Maps of Greater Tibet

15 · Maps of Greater Tibet
JOSEPH E. SCHWARTZBERG
INTRODUCTION Tibetan maps are known to exist or are likely to be dis-

covered, discuss the sources and general nature of the
Despite its limited population, the area referred to in this highly diversified cartographic corpus, point out some of
chapter as Greater Tibet has given rise to a remarkably the canons of cartography followed by monks, scholars,
rich and varied cartographic tradition. Moreover, this tra- and other artists charged with making maps, and consider
dition remains vibrant and resistant to the cultural influ- the nature of what one might call "folk cartography." I
ence of far more populous and powerful neighbors. For shall then consider cosmographic mapping, which has a
this history, Greater Tibet is taken to correspond roughly long and rich history in the region and remains an impor-
to the area between India and China proper in which a tant aspect of its culture. This will be followed by a
distinctive form of Buddhism, widely known as Lamaism, discussion of geographic maps, ranging in coverage from
is the dominant faith. But since Lamaism is a term that the entire world as known to Tibetans, through maps of
Tibetans themselves avoid, the preferred form, Tibetan regions of widely varying territorial extent, to plans of
Buddhism, will be followed throughout this work. The towns and individual shrines. The conclusion will assess
area of Tibetan Buddhism that I shall consider includes the nature of the entire cartographic corpus of the region
the whole of Tibet and Bhutan as now constituted as and seek to relate it to the traditions of other cultures.
well as adjacent territories in India (particularly Ladakh
and Sikkim), Nepal, and China (virtually all of Qinghai
THE STATE OF OUR KNOWLEDGE
and parts of Xinjiang, Gansu, Sichuan, and Yunnan) in
which Tibetan culture either predominates or is signifi- As was true of the cartography of India and Southeast
cantly represented. Asia, maps from Greater Tibet have not previously been
In areas where Tibetan Buddhism blends with other given due recognition in standard histories of the subject.
cultural traditions, the omission of maps from these other Bagrow, for example, ignores the area completely.1 San-
traditions in this chapter should not be interpreted to tarem, citing Francis Wilford, calls attention to an eigh-
mean that such works do not exist. Some of those maps- teenth-century three-dimensional map of the kingdom of
for example, several from the culturally eclectic region Nepal that was presented to Warren Hastings, but that
of Nepal-have already been discussed in volume 2, book no longer extant work does not appear to have belonged
1 of The History of Cartography, along with others that within the Buddhist tradition. 2 Similarly, Adler, citing
fall unquestionably within the Hindu tradition. However, Pulle, discusses a Nepalese map that clearly belongs more
other Nepali maps associated with Hindu patrons are
discussed here if they follow a style more consistent with
Tibet than with India. Apart from the well-known faiths 1. Leo Bagrow, History of Cartography, rev. and en!. R. A. Skelton,
of Buddhism and Hinduism, over much of Greater Tibet trans. D. L. Paisey (Cambridge: Harvard University Press; London: C.
there is continued adherence to a set of beliefs and prac- A. Watts, 1964; reprinted and enlarged, Chicago: Precedent Publishing,
tices associated with Bon, a pre-Buddhist cult, that has, 1985). Bagrow does refer, on 207, to "a Buddhist map showing the
world as a floating lotus-blossom," but what he is describing is actually
as we shall see, relevance for the history of cartography.
a Hindu cosmographic conception. See Joseph E. Schwartzberg, "Intro-
Over the centuries Bon and Buddhist practices became duction to South Asian Cartography," in The History of Cartography,
inextricably intertwined, and it will not always be pos- ed. J. B. Harley and David Woodward (Chicago: University of Chicago
sible to differentiate between the two in discussing the Press, 1987-), vol. 2.1 (1992), 295-331, esp. 300.
sources of various aspects of Tibetan cartography. Figure 2. Manuel Francisco de Barros e Sousa, Viscount of Santarem, Essai
sur f' histoire de fa cosmographie et de fa cartographie pendant fa
15.1 shows the locations of most of the places to be
Moyen-Age et sur fes progres de fa geographie apres fes grandes decou-
discussed in this chapter. vertes du xve siecle, 3 vols. (Paris: Maulde et Renou, 1849-52), 1:364.
I shall first consider the state of knowledge about the Wilford's description of the map in question is quoted at length in
indigenous cartography of Greater Tibet, note where Schwartzberg, "South Asian Cartography," 326 (note 1).
607
608 Cartography in Greater Tibet and Mongolia
INSET 1 90'
- - National boundary NEPAL Independent state
- - - Internal boundary TIBET Other political region
• National capital Amdo Historical region
Other settlement
o 100 200 miles
... Mountain peak --1.1-'1r---;Ir'1
f - I- - - - . - ,
o 100 200 300 km
80' 85'
, A N G
(I
)
C h a n 9 a n\ 9
C H
T B E T (X Z
Tlbeta n P I a t e
30'
25'
FIG, 15.1. REFERENCE MAP FOR GREATER TIBET.
to the Hindu than to the Buddhist tradition. 3 Other well- voznaniya, Antropologii i Etnografii: Trudy Geograficheskogo Otde-
known histories totally fail to note maps from Greater leniya 119, no. 2 (1910), esp. 231; Francesco L. Pulle, La cartografia
Tibet, but there is little point in providing a detailed list antica dell'India, Studi Italiani di Filologia Indo-Iranica, Anno IV, vol.
4 (Florence: Tipografia G. Carnesecchi e Figli, 1901).
of them. 4 4. The notes relative to this point in Schwartzberg, "South Asian
Perhaps the earliest outside scholarly reconstruction of Cartography," 296 and 298 (note 1), are as applicable to Greater Tibet
any form of indigenous cartography from this region is as they are to India.
that of the Augustinian missionary Antonio Agostino 5. Antonio Agostino Giorgi, Alphabetum Tibetanum missionum
Giorgi (1711-97), who in 1762 published an engraved apostolicarum (Rome: Typis Sacrae Congregationis de Propaganda Fide,
1762). The original manuscript is in the Bibliotheca Apostolica Vaticana,
Tibetan cosmography that was almost certainly copied Rome. Giorgi's detailed illustration, Cosmografia Buddhistica Tibetana,
from a local source,s Among relatively modern scholars, facing page 472 of the published version, has an explanation keyed to
the task of explicating and illustrating Tibetan cosmo- the original on 470-86. His diagram has been copied in several sub-
graphy was resumed by Waddell, whose exposition of sequent works including those by Laurence Austine Waddell, The Bud-
Tibetan Buddhism first appeared in 1895,6 Many others dhism of Tibet, or Lamaism, with Its Mystic Cults, Symbolism, and
Mythology, and Its Relation to Indian Buddhism, 2d ed. (Cambridge:
have since contributed to our understanding of this sub- W. Heffer, 1935),79 (originally published in 1895); Pulle, La cartografia
ject. One lavishly illustrated work, Ajia no kosumosu + antica dell'India, 23 (note 3); and Sugiura Kohei, ed., Aiia no kosumosu
mandara, is particularly worthy of mention in that it puts + mandara (The Asian cosmos), catalog of exhibition, "Ajia no Ucho-
the cosmography of Greater Tibet into a sweeping pan- kan Ten," held at Rafore Myojiamu in November and December 1982
Asian context? The importance of cosmography in the (Tokyo: Kodansha, 1982), 18.
