bsam yas Inscription and the accompanying edicts preserved in the Mkhas
pa’i dga’ ston, as these announce the adoption of Buddhism as the official religion
of Tibet. These are particularly difficult to determine because the two major sources
for the history of the Tibetan Empire, the Old Tibetan Annals and the Old Tibetan
Chronicle, both end during the first part of Khri Srong lde btsan’s reign. Concerning
the royal succession itself, the dates of Khri Srong lde btsan’s death are also unclear,
as are the regnal dates of his short-lived successor, Mu ne btsan/Mu ne btsan po.
Many post-dynastic histories preserve a tradition according to which Khri Srong
lde btsan abdicated in favor of his son, Mu ne btsan po.25 This is most often
reckoned to have occurred in 797. Khri Srong lde btsan’s death is generally recorded
shortly thereafter, and Mu ne btsan po is usually said to have reigned for less than
two years.26
The post-dynastic histories are hopelessly confused regarding the names of Khri
Srong lde btsan’s sons. In a long article, Haarh clarifies the tangle of names in
post-dynastic histories.27 As he points out, some of the most influential Tibetan
histories, including the Sba bzhed, Rgyal rabs gsal ba’i me long, and the Mkhas
pa’i dga’ ston, contribute to the confusion surrounding this issue by using the name
Mu tig btsan po to refer “more or less arbitrarily” to either of Khri Srong lde btsan’s
last two sons.28 Subsequent scholars, both within and outside of Tibet, have repeated
these errors. Relying mostly on a useful passage from U rgyan gling pa’s Rgyal
po bka’ thang yig, Haarh reveals that Mu tig btsan po is none other than Khri Lde
srong btsan, alias Sad na legs mjing yon.29 This is, of course, the father of Khri
Gtsug lde brtsan, alias Ral pa can. Haarh concludes that Khri Srong lde btsan had
four sons, the first being Mu khri btsan po, who died young. Despite the fact that
this name is also conflated with those of the three later sons, Haarh argues that Mu
Tibet.” The harmonious coexistence of these figures within the illustrated pages of a catalogue of Tibet’s
state-sanctioned Buddhism texts certainly indicates a milieu far more permissive of the “simultaneist”
than that depicted in later Tibetan Buddhist historiography.
25
Per K. Sørensen, Tibetan Buddhist Historiography: The Mirror Illuminating the Royal Genealogies
(Weisbaden: Harrassowitz Verlag, 1994), 403 n. 1377.
26
Sørensen, Tibetan Buddhist Historiography, 406 n. 1401.
27
Erik Haarh, “The Identity of Tsu-chi-chen, the Tibetan ‘King’ Who Died in 804 AD,” Acta
Orientalia 25 (1960): 121-70.
28
Haarh, “The Identity of Tsu-chi-chen,” 149.
29
Haarh, “The Identity of Tsu-chi-chen,” 147-48.
khri btsan po is to be identified with the son whose birth is registered in the Old
Tibetan Annals’ entry for the mouse year 760.30 The remaining three sons, according
to Haarh, are Mu ne btsan po, Mu rug btsan po, and Mu tig btsan po, alias Khri
Lde srong btsan, whose births he dates sequentially to the tiger, hare, and dragon
years 774, 775, and 776.31 Haarh worked almost exclusively from post-dynastic
histories, so these dates cannot be accepted with certainty. Employing Old Tibetan
sources alongside the ’Phang thang ma Catalogue, however, I will uphold Haarh’s
earlier conclusions regarding the names of Khri Srong lde btsan’s sons.
The matter of the identities of Khri Srong lde btsan’s sons is easily resolved in
comparison to the problem of the events surrounding their royal succession. In his
article devoted to the succession to Khri Srong lde btsan, Haarh attempted to
harmonize post-dynastic Tibetan sources with a passage in the Tang Annals and
other Chinese sources that recorded the death of a Tibetan emperor in 804. Haarh
concludes that this reflected a struggle for succession between Mu rug btsan po
and his younger brother Khri lde srong btsan, and that the former, while not a
legitimate emperor, was recognized by the Chinese as such.32 Consequently, Haarh
argues, the Chinese sources that record the death of an emperor (btsan po) in 804
in fact refer to Mu rug btsan po.33 Considering the evidence of the ’Phang thang
ma Catalogue, and that of Old Tibetan sources, I will come to similar conclusions
here.
The ’Phang thang ma Catalogue demonstrates, in fact, that the Chinese were
not alone in recognizing Mu rug btsan po/Mu rug btsan as emperor, but that he
was recognized as such in Tibet as well. A fascinating entry in the section under
the misleading heading “commentaries on sūtras and śāstras by Khri Srong lde
btsan” records a commentary to the noble Saṃdhinirmocanasūtra, attributed to
“Emperor Mu rug btsan.”34 While Mu rug btsan is mentioned in a pillar inscription
at Zhwa’i lha khang as the elder brother of Khri Lde srong btsan, the ’Phang thang
ma Catalogue is the only known Tibetan source that refers to him explicitly as an
30
Haarh, “The Identity of Tsu-chi-chen,” 149, 152. See also Jacques Bacot, Frederick W. Thomas,
and Charles Touissant, Documents de Touen-Houang Relatifs a l’Histoire du Tibet (Paris: Libraire
Orientaliste Paul Geuthner, 1940-1946), 58, 65.
31
32
33
Haarh further argues that the struggle for succession was an expression of conflict between Bon
po and Buddhist factions, but, due to the fact that this point rests on rather simplistic assumptions
concerning “pre-Buddhist Bon religion” and “royal religion,” I cannot support this conclusion. Indeed,
while there is certainly continuity in terms of the practices of those ritual specialists known as bon (po)
and gshen during the imperial period and some of the rites of the institutionalized Bon religion,
particularly those of the first four of the nine vehicles, it has never been established with any certainty
that “Bon” constituted a religion per se during the imperial period (though this depends in large part
on what one means by “religion”). The role of bon and gshen in Old Tibetan ritual texts is taken up in
detail in Brandon Dotson, “Complementarity and Opposition in Early Tibetan Ritual,” Journal of the
American Oriental Society (forthcoming).
34
’phags pa dgongs pa nges par ’grel pa’i mdo bshad pa btsan po mu rug btsan gyis mdzad pa/ (Rta
rdo, Dkar chag ’phang thang ma, 55). This was also noted by Halkias (“Tibetan Buddhism Registered,”
56-57), who further recognized its possible relation to the claim in Chinese sources regarding the death
of a Tibetan emperor in 804.
emperor. Let us now consider this new evidence as it relates to what is known of
the succession from Old Tibetan sources, turning first to inscriptional evidence.
The Kong po Inscription reveals that Khri Srong lde btsan ruled jointly with
Lde srong, his chosen heir, who, when he took the throne, was known as Khri Lde
srong btsan. This is evident from the fact that the inscription opens, “A decree
granted to Rkong kar po during the reign of the divine emperor Khri Srong lde
brtsan and Lde srong, father and son” (lines 1-3).35
The west inscription at Zhwa’i lha khang, which records grants made to Myang
Ting nge ’dzin bzang po and his clansmen in recognition for their service to Khri
Lde srong btsan, reveals that before Lde srong was enthroned as Khri Lde srong
btsan there was friction between two parties, the one being Khri Srong lde btsan,
and the other being Khri Lde srong btsan’s elder brother, Mu rug btsan, and that
this was followed by a period of upheaval. Lines nine to twelve read:
Later, after the harmful struggle between my father and elder brother, at the time
when I had not yet taken over the polity, there were also some who fomented
disturbance and civil strife... (lines 9-12).36
It is probably safe to assume that the dispute between Khri Srong lde btsan and
his son concerned succession. The identity of this quarrelsome royal scion is made
clear in lines 48-50 of the inscription, which lists some of those who swore to
uphold the grants and privileges it records:
My elder brother Mu rug brtsan, the royal ladies, minor kings, political ministers
on downwards, and indeed all greater and lesser ministers swore an oath. The
decree is given in perpetuity.37
This passage further demonstrates that Khri Lde srong btsan ruled alone at the
time of the inscription, and subordinated his elder brother Mu rug brtsan, who,
along with queens, minor kings (rgyal phran), and ministers, was bound by oath
to uphold Khri Lde srong btsan’s edict.
By the time of the later, east face inscription at Zhwa’i lha khang, dated to “the
later dragon year,”38 there is no mention of Mu rug brtsan. By naming a later dragon
35
lha btsan po khr-i srong lde brtsan dang / lde srong yab sras ky-i ring la’// // rkong kar po la
gts-igs gnang ba’//. Li and Coblin, A Study of the Old Tibetan Inscriptions, 198, 205, 208.
36
// phyis yab dang gcen thugs nongs brtud par byung ba’i rjes // nga chab sr-id ma bzhes pa’i skabsu
kha cig phan phun dang / gdon stson pa dag yod pa yang / (Li and Coblin, A Study of the Old Tibetan
Inscriptions, 264, 276). My translation differs slightly from that of Li and Coblin. For comparison, see
also Hugh E. Richardson, A Corpus of Early Tibetan Inscriptions (London: Royal Asiatic Society,
1985), 46-47. In both Li and Coblin’s translation and that of Richardson, “disagreement” is not strong
enough for thugs nongs, which, while literally meaning “harming the mind,” can also indicate killing.
37
// gcen mu rug brtsan dang / jo mo mched dang rgyal phran rnams dang / chab srid kyi blon po
man chad / zhang lon che phra kun kyang mnas bsgags te / gts-igs g.yung drung du gnang ngo /; Li
and Coblin, A Study of the Old Tibetan Inscriptions, 266-67, 279-80. My translation differs slightly
from that of Li and Coblin. For comparison, see also Richardson, A Corpus of Early Tibetan Inscriptions,
50-53.
38
’brug gi lo phyi ma.
year, the dating formula of this later inscription does indicate, however, that Khri
Lde srong btsan claimed the previous dragon year, 800, as part of his reign.39 This
does not necessarily indicate, however, that it was the first year of his reign.
The absence of Mu rug brtsan in the later inscription at Zhwa’i lha khang might
indicate that he was safely out of the picture by 812. Another inscription dating to
the reign of Khri Lde srong btsan, however, predates this inscription, and makes
no mention of Mu rug brtsan. This is the Ldan ma brag Inscription, which dates
to the monkey year 804.40 This indicates that by the year 804, for reasons that are
as yet unclear, Khri Lde srong btsan had no need to refer to or bind under oath his
troublesome elder brother Mu rug brtsan.
To summarize the inscriptional evidence, the Kong po Inscription and the
inscriptions at Zhwa’i lha khang indicate that Khri Srong lde btsan ruled jointly
with Prince Lde srong/Khri Lde srong btsan, and came into serious conflict with
Lde srong’s elder brother, Mu rug brtsan. Khri Srong lde btsan then died, and
sometime thereafter Lde srong was enthroned as Khri Lde srong btsan.
This picture is complicated somewhat by the presence in Old Tibetan sources
of another one of Khri Lde srong btsan’s brothers, Mu ne brtsan. The Royal
Genealogy (PT 1286) states that Khri Srong lde btsan and Lady Rma rgyal ldong
skar of the Tshes pong clan bore the sons Mu ne brtsan and Lde srong brtsan, and
that Mu Ne brtsan’s line was cut off (gdung chad).41 This official or quasi-official
39
In fact, the passage refers to the “benevolent later dragon year” (bka’ drin ’brug gi lo phyi ma,
lines 22-23; Li and Coblin, A Study of the Old Tibetan Inscriptions, 271, 292). Li and Coblin translate
this as “with regard to my grace, in the later Dragon year,” but this is borne out neither by the grammar
nor the punctuation. Richardson (A Corpus of Early Tibetan Inscriptions, 56-57) seems to have ignored
the presence of bka’ drin entirely in his reading of this passage. Intriguingly, the use of bka’ drin here
before a year in the reign of Khri Lde srong btsan mirrors the use of skyid rtags, which in the Lhasa
Treaty Pillar precedes dates during Khri Gtsug lde brtsan’s reign (East face inscription, lines 58, 62,
and 65; Li and Coblin, A Study of the Old Tibetan Inscriptions, 51, 98-99). As has been noted before,
“eternal happiness” (skyid rtags) was the name given to Khri Gtsug lde brtsan’s reign, following similar
Chinese practices (Li and Coblin, A Study of the Old Tibetan Inscriptions, 114). While this was thought
to be the only instance of such a Tibetan regnal name, we should not discount the possibility that the
reign of his predecessor, Khri Lde srong btsan, was referred to as the “reign of benevolence” (bka’
drin).
40
spre’u gi lo’i dbyar/ btsan po khri lde srong brtsan gy-i ring la/ (Chab spel tshe brtan phun tshogs,
Btsan po’i dus kyi brag brkos yig ris gcig gsar du mtshams sbyor zhus pa, in Bod kyi yig rnying zhib
’jug, ed. Kha sgang bkra shis tshe ring [Beijing: Mi rigs dpe skrun khang, 2003 (1988)]). See also
Yoshiro Imaeda, “Re-examination of the ldan ma drag Inscription (II) in Eastern Tibet,” in Old Tibetan
Studies Dedicated to the Memory of Professor Ronald E. Emmerick (1937-2001), ed. Cristina A.
Scherrer-Schaub, PIATS 2003: Proceedings of the Tenth Seminar of the International Association for
Tibetan Studies, Oxford 2003 (Leiden: Brill, forthcoming). See, however, the 1998 addendum to an
older article in Hugh E. Richardson, “More Early Inscriptions from Tibet,” in High Peaks, Pure Earth:
Collected Writings on Tibetan History and Culture, ed. Michael Aris (London: Serindia, 1998 [1988]),
278, where Richardson supports the later date of 816 for this inscription.
41
khri srong lde brtsan dang / tshes pong za rma rgyal ldong skar du bshos pa’-i sras/ mu ne brtsan
dang / lde srong brtsan/ mu brtsan gdung chad nas/ (PT 1286, lines 67-69). For the text, see Ariane
Spanien and Yoshiro Imaeda, Choix de documents tibétains conservés à la Bibliothèque nationale:
complété par quelques manuscrits de l’India Office et du British Museum, Tome 2 (Paris: Bibliothèque
nationale, 1979), plate 556. For transliteration, see Yoshiro Imaeda and Tsuguhito Takeuchi, Choix de
documents tibétains conservés à la Bibliothèque nationale: complété par quelques manuscrits de l’India
Office et du British Museum: Corpus syllabique, Tome 3 (Paris: Bibliothèque nationale, 1990), 16. For
royal genealogy therefore indicates that Mu ne brtsan ruled as emperor between

Khri Srong lde btsan and Khri Lde srong btsan. It says nothing, however, of Mu
rug brtsan. What are we to make, then, of Mu rug brtsan’s absence in this Royal
Genealogy when the ’Phang thang ma Catalogue explicitly designates him as an
emperor? First of all, the Royal Genealogy, as an official or semi-official document
composed near to or after the collapse of the Tibetan Empire in the mid-ninth
century, was an idealized lineage. This is not to say that the document is inaccurate
or represents a whitewash, but only to point out that its imperative was to simplify
the royal succession and emphasize the continuity of the royal line from its first
mythical ancestor until the last emperor of royal blood. This genealogy, for example,
does not mention Lha bal po, elder brother of Khri lde gtsug btsan (704-c. 754),
who seized the throne for a short while after his father’s death in 704, and was
deposed in 705.42 While the ’Phang thang ma Catalogue was also an official
document, sponsored by the Tibetan emperor, its compilers obviously felt no need
to excise from the record a composition by Emperor Mu rug btsan, and in doing
so they have handed down valuable information regarding the royal succession.
Taking into account the inscriptional evidence, the Royal Genealogy (PT 1286),
and the evidence of the ’Phang thang ma Catalogue, the events surrounding the
royal succession of Khri Srong lde btsan’s sons seems to be as follows. Khri Srong
Bacot and Toussaint’s transliteration and French translation, see Bacot et al, Documents de
Touen-houang, 82, 89. See also Brandon Dotson, “A Note on Źaṅ: Maternal Relatives of the Tibetan
Royal Line and Marriage into the Royal Family,” Journal Asiatique 292, nos. 1-2 (2004): 86-88.
42
The entry in the Old Tibetan Annals for the snake year 705 states, “At Pong lag rang, they deposed
from the throne the emperor’s elder brother, Lha bal pho” (pong lag rang du btsan po gcen lha bal pho
rgyal sa nas phab; IOL Tib J 750, line 101). See Bacot et al, Documents de Touen-houang, 19, 40-41.
Beckwith, who contends that the above passage should be read as indicating the deposal of Lha from
Bal po, finds it likely that he was the son or relative of Princess Ga tun (meaning “princess” in Turkish),
whose funeral is recorded in the winter of the snake year 708 (Christopher I. Beckwith, The Tibetan
Empire in Central Asia: A History of the Struggle for Great Power among Tibetans, Turks, Arabs, and
Chinese During the Early Middle Ages [Princeton: Princeton University Press, 1987], 73). Beckwith
believes that Ga tun was a Western Turk princess, while Uebach extends the possibility that she was
a daughter of the ’Zha Khagan (Helga Uebach, “Eminent Ladies of the Tibetan Empire According to
Old Tibetan Texts,” in Les Habitants du Toit du Monde, ed. Samten Gyaltsen Karmay and Philippe
Sagant [Nanterre: Société d’Ethnologie, 1997], 59 n. 12). As Kapstein points out, the Chos ’byung
chen mo bstan pa’i rgyal mtshan by Lde’u jo sras (Lde’u jo sras, Chos ’byung chen mo bstan pa’i rgyal
mtshan, ed. Chos ’dzoms [Lhasa: Bod ljongs mi dmangs dpe skrun khang, 1987], 120) calls Lha bal
po “the elder brother Lha bal po, child of the Pa tshab [lady]” (gcen lha bal po pa tshab tsha; Matthew
T. Kapstein, The Tibetan Assimilation of Buddhism: Conversion, Contestation, and Memory [Oxford:
Oxford University Press, 2000], 216). Whatever his actual parentage, it is evident that Princess Ga tun
sent envoys to the Tang court requesting a bride, probably for Lha bal po (Christopher I. Beckwith,
“The Revolt of 755 in Tibet,” in Contributions on Tibetan Language, History and Culture, ed. Ernst
Steinkellner [Wien: Universität Wien, 1983], 6 n. 8). With the deposal of Lha bal po, Ga tun’s candidate
lost out to the candidate of another royal lady, ’Bro Khri ma lod, who then ruled the country in the
name of her grandson Rgyal gtsug ru/Khri Lde gtsug brtsan. These events are reminiscent of the rivalry
between the sons of Glang dar ma ’U’i dum brtan – ’Od srung and Yum brtan – over one century later.
The possibility should thus not be discounted that the earlier rivalry between the two queens and their
two candidates for the throne in the early eighth century served as a template for the muddled narrative
of ’Od srung and Yum brtan in the mid-ninth century. Under the above circumstances, it is not at all
surprising that Lha bal po, and indeed Mu rug btsan, were excised from the Royal Genealogy (PT
1286).
lde btsan abdicated in favor of his son, Mu ne btsan.43 Mu ne btsan then ruled for
a short time before he died. Khri Srong lde btsan, like his ancestor, Srong btsan
sgam po (d. 649), then retook the throne following the death of his son, and groomed
his successor, Khri Lde srong btsan, with whom he ruled jointly. This upset Khri
Srong lde btsan’s elder son, Mu rug btsan, and serious conflict ensued. Khri Srong
lde btsan died, and the two sons, Mu rug btsan and Khri Lde srong btsan, disputed
the succession. Mu rug btsan briefly gained the upper hand and was recognized as
emperor for a short period of time before Khri Lde srong btsan deposed him and
took the throne. Mu rug btsan was at first subordinate to his younger brother, but
subsequently disappeared from the picture. Further, it is quite possible, as argued
by Haarh, that China recognized Mu rug btsan as emperor, and that it was his death
that Chinese sources refer to in their accounts of the year 804.44 It therefore seems
likely that Mu rug btsan indeed ruled as Tibetan emperor for a short time before
he was deposed by his younger brother, Khri Lde srong btsan.
While the order of events proposed above is quite likely, the chronology is less
certain. There are two main problems with establishing a chronology, the first
being the date of Khri Srong lde btsan’s abdication and the rule of Mu ne btsan,
as these dates are not found in Old Tibetan documents, and are therefore necessarily
based on post-dynastic sources. The second problem concerns the dates for Emperor
Mu rug btsan’s reign. While in the later, east face inscription at Zhwa’i lha khang,
dated to “the later benevolent dragon year” 812, Khri Lde srong btsan claims 800
as a year of his reign, he may well be discounting the reign of his usurper, and
counting this as his own. That is to say, 800 is not necessarily the terminus a quo
for Mu rug btsan’s reign: he could have ruled at any time between Khri Srong lde
btsan’s death and 804, the likely year of his own death, but, as in the case of the
official or semi-official Royal Genealogy (PT 1286), this would have been expunged
from the official record.
Most post-dynastic sources agree that Khri Srong lde btsan abdicated in favor
of Mu ne btsan po in 797. Of course, many of these sources also state that Khri
Srong lde btsan died shortly thereafter, but the Old Tibetan sources demonstrate
that these claims are mistaken. The early Sa skya histories, the Bod kyi rgyal rabs
of Grags pa rgyal mtshan (1147-1216) and the eponymous work of ’Phags pa
(1238-1280), both state that Khri Srong lde btsan died at age 56 in an ox year, and
that Mu ne btsan po succeeded him for either one year and nine months or for two
years.45 Given that the Old Tibetan Annals records Khri Srong lde btsan’s birth in
742, this would place his death in approximately 797. Other Tibetan histories, such
as the Rgya bod kyi chos ’byung rgyas pa of Mkhas pa lde’u46 and the Mkhas pa’i
43
It should be noted that Richardson (A Corpus of Early Tibetan Inscriptions, 44) already advanced
the first part of this hypothesis.
44
45
Guiseppe Tucci, “The Validity of the Tibetan Historical Tradition,” in Opera Minora, ed. Guiseppe
Tucci, vol. 2 (Roma: G. Bardi, 1971 [1947]), 456-57, 449.
46
Mkhas pa lde’u, Rgya bod kyi chos ’byung rgyas pa, ed. Chab spel tshe brtan phun tshogs (Lhasa:
Bod ljongs mi dmangs dpe skrun khang, 1987), 358.
dga’ ston, state that Khri Srong lde btsan died at age fifty-nine, which corresponds
to 800.47 The newly published Dba’ bzhed states that Khri Srong lde btsan died in
a horse year, which corresponds to 802.48
Regarding the death of Mu rug btsan, later sources state that Mu rug btsan was
killed by a member of the Sna nam clan in retribution for his murder of Sna nam
’U rings, son of Prime Minister Sna nam Rgyal mtshan lha snang.49 This is also
stated in the Gsang ba phyag rgya can, one of the five texts comprising the Can
lnga, which were standard historical texts that Karmay dates to the late ninth or
early tenth centuries.50 There, however, Mu rug btsan is called Mu tig btsan po,
underlining the confusion whereby the latter name is attributed sometimes to Mu
rug btsan, other times to Khri Lde srong btsan.51
According to another tradition, found in Ne’u paṇḍi ta’s Chos ’byung me tog
phreng ba, Mu rug btsan became a monk after his murder of Sna nam Dbu ring.
Further, this text also contains a tradition according to which Mu rug btsan –
referred to here as Mu tig btsan po – built Khra sna Temple (khra sna’i lha khang).
A gloss states that the temple is known as Nyams khrus rnam dag gi gtsug lag
khang, and that it was built by the divine emperor to expiate his sins from having
killed Dbu ring. The text goes on to state that he also built Kong yul rkyang bu
tshal Temple (kong yul rkyang bu tshal gyi lha khang) and became a monk.52
Despite the ample evidence concerning the death of Mu rug btsan, including a
royal tomb, Bhutanese tradition claims him as a protagonist in their hidden land
47
Dpa’ bo gtsug lag phreng ba, Mkhas pa’i dga’ ston B, 405; Mkhas pa’i dga’ ston D, 126b.5.
48
Pasang Wangdu and Hildegard Diemberger, dBa’ Bzhed: The Royal Narrative Concerning the
Bringing of the Buddha’s Doctrine to Tibet (Wien: Verlag der Österreichischen Akademie der
Wissenschaften, 2000), 92-93 n. 359. In light of the Dba’ bzhed’s evidence, Wangdu and Diemberger
(dBa’ Bzhed, 93 n. 359) suggest that after Khri Srong lde btsan died in 802, Mu ne btsan po succeeded
him, ruling until his death at the end of 803. It was this death, they suggest, that the Chinese sources
record in 804. This would be a tidy solution to the problem of Khri Srong lde btsan’s successors were
it not for the fact that it ignores Mu rug btsan entirely. Further, it is untenable in light of the order of
events determined above through an examination of Old Tibetan sources. The Dba’ bzhed is obviously
confused, as Mu ne btsan po, who presides over his father’s funeral in this text, was by this time dead.
He appears to be conflated with Khri Lde srong btsan.
49
50
Samten G. Karmay, “The Origin Myths of the First King of Tibet as Revealed in the Can lnga,”
in The Arrow and the Spindle: Studies in History, Myths, Rituals and Beliefs in Tibet, ed. Samten
Gyaltsen Karmay (Kathmandu: Mandala Book Point, 1998 [1994]), 307.
51
In fact, Dpa’ bo gtsug lag, in a passage concerning the catalogues and the translation of Buddhist
texts, cites Bu ston to the effect that the ’Phang thang ma Catalogue contains a commentary to the
Saṃdhinirmocanasūtra not by Mu rug btsan, but by Mu tig btsan po (Dpa’ bo gtsug lag phreng ba,
Mkhas pa’i dga’ ston B, 401; Mkhas pa’i dga’ ston D, 124b.4). The ’Phang thang ma Catalogue itself,
as demonstrated above, reveals Dpa’ bo’s conflation of Mu rug btsan with Mu tig btsan po to be
misleading.
52
de’i gcung mu tig btsan pos khra sna’i lha khang snyams khrus rnam dag gi gtsug lag khang zer bzhengs lha
rkang [the editor corrects this to kong] yul rkyang bu tshal gyi
btsan pa [read po] dbu ring bsad pa’i sdig sbyongs/
lha khang bzhengs nas rab tu byung ngo / (Ne’u paṇḍi ta Grags pa smon lam blo gros, Sngon gyi gtam
me tog phreng ba, in Bod kyi lo rgyus deb ther khag lnga, ed. Ldan lhun sangs rgyas chos ’phel [Lhasa:
Bod ljongs bod yig dpe rnying dpe skrun khang, 1990], 23). The small text represents glosses, which
appear in a smaller hand in the text.
(sbas yul) literature. According to this tradition, he is identified with Khyi kha ra
thod (literally, “Dog mouth Goat skull”), the offspring of Tshes pong Bza’ dmar
rgyal’s bestial liaison with a dog and a goat. Aris sees in Khyi kha ra thod an
independent figure, whose lore grew to include Mu rug btsan and the
much-maligned Queen Tshes pong Bza’ dmar rgyal.53 This may in fact have a
historical basis, as the Nyang ral chos ’byung states that Mu rug btsan – referred
to here again as Mu tig btsan po – returned from exile in Mon to take the throne
after Mu ne btsan po’s death. Upon arrival, however, he was killed by members
of the Sna nam clan.54 Mon, of course, is a vague toponym-cum-ethnonym that
denotes the south, and often includes Bhutan and Sikkim.
Given this unruly tangle of dates to choose from for Khri Srong lde btsan’s
death, it seems better to return to the Old Tibetan sources for guidance. If we
cautiously accept Khri Lde srong btsan’s claim in the east face inscription at Zhwa’i
lha khang that he ruled Tibet in the dragon year 800 (either jointly with his father
or alone), we might propose the following tentative chronology. Following
post-dynastic sources, Khri Srong lde btsan abdicated c. 797 in favor of his son,
Mu ne btsan. Mu ne btsan then ruled for two years or less, c. 797-c. 798, before
he died. Khri Srong lde btsan then retook the throne and ruled jointly with his
chosen successor, Lde srong/Khri Lde srong btsan, c. 798-c. 800. This upset Khri
Srong lde btsan’s elder son, Mu rug btsan, and serious conflict ensued. Perhaps as
a result of this conflict, Khri Srong lde btsan died, probably in 800 (i.e., at age
fifty-nine as claimed in Mkhas pa lde’u’s Rgya bod kyi chos ’byung rgyas pa). His
two sons, Mu rug btsan and Khri Lde srong btsan, then disputed the succession in
a contest that threatened to cause civil unrest. Mu rug btsan’s brief ascendancy as
emperor likely dates to c. 800-c. 802, shortly after Khri Srong lde btsan’s death.
Khri Lde srong btsan, with the help of powerful allies such as Myang Ting nge
’dzin bzang po, then gained the upper hand and retook the throne in c. 802, and
Mu rug btsan died not long after, in c. 804.
While the order of events is established almost exclusively from the evidence
of Old Tibetan sources, the chronology is less certain, and is meant only as a
working hypothesis to help guide this problem to its final resolution. Besides
resolving to a certain extent the succession at the turn of the ninth century, this
proposed chronology also suggests dates for the Kong po Inscription and the west
inscription at Zhwa’i lha khang. The former was erected when Khri Srong lde
btsan ruled jointly with Lde srong, and therefore dates to c. 798-c. 800. The latter
refers to a period in Khri Lde srong btsan’s reign before the death of Mu rug btsan,
and can therefore be placed between c. 800 and 804.
53
Michael Aris, Bhutan: The Early History of a Himalayan Kingdom (Warminster: Aris and Phillips,
1979), 74-75.
54
“Then they invited Mu ne btsan po, who had been in exile in Mon, and enthroned him. The Sna
nam clansmen said, ‘if you do this... ’ and as soon as he exposed his breast, they killed him” (de nas
mu tig btsan po mon na spyugs te yod pa khong la dbang bskur bar chad byas te spyan drangs pa dang
/ sna nam pas ’di byed na zer te/ mu tig btsan po’i brang kha bye nas krongs so/; Nyang nyi ma ’od
zer, Chos ’byung me tog snying po sbrang rtsi’i bcud, ed. Chab spel tshe brtan phun tshogs [Lhasa:
Bod ljongs mi dmangs dpe skrun khang, 1988], 414).
Conclusions
Among the considerable contributions of the ’Phang thang ma Catalogue to our
understanding of the Tibetan Empire, it adds another fold to the issue of the royal
succession at the turn of the ninth century by designating Mu rug btsan as an
emperor. Adding this new data to what we already know about the succession from
pillar edicts and Old Tibetan texts such as the Royal Genealogy (PT 1286), the
matter now seems somewhat closer to resolution. To summarize, the proposed
order of events (along with a tentative chronology) is as follows:
c. 756-c. 797 First reign of Khri Srong lde btsan.
c. 797-c. 798 Reign of Mu ne btsan.
c. 798-c. 800 Second reign of Khri Srong lde btsan; rules jointly with chosen successor Lde
srong/Khri Lde srong btsan.
c. 800-c. 802 Mu rug btsan seizes throne from Khri Lde srong btsan upon the death of their father,
Khri Srong lde btsan.
c. 802-c. 815 “Second” reign of Khri Lde srong btsan; Mu rug btsan is subordinate to his younger
brother, then dies c. 804.
When further documents come to light and provide authoritative dates for the
death of Khri Srong lde btsan and the activities of Mu rug btsan, the chronology
proposed above may require slight alterations. I can only hope that further ancient
sources, such as the Mchims phu ma Catalogue, will soon emerge and shed even
more light on such issues.
Glossary
Ka
kong po Kongpo Place
kong yul rkyang bu Kongyül Kongyül Temple
tshal gyi lha khang Kyangbutselgyi Kyangbutsel
Lhakhang Temple
dkyus kyü ad hoc Term
bka’ drin ’brug gi lo kadrin drukgi lo the benevolent later Term
phyi ma chima dragon year
rkong kar po Kong Karpo Person
skar chung Karchung Text
skyid rtags kyitak eternal happiness Term
Kha
khyi kha ra thod Khyikha Ratö Dog mouth Goat Person
skull
khra sna’i lha khang Trané Lhakhang Trana Temple Temple
khra ’brug Tradruk Monastery
khri lde gtsug btsan Tri Detsuktsen 704-c. Person
754
khri lde srong btsan Tri Desongtsen Person
khri gtsug lde brtsan Tri Tsukdetsen r. 815-41 Person
khri ’u’i dum brtan Tri Üdumten r. 841-42 Person
khri ’od srung Tri Ösung 842/43-c. Person
890
khri srong lde btsan Tri Songdetsen 742-c. Person
800
khri srong btsan Tri Songtsen Person
khrom bzher Tromzher Person
mkhas pa lde’u Khepa Deu Author
mkhas pa’i dga’ ston Khepé Gatön Text
Ga
ga tun Gatün Person
gi gu gigu Term
grags pa rgyal mtshan Drakpa Gyeltsen 1147-1216 Person
glang dar ma Langdarma Person
glang dar ma ’u’i dum Langdarma Üdumten Person
brtan
rgya bod kyi chos Gyabökyi Chöjung Text

’byung rgyas pa Gyepa
rgyal po bka’ thang Gyelpo Katang Yik Text
yig
rgyal phran gyeltren minor kings Term
rgyal rabs gsal ba’i Gyelrap Selwé Text
me long Melong
rgyal gtsug ru Gyeltsukru Person
rgyud gyü temporal lineage Term
sgom lung gomlung meditation manual Term
sgra sbyor bam po Drajor Bampo Nyipa Text
gnyis pa
Ca
can lnga Chennga Text Group
cig car ba chikcharwa simultaneist Term
bcom ldan ’das kyi chomdendekyi ringluk Term
ring lugs
bcom ldan ’das kyi chomdendekyi Term
ring lugs kyi mdun sa ringlukkyi dünsa
Cha
chos ’byung chen mo Chöjung Chenmo Text
bstan pa’i rgyal Tenpé Gyeltsen
mtshan
chos ’byung me tog Chöjung Metok Text
phreng ba Trengwa
mchims phu ma Chimpuma Mchims phu ma Text
Catalogue
Ja
jo khang Jokhang Monastery
Nya
nyang ral chos ’byung Nyangrel Chöjung Text
nyams khrus rnam Nyamtrü Namdakgi Monastery
dag gi gtsug lag Tsuklakkhang
khang
Ta
bstan ’gyur Tengyur Textual
Collection
Tha
thugs nongs tuknong lit. “harming the Term
mind”; harmful; to
die
Da
gdung chad dungché Term
ldan dkar ma Denkarma Ldan dkar ma Text
Catalogue
ldan ma brag Denma Drak Place
lde srong Desong Person
lde srong brtsan Desongtsen Person
lde’u jo sras Deu José Author
Na
ne’u paṇḍi ta Neu Pendita Author
gnubs sangs rgyas ye Nup Sanggyé Yeshé Author
shes
sna nam Nanam Clan
sna nam rgyal mtshan Nanam Gyeltsen Person
lha snang Lhanang
sna nam dbu ring Nanam Uring Person
sna nam ’u rings Nanam Uring Person
Pa
pad ma ’byung gnas Pema Jungné Person
pong lag rang Ponglakrang Place
dpa’ bo gtsug lag Pawo Tsuklak Author
phreng ba Trengwa
dpal dum brtan Peldumten Person
dpe pé source Term
Pha
pho brang ’khor gyi podrang khorgyi Term
dge ’dun gendün
phyi dar chidar Term
’phags pa Pakpa 1238-1280 Author
’phang thang Pangtang Place
’phang thang ma Pangtangma ’Phang thang ma Text
Catalogue
’phrul snang Trülnang Monastery
Ba
bandhe gzhon nu Bandhé Zhönnu Person
snying po Nyingpo
bal po Belpo Place
bu ston Butön Person
bod kyi rgyal rabs Bökyi Gyelrap Text
bon Bön Organization;
Term
bon po Bönpo Organization

dba’ dun brtan Wa Dünten Person
dba’ bzhed Wazhé Text
dbu ring Uring Person
’bras spungs Drepung Monastery
’brug gi lo phyi ma drukgi lochima Term
’bro khri ma lod Dro Trimalö Person
sba bzhed Bazhé Text
sbas yul beyül hidden land Term
Ma
mu khri btsan po Mutri Tsenpo Person
mu tig btsan po Mutik Tsenpo Person
mu ne btsan Muné Tsen Person
mu ne btsan po Muné Tsenpo Person
mu ne brtsan Muné Tsen Person
mu rug btsan Muruktsen Person
mu rug btsan po Muruk Tsenpo Person
mon Mön Place
myang ting nge ’dzin Nyang Tingngedzin Person
bzang po Zangpo
rma rgyal ldong skar Magyel Dongkar Person
Tsa
btsan po tsenpo emperor Term
rtse rje tsejé town prefect Term
Tsha
tshes pong Tsepong Clan
tshes pong bza’ dmar Tsepong Zamargyel Person
rgyal
Zha
zhwa’i lha khang Zhé Lhakhang Temple
gzhon nu snying po Zhönnu Nyingpo Person
'a
’u’i dum brtan Üdumten Person
’od srung Ösung Person
’on cang do Önchangdo Place
Ya
yum brtan Yumten Person
Ra
ra mo che Ramoché Monastery
ra sa ’phrul snang Rasa Trülnang Monastery
ral pa can Relpachen Person
Sha
sha cu Shachu Place
gshen shen Term
Sa
sa skya Sakya Organization
sad na legs mjing yon Senalek Jingyön Person
srong btsan sgam po Songtsen Gampo d. 649 Person
gsang ba phyag rgya Sangwa Chakgyachen Text
can
bsam gtan mig sgron Samten Mikdrön Text
bsam yas Samyé Monastery
Ha
hwa shang ma hwa ya Hwashang Mahayana Person
na
lha bal po Lhabelpo Person
lha bal pho Lhabelpo Person
lha bzang gsas Lhazang Setong Person
mthong
lha lung dpal gyi rdo Lhalung Pelgyi Dorjé Person
rje
lhan dkar Lhenkar Place
lhan dkar ma Lhenkarma Lhan dkar ma Text
Catalogue
A
u rgyan gling pa Urgyen Lingpa Person
Sanskrit
Asaṅga Person
bhagavat Term
Candrakīrti Person
dharma Term
spiritual friend kalyāṇamitra Term
Kamalaśīla Person
Madhyamaka Doxographical
Category
Nāgārjuna Person
Saṃdhinirmocanasūtra Text
saṃgha Term
sūtra Term
Śāntarakṣita Person
śāstra Term
Vasubandhu Person
Vimalamitra Person
Chinese
Chan Organization
Dunhuang Place
Bibliography
Old Tibetan Documents
PT 1286. The Royal Genealogy, properly a part of the Old Tibetan Chronicle.
PT 1085. A response to the town prefect (rtse rje) of Sha cu.
IOL Tib J 689 (2). A document listing various lineages of kalyāṇamitras in Tibet.
IOL Tib J 750. The second part of the Old Tibetan Annals, “civil version.”
Kong po Inscription. Concerns privileges of the land of Kong po, and dates to
c. 798-c. 800.
Inscriptions at Zhwa’i lha khang. Record grants; west inscription dates to between
c. 800 and 804, and east inscription dates to 812.
Ldan ma brag Inscription. Records promotion of monks and dates to 804.
References
Aris, Michael. Bhutan: The Early History of a Himalayan Kingdom. Warminster:
Aris and Phillips, 1979.
Bacot, Jacques, Frederick W. Thomas, and Charles Touissant. Documents de
Touen-Houang Relatifs a l’Histoire du Tibet. Paris: Libraire Orientaliste Paul
Geuthner, 1940-1946.
Beckwith, Christopher I. “The Revolt of 755 in Tibet.” In Contributions on Tibetan
Language, History and Culture, edited by Ernst Steinkellner, 1-16. Wien:
Universität Wien, 1983.
———. The Tibetan Empire in Central Asia: A History of the Struggle for Great
Power Among Tibetans, Turks, Arabs, and Chinese During the Early Middle
Ages. Princeton: Princeton University Press, 1987.
Chab spel tshe brtan phun tshogs. Btsan po’i dus kyi brag brkos yig ris gcig gsar
du mtshams sbyor zhus pa. In Bod kyi yig rnying zhib ’jug, edited by Kha sgang
bkra shis tshe ring, 86-93. Beijing: Mi rigs dpe skrun khang, 2003 [1988].
Dotson, Brandon. “A Note on Źaṅ: Maternal Relatives of the Tibetan Royal Line
and Marriage into the Royal Family.” Journal Asiatique 292, nos. 1-2 (2004):
75-99.
———. “Complementarity and Opposition in Early Tibetan Ritual.” Journal of
the American Oriental Society (forthcoming).
Dpa’ bo gtsug lag phreng ba. Dam pa’i chos kyi ’khor lo bsgyur ba rnams kyi
byung ba gsal bar byed pa mkhas pa’i dga’ ston. Beijing: Mi rigs dpe skrun
khang, 1985. Abbreviated as Mkhas pa’i dga’ ston (Khepé Gatön) B.
mKhas-paḥi-dgaḥ-ston by Dpaḥ-bo-gtsug-lag-phreng-ba. Edited by Lokesh

Chandra. Śatapiṭaka Series 9 [4]. New Delhi: International Academy of Indian
Culture, 1965. Abbreviated as Mkhas pa’i dga’ ston (Khepé Gatön) D.
Haarh, Erik. “The Identity of Tsu-chi-chen, the Tibetan ‘King’ Who Died in 804
AD.” Acta Orientalia 25 (1960): 121-70.
Halkias, Georgios. “Tibetan Buddhism Registered: A Catalogue from the Imperial
Court of ’phang thang.” The Eastern Buddhist 36, nos. 1-2 (2004): 46-105.
Imaeda, Yoshiro. “Chinese Texts on the Tibetan King Glang Dar ma.”
Bukkyô-gaku Seminâ (Buddhist Seminar) 74 (2001): 67-79 (26-38).
———. “Re-examination of the ldan ma drag Inscription (II) in Eastern Tibet.”
In Old Tibetan Studies Dedicated to the Memory of Professor Ronald E.
Emmerick (1937-2001), edited by Cristina A. Scherrer-Schaub. PIATS 2003:
Proceedings of the Tenth Seminar of the International Association for Tibetan
Studies, Oxford 2003. Leiden: Brill, forthcoming.
Imaeda, Yoshiro and Tsuguhito Takeuchi. Choix de documents tibétains conservés
à la Bibliothèque nationale: complété par quelques manuscrits de l’India
Office et du British Museum: Corpus syllabique, Tome 3. Paris: Bibliothèque
nationale, 1990.
Kapstein, Matthew T. The Tibetan Assimilation of Buddhism: Conversion,
Contestation, and Memory. Oxford: Oxford University Press, 2000.
Karmay, Samten Gyaltsen. Btsan po lha sras dar ma dang de’i rjes su byung ba’i
rgyal rabs mdor bsdus. Dharamsala: Library of Tibetan Works and Archives,
1986.
———. The Great Perfection: A Philosophical and Meditative Teaching of
Tibetan Buddhism. Leiden: Brill, 1988.
———. “The Origin Myths of the First King of Tibet as Revealed in the Can
lnga.” In The Arrow and the Spindle: Studies in History, Myths, Rituals and
Beliefs in Tibet, edited bySamten Gyaltsen Karmay, 282-309. Kathmandu:
Mandala Book Point, 1998 [1994].
———. “King Lang Darma and His Rule.” In Tibet and Her Neighbours: A
History, edited by Alex McKay, 57-66. London: Hansjörg Mayer, 2003.
Lalou, Marcelle. “Les Textes Bouddhiques au temps du Roi khri-sroṅ-lde-bcan.”
Journal Asiatique 241 (1953): 313-53.
Lde’u jo sras. Chos ’byung chen mo bstan pa’i rgyal mtshan. Edited by Chos
’dzoms. Lhasa: Bod ljongs mi dmangs dpe skrun khang, 1987.
Li, Fang Kuei and W. South Coblin. A Study of the Old Tibetan Inscriptions.
Taipei Institute of History and Philology: Academica Sinica, 1987.
Mkhas pa lde’u. Rgya bod kyi chos ’byung rgyas pa. Edited by Chab spel tshe
brtan phun tshogs. Lhasa: Bod ljongs mi dmangs dpe skrun khang, 1987.
Ne’u paṇḍi ta Grags pa smon lam blo gros. Sngon gyi gtam me tog phreng ba. In
Bod kyi lo rgyus deb ther khag lnga, edited by Ldan lhun sangs rgyas chos
’phel, 1-54. Lhasa: Bod ljongs bod yig dpe rnying dpe skrun khang, 1990.
Nyang nyi ma ’od zer. Chos ’byung me tog snying po sbrang rtsi’i bcud. Edited
by Chab spel tshe brtan phun tshogs. Lhasa: Bod ljongs mi dmangs dpe skrun
khang, 1988. Abbreviated as Nyang ral chos ’byung.
Panglung, Jampa. “New Fragments of the sGra-sbyor bam-po gnyis-pa.” East
and West 44, no. 1 (1994): 161-72.
Richardson, Hugh E. A Corpus of Early Tibetan Inscriptions. London: Royal
Asiatic Society, 1985.
———. “More Early Inscriptions from Tibet.” In High Peaks, Pure Earth:
Collected Writings on Tibetan History and Culture, edited by Michael Aris,
276-278. London: Serindia, 1998 [1988].
Rta rdo, ed. Dkar chag ’phang thang ma/ sgra sbyor bam po gnyis pa. Beijing:
Mi rigs dpe skrun khang, 2003.
Scherrer-Schaub, Cristina A. “Prières pour un apostat: fragments d’histoire
Tibétaine.” Cahiers d’Extrême-Asie 11 (1999-2000): 217-46.
———. “Enacting Words: A Diplomatic Analysis of the Imperial Decrees (bkas
bcad) and Their Application in the sGra sbyor bam po gnyis pa Tradition.”
Journal of the International Association of Buddhist Studies 25, nos. 1-2 (2002):
263-340.
Skilling, Peter. “From bKa’ bstan bcos to bKa’ ’gyur and bsTan ’gyur.” In
Transmission of the Tibetan Canon: Papers Presented at a Panel of the 7th
Seminar of the International Association for Tibetan Studies, Graz 1995, edited
by Helmut Eimer, 87-111. Wien: Verlag der Österreichischen Akademie der
Wissenschaften, 1997.
Sørensen, Per K. Tibetan Buddhist Historiography: The Mirror Illuminating the
Royal Genealogies. Weisbaden: Harrassowitz Verlag, 1994.
Sørensen, Per K. and Guntram Hazod, in cooperation with Tsering Gyalbo.
Thundering Falcon: An Inquiry into the History and Cult of Khra-’brug, Tibet’s
First Buddhist Temple. Wien: Verlag der Österreichischen Akademie der
Wissenschaften, 2005.
Spanien, Ariane and Yoshiro Imaeda. Choix de documents tibétains conservés à
la Bibliothèque nationale: complété par quelques manuscrits de l’India Office
et du British Museum, Tome 2. Paris: Bibliothèque nationale, 1979.
Stein, Rolf A. “Sudden Illumination or Simultaneous Comprehension: Remarks
on Chinese and Tibetan Terminology.” In Sudden and Gradual: Approaches
to Enlightenment in Chinese Thought, edited by Peter N. Gregory, 41-66.