6. Waddell, Buddhism of Tibet, esp. 77-111 (note 5), which include
3. Bruno F. Adler, "Karty pervobyrnykh narodov" (Maps of primitive numerous cosmographic diagrams.
peoples), Isvestiya Imperatorskogo Obshchestva Lyubiteley Yestest- 7. Sugiura, Aiia no kosumosu + mandara (note 5).
Maps of Greater Tibet 609
lives of Tibetan Buddhists cannot be overemphasized. tains no reference to an indigenous work. 17

Cosmographies form objects of contemplation in daily In addition to these wide-ranging surveys, there are a
worship and serve a variety of other, largely didactic pur- few published works by historians of cartography that
poses. deal with specific issues relating to Tibetan maps. The
With respect to geographic mapping, I noted in my oldest is a 1947 study by Nakamura of "Chinese" world
discussion of South Asian cartography that a number of maps preserved in Korea, which discusses what was in
early British travelers in the Himalayan region-beginning fact originally a Tibetan map of indeterminate age (fig.
with Francis Hamilton (formerly Buchanan) who spent 15.30 below), a copy of which was brought from China
fourteen months in Nepal in 1802-3-commented on to Japan during the ninth century.18 The map was dis-
having received "native maps," which they subsequently
used in preparing maps of their own. 8 These maps came
8. joseph E. Schwartzberg, "Geographical Mapping," in The History
from diverse ethnic groups-not always identifiable-and of Cartography, ed. j. B. Harley and David Woodward (Chicago: Uni-
probably varied in style accordingly. Some undoubtedly versity of Chicago Press, 1987-), vol. 2.1 (1992), 388-493, esp. 430-
were from Tibetans or closely related peoples. Regret- 31.
tably, none of the earliest maps Hamilton alluded to 9. The extensive journals of Hamilton in the Oriental and India Office
Collections, British Library, London, may throw some light on the
survives, and I know of no published account that
matter, however, since he had a very keen interest in cartography and
describes them. 9 Not until 1863, when the results of the discussed at great length maps he had acquired in other areas. See, for
Tibetan expedition of the Schlagintweit brothers were example, discussions and reproductions of Burmese maps in chapter 18
published, do we have any careful reproduction of a and table 18.1 below.
Tibetan geographic map (fig. 15.42 below).lO Two further 10. Hermann von Schlagintweit-Sakiinliinski, Adolphe von Schlagint-
weit, and Robert von Schlagintweit, Results of a Scientific Mission to
examples of Tibetan cartography, both relating to Lhasa India and High Asia, vol. 4 (Leipzig: Brockhaus; London: Triibner,
and its great Potala palace, were included in the work by 1863), geographical map 3.
Waddell, published several decades later. 11 Thereafter, a 11. Waddell, Buddhism of Tibet, illustrations on 40 and facing 287
variety of additional maps appeared with increasing fre- (note 5).
quency in diverse publications; but as a rule the authors 12. Barbara Nimri Aziz, "Tibetan Manuscript Maps of Dingri Valley,"
Canadian Cartographer 12 (1975): 28-38; idem, Tibetan Frontier Fam-
emphasized subjects other than cartography. Not until ilies: Reflections of Three Generations from D'ing-ri (Durham, N.C.:
1975-78, when several works by the anthropologist Bar- Carolina Academic Press, 1978), esp. 10-22; and idem, "Maps and the
bara Nimri Aziz appeared, did any published work com- Mind," Human Nature 1, no. 8 (1978): 50-59.
pare Tibetan maps and discuss their intrinsic qualities and 13. Mary Shepherd Slusser, "Serpents, Sages, and Sorcerers in Cleve-
the cognitive processes involved in their construction. 12 land," Bulletin of the Cleveland Museum of Art 66, no. 2 (1979): 67-
82; idem, Nepal Mandala: A Cultural Study of the Kathmandu Valley,
Shortly thereafter another anthropologist, Mary Shep- 2 vols. (Princeton: Princeton University Press, 1982); idem, "On a Six-
herd Slusser, published several insightful works on tra- teenth-Century Pictorial Pilgrim's Guide from Nepal," Archives of
ditional maps from the Vale of Kathmandu in Nepal. 13 Asian Art 38 (1985): 6-36; idem, "The Cultural Aspects of Newar
Since then a considerable number of publicati0ns treating Painting," in Heritage of the Kathmandu Valley: Proceedings of an
maps have appeared, mostly by art historians, students International Conference in Lubeck, June 1985, ed. Niels Gutschow
and Axel Michaels (Saint Augustin: VGH Wissenschaftsverlag, 1987),
of religion, and others with little specialized knowledge
13-27.
of cartography. Nevertheless, some of these studies are 14. Braham Norwick, "Locating Tibet-The Maps," in Tibetan Stud-
very informative and are cited in this chapter. Although ies: Proceedings of the 4th Seminar of the International Association
there is as yet no comprehensive work on the indigenous for Tibetan Studies, Schloss Hohenkammer, Munich, 1985, ed. Helga
mapping of Greater Tibet, such a volume could be rather Uebach and jampa L. Panglung (Munich: Kommission fOr Zentralasia-
tische Studien, Bayerische Akademie der Wissenschaften, 1988), 301-
substantial. A trilogy of well-researched articles by Nor- 20; idem, "Why Tibet Disappeared from 'Scientific' 16th-17th Century
wick reviews the geographic mapping of Tibet as a whole European Maps," in Tibetan Studies: Proceedings of the 5th Seminar
and discusses the ways that region has appeared on East of the International Association for Tibetan Studies, Narita, 1989, ed.
Asian, Arabic, and Western maps, as well as on a few Ihara Shoren and Yamaguchi Zuiho (Narita: Naritasan Shinshoji, 1992),
Tibetan works, since classical antiquity.14 For the much 633-44; and idem, "Modern Mapping of Tibet: A Cautionary Tale,"
presented at the 6th IATS Seminar, August 1992, forthcoming, read by
more limited area of Nepal, Gurung has brought out a me in draft form. N orwick's bibliography in the first of these three
small, well-illustrated volume, Maps of Nepal. 1s Almost papers promises to be particularly useful for future historical research.
all the maps treated within that work, however, are either 15. Harka Gurung, Maps of Nepal: Inventory and Evaluation (Bang-
European or essentially modern maps based on surveys kok: White Orchid Press, 1983).
conducted since the mid-nineteenth century, and the style 16. Schwartzberg, "Geographical Mapping," 429-35 and 455-57
(note 8).
of the few maps that might be regarded as essentially 17. Bibliographie du Nepal, vol. 3, Sciences naturelles, bk. 1, Cartes
indigenous is more Indian than Tibetan. Such works were du Nepal dans les bibliotheques de Paris et de Londres, by L. Boulnois
discussed, therefore, in relation to South Asia. 16 There is (Paris: Editions du Centre National de la Recherche Scientifique, 1973).
also a published bibliography of Nepali maps, but it con- 18. Hiroshi Nakamura, "Old Chinese World Maps Preserved by the
FIG. 15.2. SCROLL ON ANNUAL PUBLIC DISPLAY IN These are usually separated by clumps of trees or similar con-
NEPAL. This conon scroll, dating from the seventeenth or ventional devices. Thus one could construe the ensemble as a
eighteenth century, is shown hanging in the courtyard of the kind of narrative route map. Scrolls of this type have been
Guita-bahil monastery in Patan, in the Vale of Kathmandu. It recorded in the Vale of Kathmandu since about the mid-fifth
is one among a number of sacred relics that the monastery century A.D.
exhibits to the public for several days each year. The many Size of the original: unknown. Photograph courtesy of Mary
panels of the upper and lower registers of the scroll constitute Shepherd Slusser.
a series of vignettes depicting successive scenes in a narrative.
cussed in a wider context in other works by Nakamura. I9 painted scroll map hanging in a Nepali monastery court-
In a similar vein, Unno discusses numerous Chinese, yard as part of an annual display of sacred relics. Other
Korean, and Japanese maps embodying concepts origi- works are much more ephemeral. Figure 15.3 shows the
nating in India and Tibet or transferred from India to preparation of a cosmogram in the form of an elaborate
Tibet and thence transmitted farther east. 20 Finally, a mandala (Tibetan dkyil'khor) prepared with powdered
work originally written in Russian and later translated sand of various colors. Although such works may take a
into English deals with a modern recension of what is week or more to prepare, they are used only briefly in a
perhaps the most ancient Tibetan map to survive in any religious service at which, following certain offerings, the
form, allegedly a world map (fig. 15.27 below), said to mandala is swept away, reminding the faithful of the
be of an "Iranian-Tibetan" type, as well as a cosmography transitory nature of all existence. 22
of the much better known "Indian-Tibetan" tradition. 21 Although nature has been kind to the cultural legacy
In contrast to the environmental conditions in India of Greater Tibet, non-Tibetans have been less sparing.
and Southeast Asia, where humidity, vermin, and other
factors contributed to the decay and outright destruction
of indigenous maps, the prevailingly cold, dry climate of Koreans," Imago Mundi 4 (1947): 3-22. See also relevant nores in A.
Greater Tibet is conducive to their preservation. Hence, L. Mackay, "Kim Su-hong and the Korean Cartographic Tradition,"
Imago Mundi 27 (1975): 27-38.
in the numerous palaces, monasteries, temples, and other 19. See especially Hiroshi Nakamura, East Asia in Old Maps (Tokyo:
religious edifices that dot the landscape of the region one Centre for East Asian Cultural Studies, 1962), 8 and fig. 1.
finds mural paintings and manuscripts that, even after the 20. Unno Kazutaka, Chizu no shiwa (Map creases; or, Essays on the
passage of centuries, remain quite legible. Moreover, history of cartography) (Tokyo: Yilshodo Press, 1985); and idem, "The
Buddhist monasteries, like Jain monasteries in India, Asian Lake Chiamay in the Early European Cartography," in Imago et
mensura mundi: Atti del IX Congresso Internazionale di Storia della
assumed responsibility for preserving documents, and the Cartografia,3 vols., ed. Carla Clivio Marzoli (Rome: Istituto della Enci-
tasks of copying documents and adorning monasteries clopedia Italiana, [1985]), 2:287-96.
were important aspects of monastic life. Not only in 21. L. N. Gumilev and B. I. Kuznetsov, "Dve traditsii drevnetibetskoy
monasteries, however, but wherever Tibetan Buddhists kartografii (Iandschaf i etnos, VIII)," Vestnik Leningradskogo Univer-
assemble and in numerous private homes as well, one siteta 24 (1969): 88-101, translated as "Two Traditions of Ancient
Tibetan Cartography (Landscape and Ethnos, VIII)," Soviet Geography:
finds various types of cosmographies. These include both Review and Translation 11 (1970): 565-79.
abstract mandalas and other types on which the various 22. At least two examples of colored sand mandalas that have been
parts of the cosmos are identifiable, as well as maplike preserved, notwithstanding the rules calling for their destruction, are a
depictions of holy places and topographical paintings of rather small one, kept under a glass dome in the Jacques Marchais
places important in the lives of the Buddha, of numerous Center of Tibetan Art in Staten Island, New York, and a much larger
work constructed by visiting lamas at the Minneapolis Institute of Arts
bodhisattvas, and of other holy personages. in 1992. The latter work and its preparation are illustrated and described
Figures 15.2 and 15.3 show two of the many genres in Arts: The Magazine for Members of the Minneapolis Institute of
of sacred maps. Figure 15.2 depicts a small part of a long Arts, June 1992, front and back cover and 6-8.
FIG. 15.3. CONSTRUCTING A SAND MANDALA. Monks away the sand with no visible regret. The work in progress in
may work for a week or more to construct an elaborate cos- this diagram is being made at Ganden (dGa'ldan) monastery,
mographic mandala of powdered sand that, when finished, will which serves a Tibetan refugee community in southern India.
provide the focal object of a religious ceremony. Once certain By permission of Hakusuisha Publishing Co., Ltd., Tokyo.
prescribed ceremonial offerings have been made, they will sweep
Since the takeover of Tibet proper by the Communist their treasures. Fortunately, some of the greatest shrines,
regime of China, literally thousands of monasteries have in particular the Potala, the traditional residence of the
been razed or badly damaged, with incalculable loss of Dalai Lama, have been somewhat less affected, largely
because of their appeal to tourists. Moreover, in surviving

Buddhist temples and municipal and provincial museums
throughout China, Tibetan objects are held in high
regard. China's recent fostering of the tourist industry in
the once-forbidden territory of Tibet now provides a
significant incentive for cultural preservation. It is also
one among many factors-including the mystique Tibet
holds for many Westerners and the growing religious
appeal of Buddhism-that have sparked widespread inter-
est in Tibetica and a commensurate burgeoning of rel-
evant scholarship, much of it of a very high standard.
Cosmographies and other maps from greater Tibet may FIG. 15.4. SECTIONS OF ANEWARI TEXT ON ARCHI-
now be found in museums, libraries, and private collec- TECTURE. This is part of a series of drawings from an archi-
tectural manual still in use among Newaris, the dominant ethnic
tions throughout the world, including many that do not group of the Vale of Kathmandu and early recipients of cultural
specialize in Asian art. There are at least several dozen influences from India. The drawings stipulate types of floor plans
important public collections of Tibetica outside Tibet appropriate for a wide range of situations, depending on the
and other areas of China. Even in relatively small orientation of the house and on the situation of the immediate
museums, one may find significant specialized collec- neighborhood and quarter of the town, and they indicate bless-
ings and curses that may be expected from building in certain
tions-for example, in respect to Bhutanese art, at the ways.
Musee d'Ethnographie de Neuch,hel in Switzerland. The Size of the original: unknown. From Gerhard Auer and Niels
subject is stil1 very much in its infancy, and this chapter Gutschow, Bhaktapur: Gestalt, Funktionen, und religiose Sym-
can do no more than offer a sample of what exists in bolik einer nepalischen Stadt im vorindustriellen Entwick-
major collections worldwide. lungsstadium (Darmstadt: Technische Hochschule, 1974), 36.
the centuries, as have route maps intended especially for

THE DEVELOPMENT AND NATURE OF pilgrims.