Honolulu: University of Hawaii Press, 1987.
Tucci, Giuseppe. “The Validity of the Tibetan Historical Tradition.” In Opera
Minora, edited by Giuseppe Tucci. Vol. 2, 453-68. Rome: G. Bardi, 1971
[1947].
Uebach, Helga. “Eminent Ladies of the Tibetan Empire According to Old Tibetan
Texts.” In Les Habitants du Toit du Monde,” edited by Samten Gyaltsen
Karmay and Philippe Sagant, 53-74. Nanterre: Société d’Ethnologie, 1997.
Walter, Michael. “The Significance of the Term ring lugs: Religion,
Administration, and the Sacral Presence of the btsan-po.” Acta Orientalia
Scientiarum Hungaricae 51, no. 3 (1998): 309-319.
Wangdu, Pasang and Hildegard Diemberger. dBa’ Bzhed: The Royal Narrative
Concerning the Bringing of the Buddha’s Doctrine to Tibet. Wien: Verlag der
Österreichischen Akademie der Wissenschaften, 2000.
Yamaguchi, Zuiho. “The Fiction of King Dar Ma’s Persecution of Buddhism.”
In De Dunhuang au Japon: Études Chinoises et Bouddhiques Offertes à Michel
Soymié, edited by Jean Pierre Drège, 230-58. Geneva: Librairie Droz, 1996.
An Early Seventeenth-Century
Tibeto-Mongolian Ceremonial Staff

Johan Elverskog
Southern Methodist University
Abstract: This article provides a translation and study of a recently discovered

ceremonial staff housed at the Metropolitan Museum of Art. Based on the
identification of two names found in the inscription, Sodnam Ombu Taiji
(1587-1625) and the Fourth Dalai Lama (yon tan rgya mtsho, 1589-1617), it
situates the piece in the historical context of early seventeenth-century
Mongol-Tibetan relations.
Introduction
In 2005 the Metropolitan Museum of Art acquired a previously unknown socket
from a spearhead or ceremonial staff (Figure 1).1
Figure 1: Ceremonial Staff at the Metropolitan

Museum of Art.
1
All images are courtesy of The Metropolitan Museum of Art, Purchase, Gift of Arthur Ochs
Sulzberger, 2005 (2005.301.1).
1550-6363/2007/3/T3127.
© 2007 by Johan Elverskog, Tibetan and Himalayan Digital Library, and International Association of Tibetan Studies.
Elverskog: An Early Seventeenth-Century Tibeto-Mongolian Ceremonial Staff 2
It is a slightly conical, foot-long, iron tube with three half-round ring moldings,
one at the top, one in the middle, and the other at the base. Within these dividing
strips the socket has an inscription damascened in silver and gold. The top half of
the socket has eleven horizontal lines of Tibetan, and the bottom half has fourteen
lines of Mongolian.2 Although these inscriptions are fragmentary because of
corrosion, it is still possible to identify two names: Sodnam Ombu Taiji
(1587-1625)3 and the Fourth Dalai Lama (yon tan rgya mtsho, 1589-1617). This
socket therefore appears to be not only the sole known piece of ironwork with a
Tibeto-Mongolian inscription, but it is also an important textual monument from
a pivotal period in Inner Asian history.
Historical Background
On the afternoon of June 5, 1578, the leader of the Twelve Tümed Mongols met
with the abbot of ’Bras spungs Monastery. For Altan Qaγan (1508-82) this event
was simply one of many such meetings he had held over the previous decade.
Indeed, one month before his meeting with the young, relatively unknown Bsod
nams rgya mtsho (1543-88), he had entertained the famous and powerful sixteenth
abbot of the Stag lung order. The qaγan had been greatly pleased with the abbot’s
knowledge, especially on the important issues of astrology, divination, and
medicine. They had therefore agreed to meet again the following year. At that
subsequent meeting the qaγan was so taken by the abbot’s religious powers that
he gave him a silver seal, along with official documents, hats, and clothing, as well
as a large amount of silver. Most importantly, however, the qaγan bestowed on
the Stag lung abbot the title “Tathāgatha,” an epithet of the Buddha that was also
the title the Yongle emperor (1360-1424, r. 1403-24) had bestowed on the Sixth
Black Hat Karmapa (1416-53).4
What tied all of these seemingly disparate meetings of Ming dynasty emperors
and Mongol qaγans with various Tibetan religious hierarchs was the Buddhist
theory of imperial rule. The Buddha had originally preached this powerful
religio-political model;5 however, in the Chinese and Inner Asian sphere it had
become most clearly articulated in the writings of ’Phags pa bla ma (1235-80),
whom Qubilai Qaγan (1215-94) had appointed as Imperial Preceptor, or national
abbot of the Yuan dynasty (1272-1368). Indeed, during all the meetings noted
2
Donald J. LaRocca, et al., Warriors of the Himalayas: Rediscovering the Arms and Armor of Tibet
(New Haven: Yale University Press, 2006), 181 (cat. no. 85).
3
For the simplified phonetic transliteration of Mongolian I have followed Christopher P. Atwood,
Young Mongols and Vigilantes in Inner Mongolia’s Interregnum Decades, 1911-1931 (Leiden: Brill,
2002), xvi; and for Latin transcription I use the philological system of Igor de Rachewiltz and John R.
Krueger, Erdeni-yin Tobci (‘Precious Summary’): A Mongolian Chronicle of 1662, vol. 1 (Canberra:
The Australian National University, 1990), viii.
4
The biography of the Sixteenth Abbot of Stag lung is found in Ngag dbang rnam gyal’s Chos ’byung
ngo mtshar rgya mtsho (Tashijong, 1972), written between 1609 and 1626.
5
While the Buddhist model of rule and its application has been explored by numerous scholars, one
of the more important studies is Stanley Tambiah’s World Conqueror, World Renouncer (New York:
Cambridge University Press, 1976).
above everyone involved always harked back to the iconic meeting of Qubilai
Qaγan and ’Phags pa bla ma, the Buddhist “patron and priest”6 par excellence.7
They also drew more specifically on ’Phags pa bla ma’s explication of this
relationship and the nature of a proper Buddhist state, the so-called theory of “Two
Customs”: the secular realm headed by a political ruler, and the sacred realm headed
by a Buddhist teacher.8 Accordingly, only when these two realms were united,
with the king securing worldly benefits and the bla ma transcendent bliss, would
harmony and peace prevail in the empire.9
6
As David Seyfort Ruegg has pointed out, the terms “priest” and “patron” are problematic since they
often obscure the actual religio-political relation maintained between these two individuals; see David
S. Ruegg, “Mchod yon, yon mchod and mchod gnas/yon gnas: On the Historiography and Semantics
of a Tibetan Religio-social and Religio-political Concept,” in Tibetan History and Language: Studies
Dedicated to Uray Géza on his Seventieth Birthday, ed. Ernst Steinkeller (Wien: Arbeitskreis für
Tibetische und Buddhistische Studien Universität Wien, 1991), 441-53. Based on the Mongolian, it is
possible to use “alms-master” (öglige-yin ejen) and “offering-site” (takil-un oron).
7
The importance of the meeting between Qubilai Qaγan and ’Phags pa bla ma is reflected in the fact
that it is the standard point of reference in later Mongolian and Tibetan works. For example, in the
1607 history of Altan Qaγan and his descendants, the meeting between Altan and the Third Dalai Lama
is framed solely in relation to this earlier meeting. Moreover, not only do both the qaγan and bla ma
recognize their own meeting as a new manifestation of this earlier event, but they also recreate its
context by having the Third Dalai Lama initiate Altan into the Hevajra maṇḍala as ’Phags pa had done
for Qubilai:
Altan Khan then with his pure faith and reverence spoke respectfully, “Oh! All-knowing Vajradhara
Dalai Lama, since according to tradition previously Phagpa Lama bestowed the famous Hevajra
empowerment on our Supreme Sechen Khan, will you now bestow it on me?” The Supreme
All-knowing Vajradhara Dalai Lama replied, “Mighty Great Khan, your statement is very correct
… Therefore, now for you Great Khan, I will perfect myself for the empowerment of the unparalleled
Secret Tantra Vajrayāna, and bestow it on the Khan, Queen and all the believers” (Johan Elverskog,
The Jewel Translucent Sutra: Altan Khan and the Mongols in the Sixteenth Century [Leiden: Brill,
2003], 165-66).
Similarly, in Saγang Secen’s 1662 Precious Summary there is the well-known speech of Qutuγtai Secen
Qung Taiji wherein he reiterates the fact that the meeting of Altan and Bsod nams rgya mtsho
re-establishes the Tibeto-Mongol Buddhist alliance initially forged by Qubilai and ’Phags pa bla ma.
8
The theory of the qoyar yosu as envisioned by ’Phags pa bla ma is found in several of his works,
but particularly in the Shes bya rab gsal and the Rgyal bu ji big de mur la gdams du byas nor bu’i
phreng ba, see Sh. Bira, “Qubilai Qa’an and ’Phags-pa Bla-ma,” in The Mongol Empire and Its Legacy,
ed. Reuven Amitai-Press and David Morgan (Leiden and Boston: Brill, 1999), 244-9.
9
According to several scholars, this joint legitimacy is premised on the idea
that secular and spiritual salvation are something that all living beings try to obtain. Spiritual salvation
consists in complete deliverance from suffering, and worldly welfare is secular salvation. Both
depend on a dual order, the order of Religion (nom-un yosun) and the order of the State, or worldly
rule (törö-yin yosun). Just as the religious order is based on the sūtras and magic formulae (dhāranī),
the secular order rests on peace and quietness. The order of Religion is presided over by the Lama,
and the state by the Ruler. The priest has to teach religion, and the Ruler has to guarantee a rule
which enables everyone to live in peace. Religion and State are thus mutually dependent. The heads
of religion and state are equal, although each has different functions. The Lama corresponds to the
Buddha, and the Ruler to the cakravartin (Herbert Franke, “Tibetans in Yüan China,” in China
under Mongol Rule, ed. John Langlois, Jr. [Princeton: Princeton University Press, 1981], 308).
For a detailed study of the qoyar yosu/yon mchod theory of rule see David S. Ruegg, Ordre spirituel
et ordre temporel dans la pensée bouddhique de l’Inde et du Tibet, Publications de l’Institut de
civilisation indienne, Série in-8, Fasc. 64 (Paris: Collège de France, 1995).
This was, of course, the theory. Yet, as with any religio-political utopia, how
it actually played out in reality was often rather different.10 One of the most common
violations of this Buddhist ideal model was the constant shifting of alliances
between the court and various competing religious orders, as was the case with
Möngke Qaγan’s (r. 1251-59) affiliation with Karma pag shi (1206-83), and Qubilai
Qaγan’s with ’Phags pa bla ma. Moreover, as Qubilai Qaγan’s power rose, so too
did that of ’Phags pa and his Sa skya order, which acquired enormous power within
Tibet, China, and Mongolia at the expense of the other Tibetan schools.11 Even so,
this did not prevent these other schools from jockeying for position. In fact, by the
end of the Yuan dynasty, both the Rnying ma and Bka’ brgyud lineages were
prevalent at the imperial court.12 Whether or not this religious fractiousness
contributed to the fall of the Yuan can be debated, yet it is certainly a fact that the
“Two Customs” patronage model and its alignment with competing political forces
have played a significant role in the endless civil wars that have shaped much of
Tibetan and Mongol history.
This was especially the case with the 1578 meetings between Altan Qaγan and
different Tibetan Buddhist hierarchs. At the time Altan was the most powerful
ruler on the Mongolian plateau; however, he was not the recognized ruler of the
Mongols. The Mongol qaγan had to be a direct descendant of Cinggis Qaγan
(1167-1227). Altan, however, was the eldest son of the Chinggisid heir Dayan
Qaγan’s (1474-1517?) third son; and because Mongol dynastic rule was based on
primogeniture, he could claim neither Chinggisid legitimacy nor the Mongol throne.
Yet on account of his military prowess and financial resources secured through
trading rights with the Ming court he was the de facto undisputed ruler of the
10
On Mongol rule in China, its various policies, and its impact, see John D. Langlois, ed., China
Under Mongol Rule (Princeton: Princeton University Press, 1981); Hok-lam Chan and W.T. de Bary,
Yüan Thought: Chinese Thought and Religion under the Mongols (New York: Columbia University
Press, 1982); Morris Rossabi, Khubilai Khan: His Life and Times (Berkeley: University of California
Press, 1988); Elizabeth Endicott-West, Mongolian Rule in China: Local Administration in the Yuan
Dynasty (Cambridge: Harvard University Press, 1989); Herbert Franke and Denis C. Twitchett, eds.,
Cambridge History of China, vol. 6, Alien Regimes and Border States, 907-1368 (Cambridge: Cambridge
University Press, 1994), 321-664; Hok-lam Chan, China and the Mongols: History and Legend Under
the Yüan and Ming (Aldershot: Ashgate-Variorum, 1999); Paul Jakov Smith and Richard von Glahn,
eds., The Song-Yuan-Ming Transition in Chinese History (Cambridge: Harvard University Asia Center,
2003).
11
Luciano Petech, Central Tibet and the Mongols: The Yüan–Sa-skya Period of Tibetan History
(Rome: Istituto Italiano per il Medio Estremo Oriente, 1990).
12
On the Mongol-Tibet relations during the Yuan see Dieter Schuh, Erlasse und Sendschreiben
mongolischer Herrscher für tibetische Geistliche, Monumenta Tibetica Historica Abteilung 3, Band 1
(St. Augustin: VGH-Wissenschaftverlag, 1977); Dieter Schuh, “Wie ist die Einladung des fünften
Karma-pa an den chinesischen Kaiserhof als Fortführung der Tibetpolitik der Mongolen-Khane zu
verstehen?,” in Altaica Collecta, ed. Walther Heissig (Wiesbaden: Otto Harrassowitz, 1976), 209-44;
Pema Tsering, “rNin ma pa Lamas am Yüan-Kaiserhof,” in Proceedings of the Csoma de Körös
Memorial Symposium held at Matrafüred, Hungary 24-30 September 1976, ed. L. Ligeti (Budapest:
Akademiao Kiado, 1978), 511-17; and David S. Ruegg, “The Preceptor-Donor (yon mchod) Relation
in Thirteenth Century Tibetan Society and Polity, Its Inner Asian Precursors and Indian Models,” in
Tibetan Studies: Proceedings of the 7th Seminar of the International Association of Tibetan Studies,
Graz 1995, ed. H. Krasser, M.T. Much, E. Steinkellner, and H. Tauscher (Wein: Österreichische
Akademie der Wissenschaft, 1997), 857-72.
Mongolian plateau.13 It was from this position that late in life he began courting
Tibetan Buddhists.
Why exactly he did so is uncertain, especially since for the previous twenty
years during his rise to power, he had been intimately involved with Chinese
members of the White Lotus Society. This messianic Buddhist sect had been
outlawed and persecuted by the Ming court and thus many of its members had
sought safety within Altan Qaγan’s domain,14 and the qaγan had embraced them.15
They built him palaces, brought Chinese technologies such as boat-building and
farming to the steppe, and acted as advisors on how to both attack and negotiate
with the Chinese state. The qaγan also employed those members of the White
Lotus with magical abilities during his campaigns, such as one who claimed he
could chant mantras that could shatter a town’s defensive walls.16 The fact that the
chanting had no effect after two days may explain some of the qaγan’s growing
disillusionment with the White Lotus; however, why he fully abandoned them as
well as other Mongol religious specialists in favor of the tantric Buddhism of Tibet
is still unclear. It may simply have been personal piety;17 however, the chaotic
political situation in tandem with the escalating social and environmental problems
of the period certainly played a role in this turn to the bla mas.18 Indeed, if we
accept Davidson’s central insight that tantric Buddhism thrived in periods of
socioeconomic change and political fragmentation, we may also apply this model
to the Mongol conversion in the late sixteenth century.19
Regardless of the reasons, however, this new alignment between a Mongol ruler
and Tibetan Buddhism initiated by Altan Qaγan was to have enormous historical
consequences. In 1578 these repercussions were clearly unforeseen. Rather, at the
time it was seen as a revival, or a righteous return to the old order that had been
lost when the Yuan dynasty fell.20 Thus everyone claimed that Altan Qaγan had
re-established the principle of the “Two Customs” and thereby initiated the “second
13
On the history of Altan Qaγan, his rise to power, and the consequences of his rule, see Elverskog,
Jewel Translucent Sutra.
14
On the history of the Chinese state’s persecution of the White Lotus, see B.J. ter Haar, The White
Lotus Teachings in Chinese Religious History (Leiden: Brill, 1992).
15
On Altan Qaγan’s relations with the White Lotus and the larger Chinese community that numbered
50,000 in his realm, see Elverskog, Jewel Translucent Sutra, 59-62, 102-8.
16
Henry Serruys, “Chinese in Southern Mongolia during the Sixteenth Century,” Monumenta Serica
18 (1959): 67-68.
17
Karénina Kollmar-Paulenz, Erdeni tunumal neretü sudur: Die Biographe des Altan qaγan der
Tümed-Mongolen (Wiesbaden: Harrassowitz, 2001).
18
Johan Elverskog, “Tibetocentrism, Religious Conversion and the Study of Mongolian Buddhism,”
in The Mongolia-Tibet Interface: Opening New Research Terrains in Inner Asia, ed. Uradyn E. Bulag
and Hildegard G.M. Diemberger, PIATS 2003, vol. 9 (Leiden: Brill, 2007), 59-80; Johan Elverskog,
“The Legend of Muna Mountain,” Inner Asia 8, no. 1 (2006): 99-122.
19
Ronald M. Davidson, Indian Esoteric Buddhism: A Social History of the Tantric Movement (New
York: Columbia University Press, 2002).
20
On the historiography of the Mongols’ loss of Buddhism and its revival, see Dominique Dumas,
“The Mongols and Buddhism in 1368-1578: Facts – Stereotypes – Prejudices,” Ural-Altaische
Jahrbücher 19 (2005): 167-221.
conversion” of the Mongols. Yet as gloriously as this event was praised at the time,
as foreshadowed above, it also entailed the inevitable seeds of discord.
The problem was twofold. The first was that Altan Qaγan set a new precedent.
He legitimated his power not through the traditional manner of Chinggisid blood,
but through a Buddhist discourse. In a sense, he opened the door for all the
competing descendants of Dayan Qaγan, who had all been enfeoffed within their
own specific territory to rule over its inhabitants, to do the same. And they did.
Following the precedent of Altan Qaγan, they rejected the centralized authority of
the Mongol qaγan and promoted their own local interests. Instead of sending taxes
and tribute to the Mongol qaγan as they had earlier been required to do, they sought
to increase their own power and prestige by establishing trade relations with the
Ming court, Muslim Inner Asia, and the growing power of the Manchus in the
northeast. Moreover, they received religious sanctification for these economic and
political moves by meeting with Tibetan Buddhist leaders. Emulating Altan Qaγan
as well as the archetypal Qubilai Qaγan, they not only received initiation into tantric
cycles of Buddhist teachings, but they were also recognized as cakravartin, the
“wheel-turning” emperor of Buddhist political theory.21 In essence, they became
“legitimate” Buddhist rulers of their own territorial fiefdoms.22 In turn, this political
fragmentation was exacerbated by the deteriorating state of the Ming dynasty,
which could no longer maintain its military defenses on the northern border or the
attendant north-south trade networks that enabled these “khanates” to prosper. The
result of all these factors was inevitably destabilizing, and as resources became
limited a ferocious civil war erupted among these recently enthroned Mongol
princes.23
The second problem related to this turn of events was the further issue of who
was actually doing the religio-political legitimating. Different Mongol princes
allied themselves with different Tibetan Buddhist hierarchs. Of course, who was
choosing whom could be debated. The Buddhist leaders were certainly not simply
passive pawns in the unfolding Mongol civil war. Rather, from all the available
evidence it seems to have been quite the opposite. Bla mas allied themselves with
Mongol princes in order to secure their military strength and used them to defeat
their own enemies. The Mongol civil war thus became hopelessly entangled with
21
One example of this phenomenon is Bosuγtu Jinong, (1565-1624), the ruler of Ordos, who had the
famous translator Siregetü Güüsi (fl. 1570-90) prepare a translation of the Ma ṇi bka’ ’bum and in the
colophon had himself identified as Srong btsan sgam po. See Walther Heissig, “Eine kleine mongolische
Klosterbibliothek aus Tsakhar,” Jahrbuch des Bernischen Historischen Museums in Bern (1961-62):
575 n. 68; and György Kara, “Une version mongole du Mani Bka’-’bum: le colophon de la traduction
Abaga,” Acta Orientalia Hungaricae 27 (1972): 36-37 n. 35.
22
On this process and its implications see Johan Elverskog, “The Story of Zhu and the Mongols of
the 17th Century,” Ming Studies 50 (2004): 39-76; and Johan Elverskog, “Sagang Sechen on the Qing
Conquest,” in The Black Master: Essays on Central Eurasia in Honor of Professor György Kara on
his 70th Birthday, ed. Stéphane Grivelet, Ruth Meserve, Àgnes Birtalan, and Giovanni Stary (Wiesbaden:
Harrassowitz Verlag, 2005), 43-56.
23
On the civil war and its consequences see Nicola Di Cosmo, “Military Aspects of the Manchu
Wars against the Caqars,” in Warfare in Inner Asian History (500-1800), ed. Nicola Di Cosmo (Leiden:
Brill, 2002), 337-67.
the long simmering religio-political struggles of Tibet.24 And although this had
enormous consequences for the subsequent history of all the players in this drama,
from the Rin spungs rulers of Gtsang to the Bka’ brgyud and ’Bri gung pa schools,
the impact upon the Ge lugs pa was probably the most profound.
The Ge lugs pa was the youngest Tibetan order. The reformer Tsong kha pa
(1357-1419) had only founded it in the fifteenth century, and since it did not have
the political and military backing of the powerful Tibetan families it was also the
weakest politically. Nevertheless, or perhaps because of this, Altan Qaγan met the
leader of this new religious order near Köke Naγur on June 5, 1578. Yet regardless
of the exact motivating reasons behind this meeting, it was only after the ’Bras
spungs abbot gave the qaγan the illustrious title “Brahma, the Great Mighty
Cakravartin” that Altan bestowed on Bsod nams rgya mtsho the title “Dalai Lama.”
Although today the Dalai Lama is recognized worldwide as the spiritual and
political leader of Tibet, in the late sixteenth century this was not the case. Rather,
Bsod nams rgya mtsho was known for what he was, the abbot of ’Bras spungs
Monastery and leader of the fledgling Ge lugs pa order. Indeed, in the 1560s the
future of the Ge lugs pa was tenuous, since the power of the Rin spungs rulers,
who supported the rival Karma pa, had greatly expanded. In 1565 they had captured
the major city of Gzhis ka rtse and controlled all of south-central Tibet. In response,
Bsod nams rgya mtsho had traveled around Tibet seeking support. And although
he received backing from minor rulers on the Indian border and in central Tibet
as well as from the Phag mo gru pa, the weakened former rulers of Tibet, they
were no match for the Rin spungs. Indeed, when Altan Qaγan’s envoys arrived in
1572, the Rin spungs’s army had almost reached Lha sa. Bsod nams rgya mtsho’s
decision to meet with the Mongol ruler was thus not solely religiously motivated.
Of course, by meeting with the qaγan the teachings of the Ge lugs pa
subsequently did in fact spread far and wide. In A mdo the newly anointed Dalai
Lama not only founded the famous Sku ’bum Monastery but also sent gifts to
Chinese officials in Gansu. In turn, the Wanli emperor (1563-1620, r. 1573-1620)
sent the Dalai Lama gifts and a title. The Dalai Lama also strengthened the Ge lugs
pa presence in eastern Tibet by winning the support of the prince of Byang. Yet it
was among the Mongols that he achieved his greatest success. Traveling to
Mongolia upon Altan Qaγan’s death in 1582, he was invited to meet with numerous
Mongol princes such as Sengge Dügüreng Qaγan (1522-86, r. 1582-86) of the
Tümed, Abatai Qaγan (1554-88) of the Qalq-A, and Kündülen Qaγan (fl. 1570-80)
of the Qaracin. In turn, all of these local rulers were not only recognized as rightful
rulers of their own “khanates,” but they in return also promoted the Ge lugs pa
tradition among the Mongols.
24
Zahiruddin Ahmad, Sino-Tibetan Relations in the Seventeenth Century (Roma: Istituto Italiano
per il Medio ed Estremo Oriente, 1970); and Elliott Sperling, “Notes on References to ’Bri-Gung-pa -
Mongol Contact in the Late Sixteenth and Early Seventeenth Centuries,” in Tibetan Studies: Proceedings
of the 5th Seminar of the International Association of Tibetan Studies, ed. Shoren Ihara and Zuiho
Yamaguchi (Narita: Naritasan Shinshoji, 1992), 741-50.
However, four years after his arrival in Mongolia, the Dalai Lama passed away.
This unfortunate event created a power vacuum in the Ge lugs pa tradition at a
pivotal moment in its history. It also opened the possibility for the political
manipulation of the debate over the recognition of the Dalai Lama’s reincarnation.
This was nothing new. The succession of reincarnate bla mas has always been
fraught with political tension. Indeed, this long tradition continues today with the
competing claims for the rightful reincarnation of both the Karma pa25 and the
Eleventh Paṇ chen bla ma.26 Much in the same way, the death of the Third Dalai
Lama in 1588 opened the possibility of utilizing the identification of his
reincarnation for political or religious gain.
In this regard the Mongols acted first. In 1589 they recognized Altan Qaγan’s
great-grandson as the reincarnated Fourth Dalai Lama,27 and to have this confirmed
by Ge lugs pa authorities they sent Rje drung Qutuγtu to Lha sa.28 Furthermore,
Dorji Baγsi, Darqan Medegci, and others were sent to present offerings at the Jo
khang in Lha sa and to escort a Tibetan delegation back to Mongolia. In addition,
in order to meet the Tibetans, Namudai Secen (1586-1607), Altan Qaγan’s grandson
and current ruler of the Tümed, set out with the Tümed elite to the Köke Naγur
region. Having settled in for the winter, Namudai invited a Stag lung official29 to
come to Köke Naγur and celebrate the New Year of 1591. The Stag lung “Dharma
Lord” (Mon. corji, Tib. chos rje) thus led a group of Ge lugs pa and Bka’ brgyud
monks in the Smon lam chen mo ritual.30 Afterwards, in the spring, when the
Tibetan delegation arrived from Lha sa, the Tümed Mongols escorted them to
25
On the dispute surrounding the recognition of the Seventeenth Karmapa see Anonymous, The
Karmapa Papers, Paris: 1992.
26
The Dalai Lama and the Tibetan-Government-in-Exile selected one child born in Tibet, and the
Chinese Communist Party chose another. See Isabel Hinton, The Search for the Panchen Lama (New
York: W.W. Norton & Company, 2001). The boy the Dalai Lama recognized is now under house arrest
with his entire family in an unknown location, while the child selected by the Party and trained in
Beijing is now touting to the world the glorious religious freedoms of the People’s Republic of China
(“Praise for Beijing by a Lama It Appointed,” New York Times, December 27, 2005, A5).
27
One reviewer of this article noted that some Tibetan sources claim the Fourth Dalai Lama was
Altan Qaγan’s grand-nephew and wondered whether this is found in Mongolian sources. The answer
is no. In Mongolian sources the Fourth Dalai Lama’s father is Sümbür Secen Cökügür, also known as
Sümir Dayicing, who was the son of Sengge Dügüreng, Altan Qaγan’s eldest son. Mongolian sources
further note that Sümbür was Sengge Dügüreng’s fourth son. Chinese sources, on the other hand, record
him as the fifth son (Elverskog, Jewel Translucent Sutra, 195-96 n. 23).
28
Rje drung Qutuγtu, or Lha dbang chos kyi dbang phyug, was the twelfth abbot of Chab mdo’s
Byams pa gling Monastery who had escorted the Third Dalai Lama to Mongolia (Elverskog, Jewel
Translucent Sutra, 193 n. 15).
29
Although this official is otherwise unidentified, the biography of the Third Dalai Lama notes that
Stag lung chos rje kun dga’ bkra shis accompanied Bsod nams rgya mtsho to Mongolia after Altan
Qaγan’s death. This is presumably the same individual (Elverskog, Jewel Translucent Sutra, 197 n.
30).
30
Tsong kha pa had founded the Great Prayer Festival in 1409 in order to commemorate the two
weeks during which the Buddha manifested miraculous powers in order to defeat heretics; and through
the years it became a greatly contested site of power between competing factions in Tibet. The
ecumenical nature of this celebration held in Köke Naγur in 1591 therefore needs to be noted. See
Elverskog, Jewel Translucent Sutra, 60-62; and Elverskog, “Tibetocentrism, Religious Conversion
and the Study of Mongolian Buddhism.”
White Lake where the Fourth Dalai Lama had been born.31 The Tibetans then
recognized him and bestowed on the boy his new name, given to him by the head
of Dga’ ldan Monastery. The young Dalai Lama was then brought to the Tümed
capital of Kökeqota, where he was installed on the “Lion Throne.” Then, while
sitting on the skin of a black antelope,32 the four-year old Dalai Lama began teaching
the dharma. Moreover, during his tenure in Kökeqota between 1592 and 1600,
other monks translated an unidentified Prajñāpāramita Sūtra into Mongolian.
Then in 1601, according to the earliest Mongolian source detailing these events,
the Dalai Lama decided to present offerings at the Jo khang in Lha sa and also to
invite Tibetans to visit him in Mongolia. Several Mongol noblemen were thus
dispatched to Tibet and after offering gifts at numerous shrines and monasteries
they invited Tibetans to join them as they returned. The Tibetans agreed, and when
they arrived, as may have been expected, they requested the Dalai Lama to return
with them to Tibet. The Mongols, however, objected. They wanted the Dalai Lama
to stay in Mongolia. However, they finally acquiesced and allowed the young child
to be brought back to Tibet. Yet as seen in the 1607 Jewel Translucent Sūtra they
were not pleased by this turn of events.
They invited the treasurer, who had been truly important during the Dalai Lama’s
previous supreme birth, with
Others in the white court and the samgha to come [to Mongolia].
Dülwa Chöje who was supreme among the samgha at Ganden [Phodrang]
monastery,
The unhesitating messenger of the Yarlung Emperor, and the messengers of the
Three Seats,
All together with the messengers of Tibet’s greater and lesser lords,
Came in order to invite the All-knowing Dalai Lama.
Immediately after arriving, the Treasurer, Dülwa Chöje and Tibet’s wise lamas
Were suspicious and said, “Is this, or not, the incarnation of the Victorious Bodily
Incarnated Dalai Lama?”
Thereupon, when the Dalai Lama clearly showed the signs, understood the deeds
in secret,
And particularly when he suppressed all with his grandeur, everyone worshipped,
prayed to, honored and blessed him.
Afterwards the All-knowing Dalai Lama personally,

Said to all of his Alms-masters in the land of the Mongols,
“There, before the Juu Sakyamuni of the Western Eternal Land,
It is the custom that from the holder of the Dharma Throne of the Victorious
Tsongkhapa that I am to become a monk.”
31
In the biography of the Fourth Dalai Lama this name is given as “White Stūpa” (mchod rten dkar
po), yet in Mongolian sources it is “White Lake,” which lies in Ulaanchab League of Inner Mongolia
(Elverskog, Jewel Translucent Sutra, 199 n. 37).
32
Sarat Chandra Das notes that whoever sits on the skin of a black antelope remembers the vows
and duties of a bodhisattva (A Tibetan-English Dictionary [New Delhi: Gaurav, 1985 reprint], 170-71).
When this had been said,

The innumerable greater and lesser lords assembled
And discussed it. Then the Khan, Queen and everyone
Replied to the Liberator Dalai Lama’s decree.
“While being compassionate on all beings without distinction,

Immediately as you noticed those in this direction with a merciful and gracious
mind,
By being wonderfully compassionate you were born in our royal lineage.
Thus, All-knowing, deign to become a monk in front of this Juu Sakyamuni.”33
The wise Dalai Lama replied, “According to the custom, aside from the holder
of the Dharma Throne,
In this direction there is no one to be my lama!
Thus in order to help all beings I shall depart and become a monk. Please do not
try to dissuade me.
If one helps me in this deed of mine, his reward and recompense will be great.”
When he did not agree [with the Mongols] and left,
The Three West Tümen and the Three East Tümen were greatly devastated.
When he was preparing to leave the State of the two Great Nations,
The magisterial, victorious Dalai Lama, personally,
Assembled the Six Tümen in front of the wondrous Juu Sakyamuni,
And a decree that was like an indestructible vajra issued from his evenly
compassionate mind.
He thus made those Six Tümen enter into his peace by skillful means.
He reformed and corrected the State and Religion created by the ancestors,
And unhesitatingly pacified the State of the Great Nation.
This was the decree of the always All-knowing Dalai Lama.

In the Black Tiger Year [1602], all the Alms-masters agreed,
Thus on an auspicious day of the 8th month [the Dalai Lama] went towards the
Western Eternal Land. While Namudai Sechen Khan and Queen Jönggen were
unable to part from the
All-knowing Dalai Lama, who was reincarnated in their royal clan, nevertheless,
they had no choice.34
As is evident from this passage, the Tümed Mongols clearly wanted to keep the
Dalai Lama in Mongolia; however, for whatever reasons they also eventually
allowed the child to go to Tibet. Regardless, the fact that this event occurred and
that it has not been entirely expunged from the historical record, as was later to be
the case in both Mongolian and Tibetan sources, reveals the tensions behind this
decision. Indeed, the Tümed Mongols clearly knew that having a Dalai Lama
33
This refers to the Juu Sakyamuni statue Altan Qaγan had Nepalese craftsmen make and then install
in a temple in Kökeqota (Elverskog, Jewel Translucent Sutra, 174).
34
Elverskog, Jewel Translucent Sutra, 205-7.
within their family or community had powerful repercussions, especially at a time

of increasing religio-political strife.
At this point it is important to recall what happened upon Altan Qaγan’s death.
Although he had been able to maintain relative peace on the Mongolian plateau
through his charisma, financial resources, and military power, once he died none
of his heirs had the same abilities. As a result, all the Dayan Khanid princes asserted
their own right to rule by building on both Altan’s defiance of the Mongol
genealogical model of rightful rule and his newly revived Buddhist political system.
All the power, territory, and prestige Altan Qaγan had acquired was thus quickly
dissipated. This was especially the case during the reign of his eldest son and
immediate successor Sengge Dügüreng. One reason for this was no doubt that he
was a drunk. Yet this was only one of his problems.
As has often been the case in Mongol history, the collapse of the house of Altan
Qaγan was tied into the more immediate context of the family.35 And in this case
the feud was between Sengge Dügüreng and his “step-mother,” Queen Jönggen
(Jönggen Qatun, 1551-1612), Altan Qaγan’s third and favorite wife.36 These two
had long been at odds, primarily because of Altan’s treatment of his first wife,
Sengge’s mother, after he married Jönggen. As a result, Sengge had a long-running
feud with both his father and his father's younger wife. Yet upon Altan’s death,
Sengge returned from his territory in the east in order to claim his father’s territory,
and more importantly his title and seal that enabled legal trade with the Ming court.
When he returned, however, Jönggen Qatun had taken control of the official seal.
Wielding a great deal of power, she set herself up independently. At first she
planned on remaining outside the control of the other Mongol rulers, which could
have been feasible, since with the Ming seal she controlled the tribute relations.
Yet she also wanted Budasiri (1568-97), her eldest son and Altan Qaγan’s seventh
son, to be enfeoffed with Altan Qaγan’s title of Shunyi Wang, instead of Sengge
Dügüreng, the rightful heir in primogeniture. The Ming court’s decision – based
on its receipt of a petition from seventy-nine Mongol noblemen to bestow the rank
of Shunyi Wang on Sengge Dügüreng and to maintain the status quo of previous
policy – was finally to present Sengge with the title. One stipulation of this
recognition, however, was that Sengge Dügüreng had to marry Jönggen Qatun,
whom the Ming perceived as being “pro-Chinese.” They married in
October-November 1582.
Matrimonial bliss did not ensue. Instead, Jönggen Qatun continued her struggle
for control. Taking advantage of Sengge’s ineptitude on account of his alcoholism,
she took his best troops and installed herself to the west of Kökeqota in such a way
that she was independent of the qaγan. Furthermore, in order to consolidate her
power, she wanted her son Budasiri, who had been denied the rank Shunyi Wang,
35
On the problem of family feuding during Mongol successions, or what Joseph Fletcher called
“bloody tanistry,” see his “The Mongols: Ecological and Social Perspectives,” Harvard Journal of
Asiatic Studies 46 (1986): 11-50.
36
The following historical summary is drawn from Elverskog, Jewel Translucent Sutra, 36-38.
to marry Baγa Beiji. She had been married to one of Altan Qaγan’s adopted sons,
Dayicing Ejei, and when he died in 1583 she had inherited control of the capital.
Dayan Khiya (d. 1591), another of Altan Qaγan’s adopted sons and the true
powerbroker in the capital, opposed her actions. Instead Namudai Secen, the eldest
son of Sengge Dügüreng, married Baγa Beiji in 1584, thwarting Jönggen’s plans.
After Sengge’s death in 1586, the struggle for control of this title continued between
Jönggen Qatun’s son Budasiri and Sengge’s eldest son Namudai Secen. Yet, once
again, 280 Mongol noblemen, who opposed Jönggen and Budasiri wrote a petition
to the Ming court in favor of Namudai Secen. Before he received the title, the Ming
court again pressured him to marry Jönggen Qatun, which he did; and on May 3,
1587, the Ming bestowed on him the title Shunyi Wang and the privilege of trade
and tribute relations it entailed.
Yet that was not the end of these familial entanglements. After Namudai’s
divorce from Baγa Beiji and his marriage to Jönggen Qatun, her son Budasiri, as
she had earlier wished, was able to marry Baγa Beiji. She eventually gave birth to
a son, named Sodnam. It is he who is named along with the Fourth Dalai Lama in
the inscription of this ceremonial staff.
The Inscription
As noted above, the inscription consists of eleven lines of Tibetan in dbu can script
and fourteen lines of classical Mongolian in the late sixteenth- and early
seventeenth-century script37 (Figures 2-4).
Figure 2.
37
On Mongolian paleography see György Kara, Books of the Mongolian Nomads: More than Eight
Centuries of Writing Mongolian (Bloomington: Research Institute for Inner Asian Studies, 2005).
Figure 3.
Figure 4.
All of the letters are damascened in either silver or gold. In the Mongolian text
only a few important terms, such as Sodnam Ombu Taiji, are “highlighted” in gold.
On the other hand, each line of the Tibetan text alternates between gold and silver.
Moreover, in the Tibetan inscription many of the superscribed vowels are the
opposite of the line itself; thus, a golden line will have silver vowels. Unfortunately,
however, the Tibetan text is badly corroded and the transcription below is tentative
at best.
Nevertheless, what can be gleaned from the Tibetan is that it begins with a
mantra, which is not found in the Mongolian, although the Tibetan and Mongolian
are in accord with the identification of the lay sponsor of the staff. In line 9 of the
Tibetan we find patron Sönam (Sbyin bdag bso[d] na[ms]), a clear parallel to his
full name found in line 11 of the Mongolian text: Sodnam Ombu Taiji. However,
only the Mongolian has the name of the Fourth Dalai Lama, which is actually a
Mongolian translation Erdem-Ün Dalai of the Tibetan Yon tan rgya mtsho. Finally,
the Tibetan text seems to conclude with the name of the artist who made the staff,
Nasartu ([b]zo bo na sar th’i); however, since this is not corroborated by the
Mongolian it must remain a conjecture. The Mongolian inscription is better

preserved than the Tibetan, though it too has many lacunae. Even so, it does offer
us the name of the two individuals involved with the creation of the staff, and it
also serves as the basis for the translation below.
Transcriptions
Transcription Key
x – unreadable letter
[] – extrapolated letter
- - - – text missing
italics – gold letters
Tibetan Inscription
1. § ya - - - - sam sa sri s[r]i xhum / /
2. - - - - - - - - - - - - xgrub zhax x /
3. x ma x - - - ga x x - - - - x x bas gyi ’grox
4. don x - - - - - - - - - - - - - - - - - - - x x lba’i
5. x u x x x [ca]x - - - - - - - - - x x / / bka’ . x x
6. cho skya x - - - - - - - - - - - - - x rab phyugs
7. x mag x r . rgyal x mas x x sa x bsrung rg[ru] . mtsha x khor
8. x x x ca - - - - - - - - - - - - - - - x x x - - - - - - x rtan x phug ’di
9. x x x pa x sbyin bdag . bso[d] na[ms] ’ja x / / dad . ba’ x u x
10. x x x - - - - - - - - - - skya x x - - - - - - - x [sa] - - x x . mthun
11. rgyxn . ’grub . par mdzod / [b]zo bo na sar th’i . sa x la x
Mongolian Inscription
1. § --- rxx --- xxx qxxxm xxxlxxjxsaγar-a xxsu xxxdxmxx
2. ------------ xxx xin yi xxo[s]xxxi xx xxxui orui
3. ------ du[r] xxxi-i umxxxx x xxxxxx xxxxxx
4. -------------------------- tusa-yi jokiyaγci öljei-tü
5. nigülesküi-yin e[j]en qubilγan bey-e tü blam-a
6. adalidqasi ügei erdem-ün dalai yin köl
7. [tür] s[ögüd]dümüi :: tegülder nom-i t[e]ngcigülü[n]
8. xxxxxxx kixxx taγalaγulun oron-u erketü xxx
9. xxxxxx daxx --------xxx sasin-u dalai-yi sakiγcid
10. xxxx-e --- x --- x –erdeni ber cimayi xxx xxx xxxx a
11. ------------ bsodnam o[m]bu taiji xxxxx bisirel
12. x ---------------------------------- x xxxx xxxxxxxxx
13. sili[γu] ----------- x ---------- elci xx joriγ xxxx x yin
14. xxxxx ---------- xxx xxxx x x x sogdxx xxx xxxx
As is clear from the transcription, the inscription is badly damaged. Nevertheless,
from the fragmentary passages we can deduce that it is a rather standard panegyric
of a Mongol layman espousing his support of a Tibetan bla ma. From the words
that are legible one can “translate” it as follows:
I bow down to the feet of the unparalleled Yon tan rgya mtsho, the incarnated bla
ma of the Lord of Auspicious Compassion, the One who Helps [Avalokiteśvara]
… [Towards] you, the perfect Dharma evenly … love and benevolent … protector
of the ocean of Dharma … jewel … Sodnam Ombu Taiji … faith …
This may not seem like much. However, the identification of Sodnam Ombu
Taiji and Yon tan rgya mtsho, the Fourth Dalai Lama, allows us to put this
inscription and the socket itself into a historical context.
Conclusion
The socket was created during the increasing political fragmentation of the Mongols
during the early seventeenth century. As resources waned, the competing Mongol
princes tried to secure their own local power by obtaining political legitimacy
through their consecration as Buddhist rulers by various Tibetan hierarchs. In
particular, this socket reveals Sodnam Ombu Taiji’s religio-political relation with
the young Fourth Dalai Lama. It does not, unfortunately, offer us a precise date
for its creation. Yet if we return to the available historical records, some suggestions
can be offered.
When the Fourth Dalai Lama was returned to Tibet in 1602, Sodnam Ombu
Taiji was fourteen or fifteen years old. This seems a rather young age to be
proclaiming one’s faith as a patron of the dharma and having it immortalized by
means of a silver and gold damascened ceremonial staff. It is certainly not
impossible, though a more likely date seems to be 1607. By this time Sodnam was
twenty and in the prime of his life. Moreover, by this time he had indeed
consolidated the power base fostered by his grandmother Jönggen Qatun and made
himself into the most powerful ruler in Ordos. Thus when Namudai Secen Qaγan
died on May 15, 1607, Sodnam presumably believed it was finally his lineage’s
turn to take control of the Tümed by claiming Altan Qaγan’s mantle. And what
better way to cement one’s reputation than by promoting one’s Buddhist bona fides
by supporting the Dalai Lama, himself a direct descendant of Altan Qaγan? Indeed,
this was the tactic adopted by all the Ordos princes vying for power. They
commissioned translations, printed sūtras, built stūpas and monasteries, and had
elaborate histories written in which they declared themselves akin to the legendary
Buddhist kings of Tibet. This socket appears to be another monument similar to
these others.
Yet in thinking about the origin of the socket, it is also important to recall the
history of the Dalai Lama during this period of time. In 1605 the Phag mo gru pa
attacked Lha sa, and in response the Mongols marched into Central Tibet in 1607.
They were ultimately repelled by the powerful Rin spungs forces, and in the melée
the Dalai Lama fled, only to return to Lha sa in 1611. Whether Sodnam was directly
involved in these operations is unknown. However, throughout these tribulations
it is not unimaginable to think that Sodnam, the most powerful Mongol ruler of
Ordos, sent troops or other resources to help the Dalai Lama. He might have gone
to Central Tibet for faith or family: the Fourth Dalai Lama was his nephew. It
would certainly have burnished his own image as a protector of the dharma and
as a Buddhist ruler. And how better to seal such a religio-political alliance than
with a ceremonial staff or spear inscribed for eternity with a vow of allegiance
between the two parties?
Of course, this is still only a conjecture. There may have been another event or
meeting that took place between Sodnam Ombu Taiji and the Fourth Dalai Lama
that is now otherwise lost to the historical record. However, the Dalai Lama passed
away in 1617 at the age of twenty-eight. Sodnam himself never won the Tümed
throne. It went to Namudai Secen’s son, Bosuγtu Qung Taiji, who although weak
politically, outlived his rival. Sodnam died in 1625. Bosuγtu Qung Taiji then
married his widow and annexed his land and wealth. And these struggles and feuds
may have continued if it had not been for the rise of the Manchus, who, by
continuing with the Buddhist model revived by Altan Qaγan, once again reunited
the Mongols under the Qing dynasty.38
Thus even though we lack the historical precision we may desire, this iron socket
and its inscription is an important monument suggesting a heretofore unknown but
important chapter in this pivotal period of Inner Asian history.
38
On the history of the Mongols and Buddhism during Manchu rule see Johan Elverskog, Our Great
Qing: The Mongols, Buddhism and the State in Late Imperial China (Honolulu: University of Hawai’i
Press, 2006).
Glossary
Ka
karma pa Karmapa Person
karma pag shi Karma Pakshi 1206-1283 Person
sku ’bum Kumbum Monastery
Ga
dga’ ldan Ganden Monastery
rgyal bu ji big de mur Gyelbu Jibikde Murla Text
la gdams du byas nor Damdujé Norbü
bu’i phreng ba Trengwa
Nga
ngag dbang rnam gyal Ngawang Namgyel Person
Cha
chab mdo Chamdo Place
chos rje Chöjé Term
chos ’byung ngo Chöjung Ngotsar Text
mtshar rgya mtsho Gyatso
mchod rten dkar po Chöten Karpo Place
Ja
jo khang Jokhang Temple
rje drung qutuγtu Jedrung Khutughtu Person
Nya
rnying ma Nyingma Organization
Ta
stag lung Taklung Organization
stag lung chos rje kun Taklung Chöjé Künga Person
dga’ bkra shis Trashi
Da
’dul ba chos rje Dülwa Chöjé Person
Pa
paṇ chen bla ma Penchen Lama Person
Pha
phag mo gru pa Pakmo Drupa Organization
'phags pa Pakpa Person
'phags pa bla ma Pakpa Lama 1235-1280 Person
Ba
byang Jang Place
byams pa gling Jampaling Monastery
dbu can uchen Term
'bras spungs Drepung Monastery
'bri gung pa Drigungpa Organization
bla ma lama Term
sbyin bdag bso[d] Jindak Sönam Person
na[ms]
Ma
ma ṇi bka’ ’bum Mani Kambum Text
smon lam chen mo Mönlam Chenmo Festival
Tsa
gtsang Tsang Place
tsong kha pa Tsongkhapa 1357-1419 Person
Zha
gzhis ka rtse Zhikatsé Place
Ya
yar lung Yarlung Valley
yon mchod yön chö Term
yon tan rgya mtsho Yönten Gyatso 1589-1617 Person
Ra
rin spungs Rinpung Place
La
lugs gnyis luknyi Term
lugs zung lukzung Term
Sha
shes bya rab gsal Sheja Rapsel Text
Sa
srong btsan sgam po Songtsen Gampo Person
bsod nams rgya mtsho Sönam Gyatso 1543-1588 Person
Ha
lha dbang chos kyi Lhawang Chökyi Person
dbang phyug Wangchuk
lha sa Lhasa Place
A
a mdo Amdo Place
Mongolian
Wylie Phonetics English Mongolian Dates Type
Abatai Khan abatai qaγan 1554-1588 Person
Altan altan 1508-1582 Person
Altan Khan altan qaγan 1508-1582 Person
Bagha Beiji baγa beiji Person
Boshugtu Jinong bosuγtu jinong 1565-1624 Person
Boshugtu Khung Taiji bosuγtu qung taiji Person
Budashiri Budasiri 1568-1597 Person
Chinggis Khan cinggis qaγan 1167-1227 Person
corji Term
Darkhan Medegchi darqan medegci Person
dayan khiya d. 1591 Person
Dayan Khan dayan qaγan 1474-1517? Person
Daiching Ejei dayicing ejei Person
Dorji Baghshi dorji baγsi Person
erdem-ün dalai Person
Queen Jönggen jönggen qatun 1551-1612 Person
Kökenuur köke naγur Place
Hohhot kökeqota Place
Khündülen Khan kündülen qaγan fl. Person
1570-1580
Möngke Khan möngke qaγan r. Person
1251-1259
Namudai namudai Person
Namudai Sechen namudai secen 1586-1607 Person
nom-un yosun Term
alms-master öglige-yin ejen Term
Ordos Place
Khalkha qalq-a Clan
Kharachin qaracin Clan
khan qaγan Term
qoyar yosu Term
Khubilai qubilai 1215-1294 Person
Khubilai Khan qubilai qaγan 1215-1294 Person
Khutugtai Sechen qutuγtai secen qung Person
Khung Taiji taiji
Saghang Sechen saγang secen Person
Sengge sengge Person
Sengge Düüreng sengge dügüreng Person
Sengge Düüreng sengge dügüreng 1522-1586, Person
Khan qaγan r.
1582-1586
Shireetü Güüshi siregetü güüsi fl. Person
1570-1590
Sodnam sodnam Person
Sodnam Ombu Taiji sodnam ombu taiji 1587-1625 Person
Sümbür sümbür Person
Sümbür Sechen sümbür secen Person
Chüüghür cökügür
Sümir Daiching sümir dayicing Person
offering-site takil-un oron Term
törö-yin yosun Term
Tümed tümed Clan
Sanskrit
Avalokiteśvara Deity
cakravartin Term
dhāranī Term
Hevajra Deity
mantra Term
maṇḍala Term
Prajñāpāramita Text
Sūtra
stūpa Term
sūtra Term
tantra Term
tathāgatha Term
Vajrayāna Doxographical
Category
Chinese
Gansu Place
Shunyi Wang Term

Wanli 1563-1620, Person
r.
1573-1620
Yongle 1360-1424, Person
r.
1403-1424
Yuan dynasty Yuan 1272-1368 Organization
Bibliography
Ahmad, Zahiruddin. Sino-Tibetan Relations in the Seventeenth Century. Roma:
Istituto Italiano per il Medio ed Estremo Oriente, 1970.
Anonymous. “Praise for Beijing by a Lama It Appointed.” New York Times,
December 27, 2005, A5.
———. The Karmapa Papers. Paris: 1992.
Atwood, Christopher P. Young Mongols and Vigilantes in Inner Mongolia’s
Interregnum Decades, 1911-1931. Leiden: Brill, 2002.
Bira, Sh. “Qubilai Qa’an and ’Phags-pa Bla-ma.” In The Mongol Empire and Its
Legacy, edited by Reuven Amitai-Press and David Morgan, 244-49. Leiden:
Brill, 1999.
Chan, Hok-lam. China and the Mongols: History and Legend Under the Yüan
and Ming. Aldershot: Ashgate-Variorum, 1999.
Chan, Hok-lam, and W.T. de Bary, eds. Yüan Thought: Chinese Thought and
Religion Under the Mongols. New York: Columbia University Press, 1982.
Das, Sarat Chandra. A Tibetan-English Dictionary. New Delhi: Gaurav, 1985.
Davidson, Ronald M. Indian Esoteric Buddhism: A Social History of the Tantric
Movement. New York: Columbia University Press, 2002.
———. Tibetan Renaissance: Tantric Buddhism in the Rebirth of Tibetan Culture.
New York: Columbia University Press, 2005.
de Rachewiltz, Igor, and John R. Krueger. Erdeni-yin Tobci (‘Precious Summary’):
A Mongolian Chronicle of 1662. Vol. 1. Canberra: The Australian National
University, 1990.
Di Cosmo, Nicola. “Military Aspects of the Manchu Wars against the Caqars.”
In Warfare in Inner Asian History (500-1800), edited by Nicola Di Cosmo,
337-67. Leiden: Brill, 2002.
Dumas, Dominique. “The Mongols and Buddhism in 1368-1578: Facts –
Stereotypes – Prejudices.” Ural-Altaische Jahrbücher 19 (2005): 167-221.
Elverskog, Johan. The Jewel Translucent Sutra: Altan Khan and the Mongols in
the Sixteenth Century. Leiden: Brill, 2003.
———. “The Story of Zhu and the Mongols of the 17th Century.” Ming Studies
50 (2004): 39-76.
———. “Sagang Sechen on the Qing Conquest.” In The Black Master: Essays
on Central Eurasia in Honor of Professor György Kara on his 70th Birthday,
edited by Stéphane Grivelet, Ruth Meserve, Àgnes Birtalan, and Giovanni
Stary, 43-56. Wiesbaden: Harrassowitz Verlag, 2005.
———. Our Great Qing: The Mongols, Buddhism and the State in Late Imperial
China. Honolulu: University of Hawai’i Press, 2006.
———. “Tibetocentrism, Religious Conversion and the Study of Mongolian
Buddhism.” In The Mongolia-Tibet Interface: Opening New Research Terrains
in Inner Asia, edited by Uradyn E. Bulag and Hildegard G.M. Diemberger,
59-80. Vol. 9 of Proceedings of the Tenth Seminar of the International
Association for Tibetan Studies 2003. Leiden: Brill, 2007.
———. “The Legend of Muna Mountain.” Inner Asia 8, no. 1 (2006): 99-122.
Endicott-West, Elizabeth. Mongolian Rule in China: Local Administration in the
Yuan Dynasty. Cambridge: Harvard University Press, 1989.
Fletcher, Joseph. “The Mongols: Ecological and Social Perspectives.” Harvard
Journal of Asiatic Studies 46 (1986): 11-50.
Franke, Herbert. “Tibetans in Yüan China.” In China under Mongol Rule, edited
by John Langlois, Jr., 296-328. Princeton: Princeton University Press, 1981.
Franke, Herbert, and Denis C. Twitchett, eds. Cambridge History of China. Vol.
6, Alien Regimes and Border States, 907-1368. Cambridge: Cambridge
University Press, 1994.
Heissig, Walther. “Eine kleine mongolische Klosterbibliothek aus Tsakhar.”
Jahrbuch des Bernischen Historischen Museums in Bern (1961-62): 557-90.
Hinton, Isabel. The Search for the Panchen Lama. New York: W.W. Norton &
Company, 2001.
Kara, György. “Une version mongole du Mani Bka’- ’bum: le colophon de la
traduction Abaga.” Acta Orientalia Hungaricae 27 (1972): 19-41.
———. Books of the Mongolian Nomads: More than Eight Centuries of Writing
Mongolian. Bloomington: Research Institute for Inner Asian Studies, 2005.
Kollmar-Paulenz, Karénina. Erdeni tunumal neretü sudur: Die Biographe des
Altan qaγan der Tümed-Mongolen. Wiesbaden: Harrassowitz, 2001.
Langlois, John D., ed. China Under Mongol Rule. Princeton: Princeton University
Press, 1981.
LaRocca, Donald J., et al. Warriors of the Himalayas: Rediscovering the Arms
and Armor of Tibet. New Haven: Yale University Press, 2006.
Petech, Luciano. Central Tibet and the Mongols: The Yüan–Sa-skya Period of
Tibetan History. Rome: Istituto Italiano per il Medio Estremo Oriente, 1990.
Rossabi, Morris. Khubilai Khan: His Life and Times. Berkeley: University of
California Press, 1988.
Ruegg, David S. “Mchod yon, yon mchod and mchod gnas/yon gnas: On the
Historiography and Semantics of a Tibetan Religio-social and Religio-political
Concept.” In Tibetan History and Language: Studies Dedicated to Uray Géza

on his Seventieth Birthday, edited by Ernst Steinkeller, 441-53. Wien:
Arbeitskreis für Tibetische und Buddhistische Studien Universität Wien, 1991.
———. Ordre spirituel et ordre temporel dans la pensée bouddhique de l’Inde
et du Tibet. Publications de l’Institut de civilisation indienne, Série in-8, Fasc.
64. Paris: Collège de France, 1995.
———. “The Preceptor-Donor (yon mchod) Relation in Thirteenth Century
Tibetan Society and Polity, Its Inner Asian Precursors and Indian Models.” In
Tibetan Studies: Proceedings of the 7th Seminar of the International Association
of Tibetan Studies, Graz 1995, edited by H. Krasser, M.T. Much, E.
Steinkellner, and H. Tauscher, 857-72. Wein: Österreichische Akademie der
Wissenschaft, 1997.
Schuh, Dieter. Erlasse und Sendschreiben mongolischer Herrscher für tibetische
Geistliche. St. Augustin: VGH-Wissenschaftverlag, 1977.
Serruys, Henry. “Chinese in Southern Mongolia during the Sixteenth Century.”
Monumenta Serica 18 (1959): 67-68.
———. “Wie ist die Einladung des fünften Karma-pa an den chinesischen
Kaiserhof als Fortführung der Tibetpolitik der Mongolen-Khane zu verstehen.”
In Altaica Collecta, edited by Walther Heissig, 209-44. Wiesbaden: Otto
Harrassowitz, 1976.
Smith, Paul Jakov, and Richard von Glahn, eds. The Song-Yuan-Ming Transition
in Chinese History. Cambridge: Harvard University Asia Center, 2003.
Sperling, Eliott. “Notes on References to ’Bri-Gung-pa – Mongol Contact in the
Late Sixteenth and Early Seventeenth Centuries.” In Tibetan Studies:
Proceedings of the 5th Seminar of the International Association of Tibetan
Studies, edited by Shoren Ihara and Zuiho Yamaguchi, 741-50. Narita: Naritasan
Shinshoji, 1992.
Tambiah, Stanley. World Conqueror, World Renouncer. New York: Cambridge
ter Haar, B.J. The White Lotus Teachings in Chinese Religious History. Leiden:
Brill, 1992.
Tsering, Pema. “rNin ma pa Lamas am Yüan-Kaiserhof.” In Proceedings of the
Csoma de Körös Memorial Symposium held at Matrafüred, Hungary 24-30
September 1976, edited by L. Ligeti, 511-17. Budapest: Akademiao Kiado.
The Importance of the Underworlds: Asuras’ Caves in
Buddhism, and Some Other Themes in Early Buddhist
Tantras Reminiscent of the Later Padmasambhava Legends
Robert Mayer
University of Oxford
Abstract: The story of Padmasambhava taming non-human females at the Asura

Cave at Pharping is well known. Much less widely known is the wider tradition of
Asuras’ caves as the entrances to Pātāla, the magical underworlds of Asuras and
Nāgas, a colorful and often eroticized and popular belief which played a prominent
role in early Indian and Chinese Buddhist tantras. This paper surveys these now
largely forgotten beliefs, and then proceeds to raise (but not answer) the question:
might further widely attested Kriyātantra themes, such as treasure recovery, kīlas,
and water magic, have influenced the popular mythology of Padmasambhava?
Asuras’ Caves in Buddhism

This article will look at a set of practices deriving from popular Indian culture that
were once widely attested within Indian and Chinese Buddhism. However, perhaps
as a consequence of their aims and methods eventually beginning to appear
somewhat tangential to those of mainstream Vajrayāna Buddhism as it evolved
over the last centuries of the first millennium CE, they seem to have ended up
somewhat marginal within Tibetan Buddhism, despite their survival in extant Bka’
’gyur texts. Sometimes known as attainment of Pātāla (pātālasiddhi), these practices
were focused on gaining access to the subterranean kingdoms of the Asuras and
Nāgas, which were often generically referred to as Pātāla, and which the
adventurous could enter via any one of the many Asuras’ caves identified within
the sacral landscape. Once in Pātāla, the yogin could gain such boons as longevity,
magical knowledge (vidyā), fabulous material treasures, and, not least, extraordinary
pleasures, especially erotic ones.
Asuras’ Caves in the Dunhuang Text Tib J 644

Despite being somewhat marginalized in contemporary Tibetan Buddhism, we
still have substantial surviving evidence from Sanskrit, Tibetan, and Chinese
1550-6363/2007/3/T3102.
© 2007 by Robert Mayer, Tibetan and Himalayan Digital Library, and International Association of Tibetan Studies.
Mayer: The Importance of the Underworlds 2
sources that such practices were probably once more central. Some of the most
recently identified evidence in the Tibetan language can be found in a valuable
article recently published by Jacob Dalton.1 In this article, Dalton discusses the
probably tenth-century Dunhuang text Tib J 644, which describes inter alia the
three levels of holders of magical knowledge (vidyādhara) associated with the
Tantric Buddhist vehicle of Kriyātantra. Dalton writes,
According to this text, there are three vidyādhara levels that can be attained
through the practice of the Kriyātantras: the vidyādhara of accomplishments (grub
pa’i rigs ’dzin), the vidyādhara who dwells on the levels (sa la gnas pa’i rigs
’dzin), and the spontaneously accomplishing vidyādhara (lhun gyis grub pa’i rigs
’dzin).2
It is the description of the second of these, the vidyādhara who dwells on the
levels, that is relevant to this discussion. Here the text describes the Bodhisattva
Vajrapāṇi appearing and granting attainments (siddhis) to the yogin, who then
proceeds to an Asura’s cave, where an emanation of Vajrapāṇi grants him a vision.
The yogin is then able to strike his foot into a rock, as though the rock were made
of dough. From the footprint comes a sacred flow, a spring with eight streams.
One of them flows to the south face of Mount Meru, and hence is called Aśvakarṇa.
The seven others flow inside the Asura’s cave. By bathing himself in this sacred
water, the practitioner becomes purified, and achieves attainments. This is how
the yogin achieves the accomplishment of a vidyādhara who dwells on the levels.3
It is surely more a testament to the comparative marginalization of the practice
of attainment of Pātāla within mainstream Tibetan Buddhism, and not a reason for
criticism of Dalton, that even such an erudite specialist in early Tibetan Buddhism
as he was initially unaware of what exactly these rites in Tib J 644 were referring
to.4 Hence Dalton limited himself to speculations surrounding their interesting
resemblance to well-known themes from the legends of Padmasambhava, notably
his stay in the Asura Cave at Pharping in Nepal. Yet a close examination of the
Tibetan text of Tib J 644 in its full context shows that it seems to be a reasonably
standard abstract presentation of doctrine, which is quite similar to descriptions of
attainment of Pātāla as found in other Kriyātantra texts in its structure, its
grammatical use of the third person to imply “the yogin,” and its content. So rather
1
Jacob Dalton, “The Early Development of the Padmasambhava Legend in Tibet: A Study of IOL
Tib J 644 and Pelliot tibétain 307,” Journal of the American Oriental Society 124, no. 4 (2004): 759-72.
2
Dalton, “Early Development of the Padmasambhava Legend,” 761.
3
de nas phyag na rdo rje gshegs nas/ dngos grub sbyin ba dang / a su ra’i brag phug du phyin pa
dang / de na phyag na rdo rje’i sprul pa gcig bzhugs pa’i zhal mthong nas brag la rkang pa gcig brgyab
pa dang / zan la brgyab bzhin snang ngo / rjes de nas dam babs nas/ nang de na chu myig yan lag
brgyad dang ldan ba brgyad yod pa la/ gcig ni ri rab kyi lho ngos su rdol te chu myig rta rna zhes
bya’o/ bdun a su ra’i nang na ’bab pa la khrus byed cing bsgrub pa de/ sa la gnas pa’i rigs ’dzin ces
bya’o/ (Dunhuang Tibetan manuscripts held at the British Library, London: IOL Tib J 644, 2a.6-2b.1).
4
Despite their occasional appearance in Tibetan sources, topics like pātālasiddhi are not primarily
seen as the domain of Tibetologists or even of Sanskritic Buddhologists, but more of Indologists and
Sinologists. Our established traditions of academic compartmentalization – that tends to inhibit
knowledge of popular Indian religions among Buddhologists – can have drawbacks for tantric studies.
than representing an earlier version of the Padmasambhava story as such (a point

which I am certain Dalton now understands, even though his actual words are a
little ambiguous on this point), the passage more likely describes a now
long-forgotten generic Pātāla-based ritual practice that many early Tantric Buddhists
might well have attempted.
In this paper, I wish to pick up where Dalton left off, and try to contextualize
the reference to the Asura cave in Tib J 644. Not only will this help us rediscover
a little about the largely forgotten Buddhist practice of Asura caves and attainment
of Pātāla, but it also offers an opportunity to draw our attention to the noteworthy
but as yet seldom mentioned continuities between several prominent Kriyātantra
practices5 and the later legends of Padmasambhava (unfortunately, we still have
rather little idea about the historical reality of Padmasambhava).
Asuras’ Caves in the Context of Buddhist Kriyātantra

In fact, despite Tibetan Buddhism’s comparatively meager interest in the subject,
Asuras’ caves feature surprisingly prominently in Indian mythology, magic, tantric
ritual, folklore, and cosmology, where they function as the entrances or gateways
to the subterranean paradises of immense beauty, wealth, and pleasure, often
enumerated as seven in number, and often generically called Pātāla. It is within
these subterranean paradises that Asuras (along with Nāgas and various other spirits
too) are believed to dwell. Pātāla moreover became the focus of a substantial body
of magical practices in Hinduism, Jainism, and Buddhism alike. Asuras’ caves
serving as the gateway to Pātāla are thus found in the epic literature, in tantric
scriptures, in magical texts, in tantric ritual manuals, and in narratives of many
kinds. More importantly for Tib J 644’s presentation of the vidyādharas of
Kriyātantra, Pātāla, its Asura inhabitants, and Asuras’ caves as Pātāla’s entrances
are also specifically found in several canonical Buddhist scriptures still extant in
Sanskrit, Tibetan, and Chinese. In fact, as far as I am currently aware, they mainly
occur in the early Buddhist tantras, especially Kriyātantra, such as the
Mañjuśrīmūlakalpa, or the *Kaṇikrodha-vajrakumāra-bodhisattva-sādhana-vidhi
(Sheng chia ni fen nu chin kang t’ung tzu p’u sa ch’eng chiu i kuei ching),6 and
the Ārya-vajra-pātāla-nāma-tantra-rāja (’phags pa rdo rje sa ’og gi rgyud kyi
rgyal po). I have not yet heard of significant references in later Buddhist tantric
genres, although it would of course not be at all surprising if they were to turn up
there occasionally. Despite these early tantric canonical references, Pātāla and its
Asura inhabitants no longer seemed to play a very significant role in later Tantric
5
Traditional doxographical categories such as Kriyā are well known to be inconsistent and full of
anomalies. For example, most of the substantial Dhāraṇī section in the dynastic catalogues of Lhan kar
ma and ’Phang thang ma were later reclassified by the Bka’ ’gyur codifiers as Kriyātantras. However,
especially since these doxographical categories were of such importance to traditional authorities, I
cannot agree with those modern scholars who reject them as meaningless. For convenience, I use Kriyā
here as a convenient shorthand for texts containing Buddhist esoteric materials in an earlier style. Such
usage avoids the need to coin my own neologism, and permits shorter sentences. I am aware that some
of what I describe also occurs in early texts that need not necessarily be traditionally classed as Kriyā.
6
T1222a, K1355.
Buddhism, and appear of little concern to modern Tibetan Buddhism. The words
and concepts sa ’og certainly still exist, largely with reference to Nāgas, or
occasionally as a generic term for the place where Padmasambhava buries his
treasures,7 and the Asuric reference is also clearly not entirely forgotten – but as
far as I know, Pātāla is no longer a living concern of Tibetan tantric ritual, even
though attainment of Pātāla might still persist in some old tantras and lists. The
most prominent Tibetan canonical sources for Pātāla I have found so far are two
Bka’ ’gyur texts dedicated to the subject. One is the ’Phags pa rdo rje sa ’og gi
rgyud kyi rgyal po (Ārya-vajra-pātāla-nāma-tantra-rāja),8 which in its main version
is a substantial text in twenty-five chapters purely devoted to attainment of Pātāla;
and its long colophon cites no lesser a personage than the great Sa skya pa Paṇ ḍi
ta Kun dga’ rgyal mtshan, helped by the Indian Sugataśrī. Mkhas grub rje classes
it as a Kriyātantra of the vajra family, and explains that it came in two further
editions as well, one in thirteen chapters by ’Phags pa shes rab, the translator of
Zanskar (zangs dkar), and one in seven chapters by the monk Bya gdong ba can.9
The other is the Rdo rje sa ’og gi rgyud (Vajra-pātāla-tantra)10 which is counted
by Mkhas grub rje as a Caryā text of the vajra family, but which Bu ston had
viewed with suspicion.11 Little awareness of these is reflected in more recent
western Buddhological scholarship, although Davidson proposes the Sanskrit
reconstruction *asuraguhā to describe rites he had encountered in the famous
Kriyātantra root text, the Subāhuparipṛcchā.12 Comparative marginalization over
the course of time is of course true of a great deal of the magical materials found
in old texts such as the Mañjuśrīmūlakalpa, which itself is a large and somewhat
motley Indian compendium built up over a long period, closely resembling a Hindu
purāṇa, and from which Tibetan Buddhism has tended to cherry-pick chosen useful
7
See for example the preamble to the Las byang of the gter ma cycle ’Chi med srog thig that is
associated with Zil gnon nam mkha’i rdo rje, the Fifteenth Karmapa, and the late Bdud ’joms rin po
che. Here, words are put into the mouth of Padmasambhava: “…Bearing in mind sentient beings’
wanderings in the degenerate (age), (I) filled the whole of the subterranean world with treasures…”
(snyigs ’gro’i sems can la dgongs te: sa ’og thams cad gter gyis bkang:; Zil gnon nam mkha’i rdo rje,
with contributions from Karmapa XV and Bdud ’joms rin po che ’jigs bral ye shes rdo rje, ’Chi med
srog thig las byang rdo rje’i phur pa yang gsang phrin las bcud dril gyi las tshe sgrub ’chi med srog
thig, in The Collected Writings and Revelations of H. H. Bdud-’joms Rin-po-che ’Jigs-bral-ye-ses-rdo-rje,
vol. 14 [Pha] [Kalimpong: Dupjung Lama, 1979], 77). Yet I doubt any reference to the Pātālas of Asuras
and Nāgas is intended here.
8
Sde dge 744, Stog 697, Peking 403, and Ulan Bator 767.
9
F. D. Lessing and Alex Wayman, Introduction to the Buddhist Tantric Systems (Delhi: Motilal
Banarsidass, 1983), 129.
10
Sde dge 499.
11
Lessing and Wayman, Introduction to the Buddhist Tantric Systems, 207. There is also a prominent
reference to Pātāla as a heading included in the Tibetan translation of Amarasiṃha’s Amarakośa, but
this is a famous Indian lexicographical work written around 450 CE and translated into many Asian
languages, rather than a Vajrayāna treatise. This heading was cited by Mkhas grub rje (Lessing and
Wayman, Introduction to the Buddhist Tantric Systems, 79).
12
Ronald M. Davidson, Indian Esoteric Buddhism: A Social History of the Tantric Movement (New
York: Columbia University Press, 2002), 200.
items while often ignoring much of its plentiful magical materials.13 Pātāla is
perhaps more a location within popular Indian Purāṇic cosmologies rather than
within orthodox Buddhist cosmologies, and the attainment of Pātāla itself is not
necessarily consistent with core Vajrayāna concerns. Hence there might well have
been a learned Buddhist tendency to underemphasize or even expurgate references
to Pātāla, as the process of reconciling and integrating the diverse Buddhist magical
repertoire proceeded over the centuries. I will cite some examples of Asuras’ caves
and Pātāla from Buddhist Kriyātantras shortly, but first I will briefly describe the
underlying cosmological ideas about Asuras and Pātāla.
The Seven Paradisiacal Nether Worlds

Pātāla and the other subterranean paradises have been written about quite often by
western Indologists, although I am not aware of any monographs dedicated
exclusively to the subject. Pātāla is most frequently described in Purāṇic cosmology,
and those who browse Purāṇic literature are quite likely to be familiar with it from
its numerous mentions there. As one expects in anything Purāṇic, the general idea
remains similar, even if specifics vary: for example, Pātāla’s paradises can either
be below the earth but above the hells of torment (narakas); or they can also be
below the narakas. Another variation is that they can be described either as
primarily the domains of Asuras, or of Nāgas, or of both; and often of a variety of
other non-human spirits as well. A common pattern is that these subterranean
paradises are enumerated as seven in number. Deborah Soifer, in her study of the
avatars Narasiṁha and Vāmana, describes them as follows:
The purāṇas also enumerate seven worlds below the earth, variously called atalas
or regions of Pātāla. These nether regions, not to be confused with the more
numerous hells or narakas, are the dwelling places of the Asuras, the demonic
elder brothers of the gods. In beauty and luxury these residences rival the cities
and palaces of the gods.14
Woodroffe, summarizing the cosmological sections of several purāṇas, is more

specific:
Below the Hells are the seven nether worlds, Sutala, Vitala, Talātala, Mahātala,
Rasātala, Atala, and Pātāla, where, according to the Purāṇas, dwell the Nāga
serpent divinities, brilliant with jewels, and where, too, the lovely daughters of
13
The Mañjuśrīmūlakalpa has sometimes been described as a “Buddhist purāṇa.” See, for example,
Yūkei Matsunaga, “On the Date of the Mañjuśrīmūlakalpa,” in Tantric and Taoist Studies in Honour
of R. A. Stein, ed. Michel Strickmann, vol. 3 (Bruxelles: Institut Belge des hautes études Chinoises,
1985), 882: “Because it [Mañjuśrīmūlakalpa] also includes mathematical tables, astrological methods
and royal genealogies, it embodies by and large the stylistic features of a purāṇa, which is somewhat
anomalous in the Buddhist canon.”
14
Deborah A. Soifer, The Myths of Narasiṁha and Vāmana: Two Avatars in Cosmological Perspective
(New York: SUNY, 1991), 53-54.
the Daityas and Dānavas [Asura maidens] wander, fascinating even the most
austere.15
But not all purāṇas follow the sevenfold pattern. The Kūrma Purāṇa, for
example, enumerates only four Pātālas: Mahātala, where the famous Asura Bali
dwells;16 Rasātala, where Bali’s father the Asura Virocana dwells; Vitala, where
various Asuras such as Bali’s grandfather Prahlāda live, as well as great Nāgas
such as Jambhaka and Kambala; and the beautiful Talātala, where other Nāgas
dwell (Kūrma Purāṇa, Bhuvana Kośa, 15-25).17 By contrast, the Garuḍa Purāṇa
does enumerate seven, but, in agreement with the Viṣṇu Purāṇa,18 gives some of
them variant names: Atala, Vitala, Nitala, Gabhistamat, Mahākhya, Sutala, and
Pātāla; it describes them as inhabited by Nāgas and Rākṣasas (Garuḍa Purāṇa,
Bhuvana Kośa, 1-3).19
How Did the Asuras End up Underground? A Hindu and a

Buddhist Kriyātantra Version
Many might be more familiar with Asuras as rivals to the gods for possession of
the heavens, and hence as inhabitants of territories on the lower slopes of Meru or
in the oceans at its base, rather than in caverns below our continent of Jambudvīpa.
Those who are more familiar with early medieval Hinduism will be equally aware
of a large and varied corpus of mythology – Hindu, Buddhist, and Jaina alike –
which places the Asuras firmly underground, beneath our feet. There are many
examples to choose from.
In Hindu mythology, one of the best known accounts of the Asuras being driven
underground occurs in the narrative of the dwarf avatar of Viṣṇu. The Asuras, lead
by the mighty Bali, took over the heavens from the gods. The king of the gods,
Indra, was distraught. Viṣṇu recovered heaven for the gods by cunning: taking the
form of a dwarf brahmin, he obtained the offer from Bali of just as much ground
as he could cover in three steps, so that he could do a sacrifice. Recovering his
cosmic form, he covered the entire triple universe in his three strides, but a deal is
a deal, and Bali still had to give him all the land he covered. Thus Viṣṇu vanquished
the Asuras, and banished them to Pātāla far underground, to which region they
remain confined. The banishment of the Asuras to splendid confinement in Pātāla
15
John Woodroffe (aka Arthur Avalon), Tantra of the Great Liberation (Mahānirvāṇatantra) (New
York: Dover Publications, 1972), xxxvi-xxxviii.
16
For a detailed monographic study of this famous Asura still widely worshipped in South India, see
Clifford Hospital, The Righteous Demon: A Study of Bali (Vancouver: University of British Columbia
Press, 1984).
17
J. L. Shastri, Kūrma Purāṇa (Delhi: Motilal Banarsidass, 1981), 294-95.
18
H. H. Wilson, The Viṣṇu Purāṇa: A System of Hindu Mythology and Tradition, vol. 2 ( London:
Trübner & Co, 1865), 14.
19
J. L. Shastri, Garuḍa Purāṇa (Delhi: Motilal Banarsidass, 1978), 190.
is very old and occurs in virtually all occurrences of this popular story.20 There are
many other narratives, however.
A Buddhist version occurs in the well-known canonical Kriyātantra scripture,
the Mañjuśrīmūlakalpa. Here, after losing heaven to the Asuras, Indra goes for
help not to Viṣṇu, but to the sage Kāśyapa, progenitor of both gods and Asuras.
Kāśyapa teaches Indra a long wrathful mantra to invoke the Bodhisattva Mañjuśrī,
and as Kāśyapa intones it, Mañjuśrī instantly appears, making the world shake,
and demonstrating that the mantra is presided over by all the Buddhas. Indra is
impressed, learns the mantra off by heart, and decides to use it just as Kāśyapa
advises. Worshipping Mañjuśrī and reciting the mantra, Indra succeeds in defeating
the Asuras. They are banished to the subterranean paradises of Pātāla forever, and
henceforth Indra carries an image of Mañjuśrī on his banner.21
Asuras’ Caves as the Entrances to Pātāla

One of the great advantages of the subterranean paradises is that they are far more
accessible to humans than are the heavens. The late Friedhelm Hardy devoted to
Pātāla much of his chapter on cosmology, in his book The Religious Culture of
India.22 He wrote:
The nooks and crannies, the caves and cavities of earth are the openings into
splendid lower regions...In these “underworlds” (for which the generic term Patāla
is frequently used) we encounter a wide range of other non-human beings...Such
apertures into Patāla are found all over the earth – unlike the inaccessible approach
to the “heavens” via Meru...Moreover, popular literature finds it easy to talk about
the underworld and its beings.23
Hardy continues with several lengthy narratives from Prakrit and Apabhraṃśa
sources, which he describes as “travelogues”: the adventures of daring humans
who have entered into Asuras’ caves and gone to stay among the Asuras and Nāgas
of Pātāla, where endless wonders and miracles of the most marvelous kind are
encountered, and where esoteric knowledge and magical powers can be obtained
along with inconceivable wealth and sensual delights, especially of the erotic kind.
Hardy concludes:
Of interest here is not just the wider range of wonderful things to be found in
these underground realms (including vidyā or “magical sciences”) or further
species of ogre-like beings..., but also the seemingly smooth transition from the
20
Hospital, The Righteous Demon, 118.
21
Mañjuśrīmūlakalpa 3: 662 [10]; Teun Goudriaan, Māyā Divine and Human (Delhi: Motilal
Banarsidass, 1978), 77-78.
22
Friedhelm Hardy, The Religious Culture of India: Power, Love and Wisdom (Cambridge: Cambridge
University Press, 1994).
23
Hardy, The Religious Culture of India, 31.
world of man to those mysterious realms...this can be regarded as indicative of

how much the underworlds are regarded as part of the natural order.24
Asuras’ caves as the entranceways into Pātāla also occur in Buddhist

Kriyātantras. The *Kaṇikrodha-vajrakumāra-bodhisattva-sādhana-vidhi25 is a
Buddhist Kriyātantra believed to be of Indian origin, but now extant only in
Chinese.26 It gives us several good examples of Asuras’ caves. We read as follows:
Furthermore, if someone desires to dwell in the palace of the Asuras and experience
pleasures, he should go to the entrance of the cave of the Asuras and make an
aṅkuśa (hook, elephant goad) with rushes. He should recite the Vajrakumāra
mantra seven times to empower this hook, elephant goad, and then revolve it to
the right in the air near the entrance. By just reciting the mantra, he will make it
open...27
There is another rite. If the mantrin recites the mantra one hundred thousand times
in front of an Asura cave that his companions have already entered, the mantrins
who have previously entered the cave will emerge and welcome him, and then
lead him inside as far as as the palace of the Asuras...28
The root mantra of Vajrakumāra is: namo ratnatrayāya namaś caṇḍa vajra pāṇaye
mahāyakṣasenāpataye tadyathā oṃ kaṇi hūṃ phaṭ svāhā. Furthermore, there is
a second root mantra specifically to open up the palace of the Asuras: namo
ratnatrayāya namaś caṇḍa vajra pāṇaye mahāyakṣasenāpataye tadyathā oṃ
dhuna vidhuna kaṇi krodha sarva [?]yantraṇi hūṃ...29
Now I shall explain the standing mudrā to open the gates to the Asura’s cave.
The mantrin should stamp on the ground quickly, walk joyfully, leap
ferociously...This mudrā is the best of all mudrās, and it is able to break open all
the bolts of the Asura’s cave.30
24
25
T1222a; K1355.
26
Stephen Hodge, with contributions from Luke Lau, trans., “*Kaṇikrodha Vajrakumāra Bodhisattva
Sādhana Vidhi” (unpublished translation from the Chinese [London: 1989]). I am greatly indebted to
Luke Lau and Stephen Hodge for their extraordinarily generous help in translating this text from the
Chinese for me in 1989; Luke Lau first made a preliminary translation, which Stephen Hodge then
completed. I am also most grateful to Ronald Davidson for the citation locations within T1222a, and
for improving the Sanskrit renderings of the Chinese text. Hodge has expressed some uncertainty about
the title of the text, since the first element, Ka ni (= chia ni), remains obscure. The online version of
Lancaster’s catalogue simply gives the Sanskrit title as Vajrakumāra-tantra, while the Chinese title
given there is Sheng chia ni fen nu chin kang t’ung tzu p’u sa ch’eng chiu i kuei ching (聖迦尼忿怒金
剛童子菩薩成就儀軌經). Less complete versions of the same text also seem to exist in the canon. The
translation into Chinese is attributed to Amoghavajra, who came with Vajrabodhi to Lo-yang in 720
and died in 774. Note that the title is translated differently yet again in Ershi’er zhong dazangjing
tongjian (Beijing, 1997), where it is given as Ārya-ḍākinī-krodha-vajrakumāra-bodhisattva-siddhi-
kalpa-sūtra (thanks to Matthew Kapstein for this information). Among these various alternatives, I am
staying with Hodge’s rendering for now, since it seems the most reliable so far, especially in the way
it takes account of the evidence from the mantras within the text.
27
T1222a.21.102b28-102c2; Hodge, *Kaṇikrodha, 1.
28
T1222a.21.103b4-b6; Hodge, *Kaṇikrodha, 3.
29
T1222a.21.104a17-b13; Hodge, *Kaṇikrodha, 5.
30
T1222a.21.105c26-106a3; Hodge, *Kaṇikrodha, 6.
Gaining entry to Pātāla via its entry points of Asuras’ caves is clearly one of
the major concerns of this text, and the second root mantra to gain entry to the
Asuras’ palace (which is deeper within the cave) precedes a long list of branch
mantras subsequent to the two root mantras. Although dedicated to Vajrakumāra,
many of the text’s rites are attributed to Vajrapāṇi with the phrase “Thus taught
the Bodhisattva Vajrapāṇi.” Likewise, the main mantra (mūlamantra) is a variant
of the Vajrapāṇi mantra as commonly used throughout Kriyātantra texts. It is
Vajrapāṇi of course who figures in IOL Tib J 644; while another similarity this
text shares with IOL Tib J 644 is its enumeration of various vidyādhara levels
(again common to many Kriyātantra texts), but here these do not have the same
names as in the Dunhuang text.
Within his excellent introduction to Asuras’ caves and attainment of Pātāla in
Chinese Tantrism,31 R. A. Stein has pointed out several more passages from the
Chinese canon that describe gaining entry into Asura caves, with all the usual
ramifications of Vajrapāṇi, longevity, awaiting Maitreya, enjoying Asura maidens,
finding treasures, and so on and so forth.32
Extant Indian Buddhist texts also have rites to enter Pātāla, for example the
Mañjuśrīmūlakalpa.33 Likewise Sādhanamālā 172, dedicated to a form of Kurukullā,
enumerates the attainment of Pātāla as one of the eight attainments listed.34 No
doubt more such references will turn up, if scholars eventually turn their attention
to the mass of largely forgotten magical rites long buried in the early Buddhist
tantras.
Why Do Vidyādharas Want to Go to Pātāla?

One of the main attractions of Pātāla is pleasure. Returning from a visit to Pātāla,
the sage Nārada is said to have declared that he had found it much more delightful
than Indra’s heaven (Viṣṇu Purāṇa V).35 In particular, the loveliness of the Asura
and other non-human maidens is emphasized, and having sexual intercourse with
them is one of the main objectives of those (usually male) adventurers who visit
Pātāla, as, for example, in Sādhanamālā 172 cited above.
The Mañjuśrīmūlakalpa expresses such sentiments in a verse: those Buddhist
yogins accomplished in mantras and desirous of sexual intercourse with supernatural
women can invoke all kinds of Asura or other supernatural females by using their
mantras, and dwell with them in Pātāla for the duration of a complete kalpa. Then,
31
R. A. Stein, Grottes-Matrices et Lieux saints de la déesse en Asie orientale (Paris: EFEO, 1988),
especially 23 ff.
32
Stein, Grottes-Matrices et Lieux saints, 27. Stein’s main citations are as follows: [1] Taishō 901,
k.5, 833b and k.6, 837b; [2] Taishō 1096, 413c; [3] Taishō 1097, 425 b-c; [4] Taishō 1246, Taishō
1248, 327a.
33
Mañjuśrīmūlakalpa, 572, cited in Goudriaan, Māyā Divine and Human, 103.
34
Sādhanamālā, ed. B. Bhattacharya (Baroda: Gaekwad’s Oriental Series, 1925), 2:350; see also
lxxxv-vi.
35
H. H. Wilson, The Viṣṇu Purāṇa: A System of Hindu Mythology and Tradition (London: Trübner
& Co, 1865), 2:14-15.
when Maitreya finally arrives, they will hear his dharma and become enlightened.
But those who have subjugated Asura women this way should henceforth never
even touch a human woman.36 The Mañjuśrīmūlakalpa itself has numerous rites
and mantras to procure non-human females in this way.37
The *Kaṇikrodha-vajrakumāra-bodhisattva-sādhana-vidhi has very similar
rites. Having forced entry by using the mantras and gestures using the aṅkuśa
made from rushes as prescribed (see above), the vidyādhara encounters the
following scene:
Within the Asura’s cave, a great mass of fire will arise and all the male and female
Asuras will burn within it while shrieking and wailing with terror. Each one of
the female Asuras will reveal themselves and say to the mantrin, “Noble one!
Please come into our cave and enjoy yourself as you will!” When he has gained
entrance, he will be able to remain there for the duration of a kalpa, enjoying
heavenly delights (sukha).38
Another more complex passage is even more explicitly sexual. After gaining
entrance to an Asura cave, and hoping to progress from there towards the actual
Asura palace, the yogin should perform various rites, including the recitation of
the mantra of Vajrakumāra. At one stage, the following happens:
Then the Goddess Sarasvatī and the female Asuras together with their attendants
will come forth from the cave, and going up to the yogin they will speak as follows:
“Finally, we will serve you.” The mantrin should not accept, but he should speak
with a wrathful voice like thunder and recite the mantra further. The Asura maidens
will become confused and demented, and undressing, will make their bodies
naked...39
Another theme in the literature on visiting Pātāla is the quest for immortality,
often in a more alchemical sense. As David White reports, the entire Indian
alchemical tradition (rasāyana) is attributed in its origins to the Asuras: Vyāsa’s
commentary on Yoga Sūtra 4.1 cites the Asura realms as the locus for the
immortality-conferring botanicals, and this view is repeated by the ninth-century
luminary Vācaspati Miśra, who mentions that initiation into Indian alchemical
tradition is given by alluring Asura damsels.40 In the Mahābhārata, the head priest
of the Devas, Bṛhaspati, has to practice an elaborate deception on the chief sage
of the Asuras, Kāvya Uśanas, in order to wrest from him the secrets of alchemical
immortality.41 As a consequence, and in recognition of its de facto founder being
the Asura Kāvya Uśanas, medieval Indian alchemy designated the perfectly
36
Goudriaan, Māyā Divine and Human, 103, citing Mañjuśrīmūlakalpa, 572.
37
Goudriaan, Māyā Divine and Human, 101-3.
38
T1222a.21.102c2-c6; Hodge, *Kaṇikrodha, 2.
39
40
David Gordon White, The Alchemical Body: Siddha Traditions in Medieval India (Chicago:
University of Chicago Press, 1991), 58-59.
41
White, The Alchemical Body, 285-86.
accomplished alchemist a kavi. Likewise, some Bengali ayurvedic physicians are

still called kavirāj to this day, for similar reasons.42
Wealth and wisdom are other reasons to visit Pātāla. In the Jaina sources cited
by Hardy, the visitors to Pātāla encounter not only “pleasures eight times greater
than those found in the heavens,” including stunning Asura damsels, but also
treasure chests43 and magical sciences or vidyā.44
But purer and more academic motives are attributed to the great Madhyamaka
philosopher Bhāvaviveka, who, according to Hsüan-tsang, went to dwell in Pātāla
among the Asuras until the coming of Maitreya Buddha. Hsüan-tsang reports that
Bhāvaviveka did this so that he could live long enough to be able to address
particularly troubling doctrinal questions directly to Maitreya; here, no mention is
made of alluring Asura maidens. Nevertheless, Hsüan-tsang’s account mentions
Bhāvaviveka gaining entrance to the Asura kingdoms using a procedure entirely
similar to the Kriyātantra sources quoted in this article: Vajrapāṇi is invoked,
mantras and mustard seeds are thrown at a rock face to make it open, a crowd
watches the whole operation, and some of them accompany Bhāvaviveka into the
Asura’s cave.45 I am not aware of similar accounts regarding Bhāvaviveka from
Tibetan sources, so I am not sure if they might have contributed to Sa paṇ’s interests
in attainment of Pātāla. However, a quite early Central Asian Silk Route Mahāyāna
cult of suspended animation to await Maitreya is certainly reported, perhaps based
on a similar account found in very much earlier sources still that describe the great
meditator Mahākāśyapa.46
The Waters of Pātāla

Waters are of course another special feature of Pātāla, as one would expect from
the homeland of the Nāgas (in almost all accounts, Pātāla is the domain par
excellence of both Asuras and Nāgas). According to the traditional etymology,
Pātāla’s very name derives from the fact that it is well-watered: patanti alam >
pātāla.47 In other accounts, the heavenly Ganges exists in Pātāla and can be
summoned up to earth from there.48 In addition, the cosmology of Mount Meru
includes the idea of miraculous flows of heavenly waters or juices flowing down
from its sides all the way to earth and beneath; our earthly gold, for example, is
the dried residue of the rose-apple juice seeping down from the heavenly rose-apple
42
White, The Alchemical Body, 286.
43
Hardy, The Religious Culture of India, 33-34.
44
Unfortunately, Hardy chooses not expand on these magical sciences, and I lack the resources and
time to pursue them in his primary sources.
45
Stein, Grottes-Matrices et Lieux saints, 26. See also Malcolm David Eckel, To See the Buddha: A
Philosopher’s Quest for the Meaning of Emptiness (Princeton: Princeton University Press, 1994), 11-13.
46
John Jørgensen, Inventing Hui-neng, the Sixth Patriarch: Hagiography and Biography in Early
Ch’an (Leiden: Brill, 2005), 231-32.
47
Gail Hinich Sutherland, The Disguises of the Demon: The Development of the Yakṣa in Hinduism
and Buddhism (New York: SUNY, 1991), 40.
48
fruits on Meru’s peak. Perhaps this cosmology, together with the pervasive
understanding of Pātāla as well-watered and opulent, explains the occurrence of
the Aśvakarṇa springs connected with Meru in IOL Tib J 644. The myth of the
dwarf avatar’s three steps overcoming the Asura Bali that I mention above certainly
develops the theme of sacred and magical waters from Meru flowing into Pātāla;
in many cases, this became an important integral part of the myth of the Asuras’
banishment to Pātāla, since it is through the higher of his three steps that Viṣṇu
makes these ritually most significant sacred waters flow down for the very first
time.49 In the Padma Purāṇa 6.267, to take one example, the sacred stream flowing
down the face of Meru separates into three different streams, one of which becomes
Gaṅgā (i.e., the human world’s Ganges), and another of which flows directly into
Pātāla where it could be enjoyed by Bali, who, despite being an adharmic Asura,
had nevertheless shown exemplary devotion to his conqueror, Viṣṇu. In the Asura
realms, the sacred river is called Bhogavatī. A third branch of the same stream
flows into the God’s realms, where it is called Mandakinī.50 Needless to say, such
waters confer great purification and attainment.
Conclusions
Some conclusions:
Firstly, Buddhism, like the other Indian religions, has had a long and sometimes
significant relation to Pātāla.
Secondly, while it seems that the comparatively early texts we now call
Kriyātantra once considered Pātāla an important topic, that was no longer so much
the case in subsequent Indian and Tibetan Tantric Buddhism: my hunch is that
Pātāla’s non-standard doctrinal and cosmological bases caused it to be increasingly
sidelined as Vajrayāna Buddhism emerged as a fully developed tradition (which
process might also have coincided with a similar falling away of interest in Pātāla
within other Indian traditions).
Doctrinally, attainment of Pātāla, at least in its rawest form, seems to propose
spending an entire kalpa secluded in the company of delightful Asura maidens,
until Maitreya comes, thus postponing all serious practice of dharma until that
time. It says nothing about the immediate realization of emptiness and compassion.
Mainstream Vajrayāna by contrast emphasizes the perception of this very world
we are born into as the pure maṇḍala of a Buddha, to be realized through direct
perception of its emptiness. Such a realization empowers the bodhisattva to act
for the benefit of others. Hence the temporary flight from this world and the
postponement of realization suggested by attainment of Pātāla seems tangential to
49
Often, the waters descend from Viṣṇu’s foot breaking the shell of the Cosmic Egg surrounding the
universe; sometimes, the waters descend from Brahmā’s bathing of Viṣṇu’s foot as it arrives in the
highest heavens. This can sometimes become an important cosmogonic account of the birth of the
Ganges; as it is, for example, in the Padma Purāṇa 6.267 cited above. Śaiva cosmogonic accounts will
of course prefer to say that Gaṅgā flows down from Śiva’s head.
50
Hospital, The Righteous Demon, 133-34.
Vajrayāna notions of the innate purity of all phenomena, and shows little awareness
of more fully developed Vajrayāna skillful means. Little wonder it gained scant
scholastic attention in later Indian and Tibetan Vajrayāna, despite Sa paṇ’s apparent
interest.
Cosmologically, Buddhism is often not very definite about Asuras. In earlier
Buddhism, there were quite possibly only five realms or paths for rebirth
(pañca-gatayaḥ), rather than the six now enumerated (ṣaḍ-gatayaḥ), because the
Asuras were not counted as a separate realm;51 indeed, some modern enumerations
continue to have only five realms or paths. Moreover, Vasubandhu did not specify
the precise location for Asuras in his Abhidharmakośa,52 although Kongtrul does
present a sūtric source that specifies that Asuras live in crevices of Mount Meru
below the water level.53 Buddhist sources not infrequently prefer to locate Nāgas
predominantly in oceans, rather than in a subterranean underworld, although they
can also live in a subterranean underworld, especially beneath anthills, and their
splendid wealth is certainly mentioned in a similar vein to the Purāṇic accounts.
Although there are certainly some Buddhist patterns that do not contradict the
Purāṇic placement for the Asuras, we are left with the impression that the
consistently prominent mention of Asuras as a significant and distinct class of
beings and their unvarying placement in Pātāla is much more Purāṇic than Buddhist.
What is clear is that only a few ill-understood echoes of the old belief still
survive in Tibet.54 I have not heard, for example, that there is any contemporary
Tibetan understanding of the Asura Cave at Pharping as an entry to Pātāla, even
though there is some possible evidence that such an understanding might well once
have been attached to that place. For example, there is certainly a surviving oral
tradition of sacred subterranean waterways linking Pharping with distant holy
places and springs, which is suggestive of Pātāla.55
51
Jikido Takasaki, An Introduction to Buddhism, trans. Rolf Giebel (Tokyo: Tōhō Gakkai, 1987),
131.
52
Jamgön Kongtrul, The Treasury of Knowledge, book 1, Myriad Worlds (Ithaca: Snow Lion, 2003),
115.
53
Jamgön Kongtrul, Myriad Worlds, 113, 115.
54
It is not clear to me how the Pātāla cult fared in East Asia and Southeast Asia. It is not impossible
that the Chinese dragon could represent to some degree a hybridized form including both Nāga and
Asura; and there is certainly an awareness of the idea of Pātāla in Southeast Asia, although I know
little more than that. A recent study by Julius N. Tsai (Julius N. Tsai, “Opening up the Ritual Casket:
Patterns of Concealment and Disclosure in Early and Medieval Chinese Religion,” Material Religion
2, no. 1 [2006], 38-66) portrays a vigorous cult in medieval Daoism of the discovery of caskets containing
sacred texts, heavenly mandates, and other items that sometimes bear striking resemblance to the
Tibetan gter ma tradition, but I am not sure exactly how this cult in Daoism relates to the rather similar
ones in Chinese Buddhism as described by Michel Strickmann and others (see for example Michel
Strickmann, “The Consecration Sūtra: A Buddhist Book of Spells,” in Chinese Buddhist Apocrypha,
ed. Robert E. Buswell [Honolulu: University of Hawaii Press, 1990], 75-118). I am also not yet clear
how either of these Chinese treasure cults connect with the Indic ideas of Pātāla.
55
Rather less conclusively, some of the more extensive accounts within the Tibetan phur pa literature
on Padmasambhava’s taming of goddesses at Pharping seem to invoke typically Purāṇic cosmological
categories in describing those goddesses’ habitats as oceans of milk, wine, ghee, and butter. See Robert
Padmasambhava and Kriyātantra

What bearing might the Pātāla cult have on Padmasambhava? The popular
hagiographies indicate that Padmasambhava went to the Asura Cave at Pharping
more to use it as a generalized power place, rather than as a specific entry point
for him to visit Pātāla. There seems to be no immediate direct connection.
Nevertheless, there are certainly several noteworthy parallels between the
Padmasambhava legends and some typically Kriyātantra beliefs, including those
of the Pātāla type, that mostly have not up till now been remarked, as far as I am
aware. How exactly these parallels connected to the historical Padmasambhava,
or to his legend, is difficult to say with any precision. At the moment, I feel there
can be little doubt that the historical Padmasambhava was a major teacher of
Mahāyoga, as the Dunhuang Thabs zhags commentary, IOL Tib J 321, so eloquently
suggests.56 Yet Padmasambhava probably existed within a cultural environment
in which Kriyātantras were still very influential indeed. Of course, categories of
tantric doxography are also often rather loose and even inconsistent, so that
categories such as Kriyā and Mahāyoga can have considerable overlaps: Mahāyoga
grew within the milieu of and upon the basis of earlier traditions including
Kriyātantra, and therefore the later traditions still have much common ground with
the earlier traditions. One must also consider that late Dunhuang texts such as
Pelliot 44 clearly portray Padmasambhava as an exponent of all levels of tantra,
from Kriyā up to Ati – although the significance of such terms was more flexible
and varied in the tenth century than it became in later years.57 The situation is still
quite hard to assess accurately (I have some speculative hypotheses that I might
suggest elsewhere), but it is important not to read excessive specific significances
into these resemblances; on the contrary, it is better to understand them as indicating
a more general truth.
Let us look within the contents of the *Kaṇikrodha-vajrakumāra-bodhisattva-
sādhana-vidhi, a text substantially devoted to the Pātāla cult, to get some examples
of themes that resonate with the later Padmasambhava legend:
[1] There is in this particular text, of course, a central focus on the wrathful
deity Vajrakumāra, a name very well known in Tibet as the proper name of
Vajrakīlaya, one of the chosen yi dam deities of Padmasambhava and the most
important of all Rnying ma pa yi dams. However, the appearance of the deity here
is not identical to the later Tibetan or Mahāyoga forms. The main difference is that
the Tibetan Mahāyoga forms are normally presented as kāpālika deities with the
trademark kāpālika adornments of skulls, khaṭvāṅgas, cemetery ash, and so forth,58
Mayer, A Scripture of the Ancient Tantra Collection: The Phur-pa bcu-gnyis (Oxford: Kiscadale
Publications, 1996), 128ff.
56
Dunhuang Tibetan manuscripts held at the British Library, London: IOL Tib J 321. Cathy Cantwell
and I are currently engaged in a four year AHRC-funded research project on this text.
57
Dunhuang Tibetan manuscripts held at the Bibliothèque nationale, Paris: Pelliot tibétain 44.
58
These kāpālika adornments are linked in the Dharmaśāstra literature to the legal penances for
Brahmins who commit murder: banished to a graveyard, and becoming ritually impure, they must wear
which are generally linked with later strata of tantra. This Vajrakumāra, by contrast,
is of the pre-kāpālika Vajrapāṇi type, with snakes to show his control over Nāgas
and flames to emphasize his wrath, but lacking the specific kāpālika insignia (even
if some human skulls and bones are employed in some of his rites). In fact, in this
text, Vajrakumāra is explicitly understood as a form of Vajrapāṇi, which is an
identification less often made for Vajrakumāra in the Tibetan Mahāyoga literature.
Thus in the instructions on how to paint him, the text tells us:
The figure of Vajrakumāra stands alone, arising out of the waves of the ocean.
He should be the colour of vaiḍūrya, of robust appearance, having six arms with
strong shoulders. His face has three eyes which are red. He wears a jewelled crown
on his head. His eye teeth are thrust out and bite on his lower lip, while his
eyebrows are wrinkled in anger. Furthermore, a jewelled mountain should be
painted in the ocean. The figure has his right leg on this jewelled mountain, resting
on a beautiful lotus and his left leg is submerged in the water up to his knee. His
first right hand holds a three-pronged vajra as though about to throw it. The second
right hand holds a mace, and his third right hand holds an axe. His first left hand
grasps a club, his second left hand makes a gesture of warning, and his third left
hand holds a sword. He has one great snake slung from his shoulder diagonally
around his torso, and he also has all kinds of poisonous snakes as armlets and
bracelets. He also has a jewelled necklace, earrings, and his hair is bound [with
ribbons]. His waist is encircled three times with one large snake. The rear of his
body is surrounded with light and tongues of blazing fire, and flashes of lightning
shoot out from these flames.59
Elsewhere in the text, he is described as dark blue in color. The several references
to the sea and maritime products in the text and the ascription of its Chinese
translation to Amoghavajra are thought by some to indicate a South Indian
provenance for this deity, while the Tibetan Vajrakumāra is more likely North
Indian. Yet both this deity and the later Mahāyoga versions are both krodha deities
of roughly the same general type, and both primarily act as subjugators of spirits
of all kinds, and of demonic forces. As the text explains at the outset, those who
do the practice of Kaṇikrodha Vajrakumāra
will be able to subdue demons, eliminate those with false views who revile the
true dharma, and destroy such people as the icchāntikas within the country. The
power of the mantra will cause them to direct their minds to goodness. Poisonous
insects and plants will not harm them. Mantrins of other groups will not be able
to harm their practice…he will be able to gain access to treasures (nidhi), break
the sealed doors of the Asuras, dry up rivers, and stop the flow of water.60
only an animal skin skirt, they must carry with them at all times the leg of the bed (literally, khaṭvāṅga)
of their victim, and they must eat out of his skull as their only bowl. This image of the Brahmin penitent
lies at the heart of much tantric imagery, both Śaiva and Buddhist.
59
T1222a.21.106a22-b5; Hodge, *Kaṇikrodha, 7.
60
The control over water also occurs elsewhere in the text, a theme well-attested
for Padmasambhava,61 and throughout the text there are also numerous rites to kill,
subdue, and terrify enemies of every kind, both human and non-human, very much
as Padma did in the legends. Some of these employ a kīla, the favored instrument
of Padmasambhava. The text advises:
Go to a cemetery and make a figure of your śatru (enemy). Take a human bone
from a cemetery and make a kīla and bind it with red thread. Stab the figure of
your enemy with this in its heart and [that person] will immediately die.62
Further on, almost the same rite seems to be repeated:
There is another rite if you desire to vanquish an enemy (śatru). Take some human
bone and make a kīla with it. Then either paint or sculpt an image of that person
and reciting the mantra 108 times to empower the kīla, drive it into his heart and
he will be vanquished.63
And again:
There is another rite if you wish to overcome an enemy and make them die. Take
some wood left from a cremation and make a kīla with it. Anoint it with ground
purple sandalwood, and wind some thread you have got from a cemetery around
the kīla. Then drive it into the head of the image of your enemy as before. Your
enemy will then die.64
[2] There is a substantial concern with discovery of treasure (gter ma, nidhi),
one of the most important aspects of the Padmasambhava legend. As the text
explains right at the outset, in the first paragraph after the setting of the scene
(nidāna), by reciting the mantra 600,000 times, and performing various other rites,
the yogin will be able to gain access to treasures (nidhi).65 This is clearly one of
the main advantages of gaining entry to Pātāla, as understood in this text. At another
point, it is explained that
61
Hodge, *Kaṇikrodha, 2, 12. See Pasang Wangdu and Hildegard Diemberger, dBa’ bzhed: The
Royal Narrative Concerning the Bringing of the Buddha’s Doctrine to Tibet (Vienna: Verlag der
Österreichischen Akademie der Wissenschaften, 2000), 14 for a discussion of Padmasambhava’s close
involvement in controlling waters in the Dba’ bzhed. Bu ston also mentions Padmasambhava’s feats
with water, including reversing the Tsang po River, and taking the silver urn in which the Tibetan king
kept his fresh hair-washing water, and miraculously introducing the washing waters of the gods into
it. See Bu-ston Rin-chen-grub, The History of Buddhism in India and Tibet, trans. E. Obermiller, 2nd
ed. (Delhi: Sri Satguru Publications, 1986), 190. This feat of substituting the king’s washing water with
water from the gods might well represent an ancient Buddhist tradition of showing siddhi: Pali traditions
maintain that being capable of obtaining the refreshing waters from Lake Anotatta, in which the gods
themselves bathe, was a feat traditionally held to represent the very ultimate demonstration of siddhi.
See G. P. Malalasekera, Dictionary of Pāli Proper Names (London: Published for the Pali Text Society
by Luzac & Co, 1960), 96-99.
62
T1222a.21.107c29-108a2; Hodge, *Kaṇikrodha, 10.
63
64
T1222a.21.109a1-4; Hodge, *Kaṇikrodha, 13.
65
Hodge, *Kaṇikrodha, 1.
There is also a rite for those who desire to acquire treasure (nidhi). Do not select
a particular [astrological] season, day or hour, and it is not necessary to maintain
the discipline. In the vicinity of the treasure, the mantrin should raise one foot
and recite the mantra. Turning around to the right, he should gaze towards all the
four directions and take possession of the area (sīma-bandha) [use this method
also when preparing the great maṇḍala]. Carefully raising one foot, he should
recite the mantra 108 times. If the guardians of the treasure obstruct him, then
they will be burnt in a mass of fire. They will come screaming to the mantrin and
bow before him, vanquished. The mantrin should say to them, “Open this treasure
store and give me all that is therein to me!” They will then open it and give
everything to the mantrin. If they are mean-spirited and do not give it to him, then
he should say, “Brahmā, Nārāyaṇa, Maheśvara, the warrior goddesses and Durga
will come and crush your treasure store. You give it to me quickly! If you do not
do so, the wrathful Vajrakumāra will destroy all of your family!” When they have
heard what he has said, they will all obey, and say to him, “Noble One! Come
and take what you will, we will not hinder you!” Then he should say to them,
“You may open the store yourselves and give it to me!” They will then immediately
open the treasure store and give him the treasure respectfully.66
There are yet further elaborate nidhi rites in this text.67 Interestingly, one of
them embeds the recovery of treasure into a list of impressive attainments that
culminates in the most soteriological kind:
If you desire to accomplish the most excellent result, you should go to the seashore
during the waxing phase of a lunar [first, fifth, or ninth] month, and make a stūpa
out of mud and gravel, and place the pratītya verse [that is, the Dharmakāya
Śarīra verse] inside the stūpa. Place the image [of Vajrakumāra as described
above] in front of the stūpa. You should mix water with porridge and eat it. Take
sticks of arka wood and smear them with ghee, and then throw them into a homa
fire. Recite the mantra 100,000 times. When you have finished, the ground will
move and your body will become transformed and you will fly up and become
Lord of the Tuṣita Heaven. If the surrounding ground flickers with fire, then you
will become Lord of the Four World Rulers. If rain falls from a great cloud, all
the treasures hidden in the ground will burst forth at once. If rays of golden light
appear everywhere, then you will be transformed by them and become a
bodhisattva. You will live for one kalpa and nobody will be so powerful that they
can injure you. If you see the bodies of all beings giving out rays of light, then
you will gain realization of all the Buddha’s Teachings of the Three Ways and
attain bodhicitta. If the image and the stūpa emit rays of light, then you will
become lord of all vidyādharas. If light rays fill the ten directions, then you will
see the Bodhisattva Samantabhadra, and all things you desire, whether mundane
or supra-mundane, will be fulfilled…68
Yet most of the nidhi rites are more mundane:
66
67
Hodge, *Kaṇikrodha, 7, 10, 12.
68
T1222a.21.106b24-c6; Hodge, *Kaṇikrodha, 7-8.
There is another rite if you wish to locate hidden treasure. Get some yogurt from
a yellow cow, a snake skin and shark oil, as well as some arka wood and cotton.
Make a lamp with these things. Recite mantras to empower it, and then light it
at night near the place where there is treasure. You will know the amount of the
treasure that is there by the size of the flame. If you need to expel the gods who
guard the treasure and other obstructors, take a slab of rock or pebble or some
mustard seeds or some empowered water and cast it at the treasure. The obstructor
on the treasure will withdraw. If you suspect there is a large Nāga there, this will
also leave.69
In fact, as far as I am currently aware, most scriptural Tantric Buddhist nidhi

rites are found in Kriyātantras – see, for example, the Vajraśekhara Sūtra or
Chin-kang-ting ching, Fascicle 3, Chapter 4;70 or, for an Indian example, the
Mañjuśrīmūlakalpa.71 In accordance with the wider Indian treasure cult, beyond
the confines of Buddhism, in which entire works such as the Nidhidarśana of Rāma
Vājapeyin were devoted to treasure recovery, not to mention the countless nidhi
rites found in diverse types of Hindu texts and often connected with Pātāla,72 the
treasure in these Kriyātantras is of course usually more material than textual.
However that does not necessarily contradict the developed Rnying ma pa treasure
practice, within which the recovery of straightforward material wealth, sacred
elixirs, and various valuable objects have all along been very significant, occuring
throughout Rnying ma history alongside the better-known text revelation.73 One
should add, even if the Buddhist Kriyātantra texts emphasize treasures of magic
powers and material treasures, the several famous Mahāyana texts on treasure
(nidhi) studied by Paul Harrison do emphasize textual discoveries
(dharmanidhāna),74 in terms reasonably similar to those of the tantras, and quite
strikingly similar to the more developed Tibetan treasure tradition. Both these
types of Buddhist treasure narrative, the tantric and the sūtric, entered Tibetan
translation simultaneously in the early translation period. To be more specific,
unmistakable precursors of the magical particulars for treasure recovery are richly
attested in many Kriyātantra texts, including those that we know were translated
into Tibetan from early on; while doctrinal and “historical” explanation and
justification both for nidhi’s concealment, and for the manner of its concealment
and recovery, are found expounded at great length and in very great detail indeed
in the Tibetan translation of the Pratyutpanna-buddha-saṃmukhāvasthita-samādhi-
69
T1222a.21.107b27-c3; Hodge, *Kaṇikrodha, 10.
70
Rolf Giebel, Two Esoteric Sutras (Berkeley: Numata Center for Buddhist Translation and Research,
2001), 79-80.
71
Mayer, A Scripture of the Ancient Tantra Collection, 84.
72
Mayer, A Scripture of the Ancient Tantra Collection, 82-89.
73
Mayer, A Scripture of the Ancient Tantra Collection, 87.
74
Paul Harrison, “Mediums and Messages: Reflections on the Production of Mahāyāna Sūtras,”
Eastern Buddhist 25, no. 1 (2003): 125.
sūtra,75 a famous and philosophically lofty Mahāyāna text cited by Kamalaśīla in