TIBETAN MAPPING Copies of Indian Buddhist cosmographies entered
Tibet relatively early. The bhavacakra, a type of cos-
Maps from Greater Tibet have evolved over many cen- mographic diagram found today throughout Greater
turies and assume many forms. Innumerable mandalas, Tibet (e.g., fig. 15.18 below), is derived from an Indian
objects of Buddhist meditation, may be regarded as cos- prototype, arguably from one painted in either the fifth
mographic maps. Many of these are highly abstract or the sixth century on the wall at the entrance to cave
microcosmic representations of the entire universe. 17 of the great cluster of cave temples at Ajanta; but it
Others, styled as bhavacakras (Tibetan srid-pa'i-'khar- seems likely that both that and the Tibetan versions were
[a, wheels of life) are more obviously representational derived from a stil1 earlier sourceP Waddell notes that
and depict various realms and states of existence in time a copy of a bhavacakra is alleged, "with reason," to have
and space. Stil1 others envisage the universe as a geo- been brought to the country by an Indian monk in the
metrically ordered assembly of its deities. Various views eighth century.24 Whether or not this is true, it seems
represent the plane of terrestrial existence and the set of likely that some version of that cosmogram was trans-
continents therein, centered on the sacred Mount Meru, mitted to Tibet before the long period of decline of
the axis not only of the earth, but of the universe as a Indian Buddhism, which set in about the ninth century.
whole. Many focus on specific portions of the universe, A more clearly datable transmission, early in the thir-
such as its various heavens or hells. A popular cosmo- teenth century, was that of a true-scale stone model of
graphic game focuses on the paths by which salvation the renowned Mahabodhi Temple at the Indian town of
may be reached. Much-used charts, which I shall consider Bodh Gaya, where the Buddha attained enlightenment.
only briefly, are those that relate to astrology and other This relic and another construction copied from i[ were
forms of divination. And finally, paintings incorporating seen at Narthang (sNar thang) monastery in Tibet in 1936,
maplike components commonly depict the places asso- and both may stil1 be extant. 25
ciated with the lives of various deities, saints, or epic
heroes. Among geographic maps, most relate to relatively 23. John Huntington, an art historian specializing in Tibet, believes
small localities, especially towns where important mon- that both the Ajanta and Tibetan representations are based on some
unknown pre-Ajanta prototype (personal communication, 12 May
asteries are located or the monasteries themselves. But 1992).
maps of larger regions, even including at least two 24. Waddell, Buddhism of Tibet, 108 (note 5).
believed to be world maps, have also been produced over 25. For relevant details see Rahula Sankrityayana, "Second Search of
sPa-gro Byang Tshang-pa:

c H N A Au-gnon
sKyer-chu
left foot
YANG-'DUL Alung-gnon
left hand
mTha'-'dul
Mon: fHo-brag:
Yang-'dul Bum-thang MTHA'-'DUL Kho-mthing
doubtlul
left knee left elbow
Au-lag: gYu-ru:
Grom-pa-rgyang AU-GNON Khra-'brug
left hip left shoulder
...J
=> ...J
Z ;:: -<
9 => 0
:D ~ :»
9 z c :r z
<!l
z <.: <;l 6 l>: q>
« :r =>
z d
a: 0 d c
> ~
:::; z cr- r-
o 100
I I I I I heart-blood
o 100 200 300 km
right hip right shoulder
FIG, 15,5, DISPOSITION OF MAJOR TIBETAN TEMPLES. gYas-ru: AU-GNON dBu-ru:
gTsang-'gram Ka-tshal
This figure indicates the geographic distribution of twelve
important temples in three groups of four each, arranged in a right knee right elbow
rough approximation of what ideally ought to have been suc- Byang:
Pra-dun-rtse
MTHA'-'DUL Kong·po:
Bu-chu
cessively larger quadrilaterals, with the Jo-khang Temple at
right foot right hand
Lhasa at the center. The situation of Rlung-gnon Temple is Mang-yul:
especially problematic according to this schema. (The spellings Byams-sprin
YANG-'DUL mDo-khams GIO;~;b~~~~
here follow Aris.)

FIG. 15.6. IDEALIZED ARRANGEMENT OF MAJOR
After Michael Aris, Bhutan: The Early History of a Himalayan
Kingdom (Warminster, Eng.: Aris and Phillips, 1979), 16. TIBETAN TEMPLES. This figure is a closer approximation to
the idealized arrangement, according to the Tibetan derivative
of an originally Chinese model. The names in capital letters are
regional designations in three zones of diminishing civilization
Other evidence of cultural transmissions from India outward from Lhasa; those in upper- and lowercase letters are
relates to architectural and town planning. A number of the temple designations; and the text in all lowercase letters
Nepali and Tibetan manuscripts, for example, include tells the parts of the body of the underlying cosmic demoness
diagrams that set forth rules for laying out houses and to which each of the temples corresponds. Although shown
here as a square, the Tibetan model actually is intended to be
whole communities. 26 Figure 15.4 provides an example
aligned along an east-west axis, in contrast to the Chinese
from one such text. Other diagrams suggest that Chinese model. (The spellings here follow Aris.)
architectural models were also borrowedP Influences in After Michael Aris, Bhutan: The Early History of a Himalayan
respect to architectural plans were also transmitted Kingdom (Warminster, Eng.: Aris and Phillips, 1979), 16.
widely between Tibet and other Buddhist lands. 28
this volume is noteworthy for its exhaustive bibliography of primary
and secondary sources (569-79), Other illustrations from architectural
Sanskrit Palm-Leaf MSS, in Tibet," Journal of the Bihar and Orissa manuals appear in Pratapaditya Pal, Art of Nepal: A Catalogue of the
Research Society 23 (1937): 1-57; and Charles E, A, W. Oldham, "Some Los Angeles County Museum of Art Collection (Berkeley: Los Angeles
Remarks on the Models of the Bodh Gaya Temple Found at Nar- County Museum of Art in association with University of California
thang," Journal of the Bihar and Orissa Research Society 23 (1937): Press, 1985), 176-77, diagrams for the construction of SlUpas, among
418-28. others; Slusser, Nepal Mandala, vol. 2, fig. 89 (note 13), "depicting the
26. For similar South Asian examples, see Schwartzberg, "South Asian optimal city plan as a mandala of eighty-one squares upon which is
Cartography," 318-21 (note 1). The period from the eighth to the superimposed a circular plan (alavau) 'such as a serpent would make
twelfth century in particular was one in which there was a massive by bringing head and tail together'''; and Gerhard Auer and Niels
diffusion of artistic techniques and motifs from northeastern India, the Gutschow, Bhaktapur: Gestalt, Funktionen, und religiose Symbolik
last major bastion of Buddhism on the Indian subcontinent, to Nepal, einer nepalischen Stadt im vorindustriellen Entwicklungsstadium
Tibet, and Southeast Asia. This is abundantly documented in Susan L. (Darmstadt: Technische Hochschule, 1974), 39 and 87 (respectively, a
Huntington and John C. Huntington, Leaves from the Bodhi Tree: The plan for a filter system for an underground well in the shape of a
Art of Pala India (8th-12th Centuries) and Its International Legacy swastika and a set of illustrations for house construction, including
(Seattle: Dayton Art Institute in Association with the Univetsity of those shown in fig. 15.4).