the Bhāvanākrama that he specially wrote for Tibet; as in the Sarvapuṇya-
samuccaya-samādhi-sūtra, that is cited on the topic of such text revelation by
Śāntideva in his Śikṣāsamuccaya.76 Perhaps this is why the Rnying ma treasure
cult so obviously came to combine elements from the exciting magic of the Asura
75
Paul M. Harrison, The Samādhi of Direct Encounter with the Buddhas of the Present: An Annotated
English Translation of the Tibetan Version of the Pratyutpanna-Buddha-Saṃmukhāvasthita-Sūtra
(Tokyo: The International Institute for Buddhist Studies, 1990).
76
Harrison, “Mediums and Messages,” 125. As I suggested in my article of 1994, any explanation
of Tibetan gter ma that fails to take into account these more universal and Indic treasure motifs will
probably prove inadequate: the canonical descriptions of gter ma need to be factored into any analysis
of the complex social-historical and cultural situation of post-dynastic Tibet to get a complete picture
(Robert Mayer, “Scriptural Revelation in India and Tibet: Indian Precursors of the gTer-ma Tradition,”
in Tibetan Studies: Proceedings of the 6th Seminar of the International Association for Tibetan Studies,
Fagernes, 1992, vol. 2 [Oslo: Institute for Comparative Research in Human Culture, 1994], 542).
Unfortunately, a more recent work by Davidson (Tibetan Renaissance, 210 ff) shows little if any
awareness of these important Mahāyāna sūtras, nor of the several studies made of them by Paul Harrison
(Paul M. Harrison, The Tibetan Text of the Pratyutpanna-Buddha-Saṃmukhāvasthita-Sūtra [Tokyo:
The International Institute for Buddhist Studies, 1978]; Harrison, Samādhi of Direct Encounter with
the Buddhas of the Present; Harrison, “Mediums and Messages”). Likewise, Davidson shows little or
no awareness of the highly popular cults of Asura caves, Pātāla, and treasure recovery mentioned here.
The net result is that Davidson’s account fails to understand the continuities between treasure recovery
in Tibet and elsewhere. For example, Davidson writes that the distinction between sa gter and dag
snang is “a relatively modern formulation” created in Tibet, that is “nowhere to be seen” in earlier texts
(Ronald M. Davidson, Tibetan Renaissance: Tantric Buddhism in the Rebirth of Tibetan Culture [New
York: Columbia University Press, 2005], 213). In fact, precisely this distinction is made in the early
period Tibetan translation of the Pratyutpanna-buddha-saṃmukhāvasthita-samādhi-sūtra, where
Chapter 3 describes an unmistakable prototype for dag snang, and Chapter 13 describes an unmistakable
prototype for sa gter (Harrison, Samādhi of Direct Encounter with the Buddhas of the Present, 31-44
and 96-108). Likewise it is made in the early period Tibetan translation of the Sarvapuṇya-samuccaya-
samādhi-sūtra; and also in the renowned śāstra by Śāntideva, the Śikṣāsamuccaya (Harrison, “Mediums
and Messages,” 123-29). It is likely that a thorough search will locate other Mahāyāna sources as well.
These Mahāyāna texts also describe a third process of text recovery, pratibhāna, that resembles the
later Tibetan dgongs gter, which is indigenously distinguished from sa gter and dag snang alike. Again
unaware of the Indic sources, Davidson also writes that in Tibet local deities and Klu are gradually
replaced by Ḍākinīs as the normative gter ma protectors, as a result, he believes, of a progressive
attempt at Indianization of what he seems to see as the indigenously Tibetan gter ma tradition (Davidson,
Tibetan Renaissance, 217-18). In other words, Davidson assumes that having local deities and Nāgas
as gter ma protectors was non-Indic. Yet it is the Indian Mahāyāna Pratyutpanna-buddha-
saṃmukhāvasthita-samādhi-sūtra’s Chapter 13 that specifies local deities and Nāgas (klu) as the
protectors of the text-containing sgrom bu that the Buddha has had buried in the earth. Likewise, the
entire Asura cave and Pātāla tradition so prevalent in the Indian Kriyātantras is by definition posited
on Nāgas (klu) and Asuras as the main treasure protectors (in China they become dragons), not on
Ḍākinīs. Incidentally, in the same passage, Davidson comprehensively misconstrues my 1994 paper
on gter ma (Davidson, Tibetan Renaissance, 212). I took pains to locate the Indic precedents as merely
one factor (at that time still largely unknown) among many in the construction of the Tibetan gter ma
tradition, something to be factored into Snellgrove’s interesting sociological idea (David L. Snellgrove,
Indo-Tibetan Buddhism: Indian Buddhists and Their Tibetan Successors [Boston: Shambhala, 1987]),
2:396ff). I analyzed the unique Tibetan transformation of the Indic gter ma precedents as broadly
analogous to the Tibetan construction of the sprul sku system of incarnate lamas on the basis of imported
Indic Buddhist beliefs (Mayer, A Scripture of the Ancient Tantra Collection, 533, 541-42). Davidson
however misquotes me as claiming Indian antecedents to be the only significant factor in the development
of gter ma. One further point of my 1994 article was to correct Michael Aris’s at the time still influential
characterization of the basic structuring concepts of gter ma as largely derived from indigenous Tibetan
shamanism (Michael Aris, Hidden Treasures and Secret Lives [London and New York: Kegan Paul
International, 1989], 53-63).
cave and Pātāla hunt for treasure, pleasure, and supernatural powers, with high-brow
Mahāyāna scriptural orthodoxy, together forming a particularly attractive package,
becoming quite irresistible after being thoroughly integrated with Tibetan mythic
history and imperial memory.
One could likewise point to other similarities between the *Kaṇikrodha-
vajrakumāra-bodhisattva-sādhana-vidhi and the later Mahāyoga Vajrakīlaya-
Vajrakumāra and Padmasambhava tropes, including the taking of supernatural
women as consorts,77 elaborate rites for the taming of Rākṣasas to become one’s
servants,78 an extremely wrathful maṇḍala for protection from enemies containing
the goddess Ekajaṭā/Ekajatī with Vajrakumāra (she is Vajrakumāra’s main consort
of liberation in the Mahāyoga traditions),79 and the attainment of extreme longevity
or immortality while remaining in a supernatural location filled with dangerous
spirits (cf. Padmasambhava in the Copper Colored Mountain [zangs mdog dpal
ri]).80
Yet I do not believe any of these, nor even all of them together, need to indicate
any direct connection with the Padmasambhava story, unless some further quite
unexpected evidence comes to light that directly links the *Kaṇikrodha-
vajrakumāra-bodhisattva-sādhana-vidhi with Padmasambhava and Nepal. Rather,
on available evidence, I believe they collectively point to a much more general
conclusion: that in assessing the life and times of the historical Padmasambhava,
as well as the growth of the Padmasambhava legends, and moreover in assessing
the growth of early Vajrayāna in Tibet in general, we should try not to leave out
of our calculations the fact that Kriyātantra and the other “lower” tantras were
culturally significant traditions at the time that Buddhism first became established
in Tibet.81 Hence their various mythologies and preoccupations could naturally
attach to the person of Padmasambhava, even if he was himself in all probability
also an early teacher of Mahāyoga and maybe more, in addition to being
knowledgeable in earlier tantric traditions. Moreover the “lower tantras” contributed
77
78
Hodge, *Kaṇikrodha, 2, 8.
79
80
81
Some (notably Davidson, Tibetan Renaissance, 64-65 and 385) have taken an extremely conservative
view of tantra translations in the Imperial period, limiting this to none beyond the thirteen texts
mentioned in the Lhan kar ma. Perhaps such absolute certainty is premature. The combined evidence
of what we know about Tibetan-ruled areas adjoining Kashmir (for a useful review of several sources,
see Cristina Scherrer-Schaub, “Enacting Words: A Diplomatic Analysis of the Imperial Decrees [bkas
bcad] and their Application in the sGra sbyor bam po gñis pa Tradition,” Journal of the International
Association of Buddhist Studies 25, nos. 1-2 [2002]: 263-340, 271ff); the development of the Sgra
sbyor bam bo gnyis pa (see Scherrer-Schaub, “Enacting Words,” 263-340); the Sba bzhed/Dba’ bzhed;
Bu ston’s Chos ’byung (see A. Herrmann-Pfandt, “The Lhan kar ma as a Source for the History of
Tantric Buddhism,” in The Many Canons of Tibetan Buddhism, ed. Helmut Eimer and David Germano,
129-49, PIATS 2000: Tibetan Studies: Proceedings of the Ninth Seminar of the International Association
for Tibetan Studies, Leiden 2000 [Leiden: Brill, 2002]); the ’Phang thang ma and Lhan kar ma; what
we know about contemporaneous Indian and Chinese Buddhisms; and the Dunhuang finds – might
also suggest otherwise.
towards and influenced the development of later tantrism, to a degree rather greater
than is commonly analyzed by most modern scholars, even if they remain in most
cases vaguely aware of this situation. In short, it is not at all impossible that the
largely hidden, oblique and indirect cultural imprint of these “lower” tantras on
the fully developed Rnying ma pa tradition as we now know it, with its
Padmasambhava legends, its Vajrakīlaya-Vajrakumāra popular yi dam, and its
treasure traditions, is in truth more interesting than we usually choose to recollect.
It is all too easy to become somewhat beguiled by the later Rnying ma pa rhetoric
of exclusive reliance on the three inner or “higher” yogas of Mahā, Anu, and Ati,
and ignore the sometimes interesting materials thrown up by the earlier “lower”
tantras (but Matthew Kapstein has been a distinguished exception to this trend).
In that respect, even scholars of the more developed and systematized Rnying ma
pa must welcome the works of scholars like Ariane Macdonald, David Snellgrove,
Tadeusz Skorupski, Stephen Hodge, and now also Steven Weinberger, who have
all contributed towards opening up for Tibetanists this very useful field of
knowledge of the earlier tantras.
Glossary
Ka
klu lu Term
Kha
mkhas grub rje Khedrupjé Person
Ga
grub pa’i rigs ’dzin druppé rikdzin vidyādhara of Term
accomplishments
dgongs gter gongter Term
sgra sbyor bam bo Drajor Bampo Nyipa Text
gnyis pa
sgrom bu drombu treasure casket Term
Cha
chos ’byung Chöjung Text
’chi med srog thig Chimé Soktik Doxographical
Category
Nya
rnying ma pa Nyingmapa Organization
Ta
gter ma terma treasure San. nidhi Term
Tha
thabs zhags Tapzhak Text
Da
dag snang daknang Term
bdud ’joms rin po che Düjom Rinpoché Person
rdo rje sa ’og gi rgyud Dorjé Saokgi Gyü San. Vajra-pātāla- Text
tantra
Pa
paṇ ḍi ta kun dga’ Pendita Künga Person
sprul sku trülku Term
Pha
phur pa purpa Term
’phags pa rdo rje sa Pakpa Dorjé Saokgi San. Ārya-vajra- Text
’og gi rgyud kyi Gyükyi Gyelpo pātāla-nāma-
rgyal po tantra-rāja
’phags pa shes rab Pakpa Sherap Person
’phang thang ma Pangtangma Text
Ba
bu ston Butön Author
bya gdong ba can Jadongwachen Person
dba’ bzhed Wazhé Text
sba bzhed Bazhé Text
Tsa
tsang po Tsangpo Term
Za
zangs dkar Zangkar Zanskar Place
zangs mdog dpal ri Zangdok Pelri Copper Colored Place
Mountain
zil gnon nam mkha’i Zilnön Namkhé Dorjé Person
rdo rje
Ya
yi dam yidam Term
La
las byang Lejang Title
Sa
sa skya pa Sakyapa Organization
sa gter sater Term
sa paṇ Sapen Person
sa la gnas pa’i rigs sala nepé rikdzin vidyādhara who Term
’dzin dwells on the levels
Ha
lhan kar ma Lhenkarma Text
lhun gyis grub pa’i lhüngyi druppé spontaneously Term

rigs ’dzin rikdzin accomplishing
vidyādhara
Sanskrit
Abhidharmakośa Text
Amarakośa Text
Amoghavajra Person
hook; elephant goad aṅkuśa Term
Ārya-ḍākinī- Text
krodha-
vajrakumāra-
bodhisattva-siddhi-
kalpa-sūtra
Asura Term
*asuraguhā Term
Aśvakarṇa Term
a region of Pātāla atala Term
Ati Doxographical
Category
Bali Deity
Bhāvanākrama Text
Bhāvaviveka Person
Bhogavatī Term
bodhicitta Term
Bodhisattva Term
Brahmā Deity
brahmin Term
Bṛhaspati Deity
Caryā Doxographical
Category
Ḍākinī Term
Daitya Term
Dānava Term
deva Term
Dhāraṇī Doxographical
Category
dharma Term
Dharmakāya Śarīra Text
textual discoveries dharmanidhāna Term
Dharmaśāstra Term
Durga Deity
Ekajaṭā Deity
Ekajatī Deity
Gabhistamat Place
Ganges Gaṅgā River
Garuḍa Purāṇa Text

homa Term
icchāntika Term
Indra Deity
Jambhaka Deity
Jambudvīpa Place
kalpa Term
Kamalaśīla Person
Kambala Deity
Kaṇikrodha Deity
Vajrakumāra
*Kaṇikrodha- Text
vajrakumāra-
bodhisattva-
sādhana-vidhi (Chi.
Sheng chia ni fen

nu chin kang t’ung
tzu p’u sa ch’eng
chiu i kuei ching)
kāpālika Term
Kāśyapa Deity
kavi Term
kavirāj Term
Kāvya Uśanas Deity
leg of the bed khaṭvāṅga Term
kīla Term
Kriyā Doxographical
Category
Kriyātantra Doxographical
Category
krodha Term
Kūrma Purāṇa Text
Kurukullā Deity
Madhyamaka Organization
Mahā Doxographical
Category
Mahābhārata Text
Mahākāśyapa Person
Mahākhya Place
Mahātala Place
Mahāyāna Organization
Mahāyoga Doxographical
Category
Maheśvara Deity
Maitreya Deity
Mandakinī Term
maṇḍala Term
Mañjuśrī Deity
Mañjuśrīmūlakalpa Text
mantra Term
mantrin Term
mudrā Term
main mantra mūlamantra Term
Nāga Term
hells of torment naraka Term
Nārada Deity
Narasiṁha Deity
Nārāyaṇa Deity
nidāna Term
Nidhidarśana Text
Nitala Place
Padma Purāṇa Text
Padmasambhava Person
five realms or paths pañca-gatayaḥ Term
of rebirth
Pātāla Place
attainment of Pātāla pātālasiddhi Term
patanti alam Term
Prahlāda Deity
pratibhāna Term
Pratyutpanna- Text
buddha-
saṃmukhāvasthita-
samādhi-sūtra
pratītya Term
purāṇa Term
Rākṣasa Term
Rāma Vājapeyin Person
Rasātala Place
Indian alchemical rasāyana Term
tradition
Sādhanamālā Text
Samantabhadra Deity
Sarasvatī Deity
Sarvapuṇya- Text
samuccaya-
samādhi-sūtra
attainment siddhi Term
sīma-bandha Term
stūpa Term
Subāhuparipṛcchā Text
Sugataśrī Person
heavenly delights sukha Term
Sutala Place
sūtra Term
six realms or paths ṣaḍ-gatayaḥ Term
of rebirth
Śāntideva Person
śāstra Term
enemy śatru Term
Śikṣāsamuccaya Text
Śiva Deity
Talātala Place
tantra Term
Tuṣita Heaven Place
Vācaspati Miśra Person
vaiḍūrya Term
vajra Term
Vajrabodhi Person
Vajrakīlaya Deity
Vajrakumāra Deity
Vajrapāṇi Deity
Vajraśekhara Sūtra Text
(Chi.
Chin-kang-ting
ching)
Vajrayāna Doxographical
Category
Vāmana Deity
Vasubandhu Person
magical knowledge; vidyā Term
magical sciences
holders of magical vidyādhara Term
knowledge
Virocana Deity
Viṣṇu Deity
Vitala Place
Viṣṇu Purāṇa Text
Vyāsa Person
Yoga Sūtra Text
yogin Term
Chinese
Dunhuang Place
Ershi’er zhong Text
dazangjing tongjian
Hsüan-tsang Person
Lo-yang Place
Bibliography
IOL Tib J 644, IOL Tib J 321. Dunhuang Tibetan manuscripts held at the British
Library, London. The International Dunhuang Project. http://idp.bl.uk/.
Pelliot tibétain 44. Dunhuang Tibetan manuscripts held at the Bibliothèque
nationale, Paris.
Aris, Michael. Hidden Treasures and Secret Lives: A Study of Pemalingpa
(1450-1521) and the Sixth Dalai Lama (1683-1706). London and New York:
Kegan Paul International, 1989.
Bhattacharya, B., ed. Sādhanamālā. Volume 2. Baroda: Gaekwad’s Oriental
Series, 1925.
Bu-ston Rin-chen-grub. The History of Buddhism in India and Tibet. Translated
by E. Obermiller. 2nd edition. Delhi: Sri Satguru Publications, 1986.
Dalton, Jacob. “The Early Development of the Padmasambhava Legend in Tibet:
A Study of IOL Tib J 644 and Pelliot tibétain 307.” Journal of the American
Oriental Society 124, no. 4 (2004): 759-72.
Davidson, Ronald M. Indian Esoteric Buddhism: A Social History of the Tantric
Movement. New York: Columbia University Press, 2002.
———. Tibetan Renaissance: Tantric Buddhism in the Rebirth of Tibetan Culture.
New York: Columbia University Press, 2005.
Eckel, Malcolm David. To See the Buddha: A Philosopher’s Quest for the Meaning
of Emptiness. Princeton: Princeton University Press, 1994.
Giebel, Rolf. Two Esoteric Sutras. Berkeley: Numata Center for Buddhist
Translation and Research, 2001.
Goudriaan, Teun. Māyā Divine and Human. Delhi: Motilal Banarsidass, 1978.
Hardy, Friedhelm. “Information and Transformation: Two Faces of the Purāṇa.”
In Purāṇa Perennis, edited by Wendy Doniger. New York: SUNY, 1993.
———. The Religious Culture of India: Power, Love and Wisdom. Cambridge:
Cambridge University Press, 1994.
Harrison, Paul M. The Ti b e t a n Te x t of the
Pratyutpanna-Buddha-Saṃmukhāvasthita-Sūtra. Tokyo: The International
Institute for Buddhist Studies, 1978.
———. The Samādhi of Direct Encounter with the Buddhas of the Present: An
Annotated English Translation of the Tibetan Version of the
Pratyutpanna-Buddha-Saṃmukhāvasthita-Sūtra. Tokyo: The International
Institute for Buddhist Studies, 1990.
———. “Mediums and Messages: Reflections on the Production of Mahāyāna
Sūtras.” Eastern Buddhist 25, nos. 1-2 (2003): 115-151.
Herrmann-Pfandt, A. “The Lhan kar ma as a Source for the History of Tantric

Buddhism.” In The Many Canons of Tibetan Buddhism, edited by Helmut
Eimer and David Germano, 129-49. PIATS 2000: Tibetan Studies: Proceedings
of the Ninth Seminar of the International Association for Tibetan Studies,
Leiden 2000. Leiden: Brill, 2002.
Hodge, Stephen, with contributions from Luke Lau, trans. “*Kaṇikrodha
Vajrakumāra Bodhisattva Sādhana Vidhi (T1222a).” Unpublished translation
from the Chinese. London, 1989.
Hodge, Stephen. The Mahāvairocana-Abhisaṃbodhi Tantra. London:
RoutledgeCurzon, 2003.
Hospital, Clifford. The Righteous Demon: A Study of Bali. Vancouver: University
of British Columbia Press, 1984.
Kloetzli, Randolph W. Buddhist Cosmology. Delhi: Motilal Banarsidass, 1985.
Jamgön Kongtrul. The Treasury of Knowledge. Book 1, Myriad Worlds. Ithaca:
Snow Lion, 2003.
Jørgensen, John. Inventing Hui-neng, the Sixth Patriarch: Hagiography and
Biography in Early Ch’an. Leiden: Brill, 2005.
Kapstein, Matthew. “New Light on an Old Friend: PT 849 Reconsidered.” In
Tibetan Buddhist Literature and Praxis, edited by Ronald M. Davidson and
Christian K. Wedemeyer, 9-30. Leiden: Brill, 2006.
Lessing, F. D., and Alex Wayman. Introduction to the Buddhist Tantric Systems.
Delhi: Motilal Banarsidass, 1983.
Malalasekera, G. P. Dictionary of Pāli Proper Names. London: Published for the
Pali Text Society by Luzac & Co., 1960.
Matsunaga, Yūkei. “On the Date of the Mañjuśrīmūlakalpa.” In Tantric and Taoist
Studies in Honour of R. A. Stein, edited by Michel Strickmann. Vol. 3: 882-894.
Bruxelles: Institut Belge des hautes études Chinoises, 1985.
Mayer, Robert. “Scriptural Revelation in India and Tibet.” In Tibetan Studies:
Proceedings of the 6th Seminar of the International Association for Tibetan
Studies, Fagernes, 1992. Vol. 2: 533-44. Oslo: Institute for Comparative
Research in Human Culture, 1994.
———. A Scripture of the Ancient Tantra Collection: The Phur-pa bcu-gnyis.
Oxford: Kiscadale Publications, 1996.
Scherrer-Schaub, C. “Enacting Words: A Diplomatic Analysis of the Imperial
Decrees (bkas bcad) and their Application in the sGra sbyor bam po gñis pa
Tradition.” Journal of the International Association of Buddhist Studies 25,
nos.1-2 (2002): 263-340.
Shastri, J. L. Garuḍa Purāṇa. Delhi: Motilal Banarsidass, 1978.
———. Kūrma Purāṇa. Delhi: Motilal Banarsidass, 1981.

Snellgrove, David L. Indo-Tibetan Buddhism: Indian Buddhists and Their Tibetan
Successors. Vol. 2. Boston: Shambhala, 1987.
Soifer, Deborah A. The Myths of Narasiṁha and Vāmana: Two Avatars in
Cosmological Perspective. New York: SUNY, 1991.
Stein, R. A. Grottes-Matrices et Lieux saints de la déesse en Asie orientale. Paris:
EFEO, 1988.
Strickmann, Michel. “The Consecration Sūtra: A Buddhist Book of Spells.” In
Chinese Buddhist Apocrypha, edited by Robert E. Buswell, 75-118. Honolulu:
University of Hawaii Press, 1990.
Sutherland, Gail Hinich. The Disguises of the Demon: The Development of the
Yakṣa in Hinduism and Buddhism. New York: SUNY, 1991.
Takasaki, Jikido. An Introduction to Buddhism. Translated by Rolf Giebel. Tokyo:
Tōhō Gakkai, 1987.
Tsai, Julius N. “Opening up the Ritual Casket: Patterns of Concealment and
Disclosure in Early and Medieval Chinese Religion.” Material Religion 2, no.
1 (2006): 38-66.
Wangdu, Pasang, and Hildegard Diemberger. dBa’ bzhed: The Royal Narrative
Concerning the Bringing of the Buddha’s Doctrine to Tibet. Vienna: Verlag
der Österreichischen Akademie der Wissenschaften, 2000.
White, David Gordon. The Alchemical Body: Siddha Traditions in Medieval India.
Chicago: University of Chicago Press, 1991.
Wilson, H. H. The Viṣṇu Purāṇa: A System of Hindu Mythology and Tradition.
Vol. 2. London: Trübner & Co, 1865.
Woodroffe, John (aka Arthur Avalon). Tantra of the Great Liberation
(Mahānirvāṇatantra). New York: Dover Publications, 1972.
Zil gnon nam mkha’i rdo rje, with contributions from Karmapa XV and Bdud
’joms rin po che ’jigs bral ye shes rdo rje. ’Chi med srog thig las byang rdo
rje’i phur pa yang gsang phrin las bcud dril gyi las tshe sgrub ’chi med srog
thig. In The Collected Writings and Revelations of H. H. Bdud-’joms Rin-po-che
’Jigs-bral-ye-ses-rdo-rje. Vol. 14 (pha). Kalimpong: Dupjung Lama, 1979.
Re-Assessing the Supine Demoness:
Royal Buddhist Geomancy in the Srong btsan sgam po

Mythology
Martin A. Mills
University of Aberdeen
Abstract: The myth of the Chinese princess Kong jo’s geomantic divination of
Tibet prior to the founding of the Central Temple of Lhasa (lha sa gtsug lag khang)
– and in particular the striking image of the land of Tibet as a “supine demoness”
– has been the object of considerable academic comment. Generally, it has been
read as a metaphor either of monastic Buddhism’s misogynist tendencies, or of its
superposition over putative religious precursors. In this article, the difficulties
that attend these interpretations of the supine demoness image are assessed when
examined within the context of the princess’s wider divination, as presented in
Tibetan mythic histories such as the Ma ni bka’ ’bum, The Clear Mirror of Royal
Genealogy, and the Pillar Testament (bka’ chems ka khol ma), and in particular
when it is viewed within the context of the Lha sa Valley’s actual topographic
structure. In light of these, it is proposed that both the supine demoness image and
the other elements of Kong jo’s divination should be understood as it has always
been presented by Tibetan sources – as part of an established tradition of Chinese
geomancy, a tradition which has itself been reorganized as a medium for Buddhist
themes of liberation.
Introduction
The legendary account of Emperor Srong btsan sgam po’s founding of the Central
Temple (gtsug lag khang) in Lha sa in the seventh century is perhaps one of the
most famous of all Tibetan myths. Certainly, both his and the temple’s focal place
within indigenous Tibetan histories makes a clear analysis of this legend crucial
1550-6363/2007/3/T3108.
© 2007 by Martin A. Mills, Tibetan and Himalayan Digital Library, and International Association of Tibetan Studies.
Mills: Re-Assessing the Supine Demoness 2
to understanding Tibetan conceptions of political and religious identity, and of

legitimate Buddhist governance.1
This hagiographic rendition of the foundation of the Central Temple of Lhasa
– Srong btsan sgam po’s primary ritual and regal act – is found in a variety of
Tibetan texts emerging between the eleventh and fourteenth centuries, most
famously the Ma ni bka’ ’bum2 and the Pillar Testament (bka’ chems ka khol ma).
These were generally gter ma, or “hidden treasure texts” – revealed during this
period by visionary yogins who traced their own spiritual genealogy back to the
time of the First Diffusion of Buddhism to Tibet, when the texts were said to have
been initially hidden by the likes of Srong btsan sgam po, Khri srong lde’u btsan,
and his teacher and exorcist Gu ru rin po che (Padmasambhava). Out of these initial
hidden treasure texts emerged later compilations such as The Clear Mirror of Royal
Genealogy (rgyal rabs gsal ba’i me long; henceforth The Clear Mirror) by the Sa
skya hierarch Bsod nams rgyal mtshan (1312-75).3 By the time of the Fifth Dalai
Lama (1617-82), Bsod nams rgyal mtshan’s text in particular was one of the most
influential of state histories.
This legendary corpus presents a reasonably consistent picture. Under its first
emperor, Srong btsan sgam po, the political sovereignty of the Yar lung dynasty
expanded the borders of its power outwards from Central Tibet, incorporating new
provinces through military conquest and diplomatic marriage, until its armies
pounded upon the gates of imperial China and the Buddhist kingship of Nepal.
Insisting upon royal marriage as a means to augment his authority within Asia, the
Tibetan emperor demanded – and was eventually (if reluctantly) given – brides
from the Chinese and Nepalese courts, both of whom brought Buddhist statues
with them as part of their dowries. His first consort, the Nepalese princess Khri
btsun, prompted the emperor to build a royal temple at Lha sa, his regular nomadic
feeding grounds. However, supernatural obstacles from the local spirits of Tibet
conspired to destroy the temple, destroying in the night what was built in the day.
In order to subdue them, Srong btsan sgam po sought geomantic instruction from
the Chinese princess Weng chen kong jo, who divined that the land of Tibet was
like a she-demon lying on her back, filled with inauspicious elements. All of these
required ritual suppression by subsidiary temples, mchod rtens,4 and other ritual
1
See Georges Dreyfus, “Proto-nationalism in Tibet,” in Tibetan Studies: Proceedings of the 6th
Seminar of the International Association for Tibetan Studies, Fagernes 1992, ed. Per Kvaerne, vol. 1
(Oslo: Institute for Comparative Research in Human Culture, 1994), 205-18.
2
See Leonard W. J. van der Kuijp, “Tibetan Historiography,” in Tibetan Literature: Studies in Genre,
ed. José Cabezón and Roger Jackson (Ithaca: Snow Lion, 1996), 39-56. See also Janet Gyatso, “Drawn
from the Tibetan Treasury: The gTer ma Literature,” in Tibetan Literature, 147-69.
3
See Per K. Sørensen, Tibetan Buddhist Historiography: The Mirror Illuminating the Royal
Genealogies: An Annotated Translation of the XIVth Century Chronicle rGyal-rabs gsal-ba’i me-long
(Wiesbaden: Harrassowitz Verlag, 1994). For a less scholarly but more accessible treatment, see
McComas Taylor and Lama Choedak Yuthok, trans., The Clear Mirror: A Traditional Account of
Tibet’s Golden Age (Ithaca: Snow Lion, 1996).
4
A mchod rten (San. stūpa) is a monumental reliquary – often containing the remains of dead bla
mas, old texts, or other relics – and is one of the most characteristic pieces of religious architecture in
the Buddhist Himalaya.
forms that had to subjugate the malevolent forms of the landscape and “pin down”
the limbs of the demoness before the emperor’s temple could be completed.
Following the Chinese princess’s advice, Srong btsan sgam po managed to bind
down the land of Tibet and complete the temple, built around a statue of his tutelary
deity, Avalokiteśvara. The temple acted as the central state edifice (gtsug lag khang)
for the emperor’s reign. In later years it became the home of the Jo bo statue of
Śākyamuni that had been brought from China by Kong jo, which became the basis
of the temple’s most common soubriquet outside Tibet, the Jo khang or “House
of the Lord.”
Interpreting the Myth

The myth, and the texts from which it derives, have received a very considerable
quantity of academic attention, as much for the issue of their historical veracity
(or lack thereof) as for their compelling mytho-poetic vision of the early Tibetan
emperor’s battle to bring the land of Tibet under Buddhist sovereignty.5 As a
depiction of religious conversion, much academic attention has been focused on
how the supine demoness image should be interpreted.6 By and large, the tale’s
dramatic imagery of vertical suppression has invited a series of analyses that have
emphasized its role as a metaphor for wider truths about Tibetan religion and
culture, primarily ones that emphasize social stratification and violence. Thus, the
story has been read as covertly presenting either a misogynist view of Tibetan
society,7 a fundamentally phallic understanding of kingly power8 or, in a theory
more specifically contextualized to Tibetan understandings of history, a mythic
enactment of Buddhism’s subjugation of Tibet’s pre-existing religious traditions.9
Since these are increasingly influential interpretations within western academia –
but at the same time would rarely be admitted to within the Tibetan tradition itself
– they require some careful critical discussion.
Psychoanalytic interpretations of the Central Temple of Lhasa founding myth
tend to emphasize the implicit sexual dimensions of the story, in particular the
vertical pinning of the demoness. As Janet Gyatso comments:
5
See Matthew Kapstein, The Tibetan Assimilation of Buddhism: Conversion, Contestation and
Memory (Oxford: Oxford University Press, 2000).
6
See Michael Aris, Bhutan (New Delhi: Vikas, 1980); Keith Dowman, The Sacred Life of Tibet
(London: HarperCollins, 1997); Janet Gyatso, “Down With the Demoness: Reflections on a Feminine
Ground in Tibet,” in Feminine Ground: Essays on Women and Tibet, ed. Janice D. Willis, 33-51 (Ithaca,
New York: Snow Lion, 1987) and in Alex McKay, ed., The History of Tibet, vol. 1 (London:
RoutledgeCurzon, 2003); Ana Marko, “Civilising Woman the Demon: A Tibetan Myth of State,” in
The History of Tibet, ed. Alex McKay, vol. 1 (London: RoutledgeCurzon, 2003), previously published
in Social Analysis 29 (1990): 6-18; Robert Miller, “‘The Supine Demoness’ (Srin mo) and the
Consolidation of Empire,” Tibet Journal 23, no. 3 (1998): 3-22; and Robert Paul, The Tibetan Symbolic
World (Chicago: University of Chicago, 1982).
7
See Marko, “Civilising Woman.”
8
See Paul, Tibetan Symbolic World.
9
See Gyatso, “Down With The Demoness.”
Part and parcel of the relationship between the demoness land and the architectural
structures upon her seem to be certain sexual innuendoes. If the srin mo is a Mother
Earth, then the architectural structures that hold her down must be seen as overtly
masculine. At one point in the srin mo myth this is quite explicit: one of the
pinning structures is a śiva liṅga, to be set on the “earth-enemy” (sa dgra) in the
east, a place which is “like the srin mo’s pubic hair.” Vertical buildings, imposing
structures … erections; in contrast, the feminine earth is associated with fertility,
nurturing, receptivity.10
By contrast, feminist interpretations of the demoness myth concentrate primarily

on an assumed equation between the symbol of the demoness on the one hand and
the institutional status of women (vis-à-vis a predominantly celibate male-dominated
Buddhist tradition) on the other. Thus, for Ana Marko, the violence against women
implicit within the myth is at the same time a metaphor for the genesis of the
patriarchal Buddhist state within which Tibetan women must subsist:
A vast number of Buddhist myths are contained in hagiography, or sacred history

stored in textual form, the authoritative property of the monastery. Since
monasteries are predominantly male institutions they act to reproduce culturally
constituted patriarchal power where categories of gender-based experience are
contained in myth. Violence plays a specific role in recreating a mythic notion of
wholeness through the body of woman the demon as fragmented territory, a site
for the recreation of wholeness. The body of woman the demon becomes the
mythic body of the state.11
Finally, culturalist arguments assert the myth’s metaphorical rendition of social

change, a retrospective evocation of the relationship between two religious cultures
– the Buddhist and the pre-Buddhist – during the time of the First Diffusion. Here,
the fundamental argument is that the suppressed demoness in some sense represents
the autochthonous religion of Tibet. Thus, for Keith Dowman, the supine demoness
represents one of a variety of “earth mother” symbols that
reveal a primeval strata of religion, a prehistoric era of matriarchy, or, at least, a

time when the female psyche, the primordial collective anima of the people, was
the predominant religious focus … The supine demoness, gigantic in size, is
herself vast in lust and bestial desire. But as order is imposed upon the chaotic,
instinctive and intuitive feminine realm of the psyche by the disciplined intelligence
of the masculine Buddhist will, so her desire is tamed.12
By presenting this pre-existent tradition as “subjugated,” the Buddhist tradition

is in turn seen as “stealing its thunder” and borrowing its very legitimacy in order
to augment its own. As Gyatso comments:
10
Gyatso, “Down With The Demoness,” 43.
11
Marko, “Civilising Woman,” 6.
12
Dowman, Sacred Life of Tibet, 19-20.
It is a common pattern: the old site of the indigenous religion is associated with
some sort of special configuration of the land, in which the powers of the deep
are perceived as having particular force…The incoming religion seeks out those
very sites, and builds right on top of them. The new structures obliterate the old
places of worship, but gain instant history and sacred power thereby.13
Here, the sites enumerated in Kong jo’s divination of the Tibetan landscape are
treated as pre-existent genii loci, spirits of place that were worshipped (or feared)
prior to the arrival of Buddhism. The story of the “supine demoness” thus becomes
a symbolic cornerstone of a debate between two religious traditions in early Tibet.
In this form, it speaks of two possible historical transitions:
• A cultural transition, in which the myth is a symbolic (and partial)

integration of two previously distinct cosmological systems: one a
pre-existing system of earthly and local deity cults (encapsulated en masse
in the image of the demoness); the other the subduing ritual force of a
transcendent Buddhism.14
• A political transition, in which the myth is a metaphor for the factional
debates between adherents of the local ancestral and aristocratic religious
traditions that preceded Buddhism’s arrival, and impeded its growing
hegemony within the dynastic court.
The first of these two interpretations implies an endeavor to legitimize the

incorporation of indigenous cosmological systems into Buddhist ritual forms. It
speaks primarily to the argument that Tibetan Buddhism is actually a combination
of Buddhist and tantric philosophical and ritual systems on the one hand and
indigenous Tibetan shamanism (in particular the worship of local and mountain
deities) on the other.15
These kinds of interpretation are ones in which the cosmological and mythic
are primarily metaphorical representations of the socio-cultural. Attractive though
such views of myth might be, there are several respects in which – as Gyatso admits
– “the pieces don’t quite fit together.”16 Indeed, I would argue that the pursuit of
various theoretical agendas within the socio-political sciences has caused many
such arguments to misconceive this myth, either by doing violence to the integrity
of its narrative as it appears in its various formulations (generally by emphasizing
certain elements of the story whilst eliding others) or by underestimating the
polemic intentions and narrative sophistication of its authors (this is particularly
13
14
For example, Anne-Marie Blondeau and Yonten Gyatso, “Lhasa, Legend and History,” in Lhasa
in the Seventeenth Century: The Capital of the Dalai Lamas, ed. Françoise Pommaret (Leiden: Brill,
2003), 19 n. 3.
15
For lengthier discussions of this topic, see Martin A. Mills, Identity, Ritual and State in Tibetan
Buddhism: The Foundations of Religious Authority in Gelukpa Monasticism (London: RoutledgeCurzon,
2003) and Geoffrey Samuel, Civilized Shamans: Buddhism in Tibetan Societies (Washington:
Smithsonian Institute Press, 1993).
16
See Gyatso, “Down with the Demoness.”
the case with Bsod nams rgyal mtshan). In both these regards, insufficient attention
has been paid to the clear (and explicitly recognized) Chinese origins of Tibetan
geomancy, and to the place that such geomancy had within a wider Buddhist vision
of religious liberation and state legitimation.
The Historical Dynamics of Tibetan Geomancy

The myth of the building of the Central Temple of Lhasa speaks to a highly complex
science of geomancy within Tibetan culture, either at the time of Srong btsan sgam
po himself or developed in the subsequent centuries and “reflected back” to the
Yar lung emperor’s rule by later Tibetan historians. Whichever of these was the
case (and there is some evidence that both were true to varying extents), the impact
of the myth on subsequent architects of Tibetan governance (such as the Phag mo
gru dynasty and, later, the Dga’ ldan pho brang government) was clear: to model
one’s own government on that of the early imperial period was also to adopt an
established understanding of rule as geomancy.
The science of geomancy is both one of the most ubiquitous and yet obscure
traditions in Tibet. Often called sa spyad or byung rtsis, many Tibetan historians
are fairly explicit that the traditions of elemental – that is, earthly – divination were
inherited from China, as opposed to the Kālacakra-dominated astrological system,
which was imported from India. Texts such as The Clear Mirror clearly depict the
geomantic arts as primarily being brought to Tibet by figures from China (with
the principal exception of Gu ru rin po che), and linked to the creation of royal
religious space as a basis for auspicious rule.
Over the course of the post-dynastic, local hegemonic, and medieval periods,
however, geomantic traditions seem to have become widespread throughout Tibet,
becoming a standard prerequisite for the sitting of important houses, castles, and,
above all, monasteries and temples. Tibetan geomancy developed several important
and distinctive features during this long history that separated it in particular from
the practice of imperial feng shui in China: in place of the central Chinese concern
with the correct placement of ancestral funerary sites came a focus on the
vitality-place (bla gnas) of the living;17 in place of imperial regulation came a much
more devolved concern with auspicious placement; and in place of a relatively
public and professionalized system of divination, a marked institutional reticence
– indeed secrecy – surrounding geomantic divination within the institutional folds
17
Regarding burial sites, we know that the burial sites of the old Tibetan kings are still seen – like
their Chinese counterparts – as having an ongoing geomantic influence. Thus, Rene de
Nebesky-Wojkowitz, Oracles and Demons in Tibet: The Cult and Iconography of Tibetan Protective
Deities (Kathmandu: Tiwari Pilgrim’s Book House, 1993), 482, reports how Glang dar ma’s burial site
on Bya skya dkar po ri is said to continually threaten the well-being of Lha sa. However, the emphasis
appears now to be more on the positioning of mchod rtens containing the remains of high bla mas,
although to my knowledge no research has been carried out on the geomantic sitting of such mchod
rtens. Clearly, some degree of astrological knowledge is employed at funerals (see for example Stan
Mumford, Himalayan Dialogue: Tibetan Lamas and Gurung Shamans in Nepal [Madison, Wis.:
University of Wisconsin Press, 1989], chap. 10), but it remains unclear to what extent this shades into
the specifically geomantic.
of Tibetan monasticism.18 Above all of these, however, is to be found a pronounced

incorporation of geomantic relations with the landscape into the structuring of
Buddhist ritual life, as opposed to feng shui’s general domestication to the imperial
Confucian paradigm.19
Nonetheless, despite these later developments, the image of geomancy’s
importation during the grand dynastic period of Srong btsan sgam po and his
successors remains an important literary template for both the form and cultural
place of this divinatory art in Tibet, lending a certain stability to some of its key
features. In what follows, I would like to turn the examination of the entire
demoness myth in a direction concomitant with an awareness of the key place that
geomancy has in Tibetan cultural and religious history. While certain writers –
most notably Elisabeth Stutchbury – have highlighted the importance of the
geomantic traditions (including their Chinese historical origins) to the local
formations of Tibetan religious life,20 we have yet to look more deeply at what
those geomantic formations themselves tell us about how eleventh- to
fourteenth-century Tibetan Buddhist religious thought understood the “conversion”
of the dynastic state at Lha sa.
18
Prominent exceptions to this reticence include Sangs rgyas rgya mtsho’s Bai ḍūrya dkar po and
Thub bstan rgya mtsho’s much more recent Bstan pa’i rtsa ba chos sgor zhugs stangs dang / bstab
pa’i bsti gnas gtsug lag khang bzhegs thabs (see Thubten Legshay Gyatsho, Gateway to the Temple,
trans. David P. Jackson, Bibliotheca Himalayica, series 3, vol. 12 (Kathmandu: Ratna Pustak Bhandar,
1979). In a recent set of talks on the topic given by the current Twelfth Si tu rin po che, he differentiated
between the generic tactics of household and temple geomancy (which he discussed in some detail)
and the fundamental principles at work in personal geomancy – those principles which link a person’s
known place and date of birth to the very elemental forces which keep them alive (see also
Nebesky-Wojkowitz, Oracles and Demons, 481). In particular, the science of knowing a person’s bla
gnas or “vitality-place” – a feature of the landscape that contains their life-force (bla) – was one which
could be employed to assassinate that person, and thus was to be carefully guarded by lineage holders
(Situ Rinpoche, “Geomancy,” Audio Z91 [Eskdalemuir: Samye Ling Tibetan Centre, 1988]). An oral
tradition popular in Buddhist Ladakh spoke to this very principle. During the reign of the “heretic king”
Glang dar ma, the Buddhist siddha Dpal gyi rdo rje sought to end his persecution of Buddhism by
assassinating him. Seeking to avoid a direct confrontation, Dpal gyi rdo rje sought instead to cause the
king’s death magically. Bribing the king’s diviner, he found out that the king had three bla gnas – in
a mountain, a tree, and a sheep. He was successful in digging up Glang dar ma’s life-mountain and
cutting down his life-tree, and the king fell gravely ill. However, the king had cunningly hidden his
“life-sheep” amongst a flock of five-hundred other similar sheep. Rather than kill so many animals,
Dpal gyi rdo rje was forced to confront the king in person.
19
That is not to say that the Tibetan context produced a unique set of changes in this regard, but
rather that they developed further in specific directions. As I will argue below, certain strains of
geomancy in China had already taken on a distinctly Buddhist flavor. Moreover, the Indic context of
tantric rites of subjugation – many of which were clearly focused on ritual relations with the land (see
Robert Mayer, A Scripture of the Ancient Tantra Collection: The Phur-pa bcu-gnyis [Oxford: Kiscadale,
1996]) – were the clear origin of the kīla-rites mentioned in most of the Srong btsan sgam po
hagiographies as the ritual prelude to the founding of the Central Temple of Lhasa.
20
Elizabeth Stutchbury, “Perceptions of the Landscape in Karzha: ‘Sacred’ Geography and the Tibetan
System of ‘Geomancy,’” in Sacred Spaces and Powerful Places in Tibetan Culture, ed. Toni Huber
(Dharamsala: Library of Tibetan Works & Archives, 1999).
Architectures of Auspiciousness
In the myth of the founding of the Central Temple of Lhasa, a highly complex
lattice of geomantic forces and their ritual amelioration is laid out by Princess
Kong jo. In The Clear Mirror and Pillar Testament, for example, Kong jo’s
divination contains between forty-five and fifty separate geomantic elements, along
with the means either to suppress them (in the case of visible earth-enemy) or
augment them (in the case of latent rten ’brel). Since both texts are lengthy, I have
broken them down into tabular form (Tables 1 and 2).
Rather than being a simple list of malevolent forces (as it is all too often read),
Princess Kong jo’s divination is divided into several analytic categories, a pattern
characteristic of Chinese geomancy:
• A visual analysis of the negative geomantic facets that directly impinge

upon a site from its immediate surrounding area (Table 1: items 1-9; Table
2: items 1-12). Each of these facets is then given a geomantic “solution,”
a means of counteracting its negative force. Thus, for example, Dbus ri
mountain on the northern outskirts of Lha sa (local tradition has this as
located behind the present ’Bras spungs Monastery) looked like a charging
war elephant, a negative facet that needed balancing with a stone lion that
faced it (T1:9; T2:11); similarly, the Rock of Shün (shun gyi brag, T1:8;
T2:10), the prominent finger-like pinnacles directly to the west of the city,
was seen as a “demon staring at Lha sa,”21 to be faced by a (red) mchod
rten.
• A similar analysis of the positive geomantic facets present within the
immediate surrounding area. Examples of this are the glacier behind Grib
Village to the south of Lha sa, divined to be like a conch shell (T1: 17; T2:
18), and the marsh at the opening of Stod lung Valley that was “like the
wheel of doctrine on the feet and hands of the Buddha” (T1: 21; T2: 23).
These are presented as (at the time of the initial divination) overshadowed
by the preceding negative elements, which require subjugation for the
positive facets to be brought forth.
• An analysis of the wider geomantic context within which the Lha sa area
is embedded. As with the previous category, much of the art of Chinese
geomancy lies in what Needham refers to as “physiographic map-making”22
– that is, the skill of “drawing” bodily images into a specific landscape.
Here, this category is divined by Kong jo in terms of our now-famous
21
See also Matthew Akester, “A Black Demon Peering From the West: The Crystal Cave of
Suratabhaja in Tibetan Perspective,” Buddhist Himalaya: A Journal of Nagarjuna Institute of Exact
Methods 10, nos. 1 & 2 (1999).
22
Joseph Needham, Science and Civilization in China, vol. 2 (Cambridge: Cambridge University
Press, 1962), 360.
“supine demoness” image (T1: 12-23; T2: 32-49), which crosses Tibet as
a whole.23
A variety of geomantic principles are at work here: firstly, it uses techniques

of both vertical suppression and horizontal opposition; secondly, the divination
balances its diagnosis of negative elements with a prognostication of emergent
auspicious properties in subjugated landscapes; and finally, the entire divination
is organized according to a concentric arrangement. In what follows, I will look
at each of these in turn.
Techniques of Vertical and Horizontal Suppression

The first, and most obvious, geomantic technique employed within the context of
the myth is that of vertical suppression. Most famously, the twelve temples built
by Srong btsan sgam po are employed to “bind down” the arms and legs of the
supine demoness. This is a common feature of Tibetan Buddhist ritual, utilizing a
temple or mchod rten to “bind down” troublesome or labile features of the
landscape. In many cases the binding down temples in The Clear Mirror narrative
are also sites for the subjugation of Nāga water spirits. This is most obvious in the
case of the Central Temple itself (which was built on the Lake of O Plain [’o thang
gi mtsho], the “palace of the water spirits” in Kong jo’s divination), but can also
be seen at the Khra ’brug and Ka rtsal binding-temples, and at other dynastic sites
such as the courtyard of Bsam yas Temple. At Khra ’brug, tradition relates that
Srong btsan sgam po and Princess Kong jo met substantial resistance:24 as they
sought to journey south to the emperor’s palace at Yum bu lag khang, they were
stopped by a large flood at the confluence of the Yar lung and Tsang po Valleys.
Here, the emperor perceived that the source of the flood was a recalcitrant Nāga,
a serpent with five heads. At this, the emperor transformed into a fearsome
Garuḍa-bird, and swooped down on the Nāga, slicing off first three of its heads
with his beak, and then on the second pass, the remaining two.
Despite the violence of this and similar stories, the suppression of the Nāgas
and demoness is not apparently fatal, but rather debilitating.25 They are left in
submission to the emperor’s power, rather than destroyed by it. This can be seen
in the continued ritual recognition of their presence: at the Central Temple of Lhasa,
Bsam yas and Khra ’brug temples, small Nāga-shrines can be found attended
nearby by ritual wells devoted to the Nāga-lord.26
23
In the Pillar Testament, Kong jo adds a further dimension which presents Tibet qua demoness as
one area amongst a general geography of the Southern continent of Jambudvīpa (see Sørensen, Tibetan
Buddhist Historiography, 553-60). While it is worth noting its existence, I have not included this wider
geomantic tableau in Table 2; it is not present in The Clear Mirror.
24
Guntram Hazod, “Around the ‘Secret’ of Tantruk: Territorial Classifications in the Historical
Landscape of Lower Yarlung (Central Tibet)” (unpublished manuscript).
25
See also Gyatso, “Down with the Demoness.”
26
In the Central Temple of Lhasa, this “well” is simply a small podium with a two-inch wide shaft,
at which pilgrims listen in order to hear the “sound” of the Nāga-lord’s kingdom.
The second general technique evoked in the Central Temple founding story is
that of horizontal opposition: counterbalancing geomantic obstructions around the
Lha sa Valley by placing their “suppressors” opposite them on a horizontal line
of sight. There are numerous cases of this in The Clear Mirror and Pillar Testament,
and many to this day receive some degree of ritual observance. A dramatic example
from The Clear Mirror – which demonstrates how a horizontal-facing arrangement
was seen as being able to transform the landscape – can be found in the heart of
Lha sa itself. On the eastern side of Lcags po ri (referred to in many early texts as
Lcags kha ri), the hill directly to the southwest of the Po ta la Palace, a sheer
concave shouldered cliff faces directly towards the main, west-facing door of the
Central Temple of Lhasa (see Photo 4). In the middle of the cliff are the shrines
of Brag lha klu sgug Temple, the most famous of which is a ring-like chapel built
around an “self-created” image of the Buddha entitled “Able Rock Protector” (thub
pa brag lha mgon po), flanked by Śariputra and Maitreya on his right and
Maudgalyāyana and Avalokiteśvara to his left. The “self-created” Buddha is
described in The Clear Mirror as a result of the magical action of the
newly-constructed Eleven-Headed Avalokiteśvara statue immediately following
the building of the Central Temple, which faced towards it. Indeed, in The Clear
Mirror, the subsequent rock protector image is held to act as the representative of
the Central Temple’s subjugated water-spirits.27
A more complex and instructive example can be found at Grib Village, on the
far side of the Skyid chu River to the south of Lha sa. Bsod nams rgyal mtshan’s
text has the following prognostication:
In the southern direction [from Lha sa] the “terrestrial antagonist” (sa dgra) is
found, which resembles a black scorpion pouncing on [its pray] (sdig pa nag po
gzan la rub pa). It is [to be identified as] the eastern summit of Yug ma ri. [To
suppress it, af]front it [by erecting a statue of] the garuḍa-bird ke ru.28
Whilst the names of the mountains have changed, the scorpion itself clearly
refers to what is locally referred to as “Phan ju ri outcrop,” a mountain arm on the
west side of the valley (see Photo 1).29 The Phan ju ri outcrop is “faced” by a
protector shrine on the far side of the valley, dedicated to the protector deity Tsher
rdzong rtse, whose presence local tradition within Grib ascribes to Princess Kong
jo, who is said to have summoned the deity from China (see Photo 2).
27
Sørensen, Tibetan Buddhist Historiography, 297; see also Taylor and Yuthok, The Clear Mirror,
158. This legendary formulation is not found in the earlier Pillar Testament, where the protector image
is seen as one of the apotropaic forms to be built by Srong btsan sgam po. The Clear Mirror mediates
this distinction slightly, reporting that the protector image emerged spontaneously and was later carved
into sharper relief under the emperor’s orders.
28
164.
29
Both phan ju ri and tsher rdzong rtse are approximate transliterations only.
Map 1: Geomantic organization of Grib Village,

south of Lha sa.
Photo 1: Phan ju ri outcrop, identified above as

the “leaping scorpion.”
Now, Phan ju ri certainly looks like a scorpion. However, being a
three-dimensional topographic object it does so only from a certain angle,
specifically from a line leading east-southeast from Phan ju ri (see Map 1, Line
A). However, the Tsher rdzong rtse Temple itself clearly faces the scorpion along
Line B (from which angle the scorpion looks partially distorted). If this temple (or
a predecessor) represents the “garuḍa-bird eagle” of Bsod nams rgyal mtshan’s
text – designed to suppress the malignant scorpion image – then why was it placed
slightly out of alignment, when such evident care was taken over the visual
properties of such signs?
Photo 2: Tsher rdzong rtse Temple (center),

located within the shoulder outcrop of Bong po
ri Mountain.
Photo 3: Tsher rdzong rtse Temple.

Discounting the possibility of major geological shifts since the text’s authorship,
it would seem to me that another local geomantic feature needs to be incorporated:
Tsher rdzong rtse Temple is placed at the center of a west-facing shouldered outcrop
of Bong po ri Mountain (see Map 1 and Photo 2), one which directly faces Phan
ju ri. Directly behind the temple is the principal peak of Bong po ri, and the temple
is nested between two shoulder-spurs. Such sites are often used within Chinese
geomantic systems as supporting features that augment the power of particular
temples,30 and the Nepalese princess’s temple at Brag lha klu sgug is located at
the center of a similarly shouldered outcrop of Lcags po ri Hill in Lha sa (see Photo
4).
30
See Stephan Feuchtwang, An Anthropological Analysis of Chinese Geomancy (Bangkok: White
Lotus, 2002), 156.
Photo 4: Lcags po ri Hill, Lha sa – shouldered

outcrop to the left.
The placement of the Tsher rdzong rtse Temple thus implies (and we can say
no more than this) that several different geomantic features were balanced against
one another: the angle at which the scorpion can be seen versus the angle from
which it can best be suppressed. The apotropaic geomancy of this site therefore
seems to be a complex calculation, simultaneously incorporating several different
geomantic principles.
The Concentric Organization of Divination

Like many geomantic divinations, Kong jo’s diagnosis is concentrically-structured,
focused on the potential Lha sa site for the Central Temple. The nature of the
divined facets changes qualitatively as it moves outwards from the center of the
Lha sa Plain:
• Elements within or crossing the Lha sa Plain itself (T1: 1-4; T2: 1-6).
Generally, these are depicted as the habitual place of actual demonic beings:
the palace of the Nāga-king; the cave of the black Nāgas; the meeting place
of the The’u rang spirits and ghosts; the route of the Btsan spirits. These
require vertical suppression, either by temples or mchod rten.
• Elements on the visual periphery – or “sides” – of the Lha sa Plain (T1:
5-9; T2: 7-31, but 27 is an exception). These are universally mountain
features visible from the Lha sa Plain itself or its immediate tributary
valleys. Rather than being actual numinal persons, they are (auspicious or
inauspicious) visual signs,31 requiring horizontal counter-balancing when
inauspicious (see below).
• Finally, there is the general diagnosis of the wider Tibetan geography
(between Khams in the east and the borders of Mnga’ ris in the west),
which is “like” a supine demoness. This image actually interlaces with, or
is visualized “on top of,” the above two, such that the Lake of O Plain on
the Lha sa Plain is simultaneously depicted as the “heart-blood of the
demoness” and the palace of the Nāgas.
31
The main exception here being the four mines (T1: 22-25; T2: 25-28), assuming they were treated
as substantial physical objects (which they may not have been – this would require some examination).
This subtle interpretive shift as one heads outwards from the Lake of O Plain
is matched by a political shift within the image of the supine demoness itself. The
binding-down temples, organized in three concentric squares, were designated
according to their relationship with the civilizing political power of the state. Thus,
beyond the Lha sa heartlands lay the innermost square of temples (for nailing down
her hips and shoulders), called the District-Controlling temples (ru gnon gyi lha
khang; T1: 31-34; T2: 37-40); the four intermediate temples (for nailing down her
elbows and knees) were the Border-Taming temples (mtha’ ’dul gyi gtsug lag
khang; T1: 35-38; T2: 41-44); and four outer temples (nailing down her wrists and
ankles) were the Further-Taming temples (yang ’dul gyi lha khang; T1: 39-42; T2:
45-49).32
Within this configuration, therefore, ritual power is extended beyond the state’s
own borders: in The Clear Mirror, the construction of these “further-taming”
temples are described as being placed in the care of surrounding tribal groups and
leaders such as the To dkar to the south, the Mi nyag of Khams, Sba dpal dbyang
of Hor, and the Nepalese to the southwest (T1: 39-42; not in T2). Robert Miller
has since gone on to render this analysis more concrete, by noting the tendency of
Tibetan medieval historians – in particular Bsod nams rgyal mtshan’s contemporary,
Bu ston (1290-1364) – to emphasize the relationship between the placing of the
nailing-down temples and contemporary histories of revolt within the newly
expanded Yar lung empire.33 Dunhuang documents speak of revolts amongst the
Zhang zhung, the Sum pa, the Dwags po, the Rkong po, and the Myang, followed
by expansionist expeditions by Srong btsan sgam po to the north through Mal gro
and on to ’Dam – all key sites for his later “nailing-down” of the demoness. For
Miller, Bu ston’s rendition of the story is
a tale of the process of centralizing and of re-affirming the power of the Centre.
That power was threatened by revolts, possibly stimulated by the king’s flirtation
with Buddhism … The order in which the srin mo was pinned down reflect the
expansion of the Yar lung Empire under Gnam ri slon mtshan, and its consolidation
and further expansion under Srong btsan sgam po.34
In this regard, the religious image of the demoness’ suppression had a “secular”
corollary, or at least one within the practical history of Yar lung sovereignty. Thus,
a vision is created not only of a system of temples, but of a resurgent state.
32
In Bsod nams rgyal mtshan’s text, or a subsequent redaction of it, these terms have been confused,
being placed in the order border-taming, further taming, and district taming. This is at odds with every
other version of this story. Thus, following Sørensen, Tibetan Buddhist Historiography, 261 n. 770, I
place them in the standard order in Table 1.
33
Miller, “The Supine Demoness.”
34
Miller, “The Supine Demoness.”
Re-Assessing the Significance of the Central Temple Founding

Myth
At the beginning of this article, I briefly examined some of the existing scholarly
interpretations of the Central Temple of Lhasa founding myth, interpretations that
focused primarily (indeed, almost exclusively) on the image of the suppression of
the demoness. Now that we have looked at the wider context of the supine demoness
myth and its associated geomancy, we can critically examine these ideas from a
stronger position. As I suggested earlier, there are several ways in which existing
feminist and culturalist readings are inadequate.
Problems with the Feminist Analysis

At the heart of the feminist argument is the assertion that the image of the supine
demoness is a mythic formulation of the patriarchal gender categories of a celibate
monastic elite. Now, it is certainly the case that the ideology of women’s
subordination within Tibetan society is encapsulated within certain elements of
clerical Buddhism that see women as “low-born.”35 However, to equate the social
status of women with the figure of the demoness is problematic at best, requiring
the reader to “bracket out” most of the rest of the narrative content of the wider
myth and ignore much of wider Tibetan religious culture.
Firstly, the argument spotlights the vertical and gendered dimensions of the
divination – generating an exclusively (and conveniently) stratified sexual imagery
– whilst separating out the very geomantic principles that the divination was
primarily about. In particular, it separates the processes of vertical suppression
from their clearly linked processes of horizontal opposition (which are less amenable
to feminist and psychoanalytical images of sexual repression).
Secondly, this kind of feminist argument only works for the supine demoness
story if we bracket out the dynastic tale of which it is an integral part. If we look
at the supine demoness element in this wider context, then to assert that it represents
a celibate Buddhist male suppression of the feminine seems peculiar at best, for a
variety of reasons. In the first case, it ignores the fact that the entire interpretive
project – of seeing Tibet as a demoness capable of being nailed down – was
presented not by a male celibate monastic, but by a female dynastic figure – the
princess Kong jo, in answer to a request from another female dynastic figure, the
princess Khri btsun. Moreover, the binding-down temples – whilst now monasteries
that do indeed belong to the various clerical Buddhist schools, not least the wholly
celibate Dge lugs pa – were primarily placed at sites of royal marriage rather than
celibate monasticism.
Finally, the assumption that female literary and mythic figures can be seen as
“representing” the general category of women – whilst common within feminist
critiques of religion – simply does not hold in the Tibetan context. For example,
35
Barbara Aziz, “Towards a Sociology of Tibet,” in Feminine Ground: Essays on Women and Tibet,
ed. Janice D. Willis (Ithaca: Snow Lion, 1989).
Huber’s recent examination of ritual prohibitions on the sacred mountain of Tsa

ri in southeastern Tibet describes how women are excluded from particular
pilgrimage routes round the mountain because their bodies might pollute the female
divine forms manifest in its landscape. Thus, only male ascetics are allowed to
drink from the stream whose reddish waters are seen as the menstrual blood of the
goddess Vajravārāhī.36 Similarly, the rationales given for women’s exclusion from
the region usually involve them being banished precisely by female divine figures.37
So, whilst Huber’s ethnography certainly speaks of the ritual exclusion (for which
we might read “subordination”) of women from certain Buddhist sacred sites (a
common prohibition regarding powerful tantric deities),38 the divine female figures
involved are often evoked as the very reason for that very subordination, rather
than as victims of it. In light of Huber’s ethnography, we can no longer take an
equation between the supine demoness and the social class of Tibetan women as
a given; indeed, I would argue that the “supine demoness as patriarchal ideology”
thesis only makes sense if we do scholastic violence either to the story of which
it was a part or to the culture from which it emerged.
That is not to say that the story has no polemic intent, but it is difficult to make
definitive claims. It would be equally possible to argue that the initial polemic
thrust of the Srong btsan sgam po mythos could be seen as a Rnying ma
championing of a non-celibate (indeed, dynastic) Buddhist dynamic, in opposition
to the rising power of celibate monastic institutions within twelfth-century Tibet.
Difficulties surrounding this question center on the institutional affiliation of the
early treasure-revealers. For example, of the three revealers of the Ma ni bka’ ’bum,
only Nyang ral nyi ma ’od zer (1124-92?) was clearly a non-celibate Rnying ma
tantrist; Grub thob dngos grub (c. 1100s?) was also a yogin of some kind, while
Shākya ’od is sometimes referred to as a bhikṣu.39 The story’s later adoption by
the likes of Bsod nams rgyal mtshan (who certainly was a monk) occurred within
the context of their own re-appropriation of the concepts of dynastic rulership in
Tibet.40
Problems with the Culturalist Analysis

By contrast, the culturalist argument – that myths such as that of the supine
demoness represent a polemic rendering of the historical appropriation of
36
Toni Huber, “Why Can’t Women Climb Pure Crystal Mountain? Remarks on Gender, Ritual and
Space in Tibet,” in Tibetan Studies: Proceedings of the 6th Seminar of the International Association
for Tibetan Studies, Fagernes 1992, ed. Per Kvaerne (Oslo: Institute for Comparative Research in
Human Culture, 1994), 360.
37
Huber, “Why Can’t Women Climb Pure Crystal Mountain?,” 355.
38
See for example Mills, Identity, Ritual and State, 184.
39
Many thanks to the JIATS anonymous reviewers on this point.
40
The maintenance – even in the subsequent monastic context – of this dynastic and marital dimension
to the myth can be seen rather clearly at Khra ’brug, the first Buddhist temple in Tibet and one of the
inner rings of binding temples. Here, the principal Avalokiteśvara temple (located in front of the
Nāga-shrine) contains – in pride of place for both its attendant monks and visiting pilgrims – the marital
hearth of Princess Kong jo.
pre-Buddhist ritual sites by an incoming Buddhist institutionalism – suffers from

the fact that the archaeological and architectural culture of the period points in a
very different direction. To begin with, there is no archaeological evidence to
suggest that the Lha sa area prior to the founding of the Central Temple complex
was any kind of major cultic site for pre-Buddhist ritual traditions, such that
Buddhism might inherit their “history and sacred power.”41 Whilst the Lha sa
Valley is certainly depicted as a haven of Klu (nāga), The’u rang, and Ma mo
spirits in post hoc Buddhist re-writings, this can hardly be taken as an indicator
that such spirits or anything like them existed as part of any previous institutional
religion. For example, the story itself makes no mention of any shrines or temples
to them that needed destroying, or for that matter any contra-Buddhist reaction to
protect these pre-existing genii loci. Indeed, throughout texts such as The Clear
Mirror, almost no mention is given to pre-existing named tellurian deities of the
Central Tibetan area.42
Secondly, there are difficulties in conceiving this story as a “battle between the
gods” in any sense that we might normally understand it. Much of Kong jo’s
divination is not about deities or spirits at all. Certainly, shrines are built to the
Nāga water-spirits that live under the Central Temple, Khra ’brug, and later Bsam
yas, water-spirits that actively battled against the royal powers at sites such as Khra
’brug. The demonic forces of Gla ba tshal, moreover, are depicted in The Clear
Mirror as conspiring against Srong btsan sgam po. They are, in other words,
depicted as active numinal agents. However, in terms of Kong jo’s wider divination,
this is not generally the case. In the case of the scorpion at Grib, the Rock of Shün,
or the war-elephant near ’Bras spungs, the obstructive geomantic obstacles of the
Tibetan landscape are not obviously presented as spirits or deities, despite how
often they have been read as such by modern scholars.43 No shrine appears to have
been dedicated to them, and their existence is clearly registered only from a
particular physical direction, as we saw in the case of the Grib scorpion. By contrast
with the demons of Gla ba tshal, the peripheral geomantic elements divined on the
sides of the valley are depicted simply as visual signs that look like negative omens
when seen from a particular angle, rather than (even nominally existing) gods or
spirits that inhabit places. Even the demoness (who also receives no shrine) lacks
this firm numinal quality; the land of Tibet is deemed by Kong jo to be “like” or
“resembling” or “to have the shape of” a demoness, implying it was akin in this
regard to Grib’s scorpion. That is, not a deity at all in the sense that we would
understand the term, but a visualized similitude – an image.
41
42
The sole exception here being the Nāga-king Mal gro gzi can within Khri srong lde’u btsan’s
hagiography. Regarding this question, Stan Mumford’s study of religious change in the Buddhist
communities of Nepal (Mumford, Himalayan Dialogue, chap. 3) is instructive. He clearly demonstrates
how, in the modern context, named local deities are categorically “re-structured” by high bla mas,
reforming them within rites as generic categories within a wider Buddhist cosmology. This would
clearly be a suggestive argument against my position above.
43
See Gyatso, “Down With The Demoness,” 49; Sørensen, Tibetan Buddhist Historiography, 253.
We can more clearly see the problem of treating Kong jo’s divination of the
demoness as the reading of a pre-existing divine cosmology if we expand the field
somewhat and look at, for example, The Clear Mirror’s later rendition of Gu ru
rin po che’s similar (if more upbeat) geomantic divination of the landscape around
Bsam yas, performed for Emperor Khri srong lde’u btsan prior to building his
tutelary temple:
The mountain of shar-ri [around bsam-yas] resembles the king poised on a seat
(rgyal pa gdan la bzhugs pa). This is a good [sign]. The ri-chung resembles a
brood-hen covering its [young] bird (bya mas bu la sgab pa). This is a good [sign].
The sman-ri resembles a mound of jewels (ri[sic] chen spungs pa). This is a good
[sign]. Has-po-ri resembles a queen dressed in white silk (btsun ma dar dkar gyi
na bza’ gsol ba). This is a good [sign]. Ri-nag resembles an iron nail stuck into
the ground (lcags phur sa la btab pa). This is a good [sign]. Me-yar resembles a
mule drinking water (dre’u chu ’thung pa). This is a good [sign]. Dol-thang
resembles a curtain [made] of white silk drawn (dar dkar gyi yol ba). This is a
good [sign]. The site [around bsam-yas] resembles a golden tub (gzhong) filled
with saffron-flower. This is [also] a good [sign], so erect the ruler’s personal
tutelary temple here!44
It is difficult here to see any of Gu ru rin po che’s divinatory prognostications

– the iron nail, the brood hen, the mule drinking water – as referring to a pre-existing
numinal cosmology, as opposed simply to a distinctly visual interpretation of the
auspiciousness of the site. Moreover, there is no sense emerging from the text that
their quality as omens or spiritual obstructions derives from a metonymic equation
with some hidden numinal reality, as implied by Sørensen.45 More importantly, it
seems difficult to equate this particular kind of geomantic vision with the kind of
local area god worship which is often associated with pre-Buddhist mountain
worship (or even the modern Tibetan propitiation of local spirits). Indeed, local
tradition asserts that certain local deities in the area – such as Tsher rdzong rtse in
Grib Village – were imported (in this case by Kong jo herself) precisely to subdue
such (numina-less) signs.
The Auspicious Symbolism of the Lha sa Heartlands

This may well sound somewhat hair-splitting. The point, however, is that the
aesthetic of such visual “signs” is less reminiscent of “Bon” or other forms of
Tibetan ancestral cult than of those Chinese systems of geomancy that were widely
prevalent (indeed, reaching their political zenith) within Kong jo’s native China
during this period.
44
235.
45
Sørensen, Tibetan Buddhist Historiography, 253, 552; and Per Sørensen, “Lhasa Diluvium - Sacred
Environment at Stake: The Birth of Flood Control Politics, the Question of Natural Disaster Management
and Their Importance for the Hegemony over a National Monument in Tibet,” in “Cosmogony and the
Origins,” special issue, Lungta, no. 16 (Spring 2003): 88.
This can be seen most clearly in the way in which The Clear Mirror and Pillar
Testament’s rendition of geomancy is not solely focused on the subjugation of
obstacles (bar chad), but also speaks of an understanding that the subjugation of
such geomantic obstacles gives rise to the emergence of many naturally existing
auspicious signs (bkra shis rtags) within the landscape. This is an element of Kong
jo’s divination that receives little or no critical attention within the English-language
literature on the topic, despite the fact that The Clear Mirror for one outlines a
group of nineteen potentially auspicious signs within the Lha sa Valley landscape.
These can be collated into four principal groups:
• Eleven “auspicious signs” (T1: 10-21; T2: 12-23) that are, in aggregate,
an elaboration of the standard “eight auspicious signs” (bkra shis rtags
brgyad) of Buddhism, such as a mchod rten, a heap of jewels, a parasol,
twin golden fish, a treasure vase, etc., but are collectively represented as
the manifest physical presence of a Buddha-body (see below).
• Four mines (T1: 22-25; T2: 24-27) for iron, copper, silver, and gold in the
four cardinal directions.
• Four auspicious directional protector animals (T1: 25-28; T2: 28-31): a
gray she-tiger in the east, a turquoise dragon in the south, a red bird or
cockerel in the west, and a black turtle in the north.
The first group of eleven are treated by Kong jo in a particular way – as aspects
of the Buddha’s embodied presence within the subdued landscape. Thus, The Clear
Mirror version of her divination has the mountain behind, (Rgyab ri) Mal grong,
as like the fish “representing the Buddha’s eyes” (T1: 15; T2: 17 has Mount Dgos
in the east), whilst the Rock of Dangkhar (mdangs mkhar gyi brag) is shaped like
a lotus “like the tongue of the Buddha” (T1: 16; T2 has Mount Rdzong btsan); the
endless knot on Yug ma ri (T1: 19; T2 has Mount Sgo phu) is seen to represent
the Buddha’s mind; and the victory-banner shaped mountain of ’Phan dkar his
body (T1: 20; T2 has Brang phu), and so forth. While there is clear variation in
the precise sitting of these various signs, the general principle of organization –
based on the interpretive formation of a “body” – remains consistent. This
physiological imagery is important because it creates a “body of the Buddha” that
counterpoises the body of the demoness; nonetheless, both are seen as existing
within the landscape itself. However, the addition of extra symbolic elements –
the lotuses, the mchod rten, the heap of jewels, and the skull-cup (T1: 10-13; T2:
12-25) – suggests an image not simply of a Buddha, but of ritual practice focused
on that Buddha. Indeed, it is – one might suggest – an image of Buddhist
temple-worship.
The second set is equally intriguing. The inclusion of mines here is at odds with
the kind of environmentalist presentation of traditional Tibetan culture which sees
it as fundamentally ecological in the modern sense.46 Most specifically, the presence
46
Toni Huber and Poul Pedersen, “Meteorological Knowledge and Environmental Ideas in Traditional
and Modern Societies: The Case of Tibet,” Journal of the Royal Anthropological Institute, n.s., 3, no.
3 (1997): 577-98.
of mining in pre-1950 Tibet has often been either denied or taken as one of the
major “concessions” that the Buddhist state made to the necessary production of
wealth.47 For Huber in particular, the existence of mining spoke to the “gap between
the ideal and the real – i.e., what was believed and what was actually done”48 in
Tibetan society. However, the existence of mines as part of the emergent auspicious
properties of the Lha sa landscape as divined by Kong jo – which only fully emerge
once the land is properly ritually subjugated – suggests a fundamentally different
dynamic. It suggests in particular the vision of a sacrificial organization of the
state, in which precious metals were used primarily towards the production of
religious and royal artefacts. Mines, in other words, were part of the auspicious
hierarchy of offering.
The final set of benevolent geomantic properties mentioned by Kong jo are four
animals in each of the four cardinal directions. At first these seem an odd addition
– almost an appendage – were it not for their clear correspondence to a similar
symbolic set at the heart of Chinese systems of geomancy, where the white/gray
she-tiger, turquoise dragon, black turtle, and red bird represent the four “protectors”
of a particular site.49 While the other three appear to be mountains or significant
boulders, the turquoise dragon in the south appears to be a reference to the Skyid
chu River. This would mark a distinction from modern Chinese geomancy, where
the green dragon is generally a mountain;50 however, that this is a viable
interpretation in the Tibetan context can be seen from the modern use of this
terminology is Gyatsho’s 1967 manual on monastic ceremony, Gateway to the
Temple:
When establishing (planting) a Central Temple51 and so forth [one should look
for the following]: a tall mountain behind and many hills in front, two rivers
converging in front from the right and left, a central valley of rocks and meadows
resembling heaps of grain, and a lower part which is like two hands crossed at
the wrists. The good characteristics called the four Earth-pillars area wide expanse
in the east, a heap in the south, a rounded bulge in the west, and in the north a
mountain like a draped curtain. The four directional earth protectors are: in the
east a whitish path or rock is the tiger, and in this direction there must be no
ravines cutting across the lower part of the valley. By the river of the southern
direction there must be verdure, which is the turquoise-dragon, and here it is
necessary that the water does not plummet into a cavern. Red earth or rock in
the west is the bird, and here the path must not be fraught with snags or pitfalls.
47
Toni Huber, “Traditional Environmental Protectionism in Tibet Reconsidered,” Tibet Journal 16,
no. 3 (1991): 63-77; Marcy Vigoda, “Religious and Socio-Cultural Restraints on Environmental
Degradation among Tibetan Peoples – Myth or Reality?,” Tibet Journal 14, no. 4 (1989): 27.
48
Huber, “Traditional Environmental Protectionism,” 72.
49
See Feuchtwang, Chinese Geomancy, 201; see also Needham, Science and Civilization, vol. 2,
360.
50
For example, Eva Wong, Feng shui (Boston: Shambhala, 1994), 64-65.
51
Here, Gyatsho has used the term gtsug lag khang as a general term. In the post-dynastic texts, the
term is sometimes used to refer to the Lha sa Jo khang, and sometimes to specify those temples that
were an integral part of the king’s rule.
A bearded rock in the north is the tortoise, and in this direction at the stream’s
source the water must not be obstructed by seething, roiled water. If these fοur
protectors are all present, the land is perfectly endowed.52
The importance of these four cardinal animals is more clearly brought out in
the Bka’ chems ka khol ma texts. Here, they are described as the “four gods” of
the Central Temple of Lhasa (T2: 28-31). Their orientation, moreover, is depicted
as a crucial part of Lha sa’s problematic geomancy, for these four gods are
“afflicted” by Lha sa’s five “terrestrial antagonists,” which rest in the same line
of sight as the four animals (T2: 7-11).
All of this points to a fundamentally subjectivist rendering of geomantic
divination: the landscape tends overwhelmingly to be “read” from a particular
physical viewpoint (from which good or bad signs are seen) rather than from an
abstracted “mapped” perspective (in which numina are geographically located).
This is important because it divides geomantic architecture in two functionally:
just as the Tsher rdzong rtse “faces” the scorpion at Grib from a particular angle
in order to subdue it, so (for example) is the traditional meditation cave of Srong
btsan sgam po at Pha bong kha placed in a precise line of sight to see the auspicious
black turtle as a black turtle, in order to employ its auspicious qualities to best
advantage. Both functions are perspectival, not cosmological.
The Buddhist Project of Geomancy

What, however, might we read as the Buddhist significance of all this? In The
Clear Mirror, as with many of the post-dynastic mythologies of Srong btsan sgam
po, the founding of the temple – with the subduing of the recalcitrant earthly forces
that attends it – marks a turning point in the political dynamics within Lha sa. Prior
to the founding, the ministers and princesses are depicted as consumed with jealousy
over the achievements of others. They bicker and scheme against one another,
constantly obstructing the completion of meritorious Buddhist acts. Subsequent
to the founding, these disagreements are no longer to be found within the story at
all; the ministers and heirs refer to one another instead by familial terms, and
meritorious acts are described in effusive lists. The narrative thus links together
the geomantic subjugation of the landscape with the moral disciplining of the royal
court.
Elsewhere, I have discussed the impact of this principle in Tibetan monastic
life, where geomantic arts were also linked to the strategic amelioration of monastic
discipline.53 Thus, the legend exists that, when the Tibetan prelate and incarnate
bla ma ’Jam dbyangs bzhad pa II (1728-91) was staying in Bla brang bkra shis
’khyil Monastery, he was concerned with the fact that whilst the monks were highly
learned, their moral discipline was lax and many disrobed. Meditating on the matter,
he perceived that the cause was the evil influence of a nearby hill. Ordering the
52
Gyatsho, Gateway To The Temple, 29; translation adjusted from original, my italics.
53
Mills, Identity, Ritual and State, chaps. 10-13.
hill to be levelled, he built on the place a Maitreya statue (called the Gser khang
chen mo). As a result, few if any talented monks arose again in Bla brang, but
those that were there became modest and disciplined in their vows – to the
consternation of locals but the great delight of ’Jam dbyangs bzhad pa.54 In a more
modern context, we might note Elisabeth Stutchbury’s insightful examination of
local discussions concerning the impact of unbalanced geomantic alignment of
Zhabs rjes Monastery in Dkar zha on the spiritual practice and morale of its
occupants, and the ritual and architectural means employed to ameliorate the
situation.55
Photo 5: Le shan Buddha, Sichuan, China.

That such notions might have had Chinese Buddhist parallels can be seen in an
example from around the same period as Srong btsan sgam po: the building of the
Le shan Buddha over the confluence of the Minjiang and Yuexi He rivers in
southern Sichuan during the middle T’ang dynasty. This massive work was
explicitly designed to ameliorate obstructive geomantic forces, and like the protector
image at Brag lha klu sgug in Lha sa, is oriented to face a key Buddhist religious
site (in this case that of the nearby Mount Emei shan). The project was initiated
in 713 CE by the Buddhist monk Haitong during the reign of Emperor Xuanzong
(685-762) and completed ninety years later by Wei Gao, presiding governor of the
region. Wei Gao was one of the most important provincial governors of the period,
famed for both his place within the history of Chinese Buddhism and his defence
of Szechuan against Tibetan incursion.56 His rock-edict – carved to the right side
of the Le shan Buddha’s feet (see Photo 5, left) – describes the construction thus:
Religious functions are great and penetrating and a religion can only be the work
of saints. When one has freed himself from the puzzling world, what he has
understood about Buddhism will get him close to the gods and, following the
doctrine of Buddhism, he may create miracles to save people from disasters. The
carving of the Le shan buddha is a proof of this. The Mingjiang River was said
to have a very torrential section from Le shan eastwards to Jianwei. Waves washed
the cliffs on the banks and ran down the shoals, rumbling like thunders [sic].
54
See Mills, Identity, Ritual and State, 315.
55
Stutchbury, “Perceptions of the Landscape.”
56
Charles Peterson, “Court and Province in Mid- and Late-T’ang,” in The Cambridge History of
China, vol. 3, Sui and T’ang China, 589-906, Part 1 (Cambridge: Cambridge University Press, 1979),
518-19.
Rapids gouged the bank rocks and swallowed boats, causing death to numerous
people. A river like this had long caused trouble to the communication between
the rich western part of Shu and the lower reaches of Two Hu’s [sic], Wan and
Zhe, vast areas of the Wu and Chu. In the first year of Kaiyuan of the Tang
Dynasty, Monk Haitong rose to say “No!” to the monsters of the river, determining
to make it benefit the people. He thought that since the roaring rapids were mostly
below cliffs, why not dig the cliffs and let falling rocks ease the currents. If the
image of Buddha was thus carved Buddhism could be glorified [as well]. Kind
people could be mobilized and manpower gathered to establish a great figure of
Maitreya, which would bless the coming generations and tell them to go on
worshipping Buddha. With confidence, the monk started to plan for the magnificent
project. In his mind, the Buddha statue must be great and delicate, with a head
one hundred chiin in girth, its eyes two zhang in length. The building materials
must be good and time-proof. A great many people must be gathered and money
collected from worshippers. The day came when treasures, technical workers and
every other thing [was] ready. The work site became a battlefield with tens and
thousands of people, numerous hammers over heads, and rocks down into the
river, driving the monsters away from the deep water. The project was [in this
way] pushed on day in and day out. Finally, the magnificent figure of Buddha
was shaped. The sky became clear and bright as if the Buddha was giving his
light to the world. When [the] waves were calmed down and dangers eliminated,
the world became tranquil.57
Wei Gao’s description of the events surrounding the construction, and their
effects on the river, is at first sight somewhat perplexing. In particular, it is far
from clear how he understands the Maitreya statue’s effect on the river. At several
points, indeed, his description is almost wholly prosaic: it was the rapids that were
causing people to drown, and it was the accumulation of rocks from the Buddha’s
excavation in the river at the base of the cliffs that would calm those rapids. At the
same time, the rapids are caused by “monsters” swallowing the boats, denizens of
the river that are “chased away” by the falling rocks. The forces at work seem both
geological and numinal: the turbulence of the river that brought such calamity to
local trade was both fluid turbulence and “monsters,” and one cannot easily be
reduced down to another. Wei Gao does not, for example, appear to “believe” that
the cause of the turbulence was river-monsters; rather, they are that turbulence,
and are monstrous from the perspective of human suffering.
This is a similar ontology to Grib’s scorpion on the outskirts of Lha sa: Princess
Kong jo is not presented as believing that there really is a scorpion in the mountain;
it’s only deemed to look that way from a particular direction. Nonetheless, from
that direction, the impact of its misfortune is real, and thus a temple is built to
“suppress” its influence. In this sense, the malevolent nature of these geomantic
influences is presented as a “dependent origination,” in other words as dependent
57
Le shan state memorial park, Szechuan, 2004. This translation has been provided by the PRC state
authorities at Le shan for tourist consumption. In this respect it is naturally suspect and clearly incorrect
in some regards. The author’s endeavors to unearth a scholarly treatment of this edict and its context
have proven as fruitless as they were frustrating, but my thanks go to Prof. Stephan Feuchtwang and
Prof. Tim Barrett for their critical comments on the Chinese geomantic context of this site.
on the relationship of the object to the perceiving mind, rather than being inherent
within the object itself.
This dependence on perception is highlighted by Wei Gao when discussing the
manner in which the Le shan Buddha suppressed the deadly rapids:
Then, why can the power of Buddha get rid of danger and calm waves down? We
know that bad fortune is always caused by ill-will. If we can understand that we
come from quietness, we would care nothing about the ups and down in our life.
When having and not having do not weigh [upon us], whoever would take trouble
to think too much about safety and danger? In a calm observance of the world,
Buddha denies whatever [is] unreal and vulgar, [re]wards the kind and punish[es]
the evil, teaches people each according to his intelligence, and helps those who
have momentarily lost their way. If not done by such a saint as the Buddha, who
else on earth can have these turbulent waves calmed down?58
Here then, a seeming third level of interpretation is added to Wei Gao’s

description of the construction project: the waters are subdued by the Buddha statue
because they are a form of misfortune, because misfortune arises out of ill-will
and other such emotional afflictions, and because a Buddha transcends such
afflictions.
This is one amongst three different levels by which the author describes the
causality of this event, just as key events in the life of Srong btsan sgam po are
presented by Bsod nams rgyal mtshan and other writers as being perceived in three
different ways: from the perspective of the Buddhas, from the perspective of the
wise and visionary, and from the perspective of ordinary people.59 Unlike Bsod
nams rgyal mtshan, who tends to privilege one interpretation over others, Wei Gao
tends not to imply that one explanation is “true” whilst the others are false. They
are each “dependent perceptions.” Nonetheless, he clearly expends most ink on
this final discussion of the Buddha’s power to subdue the afflictive illusions of
this world, and it is here that the tract takes an explicitly didactic tone, as a way of
seeing the Buddha’s place in the world of men.
Wei Gao’s rendering of the Buddhist logic that attended the Le shan construction
makes explicit that which Tibetan literary sources tend to point towards but rarely
render explicit: a Buddhist linking between the subjugation of “afflictive emotions”
on the one hand and the subjugation of afflictive geomantic forces on the other.
For Wei Gao, the process of subjugating malevolent geomantic forces is entirely
akin to the practice of Buddhist moral discipline. Here, the “inner” mental
disciplining of the mind and the “outer” geomantic subjugation of the landscape
are equated within the Buddhist path: both are “obstructions” to spiritual awakening.
Thus, temple-founding – whether of the Central Temple of Lhasa, of Le shan
or of the Gser khang chen mo – all speak to an acknowledged principle of Buddhist
religious governance that embeds the familiar Buddhist project of the disciplining
58
See note 57.
59
For example, see Sørensen, Tibetan Buddhist Historiography, 161-62.
of afflictive emotions within the wider tableau of the realm of governing the land
itself. As can be seen from the Central Temple legend, such a principle is presented
as an integral part of kingly sovereignty. Like the monk Haitong’s vision of Le
shan and ’Jam dbyangs bzhad pa’s vision of Bla brang Monastery, the demoness
divination is part of a wider Buddhist understanding of the relationship between
moral thought and landscape: a lengthy if indirect discourse on the nature of
afflictive emotions and their subjugation, but one told at a state level.
The governmental flavor of the supine demoness story is clear, most obviously
in the sense that the temple at its heart would eventually be the emperor’s tutelary
temple, the basis of his law; but also in the manner in which Kong jo’s divinatory
skills are alluded to in texts such as the tenth-century Pillar Testament as a
“calculation of governance” (gtsug lag gi rtsis).60 Here, the term gtsug lag (literally,
“head and limbs”) – the basis of the term gtsug lag khang – was ironically seen
by early Tibetologists such as Ariane Macdonald as denoting the indigenous
pre-Buddhist religion of Tibet, but which R. A. Stein more convincingly argues
was actually denoting the “good governance of the state.”61
In this sense, the landscape within the medieval histories of dynastic Tibet acts
as a complex metaphor for the mind itself, which has the possibility to be both
driven with afflictive emotions, or to be the emergent basis for enlightenment. This
ambiguity is reflected most clearly in the image of the Lake of O Plain, at the heart
of the Srong btsan sgam po narrative. In the legend, Kong jo’s vision of the lake
as an obstruction (bar chad) to Buddhism is balanced by Srong btsan sgam po’s
vision of the lake as a nine-terraced mchod rten composed of rays of light,62 which
eventually becomes the supernatural mainstay of the Central Temple.
Whilst it is certainly possible to see this ambiguity in the presentation of the
lake as a result of the compilation of several different traditions, a more integrated
interpretation suggests itself, one consistent with other elements of the story. Here,
the “lake” – as it is within much Tibetan metaphorical literature – is a metaphor
for the mind in its various spiritual states. Before the arrival of the Buddhist dharma,
it is variously characterized as an obstructive palace of Nāgas or, even earlier, as
the lowest level of hell.63 Nonetheless, the nine-layered mchod rten – representing
the mind of the Buddha, as mchod rten classically do within the Mahāyāna – is
inherent but unmanifest, within it. This ambiguity is resolved when we understand
that, like the mind, it is only when the lake is ritually subdued – that is, brought
under the discipline of the dharma – that its positive qualities emerge.
This dynamic polarity is one that appears repeatedly in the representation of
Lha sa and Tibet as mythic landscapes. In the Pillar Testament – a “hidden treasure”
text found in the Central Temple of Lhasa in 1048 and promulgated by Atiśa
60
See Sørensen, Tibetan Buddhist Historiography, 553.
61
See R. A. Stein, “On the Word gCug-lag and the Indigenous Religion,” in The History of Tibet,
ed. Alex McKay, vol. 1 (London: RoutledgeCurzon, 2003), 530-83.
62
Sørensen, Tibetan Buddhist Historiography, 264; also Taylor and Yuthok, The Clear Mirror, 167.
63
Sørensen, Tibetan Buddhist Historiography, 113; also Taylor and Yuthok, The Clear Mirror, 79.
(982-1054) – Kong jo’s divination of the Lha sa Plain contains similar ambiguities
when she discusses the place that the Dmar po ri and Lcags po ri Hills in the center
of the Lha sa Plain have in her vision of Tibet as like a prostrate demoness:
These two, Dmar po ri and Lcags kha ri, resemble the tails of a lion and a tiger
tied together, and should be understood as the vicious disposition of the demoness.
These two hillocks are, moreover, the bones of the heart of the demoness, which
devour the life-force (srog) of sentient beings. These two mountains are in reality
the body of Srong btsan sgam po, Lord of Compassion, and in reality the body
of Nepalese princess Khri btsun, white Tārā, Bhṛkuṭi. The king must recognize
that the palaces of these two, which resemble [’dra ba] Langkapuri, city of the
rākṣasa demons, must be ritually suppressed.64
This emphasis on the importance of the process of ritual subjugation – in

bringing out the divine reality of a landscape which initially manifests as a complex
of inauspicious features – suggests that the demoness motif within the early dynastic
stories needs to be understood as part of a wider narrative in which this demonic
appearance is replaced by the emergent reality of Buddhahood. As the various
elements of the demoness are suppressed, they are replaced by the various
components of the body of the Buddha: the “fish” on the mountainside behind Mal
gro representing the Buddha’s eyes; the “lotus” at the Rock of Dangkhar
representing the Buddha’s tongue; the “protective umbrella” at ’Phan dkar gyi ri
representing his body; the “wheel of dharma” at the marsh of Stod lung representing
his hands and feet. The land of Tibet as female demonic is suppressed, revealing
instead Tibet the male figure of enlightenment.
Conclusion: Variations on a Geomantic Theme

I have devoted this entire article to the interpretation of a particular theme within
Tibetan royal myth: the geomantic examination of a future temple site. The
hagiography of Srong btsan sgam po is far from being alone in this regard: a wide
variety of Tibetan historical figures are represented as having faced similar chthonic
challenges from the forces of landscape. Probably the most famous of these is the
later Yar lung emperor Khri srong lde’u btsan who, just like Srong btsan sgam po,
finds that his attempts to build a tutelary temple to Avalokiteśvara65 – in this case
at the monastery at Bsam yas to the south of Lha sa – are hampered by the local
deities, who destroy in the night everything that was built in the day.
The narrative trope of the supine demoness has received little or no sustained
attention from a strictly Buddhalogical viewpoint. Indeed, it is seen as having little
64
From Sørensen, Tibetan Buddhist Historiography, 553; italics mine.
65
The central deity of Bsam yas is of course Vairocana. However, The Clear Mirror is specific that
the first chapel of the Central Temple to be planned – and the one whose construction was the most
directly impeded – was the Avalokiteśvara chapel. Moreover, the king’s first propitiation in this temple
led the attendant protector statue of Hayagrīva to let out a neigh heard throughout two-thirds of
Jambudvīpa, an extent which the presiding teacher Padmasambhava declared would be the extent of
the king’s sovereign rule (Sørensen, Tibetan Buddhist Historiography, 375-76).
to contribute to our understanding of the “great tradition” of Buddhism. Rather, it

has been interpreted either as a “local” or “folklore-ish” narrative, as the misogynist
prejudice of Buddhist monks, or (as we saw above) as referring to pre-Buddhist
or non-Buddhist traditions. While such interpretations have their merits, they often
fail to take seriously the simple truth universally asserted by Tibetan historians:
that such divinations were part of an imported Chinese tradition of geomancy.66
The Chinese geomantic traditions alluded to in works such as The Clear Mirror,
moreover, contain a strong Buddhist framework, one which – I have argued –
depicts the process of state formation as akin to the Buddhist project of taming the
mind. Here then, as I have argued elsewhere for modern Tibetan monasticism, the
Buddhist project of the ethical disciplining of minds is rendered inseparable from
the geomantic disciplining of places.67
66
As discussed above, however, it would be a profound error to assume – simply because there is a
clear indigenous and historical consensus on the Chinese origins of Tibetan geomancy – that Tibetan
geomancy is simply a cipher of its Chinese cousin. Tibetan geomantic systems – especially since the
medieval period – show several marked differences, especially from Chinese imperial feng shui. Here,
it is worth noting the changes that have occurred during the post-dynastic period (eleventh-fourteenth
centuries), and those characteristic of the medieval and modern period. Amongst the textual renditions
of the supine demoness story in the post-dynastic period, several important variations are striking. This
is actually a vast topic, previously covered in great detail by Per Sørensen in Tibetan Buddhist
Historiography; here, then, I will restrict my comments to distinctions between The Clear Mirror and
the Pillar Testament. Firstly, while both texts agree on the existence of the four key geomantic animals
(tiger, dragon, tortoise, and bird) to be found in the immediate vicinity of the Lha sa site (a fairly
standard Chinese usage), the later The Clear Mirror replicates these at the distant level of the intermediate
border-taming temples (T1: 35-38). Since, however, the organization of the concentric levels of these
temples in the received versions of this latter text are clearly confused (see Sørensen, Tibetan Buddhist
Historiography, 261 n. 770), it would be difficult to assert that this difference constituted a different
geomantic tradition as above bad redaction. More distinctive is The Clear Mirror’s inclusion of extra
temples to suppress astrological and elemental influences: tthe sun, moon, planets, and lunar
constellations (T1: 43); and the fire, water, and wind elements (T1: 44, 45, 46). These are also present
in Mkhas pa lde’u’s Rgya bod kyi chos ’byung rgyas pa (penned in 1261 or after, according to Dan
Martin, Tibetan Histories: A Bibliography of Tibetan-Language Historical Works [London: Serindia
Publications, 1997], 43), but not in the Pillar Testament (Sørensen, Tibetan Buddhist Historiography,
567, 637). Much more dramatic at this level are the distinctions between The Clear Mirror and the
Pillar Testament in terms of their descriptions of the ritual nature of the subduing temples. Within the
Pillar Testament, the subduing temples are designed to be built around named maṇḍalas (T2: 37-49).
This is an element which is wholly absent from The Clear Mirror. These observations are of course
merely preliminary, and would require further elaboration elsewhere, but arguably point to the emergence
of distinct lineages of geomantic interpretation during the post-dynastic period, distinctions which
probably have more to do with internal Tibetan dynamics than any cultural relationship with China.
67
See Mills, Identity, Ritual, and State, chap.10.
Tables
Table 1: Kong jo’s Divination in The Clear Mirror68

Signs of the Lha sa Heartlands (Inauspicious)
No. Geomantic Interpretation Place Name (direction) Suppressing Architecture
1. Palace of the Nāga kings and Lake of O Plain (’o thang Ra sa gtsug lag khang.
entrance to the lower realms. gi mtsho).
2. Cave below is resting place of Ra mo che Plain. Ra mo che Temple and

black Nāga demons (klu Lord (jo bo) statue.
bdud).
3. Meeting place of The’u rang At the root of the poison Dispersed by the chos
spirits and ghosts (mi ma yin). tree at the Garden of the skyong Hayagrīva and
Moon (zla ba tshal). Amṛtakuṇḍalī after the
creation of the
eleven-headed
Avalokiteśvara statue.
4. Habitual route of Btsan spirits. Between upper valley of A great mchod rten at Bar
Grib (south) to upper valley chu kha.
of Nyang bran (north).
5. A crocodile (chu srin) Ba lam grum pa ri (east). Right-whorled conch shell

pursuing a De’u stone/water (dung g.yas su ’khyil) facing
spirit. it.
6. An ogress thrusting forward Byang stod seng phug A phallic image (gsang ba’i
her genitals (’doms bzed pa). (southeast). rten) of Maheśvara facing
it.
7. Malign influence resembling Shar ri (the eastern Statue of Ke ru garuḍa

a black scorpion pouncing on mountain) of Yug ma ri facing it.
[its pray] (sdig pa nag po gzan (southwest).
la rub pa).
8. A black Bdud demon keeping Rock of Shün (shun gyi Stone mchod rten facing it.
watch. brag) (west).
9. Elephant engaged in battle. Dbus ri, central mountain A stone lion facing it.
of Nyang bran and Dogs te
(north).
Signs of the Lha sa Heartlands (Auspicious)
No. Geomantic Interpretation Place Name (direction) Suppressing Architecture
10. Mountain resembling a Ban khos bang ba ri After suppression of items

mchod rten. (east). 1-9.
11. Mountain resembling a heap Mountain behind Grib After suppression of items
of jewels. (south). 1-9.
68
See Sørensen, Tibetan Buddhist Historiography 253-63, 275-80.
12. Mountain resembling a Brang phu in Stod lung After suppression of items
skull-cup on a tripod. (west). 1-9.
13. Mountain resembling lotus in Lha’i phu’i ri of Dogs After suppression of items
bloom. te/Dge te (west). 1-9.
14. Mountain resembling an Mount ’Phan dkar ri in After suppression of items

umbrella for the head of the Nyang bran. 1-9.
Buddha.
15. Fish representing the Mountain behind Mal After suppression of items
Buddha’s eyes. grong. 1-9.
16. Lotus representing the Rock of Dangkhar After suppression of items

Buddha’s tongue. (mdangs mkhar gyi brag). 1-9.
17. Conch shell representing the Glacier at Grib. After suppression of items
Buddha’s speech. 1-9.
18. Vase representing the On Mountain of Rdzong After suppression of items

Buddha’s neck. btsan. 1-9.
19. Endless knot representing the On Yug ma ri. After suppression of items
Buddha’s mind. 1-9.
20. Victory Banner representing On Mount ’Phan dkar. After suppression of items
the Buddha’s body. 1-9.
21. Wheel representing the Marsh of Brang phu at After suppression of items
Buddha’s hands and feet. Stod lung. 1-9.
22. Iron mine. Ridge at Gar pa’i jo mo After suppression of items

ze ze. 1-9.
23. Copper mine. Ra ga brag. After suppression of items

1-9.
24. Silver mine. Rock of La dong. After suppression of items

1-9.
25. Gold mine. Rock of Lcags kha ri. After suppression of items
1-9.
25. Gray She-Tiger (stag skya Dga’ bo gdong ’og ma After suppression of items
bo). (east). 1-9.
26. Blue Turquoise Dragon (g.yu South. After suppression of items

’brug sngon mo). 1-9.
27. Red Bird (bya dmar po). Rock of Shün (shun gyi After suppression of items
brag) (west). 1-9.
28. Black Turtle (rus sbal nag Pha bong kha in Nyang After suppression of items
po). bran (north). 1-9.
Binding the Supine Demoness (1): The Lha sa Heartlands
No. Geomantic Interpretation Place Name (direction) Suppressing

Architecture
29. Heart blood of the Plain of Milk (’o ma’i thang). Ra sa gtsug lag khang
demoness. (’phrul snang).
30. Bones of the demoness’ Three mountains of Lha sa

heart (snying gi rus pa). (Dmar po ri, Lcags kha ri, and
Bha ma ri).
Binding the Supine Demoness (2): The District-Controlling Temples (ru gnon gyi lha khang)
No. Geomantic Place Name Suppressing Architecture

Interpretation (direction)
31. Left Shoulder. G.yo ru. Khra ’brug bkra shis byams snyoms; Tsan
thang temple as subsidiary; Phug po che
meditation cave as subsidiary.
32. Right Shoulder. Dbu ru, cave of Rtse Ka rtsal shrine; Rtse no gdong shrine as
no gdong. subsidiary; Sher shang gi rti meditation
cave as subsidiary.
33. Left Hip. Cave of Grom pa Grub pa rgyal dri ma rnam dag shrine; ’Bre
rgyang. shrine as subsidiary; meditation cave of
Rgyang.
34. Right Hip. G.yas ru. Btsang ’brang byang chub dge gnas; Dge
dung shrine as subsidiary; rock cave of
Rtse as meditation cave.
Binding the Supine Demoness (3): The Border-Taming Temples (mtha’ ’dul gyi lha khang)

35. Left Elbow. Lho brag. Mkhon mthing gser gyi lha khang upon
the crest of the Southern Dragon.
36. Right Elbow. Bu chu. Gong po’i bu chu shrine upon the head
of the Eastern Tiger.
37. Left Knee. Spra dun rtse, on the head of the

Northern Tortoise.
38. Right Knee. Byams chen dge rgyas, on the back of

the Red Bird of the west.
Binding the Supine Demoness (4): Further-Taming Temples (yang ’dul gyi lha khang)

39. Left Hand. Bum thang, in the Bum thang skyer chu shrine, under the
south. supervision of craftsmen of To dkar.
40. Right Hand. Glang thang sgron Shrine, under the supervision of Mi
ma (Khams). nyag craftsmen.
41. Left Foot. To the north. Tshangs pa rlung gnon, under the
supervision of Sba dpal dbyang of Hor.
42. Right Foot. In Kashmir (kha Mtshal rig shes rab sgron ma, under
che). supervision of the Nepalese.
43. Sun, moon, planets, To the east. Ka chu, Kang chu, Gling chu shrines.
and lunar
constellations.
44. Suppress the fire To the south. Snyal snang gro and Gling thang
element. shrines to Me lha (Agni).
45. Suppress the water To the west. Gu lang (Paśupatināth outside

element. Kathmandu) and Shing kun
(Svayambhūnāth), in order to secure
the border between Nepal and Tibet.
46. Suppress the wind To the north. Dge ri and Dpal ri, to bind to oath the
element. deities, Nāgas, and ogres.
Table 2: Kong jo’s Divination in the Pillar Testament69

Signs of the Lha sa Heartlands (Inauspicious)

Architecture
1. Palace of the Nāga kings. Lake of O Plain (’o thang gi Ra sa gtsug lag khang.
mtsho).
2. Like Langkapuri, city of Palaces of Srong btsan sgam

Rākṣasas. po and Princess Khri btsun.
3. Palace of the Nāgas. Cave below Ra mo che Plain. Lord (jo bo) statue in
Ra mo che Temple.
4. Sleeping place of the Ma mo Brag lha, north-west side of Image of Able Rock
spirits. Lcags kha ri. Protector (brag lha
mgon po).
5. Habitual route of Klu btsan Between upper valley of Grib A white mchod rten.
spirits. phu (south) to upper valley
of Nyang bran (north).
6. Meeting place of The’u rang Dkar chung gla ba tshal. Destroy (shig).
spirits and ’Dre/Dungeon of the
’Dre.
7. Terrestrial Antagonist: a East: Mount Ngan lam gron A conch shell (dung
water-demon (chu srin) standing pa. kha) facing (ston) it.
up.
8. Terrestrial Antagonist: an ogress East: Bye ma lung stong A phallus (liṅga) of

thrusting forward her genitals (stod). Maheśvara facing it.
(’doms bzed pa).
9. Terrestrial Antagonist: black South: Dkar chung yug ma A garuḍa facing it.
scorpion pouncing on [its pray] ri.
(sdig pa nag po gzan la rub pa).
10. Terrestrial Antagonist: a black West: summit of Rock of A red mchod rten
Bdud demon keeping watch. Shün (shun gyi brag te’u). facing it.
69
See Sørensen, Tibetan Buddhist Historiography, 553-60.
11. Terrestrial Antagonist: elephant North: central mountain of A lion facing it.
engaged in battle. Nyang bran and Dor te’u.
Signs of the Lha sa Heartlands (Auspicious)

Architecture
12. Mountain resembling heap of lotuses. East, behind mountains of 3. Combined,

these “bring
13. Mountain resembling heap of jewels. South, behind mountains of 3. together and
14. Mountain resembling erected mchod West, behind mountains of 3. assemble
rten. bodhisattvas
and
15. Mountain resembling skull-bowl on North, behind mountains of 3. magnanimous
tripod. peoples, along
with the
wealth of the
four directions
accumulating
here.”
16. Mountain resembling an umbrella Mount ’Phan dkar ri in Nyang

for the head of the Buddha. bran.
17. Fish representing the Buddha’s eyes. Mount Dgos in the east.
18. Conch shell representing the Mount Ldong brtsan in the

Buddha’s speech. south.
19. Lotus representing the Buddha’s Mount Rdzong btsan.

tongue. Combined,
these are the
20. Vase representing the Buddha’s Rock of Grib rdzong. Buddha’s
neck. body (sku).
21. Endless knot representing the Mount Sgo phu.
Buddha’s mind.
22. Victory Banner representing the Brang phu in the west.

Buddha’s body.
23. Wheel representing the Buddha’s Stod lung mda’.

hands and feet.
24. Copper mine. Rock of Raga (ra ga brag) of

Dog sde.
25. Iron mine. Sgo phu of Dog sde.
26. Silver mine. Rock of La dong.
27. Gold mine. Rock of Lcags kha ri.

28. Gray She-Tiger (stag skya bo) East of O Plain (’o thang). These animals
are the “four
29. Blue Turquoise Dragon (g.yu ’brug South of O Plain (’o thang). gods”
sngon mo). (lha bzhi) of
30. Red Bird (bya dmar po). West of O Plain (’o thang). the Ra sa
gtsug lag
31. Black Turtle (rus sbal nag po). North of O Plain (’o thang). khang. They
are depicted
as being
initially
“obstructed”
by the
“terrestrial
antagonists”
(sa dgra, see
items 8-12),
which stand in
the same line
of sight
(thad ka na).
Binding the Supine Demoness (1): The Lha sa Heartlands

Architecture
32. Heart blood of the Plain of Milk (’o ma’i thang). Central Temple of
demoness. Lhasa (lha sa gtsug
lag khang).
33. Nipples of the Three mountains of Lha sa (Dmar Central Temple of

demoness/vein of her life po ri, Lcags kha ri, and Bha ma Lhasa (lha sa gtsug
force. ri). lag khang).
34. The retinue (’khor) of the Mountains in four directions with

demoness. four summits pointing at Lha sa.
35. Mouth of the demoness. Four tortoise-like mountains

around Lha sa.
36. Heart-Bones/vicious Dmar po ri and Lcags kha ri, like

disposition of demoness. the tail of a lion and tiger tied
together.
Binding the Supine Demoness (2): District-Controlling Temples (ru gnon gyi lha khang)

37. Left Shoulder. G.yo ru. Khra ’brug Temple: maṇḍala of the
Eight Great Planets.
38. Right Shoulder. Dbu ru. Ka rtsal shrine: maṇḍala of the

twenty-one Lay Devotees.
39. Left Hip. G.yon ru Grum pa rgyal Temple: maṇḍala of

Hevajra.
40. Right Hip. G.yas ru. Btsang ’brang Temple: maṇḍala of the
Four Directional Guardian Kings.
Binding the Supine Demoness (3): The Border-Taming Temples (mtha’ ’dul gyi lha khang)

41. Left Elbow. Lho brag (southwest). Mkhon mthil Temple: maṇḍala of the
Five Buddha families.
42. Right Elbow. Sgon bu (southeast). Bur chud Temple: maṇḍala of Black
Mahākāla.
43. Left Knee. Bum thang. Bum thang skyer chu Temple:
maṇḍala of Lotus Buddha-family
(padma dbang).
44. Right Knee. Unspecified. Byang tshal phyi dbang chen Temple:
maṇḍala of the Jewel Buddha-family
(rin po che).
Binding the Supine Demoness (4): Further-Taming Temples (yang ’dul gyi lha khang)

45. Left Hand. Bal chad. Ka brag Temple: maṇḍala of

Vaiśravana.
46. Right Hand. East: Mdo khams. Klong thang sgrol ma Temple:
maṇḍala of Bdud ’dul.
47. Left Foot. Southwest. Byams pa sprin Temple: maṇḍala of

the Earth Goddess.
48. Right Foot. Northwest. Spran dun rtse Temple: maṇḍala of the
nine astrological signs (gtsug lag).
49. Inner cavity (sbugs). Tshangs pa rlung gnon Temple: no

maṇḍala specified.
Glossary
Ka
ka chu Kachu Place
ka brag Kadrak Building
ka rtsal Katsel Building
kang chu Kangchu Place
dkar chung gla ba Karchung Lawa Tsel Place
tshal
ke ru keru garuḍa-bird Term
kong jo Kongjo Person
klu lu water spirit San. nāga Term
klu khang Lukhang nāga-shrine Building
klu bdud ludü black nāga demon Term
klong thang sgrol ma Longtang Drölma Longtang Drölma Building
lha khang Lhakhang Temple
dkar chung yug ma ri Karchung Yukma Ri Place
dkar zha Karzha Place
bka’ chems ka khol Kachem Kakhölma Pillar Testament Text
ma
bkra shis rtags trashi tak auspicious signs Term
bkra shis rtags brgyad trashi takgyé eight auspicious Term
signs
rkong po Kongpo Place
sku ku Term
skyid chu Kyichu River
skye dman kyemen low-born Term
Kha
kha che Khaché Kashmir Place
khag gnon khaknön bind down Term
khams Kham Place
khra ’brug Trandruk Building
khra ’brug bkra shis Trandruk Trashi Place
byams snyoms Jamnyom
khri btsun Tritsün Person
khri srong lde’u btsan Trisong Deutsen Person
mkhas pa lde’u Khepa Deu Person
mkhon mthing gser Khönting Sergyi Building
gyi lha khang Lhakhang
mkhon mthil Khöntil Building

’khor khor retinue Term
Ga
gar pa’i jo mo ze ze Garpé Jomo Zezé Mountain
gu ru rin po che Guru Rinpoché San. Person
Padmasambhava
gu lang Gulang San. Paśupatināth Place
gong po’i bu chu Gongpö Buchu Place
gyab ri mal grong Gyapri Meldrong Mountain
grib Drip Place
grib phu Drippu Place
grib rdzong Dripdzong Mountain
grub thob dngos grub Druptop Ngödrup c. 1100s? Person
grub pa rgyal dri ma Druppa Gyel Drima Place
rnam dag Namdak
grum pa rgyal Drumpa Gyel Building
grom pa rgyang Drompa Gyang Place
gla ba tshal Lawa Tsel Place
glang thang sgron ma Langtang Drönma Place
glang dar ma Langdarma Person
gling chu Lingchu Place
gling thang Lingtang Place
dga’ ldan pho brang Ganden Potrang Organization
dga’ bo gdong ’og ma gawo dong okma Term
dge te Geté Place
dge dung Gedung Place
dge ri Geri Place
dgos Gö Mountain
rgya bod kyi chos Gyabökyi Chöjung Text
’byung rgyas pa Gyepa
rgyang Gyang Place
rgyal pa gdan la gyelpa denla zhukpa king poised on a Term
bzhugs pa seat
rgyal rabs gsal ba’i Gyelrap Selwé The Clear Mirror of Text
me long Melong Royal Genealogy
sgo phu Gopu Mountain
sgon bu Gönbu Place
Nga
ngan lam gron pa Ngenlam Drönpa Place
mnga’ ris Ngari Place
Ca
lcags kha ri Chakkhari Mountain
lcags po ri Chakpori Mountain
lcags phur sa la btab chakpur sala tappa iron nail stuck into Term
pa the ground
Cha
chu srin chusin crocodile; Term
water-demon
chos skyong chökyong dharma protector Term
mchod rten chöten San. stūpa Term
Ja
jo khang Jokhang House of the Lord Building
jo bo jowo Lord Term
’jam dbyangs bzhad Jamyang Zhepa 1728-1791 Person
pa
Nya
nyang bran Nyangdren Valley
nyang ral nyi ma ’od Nyangrel Nyima Özer 1124-1192? Person
zer
snyal snang gro Nyelnangdro Place
snying gi rus pa nyinggi rüpa bones of the Term
demoness’ heart
Ta
to dkar Tokar Clan
gter ma terma hidden treasure Term
texts
rten ’brel tendrel auspicious sign Term
stag skya bo tak kyawo Term
stod lung Tölung Valley
stod lung mda’ Tölung Da Place
ston tön facing Term
bstan pa’i rtsa ba Tenpé Tsawa Chögor Text
chos sgor zhugs Zhuktang dang, Tappé
stangs dang / bstab Tiné Tsuklakkhang
pa’i bsti gnas gtsug Zhektap
lag khang bzhegs
thabs
Tha
thad ka na tekana in the same line Term
thub bstan rgya mtsho Tupten Gyatso Person
thub pa brag lha Tuppa Draklha Gönpo Able Rock Name

mgon po Protector
the’u rang Teurang spirit Term
mtha’ ’dul gyi gtsug Tandülgyi Building
lag khang Tsuklakkhang
mtha’ ’dul gyi lha tandülgyi lhakhang border-taming Term
khang temple
Da
dar dkar gyi yol ba darkargyi yölwa curtain [made] of Term
white silk drawn
dung g.yas su ’khyil dung yesu khyil right whorled conch Term
shell
de’u deu Term
dog sde Dokdé Place
dogs te Dokté Place
dor te’u Dorteu Place
dung kha dungkha conch shell Term
dwags po Dakpo Place
dre’u chu ’thung pa dreu chutungpa mule drinking water Term
bdud dü demon Term
bdud ’dul Dündül Deity
mdangs mkhar Dangkhar Place
mdo khams Dokham Place
’dam Dam Place
’doms bzed pa dom zepa thrusting forward Term
her genitals
’dra ba drawa resemble Term
ldong brtsan Dongtsen Mountain
sdig pa nag po gzan dikpa nakpo zenla black scorpion Term
la rub pa ruppa pouncing on [its
pray]
Na
gnam ri slon mtshan Namri Löntsen r. c. Person
600-618
Pa
padma dbang pema wang Lotus Term
Buddha-family
po ta la Potala Building
dpal gyi rdo rje Pelgyi Dorjé Person
dpal ri Pelri Place
spra dun rtse Tradüntsé Place
spran dun rtse Drenduntsé Building
Pha
pha bong kha Pabongkha Monastery
phag mo gru Pakmodru Organization
phan ju ri Penjuri Mountain
phug po che Pukpoché Place
’phan dkar Penkar Mountain
’phan dkar gyi ri Penkargyi Ri Mountain
’phan dkar ri Penkarri Mountain
’phrul snang Trülnang Term
Ba
ba lam grum pa ri Balam Drumpa Ri Mountain
ban khos bang ba ri Benkhö Bangwa Ri Mountain
bar chad barché obstacle; Term
obstruction
bar chu kha Barchukha Place
bal chad Belché Place
bu chu Buchu Place
bu ston Butön 1290-1364 Person
bum thang Bumtang Place
bum thang skyer chu Bumtang Kyerchu Place
bur chud Burchü Building
bai ḍūrya dkar po Baidurya Karpo The White Beryl Text
bong po ri Bongpori Mountain
bon Bön Organization
bya skya dkar po ri Jakya Karpo Ri Mountain
bya mas bu la sgab pa jamé bula gappa brood-hen covering Term
its children
bya dmar po ja marpo red bird Term
byang stod seng phug jangtö sengpuk Term
byang tshal phyi Jangtsel Chi Building
dbang chen Wangchen
byams chen dge rgyas Jamchen Gegyé Place
byams pa sprin Jampa Trin Building
byung rtsis jungtsi elemental calculus Term
bye ma lung stong Jema Lungtong Place
(lung stod) (Lungtö)
brag lha klu sgug Draklha Luguk Building
brag lha mgon po Draklha Gönpo Able Rock Deity
Protector
brang phu drangpu Term
bla la life-force Term
bla gnas lané vitality place Term
bla brang Labrang Monastery
bla brang bkra shis Labrang Trashikyil Monastery