Washington Press, 1990). 28. See, for example, Anne Chayet, "The Jehol Temples and Their
27. The most comprehensive scholarly discussion of Tibetan archi- Tibetan Models," in Soundings in Tibetan Civilization, ed, Barbara
tecture is Paola Mortari Vergara and Gilles Beguin, eds., Dimore umane, Nimri Aziz and Matthew Kapstein (New Delhi: Manohar, 1985), 65-
santuari divini: Origini, sviluppo e diffusione dell'architettura tibe- 72. Of particular interest in this article is the way the mixed axonometric
tana / Demeures des hommes, sanctuaires des dieux: Sources, devel- and divergent perspectives of Tibetan paintings (see fig. 15.50) were
oppement et rayonnement de I'architecture tibetaine (Rome: Universita distorted by Chinese architects who used only the former and were
di Roma "La Sapienza," 1987). Apart from its numerous illustrations, unfamiliar with the underlying conventions of the originals.
Not only were the layout and external appearance of

temples of concern to the Buddhists of Greater Tibet,
but so was placing them in accordance with systems of
divination and geomancy. These systems were a "com-
plex amalgam of Indian, Chinese, and Tibetan forms"29
but were essentially Chinese in origin. It is believed that
they were introduced by a Chinese princess, Kong-jo,
who in about A.D. 640 was sent to marry the great Tibetan
monarch Songtsen Gampo (Srong-brtsan-sGam-po) and
brought with her, as a gift from her father, "a divination
chart in 300 sections executed according to the Chinese
divinatory sciences."3o One important consequence of
the belief in geomancy relates to the positioning of
important Tibetan (including Bhutanese) temples.
Although the origins of these locations are partially FIG. 15.7. FRAGMENT OF AN OLD MAP OF THE DUN-
shrouded in myth, no fewer than ten texts arrange them HUANG REGION OF GANSU. This map fragment, painted
in black ink on paper, is thought to date from the ninth century
(though with less than perfect agreement as to their names
and is the oldest known Tibetan locality map as well as the
and locations) in three sets of four each, illustrated in oldest work to show any form of architectural plan, in this case
figure 15.5. Each of these texts presumes the existence of an unspecified Buddhist monastery. The locality depicted is
in the redactor's mind of a mental map, which is the believed to be Dunhuang in the present province of Gansu. The
Tibetan analogue, shown in figure 15.6, of the prior mixture of perspectives-oblique, frontal, and planimetric-is
noteworthy. Though the text is in Tibetan, the style is essentially
Chinese scheme conveyed by figure 4.1 above (p. 76) {with
Chinese.
five rather than three concentric zones out from the cap- Size of the original: 30 X 48 cm. By permission of the Biblio-
ital).31 theque Nationale, Paris (Pelliot tibetain 933).
The origins of geographic maps of Tibetan provenance
are obscure. If Gumilev and Kuznetsov's interpretation that date. 35 Obviously, further transmission of mapmak-
of the world map discussed below (pp. 639-42) is to be ing skills among Tibetans would have been relatively easy.
credited, that work should be regarded as an "Iranian- Some such transmissions will be noted below in the dis-
Tibetan" creation reflecting Persian geographic know- cussion of topographic mapping.
ledge from the third and second centuries B.c. 32 How The peoples of Greater Tibet had thus developed a
old the Sino-Tibetan world map brought to light by Tera- keen sensitivity to place, position, and relative location,
moto might be is also open to question, though it cannot .both geographical and cosmographic. They were recep-
be later than the ninth century when, as noted above, a tive to new ideas on such matters from a diversity of
copy of it was transmitted to Japan. Very different from
either of the foregoing is the oldest known locality map, 29. Michael Aris, Bhutan: The Early History of a Himalayan King-
dom (Warminster, Eng.: Aris and Phillips, 1979), 12.
which was found on a fragment of a Dunhuang manu-
30. Quoted from a Tibetan legend in Aris, Bhutan, 12 (note 29).
script thought to date from the Tibetan imperial era in 31. Aris, Bhutan, 3-33 (note 29).
the ninth century {fig. 15.7).33 Although there would 32. Gumilev and Kuznetsov, "Two Traditions," 568 (note 21).
probably have been even earlier opportunities for Tibet- 33. This work is illustrated and discussed in Vergara and Beguin,
ans to observe and learn from the mapping activities of Dimore umane, santuari divini, 89-91 (note 27), as well as in Dieux
their Chinese and other neighbors, the earliest I can doc- et demons de I'Himalaya: Art du Bouddhisme lamarque, catalog of an
exhibition at the Grand Palais, 25 March to 27 June 1977 (Paris: Sec-
ument relates to an expedition sent by the Mongol retariat d'Etat Ii la Culture, 1977),66-67; and in Per Kvaerne, "Tibet:
emperor Kubilay Khan (r. 1260-94) to explore and map The Rise and Fall of a Monastic Tradition," in The World of Buddhism:
the source region of the Huanghe (Yellow River), which Buddhist Monks and Nuns in Society and Culture, ed. Heinz Bechert
lies in what is now Qinghai, a culturally Tibetan province. and Richard Gombrich (London: Thames and Hudson, 1984), 253-70,
esp. 258.
The oldest surviving regional map based on that expe-
34. This map is illustrated and discussed at length in Herbert Franke,
dition (not counting Mongol maps of the whole of China) "Die Erforschung der Quellgebiete des Gelben Flusses in Nordosttibet
is in a text dated 1366.34 Subsequent Chinese mapping umer dem Mongolenkaiser Qubilai," in Der Weg zum Dach der Welt,
ventures in Tibet are well documented and are discussed ed. Claudius C. Muller and Walter Raunig (Innsbruck: Pinguin-Verlag,
elsewhere in this volume. Particularly noteworthy is the [1982]),59-61.
35. This is discussed in Clements R. Markham, ed., Narratives of
Jesuits' training two lamas to be surveyors in Beijing.
the Mission of George Bogle to Tibet and of the Journey of Thomas
These surveyors submitted a map of Tibet to those Manning to Lhasa (1876; New Delhi: Maiijusri Publishing House,
fathers in 1717; it was regarded as the most accurate to 1971), Ixi-Ixii.
sources, sometimes as a consequence of Tibetan conquest

of neighboring lands and often resulting from travels by
pilgrims between Tibet and other lands to which Bud-
dhism had been disseminated. They devised numerous
ways to express cartographically their view of the uni-
verse and its constituent parts. Their appreciation of maps
also resulted in their using cartography for a variety of
didactic purposes, largely teaching cosmographic ideas,
but also, for example, in drawing anatomical and medical
diagrams, preparing architectural models, and concep-
tualizing specific intellectual problems. 36 In this chapter
I shall examine a small sample of Tibetan maps, but
before doing so, it is in order to demonstrate the exist-
ence of certain canons by which they were drawn.