’khyil
bla ma lama Term
bha ma ri Bhamari Mountain
dbu ru Uru Place
dbus ri Üri Mountain
’bras spungs Drepung Monastery
’bre Dré Place
sba dpal dbyang Ba Pelyang Clan
sbugs buk inner cavity Term
Ma
ma ni bka’ ’bum Mani Kabum Text
ma mo mamo Term
mal gro Meldro Place
mal gro gzi can Meldro Zichen Deity
mal grong Meldrong Mountain
mi nyag Minyak Place
mi ma yin mi mayin ghost Term
me yar Meyar Place
me lha Melha San. Agni Deity
myang Nyang Place
dmar po ri Marpori Mountain
sman ri Menri Mountain
Tsa
tsa ri Tsari Mountain
tsang po Tsangpo Valley
tsan thang Tsentang Building
gtsug lag tsuklak head and limbs Term
gtsug lag khang Tsuklakkhang Central Temple Building
gtsug lag gi rtsis tsuklakgi tsi calculation of Term
governance
btsang ’brang Tsangdrang Place
btsang ’brang byang Tsangdrang Jangchup Place
chub dge gnas Gené
btsan tsen Term
btsun ma dar dkar gyi tsünma darkargyi queen dressed in Term
na bza’ gsol ba napza sölwa white silk
rtse Tsé Place
rtse no gdong Tsenodong Place
Tsha
tshangs pa rlung gnon Tsangpa Lungnön Building
tsher rdzong rtse Tserdzongtsé Deity

mtshal rig shes rab Tselrik Sherap Place
sgron ma Drönma
Dza
rdzong btsan Dzongtsen Mountain
Wa
weng chen kong jo Wengchen Kongjo Person
Zha
zhang zhung Zhangzhung Place
zhabs rjes Zhapjé Monastery
gzhong zhong golden tub Term
Za
zla ba tshal Dawa Tsel The Garden of the Place
Moon
’A
’o thang Otang O Plain Place
’o thang gi mtsho Otanggi Tso Lake of O Plain Place
’o ma’i thang Omé Tang Plain of Milk Place
Ya
yang ’dul gyi lha yangdülgyi lhakhang further-taming Term
khang temple
yug ma ri Yukmari Mountain
yum bu lag khang Yumbu Lakkhang Place
g.yas ru Yeru Place
g.yu ’brug sngon mo yudruk ngönmo blue turquoise Term
dragon
g.yo ru Yoru Place
g.yon ru Yönru Place
Ra
ra ga Raga Place
ra ga brag Ragadrak Place
ra mo che Ramoché Place;
Building
ra sa gtsug lag khang Rasa Tsuklakkhang Building
ri chung Richung Mountain
ri chen spungs pa rinchen pungpa mound of jewels Term
ri nag Rinak Mountain
rin po che rinpoché Term

ru gnon gyi lha khang runöngyi lhakhang district-controlling Term
temple
rus sbal nag po Rübel Nakpo Term
La
la dong Ladong Place
Sha
shākya ’od Shakya Ö Person
shar ri Sharri Mountain
shig shik destroy Term
shing kun Shingkün San. Place
Svayambhūnāth
shun gyi brag Shüngyi Drak Rock of Shün Place
shun gyi brag te’u Shünkyi Drak Teu Rock of Shün Place
sher shang gi rti Shershang Girti Place
Sa
sa dgra sadra earth-enemy Term
sa spyad saché earth examination Term
sangs rgyas rgya Sanggyé Gyatso Person
mtsho
si tu rin po che Situ Rinpoché 1954- Person
sum pa Sumpa Place
srin mo sinmo demoness Term
srog sok life-force Term
srong btsan sgam po Songtsen Gampo Person
gsang ba’i rten sangwé ten phallic image Term
gser khang chen mo Serkhang Chenmo Statue
bsod nams rgyal Sönam Gyeltsen 1312-1375 Person
mtshan
Ha
has po ri Hepori Mountain
hor Hor Place
lha bzhi lhazhi four gods Term
lha sa Lhasa Place
lha sa gtsug lag khang Lhasa Tsuklakkhang Central Temple of Building
Lhasa
lha’i phu’i ri Lhepüri Mountain
lho brag Lhodrak Place
Sanskrit
Amṛtakuṇḍalī Deity
Atiśa 982-1054 Place
Avalokiteśvara Deity
bhikṣu Term
Bhṛkuṭi Person
Buddha Deity
dharma Term
garuḍa Term
Hayagrīva Deity
Hevajra Deity
Jambudvīpa Place
Kālacakra Term
three-sided spike kīla Term
Laṅkapuri Place
phallus liṅga Term
Mahākāla Deity
Mahāyāna Doxographical
Category
Maheśvara Deity
Maitreya Deity
maṇḍala Term
Maudgalyāyana Person
rākṣasa Term
siddha Term
Śākyamuni Deity
Śariputra Deity
śiva liṅga Term
Tārā Deity
Vairocana Deity
Vaiśravana Deity
Vajravārāhī Deity
yogin Term
Chinese
chiin Term
Chu Place
Dunhuang Place
Emei shan Mountain
feng shui Term
Haitong Person
Jianwei Place
Kaiyuan Term
Le shan Place
Minjiang River
Shu Place
Sichuan Place
Tang Dynasty
T’ang Dynasty
Two Hus Place
Wan Place
Wei Gao Person
Wu Place
Xuanzong 685-762 Person
Yuexi He River
zhang Term
Zhe Place
Bibliography
Akester, Matthew. “A Black Demon Peering From the West: The Crystal Cave
of Suratabhaja in Tibetan Perspective.” Buddhist Himalaya: A Journal of
Nagarjuna Institute of Exact Methods 10, nos. 1 & 2 (1999).
Aris, Michael. Bhutan. New Delhi: Vikas, 1980.
Aziz, Barbara. “Towards a Sociology of Tibet.” In Feminine Ground: Essays on
Women and Tibet, edited by Janice D. Willis. Ithaca: Snow Lion, 1989.
Blondeau, Anne-Marie, and Yonten Gyatso. “Lhasa, Legend and History.” In
Lhasa in the Seventeenth Century: The Capital of the Dalai Lamas, edited by
Françoise Pommaret, 15-38. Leiden: Brill, 2003.
Cornu, Philippe. Tibetan Astrology. London: Shambhala, 1997.
Dreyfus, Georges. “Proto-nationalism in Tibet.” In Tibetan Studies: Proceedings
of the 6th Seminar of the International Association for Tibetan Studies,
Fagernes 1992, edited by Per Kvaerne. Vol. 1, 205-18. Oslo: Institute for
Comparative Research in Human Culture, 1994.
Dowman, Keith. The Sacred Life of Tibet. London: HarperCollins, 1997.
Feuchtwang, Stephan. An Anthropological Analysis of Chinese Geomancy.
Bangkok: White Lotus, 2002.
Gyatsho, Thubten Legshay. Gateway to the Temple: Manual of Tibetan Monastic
Customs, Art, Building, and Celebrations. Translated by David P. Jackson.
Bibliotheca Himalayica, series 3, vol. 12. Kathmandu: Ratna Pustak Bhandar,
1979.
Gyatso, Janet. “Down With The Demoness: Reflections on a Feminine Ground
in Tibet.” In Feminine Ground: Essays on Women and Tibet, edited by Janice
D. Willis, 33-51. Ithaca, New York: Snow Lion, 1987.
———. “Drawn from the Tibetan Treasury: The gTer ma Literature.” In Tibetan
Literature: Studies in Genre, edited by José Cabezón and Roger Jackson,
147-69. Ithaca: Snow Lion (1996).
Hazod, Guntram. “Around the ‘Secret’ of Tantruk: Territorial Classifications in
the Historical Landscape of Lower Yarlung (Central Tibet).” Unpublished
manuscript, no date.
Huber, Toni. “Traditional Environmental Protectionism in Tibet Reconsidered.”
Tibet Journal 16, no. 3 (1991): 63-77.
———. “Why Can’t Women Climb Pure Crystal Mountain? Remarks on Gender,
Ritual and Space in Tibet.” In Tibetan Studies: Proceedings of the 6th Seminar
of the International Association for Tibetan Studies, Fagernes 1992, edited by
Per Kvaerne, 350-71. Oslo: Institute for Comparative Research in Human
Culture, 1994.
Huber, Toni, and Poul Pedersen. “Meteorological Knowledge and Environmental

Ideas in Traditional and Modern Societies: The Case of Tibet.” Journal of the
Royal Anthropological Institute, n.s., 3, no. 3 (1997): 577-98.
Kapstein, Matthew. The Tibetan Assimilation of Buddhism: Conversion,
Contestation and Memory. Oxford: Oxford University Press, 2000.
Marko, Ana. “Civilizing Woman the Demon: A Tibetan Myth of State.” In The
History of Tibet, edited by Alex McKay. Vol. 1. London: RoutledgeCurzon,
2003. Previously published in Social Analysis 29 (1990): 6-18.
Martin, Dan. Tibetan Histories: A Bibliography of Tibetan-Language Historical
Works. London: Serindia Publications, 1997.
Mayer, Robert. A Scripture of the Ancient Tantra Collection: The Phur-pa
bcu-gnyis. Oxford: Kiscadale, 1996.
Miller, Robert. “‘The Supine Demoness’ (Srin mo) and the Consolidation of
Empire.” Tibet Journal 23, no. 3 (1998): 3-22.
Mills, Martin A. Identity, Ritual and State in Tibetan Buddhism: The Foundations
of Religious Authority in Gelukpa Monasticism. London: RoutledgeCurzon,
2003.
Mumford, Stan. Himalayan Dialogue: Tibetan Lamas and Gurung Shamans in
Nepal. Madison, Wis.: University of Wisconsin Press, 1989.
de Nebesky-Wojkowitz, Rene. Oracles and Demons in Tibet: The Cult and
Iconography of Tibetan Protective Deities. Kathmandu: Tiwari Pilgrim’s Book
House, 1993.
Needham, Joseph. Science and Civilization in China. Vol. 2. Cambridge:
Cambridge University Press, 1962.
Paul, Robert. The Tibetan Symbolic World. Chicago: University of Chicago, 1982.
Peterson, Charles. “Court and Province in Mid- and Late-T’ang.” In The
Cambridge History of China. Vol. 3, Sui and T’ang China, 589-906, Part 1,
467-560. Cambridge: Cambridge University Press, 1979.
Sakyapa Sonam Gyeltsen. The Clear Mirror: A Traditional Account of Tibet’s
Golden Age. Translated by McComas Taylor and Lama Choedak Yuthok.
Ithaca: Snow Lion, 1996.
Samuel, Geoffrey. Civilized Shamans: Buddhism in Tibetan Societies. Washington:
Smithsonian Institute Press, 1993.
Sørensen, Per K. Tibetan Buddhist Historiography: The Mirror Illuminating the
Royal Genealogies: An Annotated Translation of the XIVth Century Chronicle
rGyal-rabs gsal-ba’i me-long. Wiesbaden: Harrassowitz Verlag, 1994.
———. “Lhasa Diluvium – Sacred Environment at Stake: The Birth of Flood
Control Politics, the Question of Natural Disaster Management and Their
Importance for the Hegemony over a National Monument in Tibet.” In

“Cosmogony and the Origins.” Special issue, Lungta, no. 16 (Spring 2003):
85-134.
Situ Rinpoche. “Geomancy.” Audio Z91. Eskdalemuir: Samye Ling Tibetan
Centre, 1988.
Stein, R. A. “On the Word gCug-lag and the Indigenous Religion.” In The History
of Tibet, edited by Alex McKay. Vol. 1, 530-83. London: RoutledgeCurzon,
2003.
Stutchbury, Elizabeth. “Perceptions of the Landscape in Karzha: ‘Sacred’
Geography and the Tibetan System of ‘Geomancy.’” In Sacred Spaces and
Powerful Places in Tibetan Culture, edited by Toni Huber, 154-86. Dharamsala:
Library of Tibetan Works & Archives, 1999.
van der Kuijp, Leonard W. J. “Tibetan Historiography.” In Tibetan Literature:
Studies in Genre, edited by José Cabezón and Roger Jackson, 39-56. Ithaca:
Snow Lion, 1996.
Vigoda, Marcy. “Religious and Socio-Cultural Restraints on Environmental
Degradation Among Tibetan Peoples – Myth or Reality?” Tibet Journal 14,
no. 4 (1989): 27.
Wong, Eva. Feng shui. Boston: Shambhala, 1994.
Modernity, Power, and the Reconstruction of Dance in
Post-1950s Tibet
Anna Morcom
Royal Holloway College, University of London
Abstract: This article traces the changes and transformations of Tibetan dance
that have occurred since 1950. It looks at how agents of change ranging from
small groups of Tibetans to the Chinese state create, negotiate, and represent
different kinds of Tibetan modernities through reconstructions and reconfigurations
of Tibetan dance, and it examines how dance is modernized, the kinds of modernity
represented, and the social and political power dynamics involved. In particular,
the article looks at the repercussions on dance of the recent state-sponsored
economic development of Tibet that has resulted in the formation of new urban
middle classes but also a dramatic growth in disparity. It also examines the impact
on dance culture of the growth of heterogeneous communities in the increasingly
mobile population of contemporary Tibet. The article further addresses the issue
of globalization, which sees new layers being added to the social and cultural
hierarchies of twenty-first century Tibet.
Introduction
During 1950-51, the People’s Liberation Army entered and gained control of Tibet,1
and it became de facto a part of the People’s Republic of China (PRC). Despite
drastically diverging views as to whether the arrival of the People’s Liberation
Army is to be considered an act of “liberation” or one of “invasion,” it is undeniable
that a new order began in Tibet at this point, a break with the past. Intensification
of contacts with an increasingly fast-changing world outside Tibet at least since
the late nineteenth century had already induced a “modernization” of Tibet in the
sense that new technical facilities began to spread and new social or political ideas
emerged.2 However, unlike early twentieth-century China, this did not result in
1
For brevity’s sake, “Tibet” refers to the culturally Tibetan areas of the PRC.
2
See Melvyn Goldstein, A History of Modern Tibet, 1913-1951: The Demise of the Lamaist State
(Berkeley, CA: University of California Press, 1989) for a general history of the first half of the twentieth
century in Tibet. Accounts of a number of early Tibetan reformists and their activities have been
1550-6363/2007/3/T3129.
© 2007 by Anna Morcom, Tibetan and Himalayan Digital Library, and International Association of Tibetan Studies.
Morcom: Modernity, Power, and the Reconstruction of Dance 2
any large-scale reform movement in Tibet.3 Therefore, a more marked onset of

“modernity” began with the arrival of the Chinese, when “modern” notions of
science, “development,” “evolution,” and “progress” were applied on an
unprecedented scale and intensity to Tibetan society and culture, and were
accompanied by road building and mass media and communications.4 This
modernity brought to Tibet was also intimately linked to a nationalist agenda and
an agenda reflecting the Chinese version of Marxism, in particular in the period
up to the 1980s. Essential to this agenda was the integration of the newly annexed
Tibet into China, or rather the “New China,” and the redefinition or subordination
of the Tibetan people as a “minority nationality.”
With the death of Mao and Maoist ideology and the end of the Cultural
Revolution in 1976, a new phase of PRC history began. A string of economic and
market reforms and the relaxation of the ban on private ownership and commerce
first instigated by Deng Xiaoping in 1978 led to the emergence of a market
economy, triggering a rapidly expanding commercial culture and proliferation of
the mass media, increasingly linking China with the rest of the world. This
transformed Tibet most dramatically after 1999, when the Western Development
Drive led to a state-fueled economic boom in Tibet and other western regions of
the PRC.5 Whilst rural-urban migration has occurred across China, in Tibet it has
published. Some well known examples of figures from different regions of Tibet are
biographies/autobiographies of Dge ’dun chos ’phel (Heather Stoddard, Le mendiant de l’Amdo [Paris:
Société d’ethnographie, 1985]), Bkra shis tshe ring (Melvyn Goldstein, William Siebenschuh, and
Tashi Tsering, The Struggle for Modern Tibet: The Autobiography of Tashi Tsering [New York: M.E.
Sharpe, 1997]), and Ba pa phun tshogs dbang rgyal (Melvyn Goldstein, Dawei Sherap, and William
Siebenschuh, A Tibetan Revolutionary: The Life and Times of Bapa Phüntsok Wangyal [Berkeley, CA:
University of California Press, 2004]). See also A blo, A blo spun mched kyi rnam thar (Beijing: Mi
rigs dpe skrun khang, 1994), a biography of the Fifth ’Jam dbyangs bzhad pa by his elder brother,
another well-known early Tibetan reformer and communist.
3
See Han Kuo-Huang, “The Introduction of Western Music in Modern Times,” in The Garland
Encyclopaedia of World Music, vol. 7, East Asia: China, Japan, and Korea, ed. Robert Provine,
Yoshihiko Tomumaru and J. Lawrence Witzeben (New York: Garland Pub., 2002), 373-77, for an
account of this period of reform in China as it related to music and the arts.
4
See Arjun Appadurai, Modernity at Large: Cultural Dimensions of Globalization (Minneapolis:
University of Minnesota Press, 1996); Carl Freedman, The Incomplete Projects: Marxism, Modernity,
and the Politics of Culture (Middletown: Wesleyan University Press, 2002); Anthony Giddens, The
Consequences of Modernity (Cambridge: Polity Press, 1990, reprint 2005); and John Thompson, The
Media and Modernity: A Social Theory of the Media (Cambridge: Polity Press, 1995) for a range of
general, theoretical readings on modernity. The arrival of the Chinese tends to be seen in a rather
oversimplified way as marking the beginning of “modernity” in Tibet due to the proclamation of it as
such by the Chinese government with the heralding of the “new society” and the vilifying of “backward
old Tibet,” which continues to this day. The idea that modernity (in a negative sense) exists in Tibet
only because of China has also been supported by the exile establishment and western world, which
romanticize a timeless, traditional Tibet that has been destroyed by China.
5
There were development initiatives in Tibetan areas before 1999 but nothing on the scale of the
Great Development of the West (xibu da kaifa), first announced in June 1999 by Jiang Zemin. See
Tibet Information Network, News Review: Reports from Tibet, 1999 (London: Tibet Information
Network, 2000), 113-17 for Jiang Zemin’s and others’ speeches on western development, and Tibet
Information Network, China’s Great Leap West (London: Tibet Information Network, 2000) for a
study of the beginnings of the Western Development Drive. See Andrew Fischer, State Growth and
Social Exclusion in Tibet: Challenges in Recent Economic Growth (Copenhagen: Nordic Institute of
Asian Studies, 2005) for a book-length study of economic development in Tibet based on both fieldwork
had an ethnic angle, since in addition to rural Tibetans, large numbers of Chinese
have also migrated to Tibetan cities for work, largely from Chinese regions
bordering the Tibetan areas.6 Chinese from the wealthy east coast and westerners
have also begun to arrive in Tibet in large numbers following the boom in tourism,
particularly in the last few years.7 From the first few years of the new millennium,
the “global flows” of “mediascapes,” “ethnoscapes,” “technoscapes,” and
“financescapes” have definitively reached Tibet,8 although “ideoscapes” remain
tightly controlled by the one-party state in its continued resistance to the trope of
“modernity as democracy”9 and the control of separatism in Tibet.
While the Maoist, revolutionary period ended (in)famously with performing
arts culture narrowed down more or less to revolutionary songs and eight model
operas, in the Tibet of the twenty-first century, Britney Spears, MLTR’s “Take
Me to Your Heart,” and Bollywood dance and songs are performed live in larger
urban centers alongside the Tibetan and Chinese pop which has existed since the
1980s, and people dance with abandon to techno in discos and nang ma (nightclub)
and analysis of macroeconomic data. See also the economy/development sections of Tibet Information
Network, Tibet 2002: A Yearbook (London: Tibet Information Network, 2003), Tibet Information
Network, News Review: Reports from Tibet, 2000 (London: Tibet Information Network, 2001), Tibet
Information Network, News Review: Reports from Tibet, 2001 (London: Tibet Information Network,
2002), Tibet Information Network, Tibet 2003: A Yearbook (London: Tibet Information Network,
2004), Tibet Information Network, “National Autonomy Law Revised to Support Western Development
Policy,” Phayul.com, http://www.tew.org/development/autonomy.law.html (originally published on
Tibet Information Network, “News Update,” March 13, 2001; site now discontinued), and Tibet
Information Network, “ADB Supports Western Development Drive, but Not in Tibet,” Phayul.com,
http://www.tew.org/development/asia.devel.bank.html (originally published on Tibet Information
Network, “News Update,” March 30, 2005; site now discontinued) for more information relating to
the Western Development Drive.
6
Fischer, State Growth and Social Exclusion. Significant numbers of Chinese have also settled in
the low, fertile rural areas of southern and eastern Tibet, Kong po and southern Khams, some from as
early as the 1950s.
7
Tourism became an economic focus in Tibet in the early 1980s, but the scope was limited due to
lack of facilities and infrastructure.
8
Appadurai, Modernity at Large.
9
The idea that modernization necessarily involves democratization and a focus on individual rights
is assumed by the western world. See, for example, the discussion of modernity and Indian modernity,
which inherited much from the western tradition, in Avijit Pathak, Modernity, Globalization and
Identity: Towards a Reflexive Quest (Delhi: Aakar Books, 2006).
bars in county towns.10 Cassette and VCD11 players are found in villages across
Tibet, with the exception of the most remote areas still not reached by electricity.12
Although before the arrival of the Chinese in Tibet a mixing of different regional
styles of song and dance and some non-Tibetan styles had occurred at least through
religiously motivated travel and migration,13 the changes since the 1950s are
undeniably of an unprecedented order.
This article traces the changes and transformations of dance that have occurred
since 1950 by looking at how agents of change ranging from small groups of
Tibetans to the Chinese state create, negotiate, and represent different kinds of
Tibetan modernities through reconstructions and reconfigurations of Tibetan dance,
and also by examining how dance is modernized, the kinds of modernity
represented, and the social and political power dynamics involved. Although this
article draws on fieldwork in a range of Tibet’s regions as well as areas of varying
levels of urbanization, given the size of Tibet and the relatively small quantity of
existing research on any aspect of Tibetan performing arts, there will inevitably
be shortfalls in this attempt to comment on major trends in contemporary Tibet.14
10
I witnessed dancing to techno in a number of county towns of Dkar mdzes Prefecture (Dar rtse
mdo, Li thang, Dkar mdzes, and the remote Gser shul), and also Rma chu County in Gannan Prefecture.
Bollywood and western pop are performed live more in the big cosmopolitan cities with Tibetan
population – Lha sa, Chengdu and Xining (all province capitals) – and also in prefecture capitals (Nag
chu, Brtse thang, and probably Gzhis ka rtse, although I did not visit bars there) as well as ’Gram (the
county on the border with Nepal). However, I know a Tibetan singer who performs western pop who
was invited to perform in a nang ma bar in a remote county of Lho kha, and who performed
spontaneously in many nang ma bars she visited with me across Khams and A mdo, and also in her
own village in A mdo. Popular and cutting edge singers (such as those who can perform well in Hindi
or English) get paid very well to make special appearances in nang ma bars in often remote places,
with owners scouting talent in Lha sa and other big cities. Networks of performers and nang ma bar
owners also fuel the spread of styles to different areas of Tibet.
11
Neither VHS nor DVD technology made significant impact in Tibet; the VCD (Video Compact
Disc) remains the standard medium of release of music after audio cassettes.
12
Not all villages in Tibet had electricity when I visited in 2004-2005, though probably the majority
did. It appears that electricity arrived in most villages during the 1980s.
13
Anna Morcom, “History, Traditions, Identities, and Nationalism: Drawing and Redrawing the
Musical Cultural Map of Tibet,” PIATS 2006 (forthcoming). This dynamism and hybridity of
“traditional” Tibetan musical culture is glossed over/denied by essentialist and politically motivated
conceptions and representations of “old Tibet” inside and outside of China.
14
I would like to warmly thank the Leverhulme Trust for their generous support of a year’s fieldwork
in Tibet from 2004-2005 under the Study Abroad Studentship scheme, which has provided the basis
of this article. I would also like to thank all the singers and dancers I learned from and spent time with
in Tibet, who made my fieldwork both a fascinating and fun experience. The research consisted of a
mixture of formal and informal meetings and interactions with dancers, musicians and people involved
with dance and music, some filming and recording, and a large degree of participant observation, the
major parts of which were learning: Lha sa nang ma and stod gzhas dance with Dpal sgron lags, formerly
a teacher from Tibet University, in Lha sa; the modern/dance troupe style of dance in Lha sa with G.yu
sgron, a dancer/choreographer/singer in Lha sa nang ma bars and member of the Lha mo tshogs pa;
new-style sgor bro over many nights in nang ma bars across Tibet and briefer periods in town squares
in Khams and A mdo; (briefly) ’ba’ gzhas in ’Ba’ thang with Blo bzang tshe ring; and A mdo rdung
len songs and mandolin with Bstan ’dzin in Lha sa and A mdo, resulting in a commercially released
VCD, Sprin gyi me tog, Anna Morcom, VCD ISRC CN-H01-06-320-00/V, 2006. I planned expeditions
to certain areas in order to study particular genres in their contemporary context and to gain a view of
rural versus urban, big city versus small town etc., and different regions. I also made more spontaneous
However, it is hoped that in particular, this article will contribute to an

understanding of the social, economic, and cultural changes taking place in Tibet’s
newest phase of modernity of the twenty-first century, and more generally, of the
possibilities and particularities of the role of performing arts in local and national
cultural change and modernity.
Post-1950s Tibet, State Modernity

Chairman Mao was well aware of music and other arts as an active force in the
creation of identity and culture, and thereby a crucial building block for the new
society. In 1942 he stated, “There is a cultural as well as an armed front,” and “a
cultural army … is indispensable in achieving unity amongst ourselves and winning
victory over the enemy.”15 The musical styles of the 1950s were a product and a
reminder of the “old society,” and distinctive ethnic styles were potential signifiers
of the cultural distinctness of these minorities and hence a fragmentary force on
the new unified PRC, especially in the case of the newly annexed Tibet. The
traditions of music and dance of all nationalities of China, including the Tibetans,
were uniformly reformed and “developed” through a centralized network of Han
Chinese training institutions and dance troupes, the basis of which were
western-style conservatoires set up during China’s turn-of-the-nineteenth-century
drive to reinvent itself as a modern nation in reaction to humiliation by foreign
powers. Ironically, this was done by adopting the “scientific,” “developed” western
techniques, styles, and training systems of these foreign powers.16 Whilst these
overarching concepts of cultural modernity originating from western ideas dated
from the beginning of the twentieth century in China, they were first
institutionalized in Tibet by the Chinese. The new style created by these institutions
in all regions of the PRC expressed and performed two points of key political
importance to the new China:
expeditions to various places led by fortuitous contacts with singers and dancers and other friends. The
article also draws on an involvement with Tibetan performing arts in exile and diaspora Tibetan
communities dating from 1990. I emphasize that the interpretations of change in Tibetan musical culture
contained in this article are my own, and my own responsibility. I would also like to thank Thierry
Dodin and Isabelle Henrion-Dourcy for their invaluable input on drafts of this article.
15
Bonnie McDougall, Mao Zedong’s “Talk at the Yan’an Conference on Literature and Art”: A
Translation of the 1943 Text with Commentary (Ann Arbor, MI: Center for Chinese Studies, University
of Michigan, 1980), 57. See Unity and Discord: Music and Politics in Contemporary Tibet (London:
Tibet Information Network, 2004), 13-30, for a more lengthy analysis of the shaping of Tibetan music
in general by Chinese nationalist and Maoist ideology. See Isabelle Henrion-Dourcy, “Ache Lhamo:
Jeux et enjeux d’une tradition théâtrale tibétaine” (PhD thesis, Université Libre de Bruxelles, 2004),
for a detailed account of Tibetan opera, including its post-1950s history. For an account of the process
of Chinese modernization of the musics of other minority nationalities of China see Helen Rees, Echoes
of History: Naxi Music in Modern China (Oxford: Oxford University Press, 2000), 10-27, 130-69, and
Rachel Harris, Singing the Village: Music, Memory and Ritual Among the Sibe of Xinjiang (Oxford
and New York: Oxford University Press for the British Academy, 2004), 156-93.
16
Kuo-Huang, “Introduction of Western Music.”
1. The “unity of the nationalities,” through the unification of historically and

ethnically diverse styles as they were all glossed with the same brush of
the Han Chinese modernization.
2. The creation of a new style for the new nation, presenting itself as
“developed” due to the use of “scientific” training systems and professional
training and performance. Following in the long tradition of the “civilizing”
and controlling of non-Han peoples by the Han Chinese,17 but fueled with
new socialist ideology, the “raw material” of folk music and dance was
turned into “art,” “developing” and “raising the standards” of the
“backward” masses/minorities.18
Through its embodiment of the explicitly named “new society,” its profound
reformation of the old, and the near-total censorship of anything else for at least
ten years during the Cultural Revolution, this style created and defined music and
dance modernity in Tibet. The state therefore created a modernity expressed as
development, improvement, refinement, civilizing, and sinicization pitted against
the unmodernized (“traditional”) and non-Han Chinese as backward.
Although there was no genre of pan-Tibetan dance prior to the 1950s – instead,
there were more or less linked regional styles – and certain broad elements common
to all regional styles were uniformly modified to create a genuinely new style and
a new set of norms.19 The key changes in terms of style were as follows:
1. The adoption of a modified vocal style, as professional Tibetan singers

were retrained in bel canto or Chinese styles.
2. The use of orchestras and large ensembles, whereas traditional Tibetan
dances are mostly accompanied by the singing of the dancers or a single
or small group of instruments.
17
Thomas Heberer, “Old Tibet a Hell on Earth? The Myth of Tibet and Tibetans in Chinese Art and
Propaganda,” in Imagining Tibet: Perceptions, Projections and Fantasies, ed. Thierry Dodin and Heinz
Räther (Boston: Wisdom Publications, 2001), 113-21, Harris, Singing the Village, 1-15, and Unity and
Discord, 13-30 discuss these issues in relation to music.
18
Terminology from Mao’s talks; see McDougall, Mao’s “Talk at the Yan’an Conference.”
19
See Isabelle Henrion-Dourcy, “Women in the Performing Arts: Portraits of Six Contemporary
Singers,” in Women in Tibet, Past and Present (London: C. Hurst, 2005), especially 198-201, and
Isabelle Henrion-Dourcy and T. Dhondup, “Tibet, §III, 1. Folk music,” in New Grove Dictionary of
Music and Musicians, ed. Stanley Sadie and John Tyrrell, 2nd ed. (London: Macmillan, 2001), 449-52
for general information on the character of traditional Tibetan performing arts.
Figure 1: Tibetan choir sings in a festival in

Dar rtse mdo, May 2005.
3. Ballet as a basis for all dance and ballet training for all dancers, resulting
in a stretched body posture, a lightness of movement, and bigger
movements, with hands reaching high above the head. With traditional
Tibetan dance, the body is relaxed, often slightly stooped, the hands do
not reach above the head,20 and the movements are heavier with a lower
center of gravity.
Figure 2a: Traditional style group dance from

a village near Dar rtse mdo.
4. Increased complexity and “sophistication”: most Tibetan traditional

dances are group-participative, performed by non-professionals and with
little variation of steps (but long song texts), making them, by the standards
of the dance troupe aesthetics, inadequate to constitute dance as “art”
since they are only “raw material” for “art.”21 A greater variety of steps
was therefore introduced in all dances, mixing the steps of one dance with
those from different regions and integrating newly composed moves and
20
The Khams pa dances of Yul shul and Rdza chu kha appear to be an exception with big arm moves
that seem to predate the post-1950s modifications according to older performers. See Henrion-Dourcy,
“Ache Lhamo,” 393-411 for a detailed description of the movements of a ce lha mo, the most detailed
description available for any Tibetan theater or dance form.
21
Mao’s terminology, see McDougall, Mao’s “Talk at the Yan’an Conference.”
ballet moves. The dances are performed in complex formations rather

than the traditional circle, and tend to be multi-sectional compositions.
Figure 2b: County-level dance troupe perform

in Nag chu, 2004.
5. The separation of song and dance: As Henrion-Dourcy and Dhondup

state, “Dances are invariably associated with singing, but not all songs
are danced to.”22 However, as Tibetan performing arts were made to fit
into the cast of western-style Chinese conservatoires, performers were
highly trained in either song or dance but not both.
6. Audience-oriented stage performance, with an emphasis on virtuosity and
slick, professional, flawless, showy performance.
Figure 3a: Heavy winter Tibetan clothes in rural

A mdo.
22
Henrion-Dourcy and Dhondup, “Tibet, §III, 1. Folk music,” 449.
7. Brightly colored, floaty costumes: traditional Tibetan clothes consist of

heavy wool, brocade, animal skins,23 and very large ornaments that can
weigh, at a guess, from one kilogram to over ten. This change in costumes
makes a “lighter” style possible, linked to the ballet aesthetic and drive
to virtuosity, though less heavy and hot clothes are also a necessity to
some extent for indoor stage performance. Uniform costumes as opposed
to clothes were also introduced.
These modifications represented

certain unprecedented changes to the
world of Tibetan music and dance as a
whole. First, whilst change in the form of
the mixing and migration of styles
(largely Tibetan inter-regional) certainly
existed in pre-1950s Tibet (though it is
not possible to say how extensively),24
the state style was the first style that
simultaneously and uniformly changed
Figure 3b: Traditional ornaments of Nag chu theoretically all regional styles at once,
nomads. in accordance with aesthetics very much
outside of the interlocking world of Tibetan regional styles.25
Second, the state modification of Tibetan dance in this way created the first
overarching concept of pan-Tibetan dance, in contrast to the linked but regionally
distinct styles and identities of pre-1950s Tibet. Apart from a few regional genres
which are more or less recognizable – ’ba’ gzhas from southwest Khams, Tibetan
lute (sgra snyan) dance from Lha sa and Dbus gtsang, Kong po Tibetan lute dance
– most of the state conservatoire Tibetan dances involve mixtures of elements from
different regions plus balletic or new Tibetan-style moves and are difficult to place
regionally. Smaller troupes at the county level do perform varying degrees of local
repertoire, though almost always in the state style.26 However, the troupes still tend
to learn a central repertoire via teachers who have been trained in the state
conservatoires, and more recently, from watching television and VCDs. Therefore,
in addition to the modification with the pan-China style, there is incomparably
23
In one of the most dramatic demonstrations of sentiment in Tibet since the late 1980s, following
a strong plea by the Dalai Lama in India in January 2006 to Tibetans in Tibet not to wear endangered
animal skins, incidents of fur-burning spread across the Tibetan areas, and by summer 2006 animal
pelts had disappeared from the clothes of ordinary Tibetans, though sheepskin dresses/coats and lamb
fleece lining are still worn. See http://www.tibetinfonet.net/ for reports and images. During summer
2006, I saw elaborate brocade borders being worn for best clothes instead of fur at the Li thang Horse
Race Festival.
24
See note 13.
25
While styles changed, migrated, and interacted with other styles, traditional regional styles are
certainly not conceived as being mixed (’dres ma). Therefore, it can be argued that the ideological
acceptance of hybridity began at this point. I am grateful to Isabelle Henrion-Dourcy for this point.
26
The county troupes may also perform dances where none existed in the area before the 1950s, such
as Rma chu county dance troupe.
more cross-regional standardization of repertoire than there was before the 1950s.
The state thus created a form of national Tibetan dance, but one that was strictly
a subset of the new China’s performing arts: a semi-homogenized block easy to
rationalize as a part of China, a clear message, to be repeatedly performed, of Tibet
as “an inalienable part of China.”27 This can be contrasted with the Tibetan national
style created in exile, the “dances of the roof of the world” phenomenon, which
performs Tibet as a unified series of regions with Lha sa at their center.28
In the embodiment of the explicitly
named “new society,” the blanket
reformation of the old, the accompanying
banning or heavy repression in Tibet of
“traditional” performing arts unless
carrying propaganda lyrics from the
1950s until 1980, and the propagation of
this style alone through the
state-controlled media (radio,
loudspeakers, film, and limited
television), the state style solely Figure 3c: Shop in Chengdu selling floaty, exotic
constituted musical and dance modernity costumes for dance troupes and nang ma bars.
in Tibet and China. This meant therefore
that until the 1980s, modernity in Tibet became sinicization, a creation of Chinese
nationalism monolithically defined by the Chinese state. All other aspects of
modernity (media, technology, communications) were used to serve the state’s
purpose and strictly censored; as China shut itself and in particular Tibet off to the
outside world, relationships with other cultures were either terminated (for example,
the centuries-old trade and religious interactions of Tibet with India, Nepal, Bhutan,
and Mongolia) or mediated strictly by the state (for example ballet, opera, and
other aspects of western and Soviet classical music and dance entered Tibet only
via Chinese state conservatoires); and Han Chinese communists or Tibetan
communists whose views were acceptable to the Han Chinese state were the only
real agents of change, with no participation of the Tibetan masses.
27
Compare this to corporate America’s representation of global diversity and harmony in the “Small
World” ride at Disneyland Paris in Laudan Nooshin, “Circumnavigation with a Difference? Music,
Representation and the Disney Experience: It’s a Small, Small World,” Ethnomusicology Forum 12,
no. 2 (2004): 236-51.
28
Morcom, “History, Traditions, Identities, and Nationalism.” The exile representation of Tibet is
beyond the scope of this article, but it must be pointed out that the exile administration is overwhelmingly
dominated by central Tibetans, and while they represent Tibet’s regional diversity in many ways in
performance, they also perform everything in a more or less Lha sa dialect, and all modern songs are
sung in Lha sa dialect. See Morcom, “History, Traditions, Identities, and Nationalism” for an analysis
of exile Tibetan style in performance by the Tibetan Institute of Performing Arts (TIPA).
State and Other Modernities in Regional and Popular Culture

in Post-1980s Tibet
While the state singularly created and controlled performing arts modernity in the
revolutionary period, by the late 1980s, the state style seemed increasingly
old-fashioned, stilted, and obsolete. Modernity had “moved on,” leaving the state
behind. With restrictions on private ownership lifted, broadcast and new forms of
mass media (cassettes and later VCDs) began to spread, though with limited effect,
especially in the countryside, until the late 1990s and the new millennium. With
the censorship on anything but the “correct” revolutionary content lifted, there
were revivals of traditional performing arts, and forms of musical performance
born out of other modernities were able to enter Tibet in a more aesthetically
pluralistic popular sphere with increasing interaction with the private,
non-governmental realm.29 In the 1980s, karaoke bars with Taiwanese and Hong
Kong pop music swept Lha sa, and in the late 1980s and 1990s, immensely popular
songs in this style were performed in the Tibetan language, the most famous singers
being Zla sgron, Byams pa tshe ring, and Bde chen dbang mo. While rock music
swept China in the 1980s, a Tibetan hard (as opposed to soft) pop style with big
vocals and a very wide range (echoing traditional Tibetan nomadic singing) was
introduced by Yar ’drong during the early 1990s. From a more rural context, in
the early 1980s came rdung len, songs in A mdo Tibetan dialect accompanied by
mandolin and increasingly nowadays with synthesized music and a beat.30 While
Tibetan pop music began in the late 1980s, it has greatly increased in scale and
become more commercialized since the late 1990s, with the economic boom in
Tibet and the proliferation of the mass media and entertainment venues (discussed
below). Other styles of popular music have also emerged in the new millennium:
more western-style rock bands particularly in Lha sa (some very similar to the
model of pop music found amongst Tibetan exiles in India and Nepal),
African-American influenced pop (such as Chung zhol sgrol ma), and a current
plethora of highly stylized and choreographed girl bands and boy bands. All these
29
See Nimrod Baranovitch, China’s New Voices: Popular Music, Ethnicity, Gender and Politics,
1978-1997 (Berkeley, CA: University of California Press, 2003) for a discussion of this with reference
to mainland China. The sense of an “independent” popular music scene in Tibet is problematic even
now. Popular and recorded music in Tibet initially emerged from government institutions in the case
of Lha sa (for the TAR) and A mdo, with the biggest stars, such as Zla sgron and Byams pa tshe ring,
on government payroll or close to the government. Many of Tibet’s most famous singers continue to
come from state troupes and conservatoires, including most of the A mdo singers. However, the market
has opened up to singers outside the government realm, and singers inside the government realm have
been free to integrate outside influences in their music, as Zla sgron did with Taiwanese pop music.
See Anna Morcom, “Getting Heard in Tibet: Music, Media and Markets,” in Consumption, Markets
and Culture (forthcoming).
30
Institutionally speaking, rdung len, like Lha sa’s first pop music, also emerged from government
recording offices and dance troupes, with government personnel acting as key composers and lyric
writers (there were no private recording studios in the 1980s). However, it was certainly a new, local
Tibetan phenomenon with close links to traditional styles rather than a Beijing creation. See Morcom,
“Getting Heard in Tibet.”
new styles – with the exception of rdung len, which is sung only in Tibetan – are
sung in both Tibetan and Chinese.
Appadurai’s words seem to encapsulate this process: “we can see that electronic
mass mediation and transnational mobilization have broken the monopoly of
autonomous nation-states over the project of modernization.”31 However, while
the monopoly of the state was indeed broken, the influence of the state has by no
means been overwhelmed. In fact, rather than becoming a specter of the discredited
revolutionary past, in many ways the state style is leading perceptions and
expressions of modernity in dance at far deeper levels of the population than in
the period prior to 1980. Paradoxically, the state is achieving an influence and
control on popular culture in the new climate of economic liberalization and
“hands-off” cultural policy that it never managed to achieve through heavy
repression and censorship.32 While the popular culture of the 1980s and early 90s
was certainly radical (sad songs, some romantic, of loss, pain, and longing versus
happy revolutionary songs, and a new lyrical style expressing Tibetan identity
grounded in Tibetan experience and emotion rather than “correct” state
representations),33 there is a large section of the popular culture from the late 1990s
that conforms to the state style.
The Meeting of State and People

While by the 1980s the revolutionary ideology may have been outdated even from
the point of view of the party, the 1950s state performing arts style (and within it
the culturally-coded ideologies of ethnicity and “development”) was able not only
to stay up to date, but to become a key reference point for core popular culture in
the twenty-first century and also for the performance of “traditional” genres in the
modern world down to the grassroots level. In a compromise and ideological
backtrack, the state has, since the end of the Cultural Revolution, started not only
to tolerate other forms of modern performing arts but to actually promote many
aspects of the new “independent” popular culture that emerged from the 1980s
through its troupes and state television. In the first instance this has given the state
a more trendy and people-friendly image. More importantly, through a process of
inclusion and appropriation, it is the state that represents and in a sense claims or
“patents” these potentially threatening forms of modernity rather than being
undermined by them. It is also in this way able to exert more direct forms of control,
through, for example, not allowing performers to address (TV or live) audiences,
and restricting live performances in particular by artists with big followings amongst
the masses.34 State dance troupes now feature people singing pop songs in rock or
31
Appadurai, Modernity at Large, 10.
32
See Unity and Discord, 137-78, for a discussion of censorship and protest songs in post-Cultural
Revolution Tibet.
33
See Henrion-Dourcy, “Women in the Performing Arts,” for a description of Zla sgron and the
significance of sad songs in the context of this era in Tibet.
34
Baranovitch, China’s New Voices. Bdud be, the unparalleled A mdo rdung len singer, sang in Lha
sa in July 2006 at the new two-thousand-person capacity venue of Thang go la. To control the emotional
Taiwan style and with Tibetan lyrics as well as Chinese (although Chinese lyrics
tend to dominate), traditional songs or songs closer to traditional styles (such as
nomad songs [’brog glu] and rdung len), and the new more global-style girl bands
and boy bands. Explicitly Chinese nationalist and political material has been greatly
reduced in the performances of state troupes, but it has not disappeared (see figures
4a and 4b, which are photos of a lengthy dramatic performance by the Dkar mdzes
Prefecture dance troupe of the crossing of the Luding bridge on the Long March
on the seventieth anniversary of this occasion).
Since around 1990, singers may be
members of dance troupes in addition to
making private albums and performing
for personal profit at festivals and on
television, and successful singers are
invited to perform for variety shows on
Tibetan TV for occasions such as New
Year or the Yogurt Festival (zho ston) in
Lha sa. The most famous pop stars who
have achieved nationwide fame (and sing
Figure 4a: Crossing of Luding bridge drama by mostly in Chinese) may occasionally
Dkar mdzes dance troupe, May 2005.
appear in the variety shows on Chinese
mainland television. These shows feature Chinese pop stars and glitzy, spectacular
dances. They are broadcast across China (including Tibet) on CCTV3 and other
channels, featuring dancers largely belonging to state institutions, though singers
may be independent. In this way, the realms of “state,” “popular,” and “traditional”
music have become blurred, and the state’s cultural activities, rather than being
explicitly and crudely political as in the revolutionary period, now seem just cultural.
However, whilst the state is vigorously
appropriating “popular” culture, “the
people” can be seen to be actively
appropriating the state style. I will
examine two examples of this, one urban
and one rural. Nang ma bars, which
started around 1997, were originally
drinking houses in Lha sa where
traditional music from Lha sa known as
nang ma was performed along with other
Figure 4b: Crossing of Luding bridge drama by
Dkar mdzes dance troupe, May 2005.
reaction of the almost entirely Tibetan audience, the consumption of alcohol was not allowed, whereas
usually nang ma bars like Thang go la (see below) are places to go to drink as much as to listen to
music. In this way, we see the political power of Tibetan music, and the policy of the authorities to
contain and control rather than censor. Even performers who do step outside the mark, for example
with lyrics, tend nowadays to be contained rather than arrested, which tends to greatly increase their
popularity (as happened with the A mdo singer Nam mkha’, arrested around 2003).
genres.35 They rapidly spread across Tibet and are now found in all big cities
(prefecture and province capitals in Tibetan areas) and some county towns.36 As
a result of their immense success, they rapidly became commercialized and
competed for better singers and dancers and better, slicker, flashier shows. While
there are countless extremely proficient non-professional performers of regional
genres in the countryside in Tibet, the state dance troupes provide a consistent
standard of staged, slick, showy, professional, costumed performance that exists
little in traditional musical culture, especially dance.37 Nang ma bars hence started
to draw from dance troupe singers, dancers, dance teachers, and choreographers,
some remaining in their dance troupe but supplementing their unlavish state salary
by performing in the nang ma bar in the evening and others moving permanently
into the private sphere.38 Shambhala, for example, the most expensive nang ma
bar in Gannan Autonomous Prefecture, Gansu Province, is staffed almost
exclusively by moonlighting members of the two Gannan Prefecture dance troupes.39
G.yung drung rgyal’s nang ma bar, which opened in 2005 and quickly achieved
the status of the best nang ma bar in Lha sa, has a large troupe of around ten female
35
See also Adams, “Karaoke as Modern Lhasa, Tibet: Western Encounters with Cultural Politics,”
Cultural Anthropology 11, no. 4 (1996): 510–46, on karaoke and nang ma bars.
36
Lha sa has many nang ma bars; Chengdu and Xining, the capital cities of Sichuan and Qinghai,
the provinces with the biggest Tibetan areas after the TAR, both have several; and there are also two
or more that I visited in Nag chu, Gzhis ka rtse, Brtse thang, Gannan, Dar rtse mdo, and Yul shul, all
capitals of Tibetan prefectures. There are also several restaurants in Beijing with nang ma bar-style
song and dance displays. Nang ma bars are less common in county towns, many of which have only
discos or karaoke bars, though several I visited apparently had nang ma bars that had closed (’Ba’
thang, Dkar mdzes, Li thang). Gser shul in Dkar mdzes Prefecture still had a nang ma bar, and one had
just opened in Rma chu when I visited in 2006. At least several of the county towns in Lha sa
municipality also have nang ma bars, plus at least one in Lho kha.
37
Of the pre-1959 genres of Tibetan non-monastic music and dance, it appears that only gar was
performed solely by professional performers (in the sense of salaried or specially trained experts).
Three other genres were performed by either amateurs or groups that contained both amateur and
full-time practitioners: nang ma and stod gzhas (within the Nang ma’i skyid sdug), the a ce lha mo
opera (within the Skyor mo lung troupe), and ral pa. The vast majority of Tibetan performing arts were
performed by amateurs. Most group dances are traditionally performed by members of the community,
particularly those who are good dancers and singers, but certainly not by paid professionals, and it is
similar for nomad songs. I am grateful to Isabelle Henrion-Dourcy for her clarification on the above
points (personal communication, September 2006).
38
While a good state salary may be about ¥2000 in Lha sa and ¥3000 a very good salary, the most
popular singers in Lha sa’s big nang ma bars were earning over ¥5000 a month in 2004-2005, though
wages seemed to have dropped for most singers in 2006 with the influx of many singers from A mdo
happy to perform for ¥1500 (an excellent salary for a rural migrant to Lha sa, as many of the singers
are). Singers also gain exposure that can help launch careers as independent artists. Dancers are paid
less, but can earn ¥1500 a month; a choreographer/coach earns significantly more. Lha sa nang ma
bars pay by far the highest salaries (partly due to the state sponsored economic boom in the Tibet
Autonomous Region), closely followed by Chengdu, and hence attract the best and most ambitious
singers. Singers in a smaller nang ma bar in Gannan, for example, were apparently only making 300
to ¥500 per month, which is a low to medium salary for that region. Representatives of one county in
Dkar mdzes Prefecture admitted that one of their troupes had closed down because so many performers
had left to perform in nang ma bars in Lha sa due to the high salaries.
39
The Zlos gar tshogs pa and the Glu gar tshogs pa. Both perform song and dance: the former, modern
styles; and the latter, distinct traditional genres (though largely in modernized forms).
and ten male dancers,40 many of whom are from the dance troupe in A mdo snga
pa Prefecture, Sichuan Province, which is G.yung drung rgyal’s native area.
The first nang ma bars in Lha sa were talked about outside Tibet as bastions of
traditional culture resisting the domination of Chinese-style karaoke bars and
discos. However, nang ma bars now contain, at a guess, a maximum of 5-10 percent
traditional Tibetan music (and much of that in modified form), and probably 80
percent Chinese language songs in all except the smallest nang ma bars. In general,
Chinese is used to introduce acts, even in small nang ma bars in county towns.
The dance style of nang ma bars now represents a wholesale adoption of dance
troupe style, with most nang ma bars, especially the bigger ones, having a dance
coach/choreographer who is either a graduate or current member of a dance troupe
or conservatoire. Thus, the junior performers at nang ma bars – even those who
are not former dance troupe members or conservatoire graduates – learn the state
dance style. Dance troupe dances are also performed in nang ma bars, being copied
from dance troupe performances (easily available on VCDs or television), taught
by a dance troupe member who is paid to teach a particular dance (another source
of extra income for dance troupe professionals), or choreographed anew but with
the same basic repertoire of steps and the same state-style staged, costumed, and
balletic aesthetic. This has resulted in nang ma bars becoming the equivalent of
private dance troupes. G.yung drung rgyal’s nang ma bar in Lha sa, in particular,
with its flawless dance performances (dancers are fined ¥10 for mistakes), is of
the standard to be expected from a prefecture level group, certainly more trained
and professional than county troupes, and is setting new standards for nang ma
bar performance. The fact that dancers from the nang ma bars of Lha sa as well
as all the official troupes had to perform in Lha sa for the show I saw on National
Day, October 1, 2004, further shows the equivalence of nang ma bar and state
dance troupe dancing and dancers. The linking of nang ma bars to the state style
and the state institutions also results in a large degree of standardization in terms
of repertoire, dance steps, and dance style in nang ma bars as far afield as Lha sa,
Gannan, Chengdu and Xining.
While the song and vocal styles of nang ma bars, not discussed in detail here,
as much reflect popular culture being appropriated by the domain of official culture
as vice versa,41 the typical performance style of the songs is identical to that of the
dance troupes. On the whole, singers sing and do not dance, and perform statically
with hand gesturing, something that has come to be incorporated in performances
of songs as diverse as Chinese/Tibetan/Western rock and pop, bel canto opera,
nomad songs, and even modernized versions of regional dance-songs. A series of
photographs from a singing competition held in Gannan Tibetan Autonomous
Prefecture in A mdo illustrates this.
40
Previously the biggest nang ma bar troupes consisted of about five female and five male dancers.
41
The far greater level of conformity to the state style in dance and presentation as opposed to song
and singing or vocal style is discussed below.
Figure 5a: Bel canto style in Gannan

competition, August 2005.
Most singers also follow the norm of walking to the back of the stage with their
back facing the audience during instrumental interludes and then turning to the
audience again when they sing again. A new trend in Tibetan-style dance songs is
to dance a few very basic steps strictly during the interludes in Tibetan songs based
on or made for dancing, still leaving the “taboo” on dancing and singing together
unbroken. However, static performances of even dance songs with dance-based
lyrics are found. Globalized patterns that have entered Tibet seem set to change
this performance style, but in 2005, this blueprint still remained very much the
default and the norm.
Figure 5b: Nomad song in Gannan competition,

August 2005.
The state style also entered the recorded popular culture scene around the late
1990s in close conjunction with nang ma bars, as Tibetan music started to be
released visually on VCD. The first Tibetan VCD was made in 1996/1997 by the
government of Dkar mdzes Prefecture. The next was released in 1998 by Thang
go la, one of the most famous nang ma bars in Chengdu, which quickly went into
music production and has become one of the leading establishments of Tibetan
popular music culture and the Tibetan music industry, launching many famous
Tibetan pop artists such as Chung zhol sgrol ma and Bkra shis nyi ma.42 There has
been much state-style/nang ma bar-style dancing in Tibetan pop VCDs since this
42
Interview with Tshe dbang, owner of Thang go la, Chengdu, June 8, 2005.
time, with Thang go la’s immensely popular “disco” medleys filmed in Thang go
la with backing dancing from the Thang go la dancers, and a variety of other
productions using dance troupe performers, again achieving polished and modern
performance.
Figure 5c: Chinese nationalist song in Gannan

competition.
In a dance competition between amateur troupes in ’Dam gzhung Horserace
Festival 2004, key aspects of the state style were adopted through aspiration to
professionalism, good performance, and impressiveness in a parallel process to
that of the transformation of the nang ma bars in urban Tibet. Eight troupes
representing townships of ’Dam gzhung County,43 part of the high and very poor
nomadic plains known as the Byang thang north of Lha sa, performed circle dance
(sgor gzhas),44 traditionally a group, participative dance accompanied only by the
singing of the dancers. The eight township groups that competed varied greatly in
their remoteness, one very close to Lha sa and touristic (Yangs pa chen), another
half an hour’s drive on the new good road from Lha sa (Gung thang), and at least
one very remote, without electricity in the villages and a long way from good roads
(Rgyal mtha’). The troupes were all amateur, made up of nomads, and generally
had started practicing for the competition just a couple of months in advance. The
competition rules were for each group to present three “new” circle dances, and
the group that performed and also looked the best would win, as judged by leaders
of the townships who vote for all the groups but their own. “New” was not defined
in detail, but at the least would involve a new melody, new lyrics, or new
choreography, albeit in existing/traditional style with existing dance steps. However,
most troupes presented many modifications of the local style in the line of the state
style (also by this time nang ma bar style) in bids to impress and be modern.
While many troupes presented one or more of their dances close to traditional
style, singing as they danced but with newly composed lyrics and/or tunes and
configurations of steps, all of them enhanced the style with an aesthetic of virtuosity
and show, a core part of the state style and its concept of “art” as opposed to the
43
The township (xiang) is the smallest administrative unit in China, consisting of several villages;
next is the county (rdzong, xian), consisting of several townships; then similarly the prefecture, then
the province, and then the nation.
44
Sgor gzhas is known in east and northeast Tibet as sgor bro.
“raw material” of “folk” music/dance. This involved higher hands and faster, more
virtuosic performances, with women jumping as well as the men, and accelerating
foot stamping which drew cheers from the crowd. There was also a move to
complexity and variety in performance, with sections of dance juxtaposed with
nomad songs (not traditionally a part of circle dance) performed by several groups.
In one case there was a narrative drama setting for the dance, and in another, a
solo cameo performance. However, none of the performers had been trained in
conservatoires, and there was no attempt to modify voices to bel canto or Chinese
style, which is difficult anyway to do without training. Similarly, despite high hand
moves, the really characteristic stretched ballet posture of the state style was also
absent, and there were no balletic moves but rather the heavy foot stamping
characteristic of nomad dance.45 The troupes wore matching clothes, but in the
sense of their best local clothes rather than special costumes, and brightly colored
floaty costumes were absent, apart from a little girl who danced solo in a very
kitsch outfit.46
Figure 6: Entry of dancers in Damshung horse

race 2004, dance competition.
The balletic postures, moves, brightly colored floaty costumes, and modified
vocal style can be seen making their presence felt strongly in amateur and
semi-professional county dance troupes,47 and are the norm by prefecture level.
In addition to the above adaptations of traditional parameters, a more
fundamental change was the separation of dance and song. Some troupes
pre-recorded their own singing of the dance song and played this as they danced
instead of singing themselves. In a bigger festival I went to (Nag chu Prefecture,
with performances by county troupes), two troupes who pre-recorded their own
singing cited concerns for audibility. At ’Dam gzhung, however, this does not
seem to have been a factor since groups sang live (which was very audible – the
45
This style of nomad dance characterized by heavy foot stamping seems to have been introduced
to communities of A mdo nomads in Gannan in the 1960s through Chinese-led propaganda troupes,
though whether it was an “invented tradition” (Eric J. Hobsbawn and Terence O. Ranger, ed, The
Invention of Tradition, Cambridge: Cambridge University Press, 1992) of that era or was imported to
A mdo from the Byang thang is not clear at present.
46
The hats in the picture, along with cowboy hats, are widespread in Tibet and are now effectively
local style, revealing the problems of searching for the “traditional” in Tibet.
47
I witnessed this at Nag chu Horserace Festival in 2004, with the performances of county troupes.
performance space was not very large and the groups were big) and also used
pre-recorded singing. All four groups I talked to, out of eight that performed at
’Dam gzhung, said they were used to singing and dancing together and felt more
comfortable doing so. The reasons for pre-recording seem to have been more to
do with wanting to focus on both singing and dancing separately to make sure that
both were good – the same division of labor logic of the state style, although the
state dance troupes do not pre-record the singing of the dancers.
Several troupes danced to pre-recorded commercial songs as opposed to their
own singing, exactly mirroring the state style. These groups appealed explicitly to
modernity in their choice to do so, stating in two cases that despite being
comfortable singing and dancing local songs together, they wanted to “keep up
with the new society.” One troupe, which I did not talk to, danced to a disco-style
song with disco-style steps, an obvious bid for modernity. Another danced to a
pre-recorded semi-traditional ’ba’ gzhas, a famous genre of Khams pa circle dance
from about one-thousand kilometers to the southeast of ’Dam gzhung, copying
exactly the steps on the VCD. They said that in addition to wanting to “keep up
with the new society,” they could not sing themselves because they did not know
the dialect of this song – a key reason for the separation of dance and song in an
increasingly widespread new practice of amateur/public/participative circle dance
in nang ma bars and town squares, described below.
One troupe, particularly keen to win the competition, had hired a new teacher
who had been trained in Lha sa’s Art School (rgyu rtsal slob grwa), hence drawing
explicitly on the professionalism of the state style as a way of enhancing and
improving their performance to put them a cut above the rest. This troupe’s last
dance was something of a tour de force, a lengthy tripartite performance that
involved simultaneous singing and dancing, dancing to a pop song, and dancing
to their voices pre-recorded, a kind of performance diagram of processes and
interpretations of modernity. The part with the pop song (a pop ’ba’ gzhas) involved
a little girl performing solo dressed in spangled, brightly colored clothes complete
with colorfully spotted tights, dancing non-local moves – a clear and somewhat
gimmicky bid for “modernity.” This was the dance they were most proud of, and
they had high hopes of winning.
The separation of song and dance, which I directly questioned the groups about,
was also interpreted in terms of modernity by two troupes that did not adopt it –
whether in the case of recording their own singing or dancing to songs from
commercial VCDs – stating that although they did not want to perform in this way,
they could see that it “fits well with the new society.” These two groups adopted
a firmly traditionalist stance. One was the group from the most remote township
without electricity in villages (Rgyal mtha’), which stated that they liked the
traditional ways and wanted to preserve them – that they didn’t mind if they won
or not, but wanted to preserve traditions. The other traditionalist group was Dbyangs
pa can, close to Lha sa, and the most economically developed and cosmopolitan
of the townships with famous hot springs which draw many tourists. They stated
that they were proud of their traditions, were comfortable with singing and dancing
together, and thought they got more in the spirit of the performance and gave a
better show overall, although it was harder work. One member, appealing to facts
of tradition in a slightly exasperated way, stated that “circle dance means dancing
and singing together.” Yangs pa chen presented one of their dances in a dramatic
setting, seeing this as a way of being original within a traditional framework, which
they interpreted as involving singing and dancing together and not using commercial
songs. It was interesting to note that Yangs pa chen was the only troupe who had
had their traditional dances valorized by outsiders (a Chinese film crew), which
had perhaps given them a vote of confidence from the modern world to have pride
in their traditions and hence to embark confidently on being traditional.
Economic Development, Being “Modern,” and Cultural

“Distinction”
This ascent, by the late 1990s, of the state style to the position of virtually
undisputed norm of Tibetan popular culture in dance, costume, and performance
style is surprising, and significantly different from the more radical musical culture
of the 1980s and early 1990s. As the cases of nang ma bars and the dance
competition at ’Dam gzhung Horserace Festival both show, there is now an
enthusiastic appropriation by the public of the state style reaching (albeit unevenly)
right down to a grassroots level. However, the state style, even when adopted in
toto in the case of nang ma bar dance via state-trained artists, is not, as far as I
have observed, seen generally by Tibetans as being the state style or sinicization,
but rather is seen as a modern, slick, urban, professional performance that is
eminently desirable.
As mentioned above, a major factor in this has been the softening of the state’s
image with its appropriation of new music fashions and less emphasis on overtly
political performances. However, this does not explain the impetus from the public
itself to see the state style as desirable modernity, leading to its appropriation by
the people rather than imposition on them. The reason for this crucial development
can be seen to emanate from a fundamental change in the socio-economic conditions
of Tibet.
From the mid-1990s to some extent, but most clearly from the turn of the
twenty-first century, there has been a dramatic increase in wealth in the “backward”
western provinces including all those containing Tibetan areas: the Tibet
Autonomous Region (TAR), Qinghai, Sichuan, Gansu, and Yunnan. This has been
caused by economic development strategies, most crucially the Western
Development Drive launched in 1999, which saw massive state investment resulting
in a (state subsidized) economic boom, with spectacular GDP growth in all the
Tibetan areas. In its wake, a new kind of middle class of wealthy Tibetans and
Chinese immigrants has grown up in Tibetan urban areas. The economic boom
has reached little into the rural areas and hence a marked urban-rural divide has
emerged, with rural Tibetans in particular marginalized in terms of wealth,

education, and healthcare.48
With this economic development reaching Tibet, albeit extremely unevenly,
there is an entirely new phenomenon of social opportunity, mobility, and aspiration:
a widespread sense of wanting to get rich and do well and its theoretical possibility.
Because the society of “old” Tibet was hierarchical, there was limited social
mobility. During the revolutionary period, however, these social stratifications
were eliminated and everyone became, in effect, a joint underclass, except those
few able to move to positions of some privilege through politics. Now, in addition
to the political structure, there is a (subsidized) capitalist hierarchical social order
in Tibet with unprecedented scope for social mobility through competition: a nomad
or farmer can rise to as much or more wealth as someone from a “good” urban
family or a party member if they are successful in business, although the odds are
in most cases stacked against them in terms of education and finance. With the
concentration of state-sponsored economic development in the big cities, such
rural Tibetans are increasingly migrating to cities in search of work. However,
although in theory there is the prospect of work and money, in practice most rural
Tibetans are singularly unequipped to find work in the cities in terms of the way
they look and dress, as well as in their lack of knowledge of good spoken and
written Chinese. Many in fact only know their local dialect of Tibetan, and thus
are initially unable to communicate with Tibetans and deal with even ordinary
tasks (buying things, finding a bus, and so forth). Rural Tibetans arrive in cities
generally after having traveled long distances, wear dirty, torn clothes, and simply
look like “fish out of water,” and urban Tibetans, urbanized rural Tibetans, and
Chinese immigrants all look down on them. Although there are strong support
networks of Tibetans from the same native area (pha yul) and of course relatives,
I have witnessed many cases of people who, whilst generously and energetically
helping and supporting new arrivals, want at the same time to keep a certain distance
from them.
48
Tibet Information Network, “Despite Economic Boom, Rural Standards of Living in the Tibet
Autonomous Region Still Below 1992 Levels” (originally published on Tibet Information Network,
“News Update,” February 6, 2003; site now discontinued, article currently unavailable); Tibet
Information Network, “Deciphering Economic Growth in the Tibet Autonomous Region.”
http://www.phayul.com/news/article.aspx?id=455&t=1&c=4 (originally published on Tibet Information
Network, “Special Report,” April 8, 2003; site now discontinued); Tibet Information Network, “The
Rich Get Richer, and the Poor? Rural Poverty and Inequality in Tibet – Indications from Recent Official
Surveys,” (originally published on Tibet Information Network, “News Update,” May 31, 2003; site
now discontinued, article currently unavailable); Tibet Information Network, “Unequal Competitors:
Despite Improvements in Education Levels, Skills of Tibetan Workforce Still Far Short of Immigrants”
(originally published on Tibet Information Network, “News Update,” February 23, 2004; site now
discontinued, article currently unavailable); Tibet Information Network, “Tibetans Lose Ground in
Public Sector Employment,” http://www.phayul.com/news/tools/print.aspx?id=8864&t=0 (originally
published on Tibet Information Network, “News Update,” January 20, 2005; site now discontinued);
Tibet Information Network, “High TAR Wages Benefit the Privileged,”
http://www.phayul.com/news/article.aspx?id=9033&t=1&c=1 (originally published on Tibet Information
Network, “News Update,” February 10, 2005; site now discontinued); Fischer, State Growth and Social
Exclusion.
While the state modernity of the 1950s promised development, improvement,

and “civilization” by “raising cultural standards,”49 it remained in the realm of
abstract ideology, as by the end of the Cultural Revolution, society had fallen into
chaos and the overwhelming majority of people into poverty. In today’s society,
however, virtually all Tibetans can see that wealth is there to be had, although only
a small minority actually get rich. In this context, one can see the concrete appeal
and relevance of a performance style which embodies and has long propagated an
ideology of development (of being a superior style, a style of higher standards and
aesthetics, a style of “proper” training, and so on), since there are now
socio-economic realities that can be conflated with this ideology. In other words,
this style, with its developmentalist ideology, forms a ready-made package for
processes of “distinction,” of expressing, demarcating, and creating social hierarchy
through culture.50 The state style therefore embodies, fuels, and helps create the
new dominant modernity of urban rather than rural, educated rather than illiterate,
professional/good/proper dance rather than unprofessional/sloppy dance, and
Chinese at least in addition to if not rather than Tibetan language – a sense of
modernity as pitted against “backwardness,” which emerges as
rural/village/illiterate/sloppy/traditional/Tibetan. Economic necessity or
competitive social aspiration means that once out of the village, people tend to
want to become modern in this sense, and now this is true even within the village,
due to recent rural penetration by the mass media, which itself is a result of the
economic boom. A few anecdotes illuminate this process of “distinction”:
1. I questioned a singer from a rural area of Khams in a nang ma bar in Lha

sa as to why he did not sing while he danced, the norm of traditional
Tibetan dances. He at first appealed to ideas of professionalism and
“proper” performance, saying that in the village they often make mistakes
in their singing or do not sing that well, and while it does not matter in
the village, it is not OK to do that in a nang ma bar when you are being
paid to perform. Hence, he concentrates on singing while he sings and
does not dance, though he does dance a little in interludes. He further
added that he did not want to look like a farmer when he performed,
singing and dancing songs together like they did in the village.
2. A very successful nang ma bar singer (who had trained and worked in
official institutions) had two friends from her village who were in Lha sa,
impoverished and desperate for work. Both have superb voices and are
excellent singers of unaccompanied nomad songs and in the male singer’s
case, rdung len. She wanted to help them and put a lot of energy into
coaching them, giving them advice on fashion and image as well as the
pop style and repertoire, all of which needed a lot of work. Their poor –
or in the female singer’s case, total lack of – Chinese language skills were
49
McDougall, Mao’s “Talk at the Yan’an Conference,” 68-69
50
Pierre Bourdieu, Distinction: A Social Critique of the Judgment of Taste (London: Routledge,
1986).
also a problem, since bad Chinese pronunciation immediately marks you

as “rural” and “backward” in a city. They therefore stuck to Tibetan songs
rather than singing in bad Chinese, although singing in Chinese is virtually
essential in nang ma bars in Lha sa. Along with a lack of familiarity with
the pop genres, knowing how to present on stage was one of the biggest
problems. The experienced singer remarked, exasperated, after the male
singer had a disappointing unsuccessful audition, that “he just looked like
a total nomad when he sang.” She clearly meant “nomad” in the sense of
rural/backward/unpolished, referring to his lack of ability to “present” the
song on stage in the standard dance troupe/nang ma bar style with hand
gestures and so forth. I later watched the female singer give what must
have been a similar audition in a nang ma bar in Chengdu, looking
awkward, nervous, and rather lost (not knowing what to do with her hands,
and so forth). The male singer has gained sufficient stage savvy through
practice in bottom-end nang ma bars and can sing well enough in Chinese
at least for smaller nang ma bars. The girl was taken under the wing of a
supportive small nang ma bar in Chengdu who recognized her potential
in terms of voice and good looks and trained her. She has since gone to
Beijing, where she is singing in a restaurant with good prospects. This
process of adapting to modernity and shaking off their rural identity –
which my friend emphasized again and again she herself had to go through,
but without the support – took around a year and a half.
3. In an A mdo village I visited, there was a sharp distinction between the
way groups of nomads who have been to school in the prefecture town
dance and the way those who have stayed in the village, largely illiterate,
dance. The former dance to pre-recorded songs or music, without singing
themselves, in dance troupe style (though without highly balletic moves
or posture) and in the dance troupe style “costumes”; the village nomads
dance in “traditional” style,51 singing as they dance with no
accompaniment, wearing best local clothes and ornaments. This dancing
took place in a show for Tibetan New Year, which was staged and had
people introducing the acts just like a nang ma bar or TV show; people
boasted that no other village in their valley could put on such a show as
theirs, and that their village was famous “far and wide.”
4. Echoing the sense of the singer from Khams who wanted to concentrate
on singing in the professional context of the nang ma bar rather than not
singing “properly” as in the village, a Tibetan I met told me how, although
he was seen as the best dancer of his village, when he went to work in a
nang ma bar in a tourist campsite with a coach trained in a Chinese
conservatoire, he was told he danced badly. His moves were too sloppy
51
This “traditional” style of nomad dance in this part of A mdo was in fact introduced by Chinese-led
propaganda troupes in the 1960s. Prior to this, people reported that there was in fact no dance in this
area; there were dances in neighboring Bla brang but the sgor bro type dances dated only from the
arrival of the Fifth ’Jam dbyangs bzhad pa in 1920 from Li thang (the gar dances seem to be older;
Morcom, “History, Traditions, Identities, and Nationalism).” See note 13.
and relaxed (the traditional Tibetan dance body language seen in

derogatory terms). He demonstrated how he was taught to dance “properly”
by this coach in a more sharp and exact way, meeting clearer lines with
his arm and leg movements, and more vertical and horizontal lines with
his arms (the “scientific,” ballet-influenced conservatoire style of dance).
5. A very talented singer of A mdo rdung len insisted she will not make a
rdung len album or sing with an also talented singer friend who sings
rdung len, but will persist in making a pop album (which she also sings
extremely well). I expressed regret at the waste of her talent in rdung len
singing and lack of loyalty for rdung len with all her energy going into
pop, but we discussed how it would of course spoil her image as a pop
singer, since the audience may not accept her if she had already established
a rural and “backward” identity through singing rdung len and singing
with a rdung len singer. Once successful in her pop career, we discussed,
she could then sing whatever she wanted, and she said she would like to
sing rdung len.
With dance, costumes, and performance style (but song and singing style to a
lesser degree, as I explain below), these modernizing trends are very consistently
mapped according to the hierarchies of officialdom, with a steady increase in
elements of the state style as you move from village or township (as in ’Dam
gzhung), to county, to prefecture (by which time you reach a total application of
the style in dance), to province, to nation, from which the style emanates. The style
is also mapped onto hierarchies of economy, smaller to larger/better established
troupes, poorer to richer/more ambitious villages, and smaller/cheaper compared
to larger/more expensive/flashier nang ma bars. There are exceptions, but the trend
is very clear in all the places I visited, illustrating this overarching socio-economic
and cultural hierarchy.
While “old” Tibet had social hierarchies, musical culture was divided more
horizontally or geographically into loosely connected regional styles and genres.52
The arrival of the Chinese brought the first widespread vertical division of musical
culture, essentially an imperialist one, which established the new Han-created
pan-Tibetan style as developed in contrast to traditional genres as backward.53
However, whilst the cultural division of the revolutionary period was essentially
52
Henrion-Dourcy notes that although both Beijing and Dharamsala present a divide between a
popular culture and a literate culture in their representations of “old” Tibet, “while holding some truth,
it is a caricatured depiction of the past.” She goes on to say,
As far as my experience in the Tibetan performing arts goes, this polarization appears more in
discourse than in the more interwoven practice. Whether as sponsors, amateurs or even performers,
high-ranking religious figures would indulge in, for example, lha mo or other songs and dances,
with absolutely no second thought for partaking in this amusement (Henrion-Dourcy, “Women in
the Performing Arts,” 198-99 n. 11).
53
See Appadurai, Modernity at Large, and Arjun Appadurai, “Grassroots Globalization and the
Research Imagination,” in “Globalization,” special issue, Public Culture 12, no. 1 (2000): 1-20, for a
critique of the concept of culture as “geographical traits” in the light of modernity. This point is discussed
further below.
between Tibetan and Chinese, that of the liberalized and post-economic boom
period is revealing more inter-Tibetan stratifications with the pan-Tibetan
urban-rural divide. In particular, the wealthiest and most educated urban Tibetans
(educated in China and more fluent in Chinese than Tibetan), and even
well-educated middle-class urban Tibetans, in many ways have more in common
socio-economically and aesthetically with Chinese elites than with rural Tibetans.
In other words, this aspect of cultural change comes as much from inter-Tibetan
dynamics as from Chinese domination over Tibetans, the predicament of the
revolutionary period, and is far more powerful in its penetration of Tibetan society.
Ironically and paradoxically, large chunks of the state style and effectively its
core developmentalist ideology are being cemented into social consciousness at a
deep level through the class system resulting from free market reforms and
capitalism as opposed to socialism, which sought to abolish those very hierarchies.
The style and much of the ideology of the dance of the communist era seems to
have found a genuine home in the capitalist era. This is, however, not so ironic
when one recalls that the ideas of “science” and “progress” adopted in the reforms
of late nineteenth/early twentieth-century China, and later seized on by the Chinese
communists from the 1950s, are the same as those of bourgeois/elite Europe, and
furthermore, the Chinese conservatoires were modeled on the ideology of intensive
training and divided labor of the European conservatoires. This loop back to the
bourgeois/elite European origins of both the ideology and much of the content of
the state style also brings the state style in line with the now international consensus
of slick, professional, staged/audience-oriented performance on MTV, mainstream
pop albums, big concert halls, etc., the strongest institutional strand of which is
the European tradition. This global hegemony is a further reason for the strong
desirability of aspects of the state style and state trained professionals in the period
since 2000, which has seen global culture enter Tibet.
There are two major forces that run counter to this form of modernity. One is
globalization, discussed below. The other is the traditionalist tropes, seen in ’Dam
gzhung, that break the association of traditional song and dance with backwardness:
although traditional music and dance is unmodernized in the sense of the
modernizations we have been talking about, this does not result in it being perceived
as backward, but rather as something good and to take pride in. Traditionalism in
Tibet always has the potential danger of being seen as latent separatism. Although
it is not actually illegal or stamped out in principle, it is rarely supported by the
government on a large level.
Such traditionalism can be seen at a grassroots level (’Dam gzhung), and also
at higher levels. I once asked a top Tibetan composer why dance troupes never
sing and dance simultaneously even though that is a constant of all Tibetan
traditional dance. He replied surprisingly bluntly (he was drunk at the time), “That’s
because of the Chinese,” and then added, “It’s going to change.” He also launched
into a rapturous appreciation of Tibetan traditional music, and commented that he
was not proud of much that he had done (i.e., all his modern style compositions).
What will happen in this space remains to be seen. In the dance competition in
’Dam gzhung, the group that tried so hard to be modern, finishing with the tripartite
piece, did not even finish first or second. The winner was the most traditional
troupe. The rules were for marks to be awarded by the jury of township leaders on
the grounds of quality of performance, with no preference for style. Although it is
not possible to tell if there was an element of marking up more traditional
performances, the modern performances certainly failed to score on modernity
alone. It will be interesting to see how future competitions fare. In ’Ba’ thang,
where there is strong support for the local dance ’ba’ gzhas, famous across Tibet
and also relatively well-known in China, in the dance competitions (organized by
the county authorities), more marks are given for traditional dances or
traditional-style dances. However, this is the only example I have seen of dance
in a form so far towards the traditional end of the modified-traditional continuum
being preserved by the authorities. Other dances gain much support from local
governments, such as Dkar mdzes rdo bro and Yul shul dance, but largely in
modified forms.54 The Lha sa and Gzhis ka rtse traditions of Tibetan lute songs
and dances, including nang ma and stod gzhas, were taught at Tibet University for
some years after the end of the Cultural Revolution but are no longer taught.
Western instruments such as piano and violin are taught currently. The gar
repertoire of the Dalai Lama’s court had an even shorter revival.55
Assertions of traditionalism or Tibetan identity, though powerful, may also be
applied very unevenly, relying as they do on symbols as much as facts of history.
In 2005 there was a big festival in the A mdo area of Rma chu County (Gannan
Prefecture, Gansu Province), where, as far as I know, for the first time in a Tibetan
area, everyone had to sing in Tibetan, even some nationwide stars who virtually
always sing in Chinese (for example G.yung drung rgyal). The rapidly increasing
tendency of Tibetan singers to sing in Chinese, not Tibetan, is beginning to be
noticed by people across Tibet as an undesirable trend of modernization, and
assertions of Tibetan identity are beginning to be made to counter it. Another
favorite symbol of ethnicity, a musical instrument, is also being seized upon in A
mdo as an expression of Tibetan identity, with Rma chu dance troupe now only
allowing people to play the Tibetan lute rather than the ubiquitous mandolin. This
is soon to become a rule in the biggest A mdo-wide rdung len competition.
However, the dance performance by Rma chu dance troupe that formed a backdrop
to Tibetan language songs and some performances on the Tibetan lute at the festival
in 2005 was all mainstream state style with bright, floaty costumes. There were
54
Dkar mdzes rdo bro is performed in the county to a pre-recorded, big orchestral accompaniment.
In the villages it is presumably performed without accompaniment and is much more like the sgra
snyan dance traditions of Lha sa, which is what it has in fact been developed from (Morcom, “History,
Traditions, Identities, and Nationalism”). Yul shul dance, having become significantly faster and more
virtuosic, is now performed, even by many small, rural troupes, to the accompaniment of separate
singers and musicians, rather than with the dancers singing themselves. In Yul shul Prefecture in
northern Khams, Qinghai Province, a drive to speed and virtuosity in the famous local dance style has
led to an established norm of the dancers not singing, but dancing to the accompaniment of separate
singers and musicians.
55
See Unity and Discord, 50-51.
also a number of singers performing in Chinese or even bel canto vocal styles,
albeit in Tibetan language.
The Rma chu festival focus on language and instruments but not dance as an
expression of Tibetanness also reveals another important trend in the modernizing
of Tibetan musical culture: whilst the state style of dance is with few exceptions
being adopted at all levels, singing is a more complex situation, representing much
more localization and new styles that come from the people and regional traditions
and are appropriated by the state rather than the other way round. A few troupes
are now not sending singers for training in Chinese schools. Dancers of these
troupes, however, are all sent to China for training. Rdo re of Rma chu dance troupe
commented that with dance you can learn the conservatoire style and still be able
to dance the traditional Tibetan way, but the vocal training is “one way,” or worse
still, some singers are left unable to sing at all.56 There is also pride in the vocal
style of particularly the nomadic Tibetan areas where traditional voices are
extremely powerful, capable of intricate ornamentation, and have very wide ranges.
Though beyond the scope of this article, the processes of modernization in A mdo,
which has countered the pan-Tibetan trend in many ways (though not in dance),
is key to these assertions of Tibetan musical and language identity. However,
although vocal and song style involves more traditionalist, Tibetan-identity driven
motivation, the need to be able to present in the right way on stage is absolutely
essential (as we saw with the two A mdo singers being groomed for nang ma bar
performance) and the dance troupe style still leads here. This lack of traditionalist
focus on dance except in a few cases probably represents the lack of attention given
to dance in Tibetan musical culture as a whole: where good singers are admired,
the ability to dance is very much taken for granted.57
Mass Media, Migration, and Social Heterogeneity

The above discussion has focused on the adoption of the “state style” in popular
and rural dance through the conflation of economic, social, and taste hierarchies
in the public as well as the new cultural inclusiveness of the state, and facilitated
by increased penetration of the mass media and the visual media from the late
1990s. I would like to turn now to another form of dance in Tibet to look at how
aspects of the state style are being cemented by other features of modernity.
56
The famous a ce lha mo singer A ma lhag pa (1909-97) lost her voice permanently after being sent
to Shanghai for vocal training (Henrion-Dourcy, “Women in the Performing Arts,” 213). I also met a
talented singer who has lost a powerful and versatile voice after being taught how to sing “properly”
in the “artistic” and bel canto style in a Chinese academy. In the performing arts world, most people
know or know of such vocal casualties. The famous singer Bsod nams dbang mo, who shot to fame in
a nationwide competition in 2002 singing a modern arrangement of a local nomad song but in a
traditional nomad voice, has since joined the army song and dance troupe in Beijing, where her voice
has become modified; I have heard people say – and say that “people say” – that her voice is now not
as good as it was before.
57
I am grateful to Isabelle Henrion-Dourcy for drawing my attention to this key point (personal
communication, February 2005).
We have been looking at adaptations of circle dance by troupes in formal

performances. Circle dance as it was before the 1950s is much more participative,
performed by amateur members of the community for themselves; the whole sense
of a chosen troupe competing, let alone performing on stage, however “traditional”
or unmodernized their style, is a modern change, again echoing the dance troupe
norm of professionalizing performance. However, a public, participative,
non-audience-oriented but modern form of circle dance also exists and is a standard
of nang ma bars, dating at least from 2000, when the VCD of “a ba sgor bro”58 –
still the most widely known and danced circle dance (sgor bro) of nang ma bars
– was released.59 It is also danced in town squares in Dkar mdzes prefecture
(Khams), Chengdu (Shuanlu County, initiated by the Khams pa Tibetan population),
Xining, and probably other areas too, first beginning, it seems, in Chengdu around
1998.
Stylistically, this participative circle
dance also follows the state model of
dancing to pre-recorded commercial
songs (often recorded by dance troupes)
and not singing. Here, however, the
separation of singing and dancing is
certainly not related to concerns over
professionalism/show/slick
performance/virtuosity so important to
the rural troupes discussed above, since
Figure 7a: Sgor bro in Shuanlu county square, hilariously bad dancing is quite
Chengdu, August 2005.
acceptable! The ethos of this form of
circle dance is in many ways opposed to the dances of the professional troupes.
One of the promoters of public circle dance in Kangding (Dkar mdzes Prefecture),
when I asked her if she looked to dance troupes for ideas, said that although she
liked what they did, they were distant from the public in their professionalism, and
she liked to promote the public, participative circle dance, which anyone could
join in. Ideas of community-building and keeping old people in particular from
“just sitting around playing mahjong” were all key ideas expressed by the people
in Kangding and Chengdu who started the circle dance crazes there. These circle
dances as danced in town squares are also a hybrid phenomenon of Tibetan dance
with the Chinese practice of mass exercise, another state legacy. Many older people
participate in public circle dance as a form of fun, socializing, and gentle exercise,
58
Bod dang chang rigs kyi sgor bro, VCD ISRC CN-G12-00-012-00/V.J6, 2000.
59
Other public dance in nang ma bars includes disco; waltzing to any song with the right kind of
tempo (I saw people waltzing to a slow ballad by the trendy Lha sa rock band Gzi mig dgu pa as they
performed in a bar in Lha sa in summer 2004); and step dancing routines, which work with 4-by-4 beat
pop songs though not with Tibetan styles, as this would evoke sgor gzhas (Britney Spears’ “Hit Me
Baby One More Time” worked well, for instance). While it is beyond the scope of this article, it can
therefore be said that there are a number of forms of public, participative dance in music venues in
Tibet.
though younger people also dance. There are also many Chinese people who dance
in Kangding, Chengdu and Xining in particular.
The separation of singing and dancing in public circle dance can be seen most
obviously as emerging as part of the package of dancing to pop songs à la dance
troupes out of the hegemonic appeal of modernity and the will not to look backward
in the urban contexts of town squares and nang ma bars. In the nang ma bars in
particular, as stated above, the dance troupe model is now the unquestioned norm
of dance performance, and anything else would be surprising. Even with the public
circle dance in nang ma bars, at least some of the paid performers lead, ensuring
a critical mass of dancers who know the steps and thus fusing the public circle
dance to the staged circle dance performances.60
Figure 7b: Sgor bro in Dar rtse mdo square,

Dkar mdzes Prefecture, May 2005.
Figure 7c: Sgor bro in Lithang county square,

Dkar mdzes Prefecture, May 2005.
However, there is further reason for the lack of simultaneous singing and dancing
in public circle dance that relates to a key aspect of modernity other than the musical
stylistic and aesthetic modernity set by the state. As Appadurai describes, the
correlation of place, ethnicity, language, and culture is being weakened in global
modernity. With increased movement of people across and within national and
other borders, heterogeneous communities are increasingly common, as are
60
The county dance troupe also leads the public sgor gzhas in Li thang, and members have to dance.
The public sgor gzhas in Kangding and Chengdu, however, are not related to official dance troupes.
diasporas and refugees.61 Chinese immigration into Tibet increased following the
lifting of restrictions on immigration in 1994,62 and it further intensified beginning
in 1999 with the Western Development Drive, accompanied by large numbers of
rural Tibetans migrating from different regions to Tibetan cities. Thus with public,
urban circle dance, the entire group is usually unable to sing along since the songs
are not in everyone’s dialect or native singing style. This delinking of people, place,
and style is made stronger by the mass media, where, as we saw in the case of the
’ba’ gzhas danced in ’Dam gzhung, people have access to music and dance (through
the audio-visual VCD) from far away regions.
This delinking, reconfiguration and
hybridization of people, place, identity,
style and language is a feature of
modernity and globalization in general.
However, it is even more sharply a fact
of Tibetan modernity due to the rapid
influx to Tibetan areas of Chinese
immigrants who neither know nor
subsequently learn Tibetan, arriving to
make money from the economic boom
which has, along with the poor level of Figure 7d: Sgor bro in nang ma bar in Gannan
education in Tibet, resulted in a greater prefecture town, June 2005.
demand for skilled labor than local supply can fill.63 Speaking Chinese is now an
economic and practical necessity in Tibetan cities, while speaking Tibetan is not.
Furthermore, the vast majority of Chinese migrants to Tibetan areas do not pick
up the local language due to still prevalent cultural attitudes toward minorities as
inferior/backward and so forth, compounded by zero pressure from the government
for Chinese migrants in Tibetan areas to learn the indigenous language.64 Therefore,
Chinese participants in circle dance (and there are many, especially in town squares)
cannot sing along.
The continuing de-emphasis of Tibetan language in education in Tibetan areas,
where Tibetan is generally a subject taught in a Chinese-medium school, has also
resulted in the phenomenon of many well-educated Tibetans being more comfortable
61
Appadurai, Modernity at Large, 1-23. Ramnarine critiques the trope of a geographically or ethnically
bounded concept of “culture” in her article on music and “diasporic imagination” in Trinidad (Tina
Ramnarine, “Music in the Diasporic Imagination and the Performance of Cultural (Dis)placement in
Trinidad,” in Island Musics, ed. Kevin Dawe (Oxford and New York: Berg, 2004), 166-67).
62
See Tibet Information Network, trans., “Third Work Forum on Tibet,” 1994, translated excerpts
available on http://www.tibetjustice.org/materials/china/china10.html.
63
Chinese immigration into Tibet is also desired by the government for reasons of stability. I was
told by a businessman setting up a factory in Lha sa that he thought he could get away with hiring 80
percent Tibetans. I asked if it would be okay to hire 100 percent and he said no. I asked if it would be
okay to hire 100 percent Chinese and he said “absolutely no problem.”
64
In contrast, Indians and Nepalese living in or working in Tibetan exile communities speak at least
some Tibetan.
with Chinese.65 The lack of knowledge of written Tibetan in particular makes it

far more difficult for Tibetans to pick up other Tibetan dialects, since it is the
written language which is the most important common link between the different
dialects. Hence, it becomes easier to use Chinese, even amongst Tibetans who do
not like doing so because of concerns for Tibetan language and identity.66
Another parallel factor to dialect and language problems that compounds the
inability of groups to sing to public circle dance is the fact that the circle dance
craze is very much a Tibetan phenomenon, and the songs that are thought
appropriate for dancing circle dance are virtually always Tibetan songs – generally
songs with a strong ethnic flavor which is brought about through closeness to
regional singing styles. The songs are grounded in particular regional vocal styles
which Tibetans from other areas tend not to identify with, and are often rural styles
which urban Tibetans tend not to identify with in terms of actually singing. In some
cases, such as with Nag chu circle dance and pop rdung len (also danced as circle
dance), it is virtually impossible to sing along with the distinct vocal style of the
songs unless one has grown up as a nomad. One sometimes sees people in nang
ma bars singing along to certain songs from their native area (Nag chu or Khams
pa ’ba’ gzhas, for example), or to some particularly famous songs.
There is certainly a sense of loss or breakdown of Tibetan community in the
voiceless dancing of public circle dance.67 However, public circle dance is also a
strong show of Tibetan-ness, and a display of the variety and richness of Tibet’s
regional dance songs, despite centralizing/homogenizing tendencies such as a
significant degree of standardization of dance steps in dances from different regions.
In Kangding, one of the organizers of the public circle dance explained that they
seek to collect a variety of dances from different areas of Dkar mdzes Prefecture
(of which Kangding is the capital), learning dances from people from that region.
She also said they were planning to send people to certain villages to learn particular
dances. Dances are also learned from VCDs of dance troupe performances, and
pop circle dance has become popular as a result of, and in turn catering to, the
craze in town squares and nang ma bars.
The dance moves of public circle dance are also very Tibetan, though with
extensive mixing and matching of moves from different regions and newly-created
Tibetan-style moves. There are generally no balletic movements, the posture is the
65
There is also a strong tendency of educated Tibetans to speak in Chinese to show off their education
and status, in a parallel situation to the heavy use of English, the language of the (former) rulers, in
contemporary India. Tibetan language use amongst intellectuals in A mdo is far stronger than in central
Tibet and Khams due to more focus on Tibetan language in schools there.
66
Pronunciation in different dialects varies drastically, but it is consistent with spelling, so if the
spelling of a word is known, so is the phonetic transformations of it in different dialects. Literary
Tibetan also contains much of the localized vocabulary of different dialects. The difference between
Lha sa and a Khams pa or an A mdo dialect is roughly equivalent to perhaps French and Italian or
Spanish, or Hindi and Gujurati or Punjabi, initially largely mutually unintelligible and normally taking
a few months to acquire basic communication skills.
67
Singing while dancing is the norm of both the professionals and the public in sgor gzhas in exile,
which represents a much more homogenous and homogenized central-Tibetan community.
relaxed stance of traditional Tibetan dance, and the body language of the moves
is very Tibetan, apart from Chinese participants who are always easy to spot.68
Some of the professional dancers in nang ma bars can also be seen to code-switch
into a more relaxed dance posture for dancing circle dance than their formally
performed dances. There are other developments too, such as dancing the Tibetan
steps with more of a disco feel; subdividing the pop beats of the Tibetan songs in
ways that do not exist in traditional Tibetan dance; and a certain amount of humor
and fun, such as that arising from “funny dancing,” in which untrained and (in the
case of nang ma bars) often drunk individuals dance, or, as I have seen on a few
occasions, dancing circle dance to non-circle dance pop songs (in any language).
Global Modernities and a New Order of the Struggle for

Hegemony
We have seen a widespread acceptance of many aspects of the state style due to
the new cultural inclusiveness of the state, and also upward mobility and social
aspiration in the realm of the public. This process has involved a struggle for
hegemony between people and state. Since the turn of the twenty-first century,
however, a steady flow of global musical and dance modernities has reached Tibet
via the global media and migration of people, and a hegemonic struggle of a new
order has therefore begun with forces above the state (in geo-political and economic
terms) and also more lateral to it (such as India and Nepal). The state, ever aware
of the social and political implications of style and identity in music, continues to
accept/appropriate globalized forms into dance troupe performance and TV variety
shows. However, there are some interesting reconfigurations of song and dance
emerging that do suggest change of a new order again, and new possibilities for
the ways modernity is conceived.
Figure 8a: PVC trousers and Tibetan shirt in Figure 8b: Gold PVC suit, compere in nang ma
nang ma bar, Lhasa. bar, Nag chu, August 2004.
68
The only place I have seen a very un-Tibetan feel to sgor gzhas in terms of songs and dance was
in Chengdu (Shuanlu County), where a very large dancing group had been started by a party-loving
Tibetan-Chinese woman with overwhelmingly Chinese participants, and many Chinese language
propaganda songs were played. The dance was far more hybrid, involving very dance-troupe style
balletic moves and many “uplifting” moves. There were, however, more typically Tibetan groups in
smaller squares nearby.
One trend, seen across much of the “developing” world, is the resurgence of
ethnic sounds, though in a form combined with global sounds, thereby overcoming
associations of backwardness. There are also moves toward the recoupling of
Tibetan song and dance, the separation of which is one of the most ingrained
legacies of the state style. Dbyangs can lha mo69 shot to fame around 2003 with
her pop ’ba’ gzhas albums, on which she sings and dances accompanied by a troupe
of dancers who do not sing but dance with abandon in a Tibetan landscape. The
dancers perform largely in a circle, and she sometimes performs with them and
sometimes in the center (more the soloistic star); shots of Dbyangs can lha mo in
various Tibetan scenes (monasteries, Tibetan homes, with sheep, yak, and so forth)
are also intercut with the hillside scene. In the wake of these massively successful
albums, themselves almost certainly building on the earlier circle dance craze of
a ba sgor bro-type circle dance, more similar albums have been released from
other areas of Tibet. The public circle dance craze, in a sense a craze for Tibetan
ethnic songs, is also a feature of this globalization of Tibetan sounds, and these
public circle dance or similar songs may also be performed solo in nang ma bars
and festivals by singers who dance a bit in the interludes, as explained above. A
new look also seems to be emerging on the lines of this new sound, less Chinese
kitsch and more a fusion of Tibetan with western, with a heavier look of Tibetan
traditional textiles and deeper colors mixed also with fabrics like PVC.
Many of the newest pop stars owe as much to Afro-American performance style
as to dance troupe style, and with this, there is increased importance on dance in
performance and another kind of recoupling of song and dance. There are the
highly choreographed girl bands and now boy bands, the girl bands in particular
involving some Tibetan style moves. These bands tend to adopt a highly ethnic
but globally hip and sexy image (for example, off-the-shoulder Tibetan dresses),
but there is still a lot of kitsch.
Figure 9a: Girl band in nang ma bar, Lhasa, Figure 9b: Girl band perform in festival in Dar
October 2005. rtse mdo, May 2005.
There are also solo artists who dance in a more western pop/soul/rap style as a
part of singing performance, such as Chung zhol sgrol ma, who launched a female
soul style of singing in Tibet. These global dance (and vocal) styles also undermine
69
Not to be confused with Real World’s Dbyangs can lha mo, who is from the exile community
(though born in Tibet).
the light and stretched balletic posture and aesthetic of the dance troupes, since
they adopt a relaxed posture in fact more in line with that of traditional Tibetan
dance. For one of the really top-end nang ma bars, there is now an expectation of
something more than the conventional static performance style (which, indeed, is
becoming sidelined in some dance troupes too, and especially on Chinese television
where performers move and dance a lot now). One of the performers who integrated
song and dance most and never failed to get the most enthusiastic reception in
nang ma bars was a Tibetan girl who performed Britney Spears and Tibetan and
Chinese pop full of moves learned from Britney Spears and Jennifer Lopez videos
(which are widely available on VCD).
So while it may be perceived as backward for a singer to dance and sing together
“like a farmer” in a bar in Lha sa, when there is the polished, global hip of some
of these styles, there is absolutely no question of backwardness whatsoever, but
rather a new sense of ethnic identity and another level to the hierarchy of taste.
But to keep the new upsurge of indigenous sounds in context it must be emphasized
again: while ethnic is trendy, traditional is not. Rural Tibetans, however good their
voices are and however good-looking they are, will not get work in a nang ma bar
unless they can adapt to the modern, urban, staged, slick performance style, whether
dance troupe or global. In an echo back to Mao, such performers can at best be
seen as “raw material,” to be trained, changed, “developed,” and made acceptable
for the new modernity of twenty-first century Tibet.
Glossary
Ka
kong po Kongpo Place
dkar mdzes Kandzé Place
dkar mdzes rdo bro Kandzé dodro Term
bkra shis nyi ma Trashi Nyima Person
bkra shis tshe ring Trashi Tsering Person
skyor mo lung Kyormolung Organization
Kha
khams Kham Place
khams pa Khampa Ethnicity
Ga
gar gar Term
gung thang Gungtang Place
glu gar tshogs pa Lugar Tsokpa Organization
dge ’dun chos ’phel Gendün Chömpel Person
'gram Dram Place
rgyal mtha’ Gyelta Place
rgyu rtsal slob grwa Gyutsel Lapdra Art School Organization
sgor bro gordro circle dance Term
sgor gzhas gorzhé circle dance Term
sgra snyan dranyen Tibetan lute Term
Cha
chung zhol sgrol ma Chungzhöl Drölma Person
Ja
’jam dbyangs bzhad Jamyang Zhepa Person
pa
Ta
stod gzhas tözhé Term
bstan ’dzin Tendzin Person
Tha
thang go la Tanggola Building
Da
dar rtse mdo Dartsedo Chi. Kangding Place
bdud be Dübé Person
bde chen dbang mo Dechen Wangmo Person
’dam gzhung Damzhung Place
’dres ma drema mixed Term
rdung len dunglen Term
rdo re Doré Person
Na
nag chu Nakchu Place
nang ma nangma nightclub Term
nang ma’i skyid sdug Nangmé Kyiduk Organization
nam mkha’ Namkha Person
Pa
dpal sgron lags Peldrönla Person
Pha
pha yul payül native area Term
Ba
ba pa phun tshogs Bapa Püntsok Person
dbang rgyal Wanggyel
byang thang Jangtang Place
byams pa tshe ring Jampa Tsering Person
bla brang Labrang Place
dbus gtsang Ütsang Place
dbyangs can lha mo Yangchen Lhamo Person
’ba’ thang Batang Place
’ba’ gzhas bazhé Term
’brog glu droklu nomad song Term
Ma
rma chu Machu Place
Tsa
brtse thang Tsetang Place
Tsha
tshe dbang Tsewang Person
Dza
rdza chu kha Dzachukha Place
rdzong dzong county Chi. xian Term
Zha
zho ston Zhotön Yogurt Festival Festival
gzhis ka rtse Zhikatsé Place
Za
zla sgron Dadrön Person
zlos gar tshogs pa Dögar Tsokpa Organization
gzi mig dgu pa Zimik Gupa Organization
Ya
yangs pa chen Yangpachen Place
yar ’drong Yardrong Person
yul shul Yülshül Place
g.yu sgron Yudrön Person
g.yung drung rgyal Yungdrung Argya Person
Ra
ral pa relpa Term
La
li thang Litang Place
blo bzang tshe ring Lopzang Tsering Person
Sa
gser shul Sershül Place
bsod nams dbang mo Sönam Wangmo Person
Ha
lha mo lhamo Term
lha mo tshogs pa Lhamo Tsokpa Organization
lha sa Lhasa Place
lho kha Lhokha Place
A
a ce lha mo Aché Lhamo Term
a mdo Amdo Place

a mdo snga pa Amdo Ngapa Place
a ba sgor bro aba gordro Term
a ma lhag pa Ama Lhakpa 1909-1997 Person
Chinese
Beijing Place
Chengdu Place
Gannan Place
Gansu Place
Jiang Zemin Person
Luding Place
Qinghai Place
shang shang xiang Term
Shuanlu Place
Sichuan Place
Great Development xibu da kaifa Term
of the West
Xining Place
Bibliography
A blo. A blo spun mched kyi rnam thar. Beijing: Mi rigs dpe skrun khang, 1994.
Adams, Vincanne. “Karaoke as Modern Lhasa, Tibet: Western Encounters with
Cultural Politics.” Cultural Anthropology 11, no. 4 (1996): 510–46.
Anonymous. Unity and Discord: Music and Politics in Contemporary Tibet.
London: Tibet Information Network, 2004.
Appadurai, Arjun. Modernity at Large: Cultural Dimensions of Globalization.
Minneapolis: University of Minnesota Press, 1996.
———. “Grassroots Globalization and the Research Imagination.” In
“Globalization.” Special issue, Public Culture 12, no. 1 (2000): 1-20.
Baranovitch, Nimrod. China’s New Voices: Popular Music, Ethnicity, Gender
and Politics, 1978-1997. Berkeley, CA: University of California Press, 2003.
Bod dang chang rigs kyi sgor bro. VCD ISRC CN-G12-00-012-00/V.J6, 2000.
Bourdieu, Pierre. Distinction: A Social Critique of the Judgment of Taste. London:
Routledge, 1986.
Dodin, Thierry, and Heinz Räther, eds. Imagining Tibet: Perceptions, Projections
and Fantasies. Boston: Wisdom Publications, 2001.
Fischer, Andrew. State Growth and Social Exclusion in Tibet: Challenges in
Recent Economic Growth. Copenhagen: Nordic Institute of Asian Studies,
2005.
Freedman, Carl. The Incomplete Projects: Marxism, Modernity, and the Politics
of Culture. Middletown: Wesleyan University Press, 2002.
Giddens, Anthony. The Consequences of Modernity. Cambridge: Polity Press,
1990, reprint 2005.
Goldstein, Melvyn. A History of Modern Tibet, 1913-1951: The Demise of the
Lamaist State. Berkeley, CA: University of California Press, 1989.
Goldstein, Melvyn, Dawei Sherap, and William Siebenschuh. A Tibetan
Revolutionary: The Life and Times of Bapa Phüntsok Wangyal. Berkeley, CA:
University of California Press, 2004.
Goldstein, Melvyn, William Siebenschuh, and Tashi Tsering. The Struggle for
Modern Tibet: The Autobiography of Tashi Tsering. New York: M.E. Sharpe,
1997.
Harris, Rachel. Singing the Village: Music, Memory and Ritual Among the Sibe
of Xinjiang. Oxford and New York: Oxford University Press for the British
Academy, 2004.
Heberer, Thomas. “Old Tibet a Hell on Earth? The Myth of Tibet and Tibetans
in Chinese Art and Propaganda” In Imagining Tibet: Perceptions, Projections
and Fantasies, edited by Thierry Dodin and Heinz Räther. Boston: Wisdom
Publications, 2001.
Henrion-Dourcy, Isabelle. “Women in the Performing Arts: Portraits of Six
Contemporary Singers.” In Women in Tibet, Past and Present, 195-258.
London: C. Hurst, 2005.
———. Ache Lhamo: Jeux et enjeux d’une tradition théâtrale tibétaine. PhD
thesis, Université Libre de Bruxelles, 2004.
Henrion-Dourcy, Isabelle and Dhondup, T. “Tibet, §III, 1. Folk music.” In New
Grove Dictionary of Music and Musicians, edited by Stanley Sadie and John
Tyrrell. 2nd edition, 449-52. London: Macmillan, 2001.
Hobsbawn, Eric J., and Terence O. Ranger, eds. The Invention of Tradition.
Cambridge: Cambridge University Press, 1992.
Kuo-Huang, Han. “The Introduction of Western Music in Modern Times.” In The
Garland Encyclopaedia of World Music. Vol. 7, East Asia: China, Japan, and
Korea, edited by Robert Provine, Yoshihiko Tomumaru and J. Lawrence
Witzeben, 373-77. New York: Garland Pub., 2002.
Morcom, Anna. “History, Traditions, Identities, and Nationalism: Drawing and
Redrawing the Musical Cultural Map of Tibet.” PIATS 2006 (forthcoming).
———. “Getting Heard in Tibet: Music, Media and Markets.” In Consumption,
Markets and Culture. Forthcoming.
McDougall, Bonnie. Mao Zedong’s “Talk at the Yan’an Conference on Literature
and Art”: A Translation of the 1943 Text with Commentary. Ann Arbor, MI:
Center for Chinese Studies, University of Michigan, 1980.
Nooshin, Laudan. “Circumnavigation with a Difference? Music, Representation
and the Disney Experience: It’s a Small, Small World.” Ethnomusicology
Forum 12, no. 2 (2004): 236-51.
Pathak, Avijit. Modernity, Globalization and Identity: Towards a Reflexive Quest.
Delhi: Aakar Books, 2006.
Sprin gyi me tog. Anna Morcom. VCD ISRC CN-H01-06-320-00/V, 2006.
Ramnarine, Tina. “Music in the Diasporic Imagination and the Performance of
Cultural (Dis)placement in Trinidad.” In Island Musics, edited by Kevin Dawe,
153-70. Oxford and New York: Berg, 2004.
Rees, Helen. Echoes of History: Naxi Music in Modern China. Oxford: Oxford
Shakya, Tsering. The Dragon in the Land of Snows: A History of Modern Tibet
Since 1947. London: Pimlico, 1999.
Stoddard, Heather. Le mendiant de l’Amdo. Paris: Société d’ethnographie, 1985.

Thompson, John. The Media and Modernity: A Social Theory of the Media.
Cambridge: Polity Press, 1995.
Tibet Information Network. Tibet 2003: A Yearbook. London: Tibet Information
Network, 2004.
———. Tibet 2002: A Yearbook. London: Tibet Information Network, 2003.
———. News Review: Reports from Tibet, 2001. London: Tibet Information
Network, 2002.
Network, 2001.
———. China’s Great Leap West. London: Tibet Information Network, 2000.
Network, 2000.
——— “ADB supports Western Development Drive, but Not in Tibet.”
Phayul.com, http://www.tew.org/development/asia.devel.bank.html. Originally
published on Tibet Information Network, “News Update,” March 30, 2005,
site now discontinued.
———. “High TAR Wages Benefit the Privileged.”
http://www.phayul.com/news/article.aspx?id=9033&t=1&c=1. Originally
published on Tibet Information Network, “News Update,” February 10, 2005,
———. “Tibetans Lose Ground in Public Sector Employment.”
http://www.phayul.com/news/tools/print.aspx?id=8864&t=0. Originally
published on Tibet Information Network, “News Update,” January 20, 2005,
———. “Unequal Competitors: Despite Improvements in Education Levels, Skills
of Tibetan Workforce Still Far Short of Immigrants.” Originally published on
Tibet Information Network, “News Update,” February 23, 2004, site now
discontinued (article currently unavailable).
———. “The Rich Get Richer, and the Poor? Rural Poverty and Inequality in
Tibet – Indications from Recent Official Surveys.” Originally published on
Tibet Information Network, “News Update,” May 31, 2003, site now
discontinued (article currently unavailable).
———. “Deciphering Economic Growth in the Tibet Autonomous Region.”
http://www.phayul.com/news/article.aspx?id=455&t=1&c=4. Originally
published on Tibet Information Network, “Special Report,” April 8, 2003, site
now discontinued.
———. “Despite Economic Boom, Rural Standards of Living in the Tibet

Autonomous Region Still Below 1992 Levels.” Originally published on Tibet
Information Network, “News Update,” February 6, 2003, site now discontinued
(article currently unavailable).
———. “National Autonomy Law Revised to Support Western Development
Policy.” Phayul.com, http://www.tew.org/development/autonomy.law.html.
Originally published on Tibet Information Network, “News Update,” March
13, 2001, site now discontinued.
———, trans. “Third Work Forum on Tibet.” Tibet Justice Center,
http://www.tibetjustice.org/materials/china/china10.html. 1994
Book Reviews
Review of Holy Madness: Portraits of Tantric Siddhas,
edited by Rob Linrothe
Review of The Flying Mystics of Tibetan Buddhism, by

Glenn H. Mullin
Serinity Young
American Museum of Natural History
Rob Linrothe, ed., Holy Madness: Portraits of Tantric Siddhas.