CANONS OF CARTOGRAPHY
FIG. 15.8. PREPARATION OF A MANDALA AS A PART OF
One is struck by the amount of conventionalization that THE INITIATION RITUAL FOR A TIBETAN MONK. Here
characterizes so many Tibetan maps, though not to the a novice monk at Chimre monastery is seen laying out the basic
plan of what will probably become an elaborate mandala fol-
extent of denying them their originality. Much has been lowing well-established conventions for the proportions of the
written by art historians about the stylistic and icono- elements. Other conventions will guide him in the use of color,
graphic canons of Tibetan painting, sculpture, and archi- the choice of signs, and so forth.
tecture; and it is clear that the rules that have evolved From Manfred Gerner, Architekturen im Himalaja (Stuttgart:
(and continue to evolve) are also largely applicable to Deutsche Verlags-Anstalt, 1987),48.
maps, both cosmographic and geographic, assuming that
And so on for several more pages.
those maps are made by monks and others whose training
Though usually prepared in two dimensions, mandalas
requires learning the conventions.
are always perceived as three-dimensional in the act of
Starting with the most abstract of cosmographic maps,
meditation. Among actual three-dimensional structures
a mandala, we see in figure 15.8 a monk demonstrating
that may also be seen as mandalas (though their principal
the laying out of the basic structure of what will ulti-
use is as reliquaries) are chortens, the term by which stu-
mately become a rather complicated design (compare fig.
pas are known in Greater Tibet. Every structural element
15.3). Constructing a mandala is an important exercise in
of a chorten has a specific cosmographic meaning. 38
the initiation of many novice monks, and learning the
requisite discipline takes much practice. Long poems, 36. Diagrams of various types of systems in the human body and of
comprising a series of rhyming verses, provide mnemonics the method of applying a medical treatment known as moxibustion
appear in Tresors du Tibet: Region autonome du Tibet, Chine (Paris:
to assist in this endeavor. Brief excerpts from one such Museum National d'Histoire Naturelle, 1987), fig. 89. Other anatomical
poem, in translation, follow: tables are illustrated in Liu Lizhong, Buddhist Art of the Tibetan Pla-
teau, ed. and trans. Ralph Kiggell (Hong Kong: Joint Publishing, 1988),
Having thus completed these preliminary rites,
314-15; Gunter Schuttler, Die letzten tibetischen Orakelpriester: Psy-
one must construct the place for the temple [for chiatrisch-neurologische Aspekte (Wiesbaden: Franz Steiner, 1971), fac-
which the mandala is a metonym] ... ing 65. Architectural models have been discussed above. A well-
Take light-coloured soil which is viscous, pliable illustrated discussion of what may be construed as a road map used for
and soft, ... didactic purposes appears in Charles Genoud, Buddhist Wall-Painting
Sprinkle it and make it completely smooth with of Ladakh, trans. Tom Tillemans (Geneva: Edition Olizane, 1982),
pure consecrated water and sweet-smelling painting no. 1 on unnumbered page in section on Spituk, and 103-6.
medicinal incense. A smaller photograph of the same map and a briefer explanation appear
Smear conon threads with white and red colouring in Richard Gombrich, "The Buddhist Way," in The World of Bud-
and consecrate them as Method and Wisdom dhism: Buddhist Monks and Nuns in Society and Culture, ed. Heinz
Bechert and Richard Gombrich (London: Thames and Hudson, 1984),
possessed of no duality.
9-40, esp. 28, pI. 22.
Cover the sphere of the void (viz. the space for the 37. David L. Snellgrove, ed. and trans., The Nine Ways of Bon:
ma1Jejala) with rays (viz. lines) of white and red, Excerpts from "gZi-brjid" (London: Oxford University Press, 1967),
(the four) bordering lines (of the square), 199.
(the four) crossing lines (two diagonal and two 38. A well-illustrated, though very speculative, study of the subject
straight across), for South, Southeast, and East Asia is Adrian Snodgrass, The Symbolism
the encircling line (inside the square),37 of the Stupa (Ithaca, N.Y.: Southeast Asia Program, Cornell University,
FIG. 15.10. THE POTALA OF LHASA. This detail from a nine-

teenth-century native drawing of Lhasa demonstrates the diver-
gent perspective so common on Tibetan paintings and drawings,
which permits the viewer to take in much more of the object
being presented than would the convergent perspective that
. characterizes Western an. Of panicular note is the depiction
of the complete path around the Potala, which, as the residence
of the revered Dalai Lama, is a shrine wonhy of circumam-
bulation by devout Buddhist pilgrims. In actuality, no more than
half of this path would be visible at anyone time. Several
pilgrims, one prostrate, are shown along the path.
Size of the original: unknown. From Laurence Austine Waddell,
The Buddhism of Tibet, or Lamaism, with Its Mystic Cults,
Symbolism, and Mythology, and Its Relation to Indian Bud-
dhism, 2d ed. (Cambridge: W. Helfer, 1935), facing 287.
FIG. 15.9. SYNTHETIC LANDSCAPE EMBODYING A DI- Finally we come to conventions followed in many geo-
VERSITY OF CONVENTIONAL ELEMENTS EMPLOYED graphic maps as well as in various types of hagiographic
IN TIBETAN MAPS AND OTHER FORMS OF PAINTING.
From David P. Jackson and Janice A. Jackson, Tibetan Thangka and narrative painting that commonly embody maplike
Painting: Methods and Materials (London: Serindia Publica- depictions of landscape. Figure 15.9 is a synthetic diagram
tions, 1984), 154, drawn by Robert Beer. incorporating a large number of the conventional land-
scape signs used in such works. It is taken from a book
that discusses the theory and practice of painting various
forms of thankas (thang-kas, painted cloth hangings), in
Though chaTtens are architectural features and hence, respect to composition, iconometry, color, shading, out-
strictly speaking, outside our purview, plans for con- lining, and so forth, specifically from an appendix that
structing them that appear in various manuals may be deals with motifs and signs. Within that appendix we find
considered cartographic. Such paradigmatic drawings numerous (as many as several dozen) acceptable ways to
have been reproduced in a number of published works. 39 draw trees, rocks, clouds, "cloud thrones," and other
The rules for construction are set forth by Olschak and
Thupten Wangyal, who introduce the subject by noting,
"The perfect proportions of the Buddha's body corre- 1985). A more complete and reliable study is The Stiipa: Its Religious,
Historical and Architectural Significance, ed. Anna Libera Dallapiccola
spond to the design of religious monuments."40 At a more in collaboration with Stephanie Zingel-Ave Lallemant (Wiesbaden:
elaborate level of detail are the rules for drawing bhava- Franz Steiner, 1980).
cakras (compare fig. 15.18). Although these rules will not 39. For example, Pal, Art of Nepal, 176-77 (note 27); and Blanche
be discussed, I note them here simply to demonstrate Christine Olschak and Geshe Thupten Wangyal, Mystic Art of Ancient
that well-established conventions apply to drawings of Tibet (1973; Boston: Shambhala, 1987), 18.
40. Olschak and Thupten Wangyal, Mystic Art, 18 (note 39).
considerable complexity.41 As with all such rules, how- 41. For further details, see Geshe Sopa, "The Tibetan 'Wheel of Life':
ever, exceptions developed that in time themselves Iconography and Doxography," Journal of the International Associa-
became conventions. tion of Buddhist Studies 7, no. 1 (1984): 125-45.
features that characterize Tibetan maps and other repre- widespread Tibetan conventions. Among these are the
sentational paintings. 42 use of a divergent (rather than a convergent) perspective,
Distinctive conventionalized ways of depicting settle- the inclusion of more features than the eye could possibly
ments and individual works of architecture have also take in at a single viewing, the essentially pyramidal
evolved in Greater Tibet. Some of these relate to colors arrangement of the composition (compare fig. 15.48), and
used for particular kinds of edifices (e.g., red is often the use of off-white throughout for residences and of red
selected to represent religious shrines and white for ordi- for temples.
nary residences), while others have to do with visual per- But not all Tibetan folk cartography reflects the seem-
spective. Figure 15.10, for example, depicting the famous ing familiarity with more schooled cartographic creations
Potala palace in Lhasa, illustrates what has been called found in the three examples previously discussed. A more
"divergent perspective," which is almost opposite to the naive, but nevertheless very informative, map of Ding-ri
optically "correct" Western perspective, with one or by a trader from that region has also been reproduced
more vanishing points. Here we see several sides of the and discussed by Aziz. 45 This work (fig. 15.11) is more
Potala represented simultaneously, which one could not idiosyncratic than the others. It was drawn spontaneously
normally see from a single location on the ground. But at Aziz's request, and the artist, then about forty years
to Tibetans familiar with the edifice, this approach makes old, had no familiarity with modern mapping. He had,
eminent sense, since it conveys more of the information however, traveled widely in Tibet and had probably seen
they want than would a Western view. various traditional maps from that region. Although his
map does incorporate certain common premodern con-
FOLK CARTOGRAPHY
ventions, such as placing the artist in the center of the
area portrayed and arranging various features and text as
Although most maps from Greater Tibet that have come if seen from that vantage point, the sorts of specifically
to light were made by specially trained individuals who Tibetan iconographic conventions displayed in figure 15.9
learned to follow the cartographic canons just noted, are nowhere evident. Compared with maps of individual
others are "naive" productions that nevertheless dem- settlements, generally drawn from an oblique perspective,
onstrate impressive cartographic skills. Many, perhaps essentially planimetric maps of regions are relatively rare
most, such maps were made at the request of foreign in Tibet. Nevertheless, if one compares figure 15.11 with
travelers and scholars. I have already noted that a number figure 15.12, one can readily discern many features on
of European travelers in the Himalayas benefited from the trader's map that reflect the topography shown on a
maps made by local informants; but I can recall no state- modern map. It would probably be difficult to obtain a
ment by a Himalayan traveler who had tried to obtain a comparably detailed and topographically reliable ren-
map locally and failed. The map obtained by the Schlag- dering of the home region by an untrained layperson in
intweit brothers during their expedition of 1854-58 (fig. most parts of the world.
15.42 below) is a good example of what may be regarded Whether living in an environment of high mountains
as folk cartography. Made by a Bhotia chief who had or largely barren plateau such as characterizes most of
evidently traveled much and seen many other works Greater Tibet somehow enhances cartographic ability or
employing the conventions discussed above, this map conduces to a particular approach to mapmaking is a
does embody some of them in simplified form; but overall question worthy of scholarly investigation. Some relevant
it has a spontaneous, untutored appearance. The same experiments, admittedly on a very modest scale, have in
may be said of a map painted by a Sherpa a century or
so later. 43 This charming work depicts the artist's home
42. See David P. Jackson and Janice A. Jackson, Tibetan Thangka
village at the foot of a mountain bearing rock formations Painting: Methods and Materials (London: Serindia Publications,
and surmounted by clouds strikingly like some of those 1984), 150-72. The conventions illustrated are even in contexts that
in figure 15.9. Within and around the village are men and have nothing to do with the landscape of Greater Tibet. On a visit to
women, represented as taller than their houses, going the Institut fOr Tibetologie und Buddhismuskunde in Vienna in 1983,
I was given a rather detailed map of that city prepared for a conference
about their day-to-day activities (one of which is praying
on Tibet by a visiting Tibetan scholar in a style that was immediately
together with a lama), as well as animals drawn in a sim- and unmistakably identifiable as Tibetan.
ilarly exaggerated size. 43. Herbert Tichy, Himalaya (Vienna: Anton Schroll, 1968), fron-
A third example of the partial absorption of Tibetan tispiece and relevant text.
cartographic conventions by folk artists appears in a 44. Aziz, "Tibetan Manuscript Maps," 35, with explanatory text on
34 and 36-37 (note 12); also reproduced in Aziz, Tibetan Frontier
painting of a townscape brought to light by Aziz. 44 This
Families, among unnumbered plates between 96 and 97 (note 12).
remarkably detailed map, painted in 1970 by an elderly 45. Aziz, "Tibetan Manuscript Maps," 32, text on 34 and 36-37
artist of Ding-ri Gang-gar, the sole town of the Ding-ri (note 12). Some of the details relating to this map were conveyed to
region, north of Mount Everest, also adheres to some me by Aziz on 12 November 1992.
_ >18,000 feet
o 16.000-18.000 feel
o <16,000 feel
o 5 10 miles
I I
I
I
o 10 km
.Chudrug
28"'5'
.v
FIG. 15.11. INDIGENOUS MAP OF DING-RI VALLEY,

TIBET. This is a map of the region of Ding-ri drawn from
memory in 1970 by Dinggang Nima Woser, a local Tibetan
trader, who had never seen a modern surveyed map. Drawn in
FIG. 15.12. MODERN MAP OF DING-RI VALLEY, TIBET.
ink and various pencils on paper. The orientation of map fea-
tures and text is outward from the center, where the town of This map covers the same area as figure 15.11 and is provided
Ding-ri Gang-gar is situated. The meanings of the signs em- for comparison.
ployed by the artist are, by and large, self-evident. The map is After Barbara Nimri Aziz, "Tibetan Manuscript Map of Dingri
oriented here with north at the top. Mount Everest is in the Valley," Canadian Cartographer 12 (1975): 28-38, esp. fig. 4.
lower right corner, and Rongphun is the circle directly above
it. Langkor is shown as the central circle in the range that runs
north-south on the left side of the map. Yoldon is the circle
southeast of the northernmost circle and is connected to it by
a dashed line.
Size of the original: ca. 48 X 30 em. Photograph courtesy of
Barbara Nimri Aziz.
fact been carried out. In Nepal the cognItive skills of The maps of all but the youngest Sherpa children ...
schoolchildren from various ethnic groups have been are generally more sophisticated than similar maps
evaluated by testing their ability to draw a map of the drawn by Newar, Limbu, Chhetri, and Gurung chil-
route between home and school. Similar tests have been dren ....
conducted among American students. Tests of Sherpa The maps of seventh-grade children [three out of
children among others were carried out and analyzed by four of which Fisher illustrates] attempt to symbolize
spatial relationships abstractly and clearly. All three
an American anthropologist, James F. Fisher, who has
maps not only show the house and school as recog-
had much experience working in the Himalayas. They
nizable buildings but keep them relatively small, much
yielded a total sample of twenty-eight maps from pupils more nearly proportional to the distances involved.
in grades four through seven from a single school in the The number of landmarks increases, and everything
Solo-Khumbu region of Nepal, virtually in the shadow is labeled so that a stranger looking at the map will
of Mount Everest. Of this sample Fisher reports: understand its symbols. Where trail intersections
might cause confusion, these older students are careful consisting of a lowermost realm, the Kamadhatu, or
to put in the other trail. That correct choices must be Realm of Desire; a middle realm, the Riipadhatu, or
made to arrive at the school, not somewhere else, is Realm of Form; and a uppermost realm, the Arupyad-
understood. 46 hatu, or Realm of Nonform. 48
4. The horizontal terrestrial plane, within the Kamad-
Noting that the maps of American children may seem hatu, constitutes only an insignificant part of the cosmos
to be more accurate (Fisher again uses the term "sophis- as a whole.