New York: Rubin Museum of Art & Chicago: Serindia
Publications, 2006. 454 pp., $150.00 hard.
This volume puts mahāsiddha studies on an altogether different plane. It is,
therefore, essential reading. It is much more than a catalog in that Rob Linrothe’s
introduction and the accompanying essays by noted scholars, though short, combine
rich textual evidence with an outstanding assortment of mahāsiddha images. As
Linrothe rightly argues in his introduction, earlier scholarship on this topic has
been divided
into textual and visual camps, with textual scholars focusing on the same sets of
texts and the art historians depending on a small number of textual sources. Neither
has made as much use of the evidence of inscriptions on the works of art as they
deserve nor have they collaborated to the degree that is desirable. Here we have
tried to mine the inscriptions as well as lesser-known textual lists of mahāsiddhas
and consider them in relation to the visual dimensions (25).
Additionally, many of the insights contained in this book came about by bringing
together and comparing complete and incomplete sets of mahāsiddha paintings
from which certain historical tendencies could be deduced, such as the trend toward
characterizing individual siddhas in order to make them immediately recognizable.
Thorough descriptions of these often densely populated paintings present a wealth
of information, as do their historical contextualizations. The footnotes richly
document recent and classical studies of the mahāsiddhas in text and iconography.
They are supported by bibliographies of works in English and Tibetan, as well as
appendices listing the mahāsiddha names in various lists, and a glossary. The
1550-6363/2007/3/T3132.
© 2007 by Serinity Young, Tibetan and Himalayan Digital Library, and International Association of Tibetan Studies.
Young: Reviews of Linrothe and Mullin 2
reproductions are sharp and clear and the overall construction, look, and feel of
the book is very attractive.
The Essays
In his introduction Linrothe explains that the catalog celebrates “the powerful
appeal of siddha legends [while acknowledging] that the actual history of the
[eleventh- and twelfth-century] siddhas is unrecoverable” (15). This means that
the portraits are not necessarily those of historical people but more often portraits
of types – the mad holy men and women of the Himalaya region, though a madness
suggestive of a higher sanity that sees absolute reality. He also provides a brief
historical overview of visual representations of the siddhas, especially those found
in complete sets of multiple paintings.
The first three essays establish some historical and literary contexts for the
siddhas and begin to mark their enduring legacy. Geoffrey Samuel begins the essay
section by establishing the cultural context for mahāsiddhas of all stripes: Buddhist,
Hindu, Sikh, and so on. This is a theme that will continue throughout the volume,
which presents essays on siddhas of many different faiths. Matthew T. Kapstein’s
essay focuses on the literary legacies of the mahāsiddhas, the systems of tantric
meditation and yoga they propagated, as well as the songs and poems attributed
to them. The first category includes a large part of the corpus of Buddhist tantric
texts, including, for example, the Lam ’bras system of the Sa skya that is attributed
to the Indian siddha Virupa. Siddha songs and poems were used by later generations
as themes for meditation, to sustain visionary practices, and as initiations into their
teachings. Kapstein, however, is mainly interested in the interaction of these two
legacies. Toward this end, he translated several songs from a Tibetan anthology,
Eight Dohā Treasures, and traces the rationalization of these ecstatic poems of
realization into particular traditions of spiritual discipline. E. Gene Smith’s essay
focuses on the groupings of the mahāsiddhas in Tibetan art and literature into
specific teaching lineages, the gradual inclusion of Tibetan siddhas, and the disciples
of Padmasambhava as siddhas.
The next group of essays emphasizes siddhas in art, though the textual context
is always incorporated, as each author deals with a different combination of visual
and textual evidence from a variety of periods and geographical settings. The first
three essays deal with Himalayan art. Christian Luczanits discusses early Tibetan
paintings of the eight (sometimes nine) great siddhas who are commonly shown
on the sides of a main image. For example, the main image in a fourteenth-century
painting (catalog no. 7) is the siddha Jñātāpa surrounded by smaller portraits of
the eight siddhas and eight teachers. This arrangement established a direct lineage
from the siddhas to these particular teachers, thus authenticating their teachings.
Luczanits then goes through several other paintings, arguing that this authentication
of teaching lineages was the role of the eight siddhas, despite inconsistencies in
their visual representations and their names. David Jackson continues the discussion
of the eight siddhas through the eighteenth-century paintings of Si tu paṇ chen
Chos kyi ’byung gnas (1700-74) of Sde dge, and some of the countless copies that
were made from them. Details about the circumstances under which the paintings
were executed are contained in Si tu paṇ chen’s autobiography, which enables
Jackson to contextualize them within the changing political situation of
eighteenth-century Khams. These were individual portraits painted in sets of nine,
with Padmasambhava being the ninth siddha, that were scattered hither and yon
over the centuries. Jackson traced a set of nine paintings in the Rubin collection
(Figs. 5.2-5.10) back to Si tu paṇ chen’s originals, though they were executed by
different painters at different times.
Dan Martin concentrates on Tibetan representations of one South Indian siddha,
Pha dam pa sangs rgyas (d. 1117), a teacher of Ma gcig labs sgron and therefore
important to the Gcod tradition. Martin attempts to understand what he stood for
in “the minds of artists and their communities” of the twelfth century (109) rather
than what we today think about them. Pha dam pa sangs rgyas made three trips to
Ding ri, where he practiced meditation, lived and taught Mahāmudra, and founded
the Peacemaking (Zhi byed) school. Martin’s main source for Pha dam pa sangs
rgyas is what he calls the Zhijé Collection, a four-volume thirteenth-century
manuscript that includes a biography of Pha dam pa sangs rgyas with illustrations.
His essay reproduces some of its images and he translates passages that elaborate
on Pha dam pa sangs rgyas’s appearance, comparing them to the illustrations in
the biography. He then contrasts these visual representations with those from the
Gcod tradition. While his physical characteristics remain consistent, in the Zhi
byed Collection Pha dam pa sangs rgyas’s hands are usually empty, but in Gcod
representations his right hand holds the ḍamaru and the left either a bell or
thigh-bone trumpet.
The next four essays concentrate on Indian art, beginning with Rob Linrothe’s
exploration of the holiest site in South India for Buddhist and Hindu siddhas,
Śrīśailam, with its early sixteenth-century sculpted walls containing hundreds of
images of sādhus and siddhas. Through these images and texts written in Tibetan
and Sanskrit he demonstrates that Hindu and Buddhist siddhas shared narratives
and visual imagery, and further, that similar visual imagery can be found in Assam
and Tibet.
Debra Diamond brings out the siddha’s connection with sovereignty through
the textual and visual relationship of the siddha Jālandranāth and the Mahārāja of
Jodhpur, Mān Singh (r. 1803-17), a connection enhanced by the expression of
darśan within the paintings. Mān Singh’s ascension to the throne took place in the
midst of intrigues and assassinations. Later his rule was threatened by British
ambitions. At various periods, Mān Singh commissioned paintings of himself
worshipping Jālandranāth, presenting him as the most important religious figure
of the period, and demonstrating his own favored relationship with him (Figs.
8.3-8.5). In this way Mān Singh represented himself as a king chosen by the divine,
thereby bolstering his claim to the throne.
Tushara Bindu Gude discusses a similar group of visual and textual encounters
between two men, in this case Guru Nāth, the founder of Sikhism, and Gorakhnāth,
well known in Hindu sources as a leading kānphaṭa yogī and in Buddhist sources
as a mahāsiddha. Such a meeting was a historical impossibility since they lived

centuries apart, which leads Gude to suggest that these fictive meetings describe
the pervasiveness of kānphaṭa yogīs in the area of the Punjab where Guru Nānak
taught, requiring him to be seen engaging their leading exponent even if he was
long dead.
It is appropriate that Sondra L. Hausner gets the final word in the last essay as
she contextualized the mahāsiddhas in several profound ways that wrap up many
of the ideas and concepts of the preceding essays. She notes the similarity between
mahāsiddha practices and those of contemporary yogīs which emphasize the use
of the body. Hausner beautifully describes the material world, nature, and the
human body as their laboratory; all are composed of the five elements fire, water,
earth, air, and space. “The five elements constitute all things material, and act as
the aligning forces and building blocks of both the inner body and the outer world,
which mirror one another. Balancing the five elements in one’s own environment,
in part through finding the right location for practice, is critical for attaining
realization” (166-67). The physical place of practice and the actual seat of the yogī
gains importance through the concentrated energy of the yogī and that of those
who have meditated there in the past. Further, these seats of practice are connected
with a web of pilgrimage sites containing additional seats of practice. Yet, these
pilgrimage sites are also within the yogī’s body. In much the same way, siddhas
of the past and yogīs of today maintain a tension between the outer and inner social
processes; they meditate in solitude but gather together for festivals and to
acknowledge lineages and brotherhoods. “The life of a yogī centers on juggling
the mundane, everyday aspects of existence – bodies and places and people – with
an attempt to transcend physical limitations and achieve a level of enlightened
bliss” (176). The same can be said of the medieval mahāsiddhas.
The Catalog
The catalog is divided into sections, such as sets of eighty-four mahāsiddhas,
mahāsiddhas organized by lineages, and so on, that add to our knowledge of these
figures. For example, paintings that placed the eighty-four mahāsiddhas around a
central figure, whether a particular teacher or tutelary deity, bestowed authenticity
on that teacher or deity. This is a point made in Luczanits’ essay. This is the case
whether all the siddhas are represented in one painting or in sets of various numbers,
for instance four siddhas to a painting with the central figure alone in a separate
painting. Their representation in paintings that depict particular lineages continues
the theme of their authenticating presence in representing chains of transmission
and also depicts them as “powerful spiritual figure[s] accessible in the continual
presence” (296). In other words, they are not just historical authenticators of a
tradition but inspirational beings who can be accessed by intense devotional and
meditative practices.
There is also an interesting discussion of representations of mahāsiddhas on
manuscript covers. The catalog then moves on to painted images of later Tibetan
siddhas (both Buddhist and Bön), Śaivites, Muslims, and Sikhs. This shows the
enduring force of the siddha ascetic ideal before concluding with photographs of
Indian siddhas dating from the late-nineteenth to twentieth centuries.
My only criticism is that the one-page commentary on female mahāsiddhas is
disappointing, especially as it argues for the inclusion of women consorts as
mahāsiddhas. While this may or may not be the case, it begs the question of the
four named female mahāsiddhas in Abhayadatta. Even the discussion of female
consorts could have featured more images and a deeper analysis. As it is, only
three painted images of female mahāsiddhas, one of a consort, and a statue of
another female mahāsiddha are discussed, though there are additional female
images in the catalog. The catalog is, as it should be, almost overwhelmed with
what is and does not ask about what is absent. It does not ask why a tradition that
requires human and divine females in order to function gives no prominent place
to women. It does not ask why women, with few exceptions, did not achieve success
in this tradition. Surely the implications of their absence are fraught with meaning.
Many of the essays point to new directions in studying issues and topics involving
mahāsiddhas, but none more so than this one.
Glenn H. Mullin, The Flying Mystics of Tibetan Buddhism. New

York: Rubin Museum of Art & Chicago: Serindia Publications,
2006. 248 pp., $50.00 hard.
On a completely different note, there is a certain lightness to Glenn Mullin’s catalog
that seems to go along with his elusive aerial subjects. It is written in an easy-going
and personable style – he often refers to his former Tibetan teachers, but not always
by name – with occasional mentions of esoteric matters and texts, none of which
are cited in either Tibetan originals or English translations, even in the case of
direct quotes. The few footnotes listed are off – for example, on page 27 – they do
not refer to the subject at hand. Based on the number of typos in the text, I assume
the footnote numbers are also typos. There is no bibliography. There are, however,
many, many long quotes, most of them from mystical poems, but they come across
more as page-filler than explanation, and, again, they are not documented.
Mullin has certainly written scholarly, well-documented studies of Tibetan
Buddhism and done important translations, so he seems to have made a conscious
choice to write a rather unsubstantial book to accompany what was a rather
unsubstantial exhibition – so much so that he does not contextualize flying as one
of the siddhis acquired through ascetic practice or the ubiquity of magical flight
among Asian shamans. Instead, he says, “our basic thesis here is that Buddha could
fly, as could many of his immediate disciples, as well as the Indian Buddhist masters
of later generations. The tradition was continued with Buddhism in Tibet, within
both the old and new schools. This is proclaimed in many Buddhist texts and oral
traditions. In addition, the theme is ubiquitous in Tibetan art” (26).
His discussion of the origins of ḍākinīs – a topic truly lost in the mists of time
– is exemplary of bad scholarship. In between telling us that it is a Dravidian term
pre-existing the Aryan invasion and that the Dravidians probably came from Africa
before settling in India, there is an unidentified legend that states: “the first humans
on earth came from another planet and originally had the ability to fly” (224).
However, they ate some coarse food and lost that ability. I fail to see the point of
exaggerating the Visuddhimagga (chapter 13), which simply has heavenly beings
eating pieces of earth and as a consequence developing sexual characteristics.
Mullin goes on to say that ḍākinīs were Dravidian high priestesses who the Aryans
persecuted as witches, at least until the Buddha came along “and gave many of his
tantric lineages exclusively to female disciples” (226) whom he called ḍākinīs.
Two questions come to mind: for whom is Mullin writing, and what do these
speculations have to do with the paintings? Especially since, despite the excellent
example he proposes of the Buddha’s high regard for women, Mullin’s treatment
of flying female mystics leaves a great deal to be desired. The only female images
are one of Ye shes mtsho rgyal, but it is a detail from a larger painting (Plate 30),
and one painting of Ma gcig labs sgron (Plate 31).
There are 47 color plates and an abundance of details from them. Visually, it
is clear that flight was an important theme in Buddhist art, one worthy of more
serious study.
Glossary
Kha
khams Kham Place
Ca
gcod Chö Religious
Practice
Da
ding ri Dingri Place
sde dge Degé Place
Pha
pha dam pa sangs Padampa Sanggyé d. 1117 Person
rgyas
phyag rgya chen po Chakgya Chenpo San. mahāmudra Doxographical
Category
Ba
bon Bön Organization
Ma
ma gcig labs sgron Machik Lapdrön Person
Zha
zhi byed zhijé Peacemaking Doxographical
Category
Ya
ye shes mtsho rgyal Yeshé Tsogyel Person
La
lam ’bras Lamdré Doxographical
Category
Sa
si tu paṇ chen Situ Penchen Person
si tu paṇ chen chos kyi Situ Penchen Chökyi 1700-1774 Person

’byung gnas Jungné
Sanskrit
Abhayadatta Person
darśan Term
ḍākinī Term
ḍamaru Term
Jālandranāth Person
Jñātāpa Person
kānphaṭa Term
Mahārāja Term
mahāsiddha Term
Mān Singh r. Person
1803-1817
Padmasambhava Person
sādhu Term
siddha Term
siddhi Term
Śrīśailam Place
Virupa Person
Visuddhimagga Text
yogī Term
Review of Thundering Falcon: An Inquiry into the History
and Cult of Khra ’brug, Tibet’s First Buddhist Temple, by
Per K. Sørensen and Guntram Hazod, with Tsering Gyalbo
Bryan Cuevas
Florida State University
Per K. Sørensen and Guntram Hazod, with Tsering Gyalbo,

Thundering Falcon: An Inquiry into the History and Cult of Khra
’brug, Tibet’s First Buddhist Temple. Denkschriften der
philosophisch-historische Klasse 333. Beiträge zur Kultur- und
Geistesgeschichte Asiens 46. Wien: Verlag der Österreichischen
Akademie der Wissenschaften, 2005. 432 pp., color photos, 1 fold,
b&w plates, maps, 1 videodisk (DVD). €99.36/$160.00.
Thundering Falcon is the second of two significant ethno-historical studies of
the Yar lung region and neighboring valleys of Central Tibet by the research team
of Per K. Sørensen, Guntram Hazod, and Tsering Gyalbo. The first work in the
series, Civilization at the Foot of Mount Sham-po (Verlag der Österreichischen
Akademie der Wissenschaften, 2000), focused on the local history of Upper Yar
(yar stod). In this companion study, Thundering Falcon, Sørensen, Hazod, and
Gyalbo turn their attention to the Lower Yar lung region, long acknowledged as
the cradle of Tibetan civilization. As an entry point into the rich history of this
area, the authors concentrate on the royal shrine of Khra ’brug or “Thundering
Falcon,” Tibet’s oldest Buddhist temple. Khra ’brug was built in the seventh century
during the reign of King Srong btsan sgam po (d. 649) as the first of several
“border-suppressing” (mtha’ ’dul) sanctuaries, and is also one of the three earliest
and most sacred pilgrimage sites (chos ’khor gnas gsum) in Central Tibet; the two
others are, of course, the Jo khang and Bsam yas. The study is arranged in three
lengthy sections: (1) Introduction (3-38); (2) Sources for the History of Khra ’brug
(39-168); and (3) Appendices (169-333).
In the introduction, Sørensen and Hazod begin with the early history of the Yar
lung and ’Phyong po Valleys, suggesting that the two actually formed a
“historico-geographical unity” during the time of Tibet’s first kings. With the larger
1550-6363/2007/3/T3130.
© 2007 by Bryan Cuevas, Tibetan and Himalayan Digital Library, and International Association of Tibetan Studies.
Cuevas: Review of Thundering Falcon, by Per K. Sørensen et al. 2
geographical context established, the authors then briefly describe the temple of
Khra ’brug, its foremost position in this central region, and the successive phases
in the history of its construction. This largely descriptive opening section is followed
by a series of very detailed historical presentations beginning with a discussion of
the temple’s founding, the political atmosphere during the so-called “Dark Period,”
and the special administrative bonds that existed between Khra ’brug and the local
ruling clan, the Dgyer (var. Sger, [S]gyer mi). The rest of the introduction takes
up the religious significance of the temple and the shifting political climate in each
stage of its history from the early eleventh through the mid-eighteenth century,
shortly after the Fifth Dalai Lama had transformed Khra ’brug into a national
institution.
Part two of the book consists of annotated translations and brief synopses of
six primary sources relevant to the history and cult of Khra ’brug, one of which
has not been published until now. This is the Guidebook to the Pilgrimage Site of
Trandruk (khra ’brug gnas bshad), a local text compiled and set to blocks in the
early 1920s and acquired by Giuseppe Tucci in 1948 during his expedition to
Central Tibet. A clear and very readable facsimile of the xylograph print is included
as an appendix (337-53). The other key sources include brief narrative passages
from three celebrated Tibetan historiographical works (Bka’ chems ka khol ma,
Rgyal po bka’ thang, and Mkhas pa’i dga’ ston), the ninth-century inscription of
the temple bell at Khra ’brug, and an eighteenth-century eulogy (bskang gso) to
the temple’s main protector deities, Tshangs pa, Bse sku, and so forth.
The largest section of the book, part three, is made up of four substantial
appendices, each meticulously researched and thoroughly documented. As in their
previous work, these appendices by Sørensen and Hazod constitute the real heart
of their study and are based on a detailed reading of an impressive range of primary
textual sources combined with insights gleaned from local oral traditions. The
ethnographic data was gathered from several extended field surveys in and around
Yar lung between 1998 and 2002. In Appendix One (“The Tradition Concerning
Tibet’s First Royal Temples,” 171-215), the authors discuss the historical origins
of the royal temple complexes and assess their traditional geomantic significance.
The familiar myth of the supine demoness (srin mo gan rkyal du nyal ba) is revisited
here and examined in great detail, with special attention given to a marvelous
eighteenth-century painting (reproduced in color fold-out) of the temple-pinned
demoness. The foundation myths, ancestral legends, and religious history of Khra
’brug itself is the subject of Appendix Two (“The Falcon and the Lizard: Yar-lung
and the Cultic History of Its Royal Temple, the ‘Thundering Falcon,’” 217-96).
Included in this two-part essay by Guntram Hazod are insightful discussions of
the royal ancestor cults of Yar lung, the structures of territorial power of the early
Yar lung Kingdom, classifications of the major protector deities of the area, and
religious life at Khra ’brug. The final two appendices provide chronological tables,
genealogies of the lords of Yar lung (yar lung jo bo, eleventh-fourteenth century),
and architectural drawings (by Reinhard Herdick).
An added bonus to this valuable study is an accompanying DVD that includes

two short animated “bird’s-eye” view satellite films of the Yar lung Valley (approx.
five minutes each), a twenty-four minute documentary tour of Khra ’brug (recorded
by Sørensen, October 2002), and a twenty-six minute slideshow of the annual
temple festival, the Festival of the Flower Offerings (me tog mchod pa; recorded
by Tsering Gyalbo, May 2001). The disk also contains photographic reproduction
of the main liturgical text, Tshangs pa mchod bstod (ms 1v.1-85v.2), which is
summarized in part two of the study (Text F, 162-68).
In the end, Thundering Falcon proves to be a deeply informative study of Central
Tibet’s earliest Buddhist temple and an indispensable contribution to our
understanding of the local history of the Yar lung Valley, a vital geocultural region.
The work is yet another fine testament to the fruits of the ever increasing
collaborative bonds between scholars in Europe and Tibet, and particularly between
the Tibet Academy of Social Sciences (TASS) and the University of Vienna. We
eagerly await the next volume in this series on Tshal gung thang by the same
impeccable team of researchers.
Cuevas: Review of Thundering Falcon, by Per K. Sørensen et al. 4
Glossary
Ka
bka’ chems ka khol Kachem Kakhölma eleventh Text
ma century
bskang gso kangso eulogy Term
Kha
khra ’brug Trandruk Temple
khra ’brug gnas bshad Trandruk Neshé Text
mkhas pa’i dga’ ston Khepé Gatön sixteenth Text
century
Ga
dgyer Gyer Clan
rgyal po bka’ thang Gyelpo Katang fourteenth Text
century
sger Ger Clan
sgyer mi Gyermi Clan
Cha
chos ’khor gnas gsum chönkhor nesum three sacred Term
pilgrimage sites
Ja
Tha
mtha’ ’dul tandül border-suppressing Term
Pha
’phyong po Chongpo Valley
Ma
me tog mchod pa Metok Chöpa Festival of the Festival
Flower Offerings
Tsha
tshangs pa Tsangpa Deity
tshangs pa mchod Tsangpa Chötö Text

bstod
tshal gung thang Tsel Gungtang Monastery
Ya
yar lung jo bo Yarlung jowo lords of Yarlung Term
yar stod Yartö Upper Yar Place
Sa
srin mo gan rkyal du sinmo genkyeldu supine demoness Term
nyal ba nyelwa
srong btsan sgam po Songtsen Gampo d. 649 Person
bse sku Seku Deity
Review of Tibetan Songs of Realization: Echoes from a
Seventeenth-Century Scholar and Siddha in Amdo, by
Victoria Sujata
Lauran Hartley
Columbia University
Victoria Sujata, Tibetan Songs of Realization: Echoes from a

Seventeenth-Century Scholar and Siddha in Amdo. Leiden &
Boston: Brill, 2005. xvi + 455 pp., CD-ROM. $149 hard.
Tibetan Songs of Realization presents with meticulous care and annotation a
critical edition of songs of realization (mgur) composed by Shar Skal ldan rgya
mtsho (1606-77), one of the most influential bla mas in the religious history of A
mdo. Drawing on four extant versions (woodblocks were first carved in 1756),
Victoria Sujata renders selected verses in English with a penchant for literal
accuracy – a remarkable feat given the colloquial nature of his songs, which often
employ the A mdo vernacular. Her work marks an important contribution in Tibetan
religious and literary studies, above all for its heretofore unavailable translations
of these influential songs. In addition to excerpts cited throughout her text, Sujata
translates in full twenty-two songs of realization (from a total of 242), noting that
only two songs have otherwise been translated (see Thupten Jinpa and Jaś Elsner,
Songs of Spiritual Experience [Boston: Shambhala, 2000]). Skal ldan rgya mtsho’s
verses are still sung by local villagers for rituals and festivals in the teacher’s
hometown of Reb gong (Tongren), as documented on the accompanying
high-quality CD-ROM audio recording (75 min).
Sujata provides a close reading of these translations, elucidated with references
to the biographies of Shar Skal ldan rgya mtsho and his older brother, Chos pa Rin
po che (1581-1659). Her analytical pursuit is threefold. In Part One (1-74), she
seeks “to show how he [Skal ldan rgya mtsho] felt about selected topics such as
his times in Amdo under Mongol dominance, how he viewed Central Tibet as the
source of the teaching, and the spread of Buddhism from there to Amdo” (xii). In
Part Two (75-93), she introduces the genre of songs of realization, with a focus
on the motivation and content of Shar Skal ldan rgya mtsho’s songs in particular.
In Part Three (94-264), she examines his songs for rhetorical tropes (e.g.,
1550-6363/2007/3/T3131.
© 2007 by Lauran Hartley, Tibetan and Himalayan Digital Library, and International Association of Tibetan Studies.
Hartley: Review of Tibetan Songs of Realization, by Victoria Sujata 2
conversations with animals) and poetic conventions, including metrics, stanza

patterns, and poetic figures. While some verses “show the direct influence of
Daṇḍin’s Kāvyādarśa,” Sujata concludes that “the indigenous poetic figures
[metaphors, parallelism, antithesis, verb repetition, etc.] in his mgur far outweigh
the classical ones” (262). Likewise, she argues convincingly that the siddha
primarily drew on folk songs for a variety of meters and likely “sang his mgur in
folk melodies” (262). The effect is startling, as illustrated in the following verse.
Here Sujata renders the colloquial “e go” (lit. “do you understand/hear?”) as “get
it?”
Oh yeah! The Gradual Stages of Englightenment
Hey you! is the essence of religion. [Ya!]
Hey! Practice in accordance with [its] manner!
Ha ha! The mind is happy and glorious, get it?
Ya yi ya yi! (297)
It seems possible that such phrases as “ya yi ya yi” and “a li lo mo a li” (234)
were the siddha’s notation of the melodic but nonsense syllables that still frequent
A mdo folk songs.
In addition to offering religious instruction, these songs frequently raise other
themes. One highlighted by Sujata is the distinction between Central Tibet (dbus)
as a “source of discipline and other teachings” (17) and A mdo as a place of
“confusion” (6) where “because all the monks and lay people/ Seek wealth and
goods, [they] waver” (7). (This is one of the few verses I would have rendered
differently – as “all the monks and lay people are preoccupied with searching for
material wealth” – where mgo bo ’khor means “preoccupied” in the A mdo dialect
and zang zing serves as an adjective). In 1617, Shar Skal ldan rgya mtsho was sent
to Lha sa for religious study at Dga’ ldan Monastery. He returned to A mdo some
ten years later (24), but began writing songs of realization in earnest only in 1647
(87). In the intervening years, Gushri Khan of the Qoshot Mongol tribe had
consolidated his base in A mdo “from which to lead a successful attack in Central
Tibet on the enemies of the Dge lugs pa school, ultimately installing the Fifth Dalai
Lama on the throne of Tibet in 1642” (1). Sujata focuses on several songs, such
as one written in 1662, which unabashedly criticize the excesses and financial
drain of the Mongol presence in A mdo. Other songs of historical interest celebrate
the visit of the Dalai Lama to A mdo in 1652 (16) and praise the First Paṇ chen
bla ma Blo bzang chos kyi rgyal mtshan (1570-1662), who ordained Skal ldan
rgya mtsho in the Jo khang Temple (370-71).
One group targeted by the Qoshots was the Bka’ brgyud school (47). It is thus
fascinating to note that the Dge lugs pa-affiliated Skal ldan rgya mtsho penned
multiple devotions to Mi la ras pa (1040-1123), an iconic figure in the Bka’ brgyud
school and Tibet’s most famous composer of songs of realization (whether by fact
or attribution). Sujata explains this seeming contradiction by noting that the siddha’s
older brother, Chos pa Rin po che, was considered Mi la ras pa’s reincarnation
(48). While the younger brother also sought a hermit’s life, he was instead urged
by the older to raise funds for the founding of a philosophical college at Rong bo
Monastery in Reb gong and a seminary for tantric studies at Bkra shis ’khyil (26).
The next forty years of Skal ldan rgya mtsho’s life, according to his songs of
realization, were largely occupied with such affairs, and it seems his search for
patrons may have involved moral compromise. Skal ldan rgya mtsho confesses,
“I formerly deceived many monks and lay people/ For the purpose of mere wealth
and servants” (29). In one verse, he calls himself a “robber” (340-41). The regret
Skal ldan rgya mtsho feels over what he terms such “shameless deeds” prompts
his desire to live in seclusion for the remainder of his life – an ideal frequently
expressed in his songs of realization:
If you delight in solitary forests,
Amidst a luscious green forest
There is a cool house of the trees’ leaves.
If [you] depend on a place like that, [your] joy will increase. (331)
Sujata worked closely with esteemed teachers of this liturgy tradition in A mdo.
At times, their explanations are awkwardly evident, as when the term “dran” –
easily translated in English as “missed” or “longed for” – is glossed here as “went
around and around in [my] mind” (9). But such instances are few and point not to
inaccuracies but Sujata’s thorough approach. Intensive fieldwork in Reb gong (in
1993, 1996, 2000, 2001, and 2002) also imbues her study with refreshing
appreciation for the lived and performed aspect of the text.
Sujata provides helpful appendices, including a catalogue of Skal ldan rgya
mtsho’s Collected Works with details from each colophon; and a concise biography
(5 pages), based on the rnam thar written by Byang chub mi la ngag dbang bsod
nams (1636-1716). Not least, the CD-ROM features twelve separate verses sung
by soloists from villages (e.g., Hor nag), accompanied by local choruses of up to
seven-hundred male and female villagers. Appendix C provides synopses of the
twelve songs of realization, and full lyrics for most of the recordings can be found
in the book.
Sujata’s approach could be characterized as one of careful scholarly reverence
for her subject. To savor this volume and its recordings is to enjoy both the value
of old-fashioned philology and the energy of liturgical devotion. At the same time,
her analysis could be bolstered by a more critical approach. When Sujata notes
that “the lives Chos pa Rin po che and Skal ldan rgya mtsho led were often similar
to the way Mi la ras pa is said to have lived” (59), one wonders if perhaps the
presentation of their lives (and aspirations) shared similarities due to hagiographic
conventions or even a wish to associate with Mi la ras pa, especially in the case
of Chos pa Rin po che. After all, it was only in the final seven years of his life that
Skal ldan rgya mtsho managed to lead a sustained life of seclusion at Bkra shis
’khyil hermitage (374). Similarly, the author takes at face value Skal ldan rgya
mtsho’s claim that Chos pa Rin po che assumed upon death the posture of Mi la
ras pa when she states this “indicates that Chos pa Rin po che had a very deep
connection with Mi la ras pa” (56). To recognize the possibility of rhetorical
flourishes or to inquire more about the aforementioned “deceptions” would weaken
neither the significance of this great teacher nor this excellent study. This book
and precious recording are essential materials for students of Tibetan literature,
poetics, and ethnomusicology. Tibetan Songs of Realization also offers interesting
details on seventeenth-century history and religious developments in A mdo.
Glossary
Ka
bkra shis ’khyil Trashikhyil Monastery
skal ldan rgya mtsho Kelden Gyatso 1606-1677 Person
Ga
dga’ ldan Ganden Monastery
mgur gur song of realization Term
mgo bo ’khor gowokhor preoccupied Term
Cha
chos pa rin po che Chöpa Rinpoché 1581-1659 Person
Ja
Da
dran dren missed, longed for Term
Na
rnam thar namtar Term
Ba
byang chub mi la Jangchup Mila 1636-1716 Person
ngag dbang bsod Ngawang Sönam
nams
bla ma lama Term
blo bzang chos kyi Lozang Chökyi 1570-1662 Person
dbus Ü Place
Ma
mi la ras pa Milarepa 1040-1123 Person
Za
zang zing zangzing material Term
Ra
reb gong Repgong Chi. Tongren Place
rong bo Rongwo Monastery
Sha
shar skal ldan rgya Shar Kelden Gyatso 1606-1677 Person
mtsho
Ha
hor nag Hornak Village
lha sa Lhasa Place
A
a mdo Amdo Place
Sanskrit
Daṇḍin Person
Kāvyādarśa Text
siddha Term
News
JIATS 3 News
Note: News items for the next issue should be sent to the editor, Prof. David
Germano .
Publications
PIATS 2003
Ten volumes of the Proceedings of the Tenth Seminar of the International
Association of Tibetan Studies (2003) are now available from Brill.
Jonang Publication Series

The first set in the Jonang Publication Series (jo nang dpe tshogs, mi rigs dpe skrun
khang, Beijing, 2007) is now available for purchase from the Jonang Foundation
website. This annual series of select Tibetan Buddhist philosophical classics features
important works by major Jonang authors from all genres of sūtra and tantra. Each
work in this ongoing series, in dbu can, is chosen from the corpus of Jonang
literature to reflect the contemporary scholastic curriculum within Jonang monastic
universities inside Tibet. To view the inaugural set of 8 books (10 volumes), and
to purchase, visit the Publications page under “Resources” on the Jonang Foundation
website: http://www.jonangfoundation.org/publications-tibetan-text
Conferences
Ladakh Studies Conference

The 13th colloquium of the International Association of Ladakh Studies (IALS)
took place at the La Sapienza University in Rome, September 7-11, 2007. There
were panels on religion and society, history, history of art, medicine, ecology, and
development. For further details please see: www.ladakhstudies.org, or contact
John Bray.
Young Tibetologists Conference in London

The first meeting of the International Seminar for Young Tibetologists (ISYT) was
held at SOAS, London, August 9-13, 2007. ISYT was born of the IATS at the 2006
meeting in Königswinter, and now has approximately 40 delegates attending from
around the world. The group has been given generous funding from the AHRC
(Arts and Humanities Research Council, UK) to bring delegates from India, China,
1550-6363/2007/3/T3298.
© 2007 by Tibetan and Himalayan Digital Library and International Association of Tibetan Studies.
News 2
Tibet, and the other Himalayan States. Further details can be found on the website:
www.ISYT2007.com.
Buddhism Conference at Visva Bharati

The Centre for Buddhist Studies at Visva Bharati University held an international
conference on the topic of “Buddhism and Its Social Significance for the Asian
World,” September 25-28, 2007, in Santiniketan, West Bengal, India. For further
details contact the convener, Dr. Andreas Loseries.
Traditional Asian Medicine Conference: Call for Papers

The International Association for the Study of Traditional Asian Medicine
(IASTAM) will hold its Seventh International Congress (ICTAM) September 7–11,
2009, at the Institute of Traditional Medicine Services in Thimpu, Bhutan. The
theme for the Seventh Congress is “Asian Medicine: Cultivating Traditions and
the Challenges of Globalisation.” IASTAM invites proposals for papers on the
cultivation of the body, plants, and traditional knowledge and practice. IASTAM
also invites proposals on the challenges faced by the globalization of traditional
Asian medicines, tensions between local and global production, interpretation,
professionalization and evaluation, as well as issues of trade, economy, and ecology.
IASTAM encourages papers from scholars from all science and humanities
disciplines as well as from practitioners of traditional Asian medicine (Unani,
acupuncture, Sowa Rigpa, Ayurveda etc.). Proposals on any theme or topic are
welcome, but papers that address the themes of cultivation and globalization will
be given preference. Proposals for organized panels will also be given preference
over individual papers. A list of panels and additional information about registration,
etc., will be available on the conference website
http://www.iastam.org/conferences.htm. Panel and paper abstracts of not more
than 250 words should be sent as e-mail attachments to Emma Griffin. Deadline
for abstract submission is January 1, 2008.
Exhibitions
Bhutanese Art
The Honolulu Academy of Arts will be hosting an exhibition, “The Religious Art
of Bhutan,” that will begin February 24, 2008 and run for six months. The show
will then travel to the Asian Art Museum of San Francisco and to the Rubin
Museum of Art in New York. A catalogue will be published by Serindia.
Bon Art
The Rubin Museum of Art in New York is hosting the first exhibition of the art of
Bon, “Bon: The Magic Word,” from October 5, 2007 until January 14, 2008. A
catalogue will be published by Wilson (U.K.). For announcements about this and
other exhibitions at RMA, please click here.
Photography Exhbition on Newars in Nepal

A selection of images from Michael Muehlich’s photo-documentation of “Buddhist
Initiation of the Newars in Nepal” is currently on display at Tibethaus, Frankfurt,
presented with English as well as German captions. A small catalogue, published
in 2004 by the Lumbini International Research Institute, on the documentation of
the ritual is also available (ISBN 99946-933-0-1, liri [at] mos [dot] com [dot] np).
Visitors interested in relations between Tibet and Nepal are welcome for a reception
and talk on September, 28, 2007 (pre-register by September 25). Details about
hours and other information can be found at: www.tibethaus.com.
New Academic Programs
New Diploma Program in Tibetan Buddhist Studies in India

In 2007-08, the Department of Chinese and Tibetan Languages of Panjab University,
Chandigarh, India will begin offering a PG Diploma in Buddhist Studies, with
Tibet and the Himalayas as one of several areas of specialty.
New Doctoral Program and Language Classes in Scotland

As part of their growing postgraduate-taught program, the Scottish Centre for
Himalayan Research at the University of Aberdeen announces the introduction of
a new doctoral program in Himalayan Research designed to allow students to
continue forward from the Centre’s existing masters programs in Himalayan Studies
and Himalayan Ethnobotany. Each program includes recently introduced language
courses in Nepali and Tibetan.
Other News Items
New Group Formed at AAS

A new Tibetan Studies Group affiliated with the Association of Asian Studies
(AAS) was founded at the San Francisco AAS meeting in 2006. The group plans
to have regular meetings at the AAS and hopes to advocate for the growth and
support of Tibetan Studies within (and beyond) the AAS. With the support of
David Germano and the staff at the Tibetan and Himalayan Digital Library, an
electronic mailing list has been established to facilitate communication and
exchange by faculty, independent scholars, and graduate students across disciplines
with specialized interest in Tibetan Studies who belong to AAS. Those wishing to
subscribe may do so at: https://list.mail.virginia.edu/mailman/listinfo/aastibet.
Those who have subscribed are free to directly post messages to the list-serv. Profs.
Gray Tuttle and Carole McGranahan request that IATS members share this
information with anyone who may be interested.
Tibetan Text Input Service

The International Trust for Traditional Medicine (ITTM), a registered non-profit
organization based in Kalimpong, India (www.ittm.org), is now offering to scholars
News 4
worldwide the service of transliterating Tibetan texts into Wylie. Two full-time
Tibetan data input operators, each with seven years of experience, have almost
completed the transliteration of twenty-one seminal Tibetan medical texts. To raise
funds to complete the final proofreading of these texts and to continue giving
employment to these two Tibetan women, ITTM is offering transliteration services
of Tibetan texts to individual scholars. Prices are determined by the size of the
resulting text file in kilobytes (KB). Student discounts are available. Contact
infoittm [at] gmail [dot] com for details.
Abstracts
JIATS Abstracts
Issue 3, December 2007
A Look at the Diversity of the Gzhan stong Tradition
by Anne Burchardi
This article introduces two studies by classical Tibetan Buddhist scholars that
explain the range of meanings of the term empty of other (gzhan stong). The two
texts – one by Padma bi dza (twentieth century) (twentieth century), the other by
Tāranātha (1575-1634) – are analytical studies that summarize and compare the
various views of previous scholars who wrote on empty of other. Such interpretive
studies are valuable in that they present us with different ways of interpreting the
heterogeneous material classified under the rubric “empty of other.” They also
suggest ways of contextualizing the different levels of discourse found within this
material.
Beyond Anonymity: Paleographic Analyses of the Dunhuang
Manuscripts
by Jacob Dalton, Tom Davis, and Sam van Schaik
This article presents a new paleographic approach to the Tibetan manuscripts from
Dunhuang. By adapting the techniques of forensic handwriting analysis to the
Tibetan alphabet, we can identify groups of manuscripts written in the same hand.
After introducing this new approach, the present paper applies it to the works of
a single scribe, taken as an initial example. Once this particular group of manuscripts
has been identified, a range of further insights into this person emerge – his many
connections to the kingdom of Khotan, his unique writing style, and his interest
in the external ritual practices relating to water and fire offerings, stūpas, rosaries,
and the like. This new approach promises to alter significantly our understanding
of the Tibetan Dunhuang documents. No longer are we confronted with a mass of
undigested material; now we can begin to date and ascribe names to whole swathes
of the collection.The authors thank the AHRC, the International Dunhuang Project,
the American Philosophical Society, and the American Academy of Religion for
their support in funding this collaborative project. The original research began
during a three-year AHRC-funded project on the tantric manuscripts from
Dunhuang, published in Jacob Dalton and Sam van Schaik, Tibetan Tantric
Manuscripts from Dunhuang: A Descriptive Catalogue of the Stein Collection at
the British Library (Leiden: Brill, 2006).
“Emperor” Mu rug btsan and the ’Phang thang ma Catalogue
by Brandon Dotson
Confusion shrouds the events surrounding the death of Emperor khri srong lde
btsan (742-c. 800) and the succession of his sons at the turn of the ninth century.
Tibetan religious histories, Old Tibetan sources, and Chinese sources offer
conflicting pictures of the order of events and the identities of those involved.
Fortunately, a newly published source, the ’Phang thang ma Catalogue (’Phang

thang ma), throws new light on the royal succession by referring to Mu rug btsan,
the elder brother of khri lde srong btsan, as nothing less than an emperor (emperor
(btsan po)). Considering this new information alongside Old Tibetan inscriptional
evidence, this article attempts to establish the order of events around khri srong
lde btsan’s abdication and death, and to locate “Emperor Mu rug btsan” within
them.
An Early Seventeenth-Century Tibeto-Mongolian Ceremonial Staff
by Johan Elverskog
This article provides a translation and study of a recently discovered ceremonial
staff housed at the Metropolitan Museum of Art. Based on the identification of
two names found in the inscription, sodnam ombu taiji (1587-1625) and Yon tan
rgya mtsho (Yon tan rgya mtshoparenval, 1589-1617), it situates the piece in the
historical context of early seventeenth-century Mongol-Tibetan relations.
The Importance of the Underworlds: Asuras’ Caves in Buddhism, and
Some Other Themes in Early Buddhist Tantras Reminiscent of the
Later Padmasambhava Legends
by Robert Mayer
The story of Padmasambhava taming non-human females at the asura Cave at
Pharping is well known. Much less widely known is the wider tradition of Asuras’
caves as the entrances to Pātāla, the magical underworlds of Asuras and Nāgas, a
colorful and often eroticized and popular belief which played a prominent role in
early Indian and Chinese Buddhist tantras. This paper surveys these now largely
forgotten beliefs, and then proceeds to raise (but not answer) the question: might
further widely attested kriyātantra themes, such as treasure recovery, kīlas, and
water magic, have influenced the popular mythology of Padmasambhava?
Re-Assessing the Supine Demoness: Royal Buddhist Geomancy in the
Srong btsan sgam po Mythology
by Martin A. Mills
The myth of the Chinese princess Kong jo’s geomantic divination of Tibet prior
to the founding of the Central Temple of Lhasa (lha sa gtsug lag khang) – and in
particular the striking image of the land of Tibet as a “supine demoness” – has
been the object of considerable academic comment. Generally, it has been read as
a metaphor either of monastic Buddhism’s misogynist tendencies, or of its
superposition over putative religious precursors. In this article, the difficulties that
attend these interpretations of the supine demoness image are assessed when
examined within the context of the princess’s wider divination, as presented in
Tibetan mythic histories such as the Ma ni bka’ ’bum, The Clear Mirror of Royal
Genealogy (Rgyal rabs gsal ba’i me long), and the Pillar Testament (Bka’ chems
ka khol ma) (bka’ chems ka khol ma), and in particular when it is viewed within
the context of the Lha sa’s actual topographic structure. In light of these, it is
proposed that both the supine demoness image and the other elements of Kong
jo’s divination should be understood as it has always been presented by Tibetan
sources – as part of an established tradition of Chinese geomancy, a tradition which
has itself been reorganized as a medium for Buddhist themes of liberation.
Modernity, Power, and the Reconstruction of Dance in Post-1950s
Tibet
by Anna Morcom
This article traces the changes and transformations of Tibetan dance that have
occurred since 1950. It looks at how agents of change ranging from small groups
of Tibetans to the Chinese state create, negotiate, and represent different kinds of
Tibetan modernities through reconstructions and reconfigurations of Tibetan dance,
and it examines how dance is modernized, the kinds of modernity represented, and
the social and political power dynamics involved. In particular, the article looks
at the repercussions on dance of the recent state-sponsored economic development
of Tibet that has resulted in the formation of new urban middle classes but also a
dramatic growth in disparity. It also examines the impact on dance culture of the
growth of heterogeneous communities in the increasingly mobile population of
contemporary Tibet. The article further addresses the issue of globalization, which
sees new layers being added to the social and cultural hierarchies of twenty-first
century Tibet.
Contributors
JIATS Contributors
Issue 3, December 2007
Articles & Essays
Anne Burchardi is an Associate Professor at The Tibetan Section of The Asian
Department, Institute for Cross Cultural and Regional Studies at the University of
Copenhagen; and a Research Librarian at The Department of Orientalia and Judaica,
The Royal Library of Copenhagen. Her publications include “Kongtruls Fem
Samlinger - En oversigt,” in Fund og Forskning i Det Kongelige Biblioteks
Samlinger (København: Det Kongelige Bibliotek, 2001) and, with Ari Goldfield,
Gaining Certainty about the Provisional and Definitive Meanings in the Three
Turnings of the Wheel of Dharma, the Two Truths, and Dependent Arising
(Kathmandu: Marpa Institute for Translation). Her research interests are Tibetan
Buddhist philosophy and epistemology.
Jacob Dalton is an Assistant Professor of Religious Studies at Yale University.
He received his doctorate in Buddhist Studies from the University of Michigan.
His recent book (with Sam van Schaik) is entitled Tibetan Tantric Manuscripts
from Dunhuang (Leiden: Brill, 2006). He is now working on two books, one on
the role of violence in the Tibetan assimilation of Buddhism, and the other on the
ritual history of early tantric Buddhism. His research focuses on developments in
tantric ritual in early medieval India and throughout later Tibetan history.
Tom Davis is a Lecturer in English at the University of Birmingham, specializing
in the bibliography and paleography of English literary documents. In 1974 he
began to use his paleographical skills for forensic purposes, and since then has
written reports on the handwriting of thousands of suspect documents and given
evidence in court many times (from magistrates courts all the way up to the Court
of Appeal before the Lord Chief Justice). He has conducted extensive research
into the analysis of modern handwriting, including five major research projects.
Currently he is engaged in the application of forensic identification practice to the
inscriptions on cuneiform tablets.
Brandon Dotson is Lector in Tibetan at the School of Oriental and African Studies.
He received his doctorate in Oriental Studies from the University of Oxford. He
is co-editer, with Matthew Kapstein, of Contributions to the Cultural History of
Early Tibet (Leiden: Brill, 2007), which contains his article, “Divination and Law
in the Tibetan Empire: The Role of Dice in the Legislation of Loans, Interest,
Marital Law and Troop Conscription.” He has reinaugurated the Seminar of Young
Tibetologists as the International Seminar of Young Tibetologists (ISYT). His
work generally focuses on the intersection of history, ritual, and narrative in and
around the period of the Tibetan Empire.
Johan Elverskog is an Associate Professor of Religious Studies at Southern
Methodist University. He received his doctorate in Central Eurasian Studies from
Indiana University. His recent publications include Our Great Qing: The Mongols,
Buddhism and the State in Late Imperial China (Honolulu: University of Hawai’i,
2006) and The Pearl Rosary: Mongol Historiography in Early Nineteenth-Century
Ordos (Bloomington: The Mongolia Society, 2006). He is currently working on
the social history of Qing Buddhism and Buddhism and Islam in Inner Asia.
Robert Mayer is a Researcher in South Asia in the Oriental Studies Department
at the University of Oxford. He received his doctorate from the University of
Leiden. Recent publications, with wife Cathy Cantwell, include The Kīlaya Nirvāṇa
Tantra and the Vajra Wrath Tantra: Two Texts from the Ancient Tantra Collection
(Vienna: Verlag der Österreichischen Akademie der Wissenschaften, 2007) and
“Two Proposals for Critically Editing the Texts of the rNying ma’i rGyud ’bum,”
Revue d’Etudes Tibétaines 10 (April, 2006): 56-70. His current projects include a
detailed study of the complete Dunhuang materials.
Martin Mills is Senior Lecturer in the Anthropology of Religion at the University
of Aberdeen and co-director of the Scottish Centre for Himalayan Research. He
received his doctorate in Social Anthropology from the University of Edinburgh.
He is the author of Identity, Ritual and State in Tibetan Buddhism: The Foundations
of Authority in Gelukpa Monasticism (London & New York: RoutledgeCurzon,
2003). His research interests include Tibetan monasticism and government, Buddhist
ritual and geomancy, and anthropological theories of religion.
Anna Morcom is an Academic Fellow in the Music Department at Royal Holloway
College, University of London. She received her doctorate in Ethnomusicology
from the School of Oriental and African Studies. Her publications include Hindi
Film Songs and the Cinema (forthcoming) and “An Understanding between
Hollywood and Bollywood? The Meaning of Hollywood-style Music in Hindi
Films,” in Music and Meaning (British Journal of Ethnomusicology 10, no. 1
[2001]: 63-84). Her research interests are, generally stated, music/dance in the
context of a rapidly changing modern world.
Sam van Schaik is a Research Project Manager for the International Dunhuang
Project at the British Library. He received his doctorate in Buddhist Studies from
the University of Manchester. His recent publications include a number of articles
and two books: Approaching the Great Perfection: Simultaneous and Gradual
Methods of Dzogchen Practice in the Longchen Nyingtig (Boston: Wisdom, 2003),
and, with Jacob Dalton, Tibetan Tantric Manuscripts from Dunhuang (Leiden:
Brill, 2006). He is currently engaged in a three-year project on the paleography of
the Tibetan and Chinese manuscripts from Dunhuang.
Reviews
Bryan Cuevas is an Associate Professor of Religion at Florida State University.
He received his doctorate in History of Religions and Tibetan Studies from the
University of Virginia. He is the author of The Hidden History of the Tibetan Book
of the Dead (Oxford: OUP, 2003). He is currently working on a study of Tibetan

sorcery and the politics of war magic from the sixteenth to eighteenth centuries.
Lauran Hartley is the Tibetan Studies Librarian for the Starr East Asian Library
at Columbia University. She received her doctorate in Tibetan Studies from Indiana
University. Her publications include an article in Contemporary Tibetan Literary
Studies (Leiden: Brill, 2006), and she is co-editor of Modern Tibetan Literature
and Social Change (Duke University Press, forthcoming).
Serinity Young is a Research Associate in the Division of Anthropology of the
American Museum of Natural History in New York City. She received her doctorate
in Comparative Religion from Columbia University. She is the author of Dreaming
in the Lotus: Buddhist Dream Narrative, Imagery and Practice (Boston: Wisdom
Publications, 1999) and Courtesans and Tantric Consorts: Sexualities In Buddhist
Narrative, Iconography, and Ritual (London: Routledge, 2004).

JIATS_03_2007

Uploaded by

Copyright:

Available Formats

JIATS_03_2007

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

JIATS_03_2007

Uploaded by

Copyright:

Available Formats

Journal of the

Issue 3 — December 2007

• A Look at the Diversity of the Gzhan stong Tradition (24 pages)

Contributors to this Issue

1. A general presentation of the different opinions.

A General Presentation of the Different Opinions

1. The omniscient Jo nang, (Dol po pa shes rab rgyal mtshan [1292-1361])

Another question reads:

A Specific Description of the Jo nang Intention

A Presentation of the Validity of the Two Kinds of Madhyamaka

1. The differences between the two types of Madhyamaka.

The Differences Between the Two Types of Madhyamaka

and intention of Sūtra[-yāna] (Path of the Buddha’s Words) and Mantra[-yāna,

1. Difference in terms of subject matter.

Difference in Terms of Subject Matter

Difference in Terms of Manner of Ascertainment

How These Two Ultimately Do Not Contradict Each Other

Tāranātha presents Shākya mchog ldan’s position first, referring to him as

Shākya mchog ldan’s position is depicted as sympathetic to gzhan stong while

on the status of tathāgatagarbha, and on their understanding of Mantra[-yāna],

dbyings ye gnyis ka ying yé nyika both the sphere of Term

shes rab kyi pha rol tu Sherapkyi Paröltu Text

secret syllables mantra Term

———. “Tāranātha’s ‘Twenty-one Differences with Regard to the Profound

Wangchuk Dorji. “The rÑiṅ-ma Interpretations of the Tathāgatagarbha Theory.”

Sam van Schaik

Abstract: This article presents a new paleographic approach to the Tibetan

An Introduction to Forensic Handwriting Techniques

1. An introduction to the basic practice of forensic handwriting analysis and

letter considered abstractly. A reader sees a graph (a unique mark on a piece of

time, place, and by a particular kind of writer. It involves a knowledge of all of

A Forensic Approach to Paleography

The Adaptation of Forensic Techniques to the Tibetan Script

Building a Tool for Forensic Examination

For help in this regard we utilized contemporary Tibetan handwriting guides,

result of which can be found at the following URL: http://idp.bl.uk/handwritings/

Small Heads and Long Legs

Images 1 & 2: A na graph from ITJ425 and

Images 3 & 4: A zha graph from ITJ425 and

Images 5 & 6: A da graph from ITJ425 (with

Images 7 & 8: A ga graph from ITJ425 and

The Grapheme Nya

Images 9 & 10: A nya graph from ITJ425 and

Writing Practices around Dunhuang

An Example Scribe: Supporting Internal Evidence

The Identified Manuscript Group

Thematic Links between the Manuscripts

Multiples and Revisions

Possible Headed Manuscripts

Canonical Primary Sources

through which to triangulate regnal dates, some tenacious problems persist

The ’Phang thang ma Catalogue

Contributions of the ’Phang thang ma Catalogue to the History

royal genealogy therefore indicates that Mu ne brtsan ruled as emperor between

rgya bod kyi chos Gyabökyi Chöjung Text

bon po Bönpo Organization

mKhas-paḥi-dgaḥ-ston by Dpaḥ-bo-gtsug-lag-phreng-ba. Edited by Lokesh

to Enlightenment in Chinese Thought, edited by Peter N. Gregory, 41-66.

Tibeto-Mongolian Ceremonial Staff

Abstract: This article provides a translation and study of a recently discovered