ticated") in that they are closer to the model Western 5. Around Mount Meru on earth are symmetrically
planimetric ideal, Fisher observes that the differences
arranged continents and seas, differently shaped in each
can be explained by cultural and environmental dif- of the four cardinal directions.
ferences between the two groups. Sherpas are used to 6. Each continent is of stupendous extent and is char-
an environment of steep slopes with relatively few flat acterized by a unique set of denizens, some divine and
places and must therefore handle spatial relations in others not.
three, rather than only two, dimensions.... To repre- 7. Neither the cosmos nor the terrestrial plane is
sent height on a two-dimensional surface is a chal- anthropocentric, since the southern continent, Jambii-
lenging topographic problem. In a bird's-eye view of dvlpa, is the only one where human beings live.
a steep slope ... , two points may look close together 8. Above and below the terrestrial plane within the
although they are actually far apart-but on a vertical, Kamadhatu are other planes of heavens and hells, where
not a horizontal, plane. The problem that Sherpa and
still other beings live.
most Nepalese children (and adults, for that matter)
9. The universe, like all animate creatures within it,
face is that faced by any mountain people explaining
how far away a place is. A destination may be only goes through cycles of emergence, decline, dissolution,
half a mile away but up a steep and difficult slope. and rebirth.
Therefore, as any trekker in Nepal knows, distance is 10. The most exalted cosmic state is nirvana, and
measured in time, not linear units. attaining it frees souls from the painful process of rebirth;
The Sherpa children thus tend to construct their but this state lies beyond the three levels of existence
maps to show the relation of higher and lower, sac- and is theoretically without form and therefore dimen-
rificing that of depth and width, so that the map repre- sionless.
sents a vertical cross section rather than a bird's-eye Much of the reality just described is represented graph-
view. 47 ically, and also in three dimensions, in a great variety of
The probable validity of the foregoing insights will, I ways, some of them quite abstract and others rather
believe, be demonstrated repeatedly in the analysis of a obviously representational. In this section I shall consider
variety of maps from Greater Tibet in the rest of this several of those ways, beginning with relatively abstract
chapter. mandalas and proceeding to genres, also largely con-
strued as mandalas, in which the component elements
are increasingly discernible to laymen and even to unin-
COSMOGRAPHIC MAPS itiated non-Buddhists. I shall also briefly illustrate how
maplike cosmographic images are incorporated into
In chapters 8 and 17 of this book, as well as in the chapter paintings of deities, saints, and epic heroes.
on cosmographic mapping in South Asia in volume 2,
book 1 of The History of Cartography, cosmographic ABSTRACT MANDALAS
ideas associated with various forms of Buddhism prac-
ticed outside Greater Tibet have been illustrated and dis- Although there are many varieties of mandalas, all have
cussed. Space limitations preclude comparably detailed in common the idea that they are a means of recognizing,
coverage of the exceedingly elaborate cosmography of through meditation, a divine palace, the abode of oneself
Greater Tibet, but in brief we may note that the region conceived as a Buddha. Abstractly a mandala may denote
shares with other Buddhist areas the following general
characteristics (most of them also applicable to Hinduism 46. James F. Fisher, Sherpas: Reflections on Change in Himalayan
and Jainism). Nepal (Berkeley and Los Angeles: University of California Press, 1990),
1. Our universe is but one among many millions exist- 82-89, quotations from 83-84.
ing in infinite space and in time measured in many millions 47. Fisher, Sherpas, 84-85 (note 46).
48. See table 17.1, p. 718, relating to the vertical ordering of the
of years.
Hinayana Buddhist cosmos and the placement therein of the major
2. Each universe is vertically ordered and centered on realms and subrealms. Although there are differences between this
an axis mundi, Mount Meru (Sumeru or Ri-rab). schema and that of Tibetan Buddhism, there is an overall similarity of
3. The essential structure of the universe is tripartite, structure.
West
remarkable accord between the Tibetan schema and the
alchemical diagrams of the elements found in Western
medieval and Renaissance astrological texts. The prin-
cipal difference is that the latter assigns a central position
to the human world, whereas the ineffable element
"ether" occupies that position in the Tibetan view. In
fact, the "human" world is not even an element in the
type of Tibetan view presented here. The connections
between the two call for additional research. 51 Only in
East
respect to the basic elemental structure are the spatial
and directional aspects of the mandala obvious; but it
would be a mistake to adopt a purely materialist per-
Amitabha spective and suppose that the remaining views are aspa-
ccs
>
ccs
..c: tial. Rather, one should see each view as suffusing all the
.0
E others in a complex, multivalent cosmic whole. But even
ccs
(/)
ccs at the elemental level the parts are regarded as being
c
Cti related in the ways they are successively subsumed into
ex: one another until only the psychic center representing
Ak~obhya
awareness remains. Lauf illustrates this point in a
sequence of diagrams that begins with the "Unity of All
Elements" and then proceeds successively with the dis-
solution of earth as it sinks into water, the dissolution
Wisdom of
Clear Sight
Vac of water as it sinks into fire, the dissolution of fire as it
sinks into air, and at the very end, the dissolution of air
as it sinks into awareness, the ultimate reality. This is
what is said to occur in the "process of death," which
initiates each stage of transmigration toward or away
Mirror-like from the ultimate goal of attaining nirvana. 52
Wisdom
The earliest surviving Nepali mandala, dating from the
thirteenth century, is one of only three known examples
of Nepali paintings predating the fourteenth century
FIG. 15.13. BASIC SCHEMA OF THE TATHAGATAMAN- (plate 33). The structure of this work is fundamentally
QALA. Six essential and inherently inseparable aspects of this the same as that of the model presented in figure 15.13
particular mandala, associated with the process of rebirth and
(or the one being created in fig. 15.3).
the attainment of nirvana, are shown in this diagram. It illus-
trates the essential plan of numerous other mandalas that are Although most of the hundreds of mandalas that one
also used as objects of meditation by Tibetan Buddhists, though
there are numerous differences of detail.
After Detlef Ingo Lauf, Secret Doctrines of the Tibetan Books 49. Giuseppe Tucci, The Theory and Practice of the Ma1J4ala, with
of the Dead, trans. Graham Parkes (Boulder, Colo.: Shambhala, Special Reference to the Modern Psychology of the Subconscious, trans.
1977), 111. Alan Houghton Brodrick (New York: Samuel Weiser, 1970; first pub-
lished by Rider, 1961), 25. This work provides an extensive, though
often controversial, discussion of the many forms and meanings of
Tibetan mandalas and both supplements and complements the better
the divinization of our environment. This is true irre- known, but also controversial, work by Carl Gustav jung, Mandala
spective of the scale or form in which the mandala is Symbolism, trans. R. F. C. Hull (Princeton: Princeton University Press,
perceived. Tucci defines a mandala as "a psychocosmo- 1972). Many individual mandalas are discussed by Tucci in his magnum
gram."49 Figure 15.13 shows a form of mandala associ- opus, Tibetan Painted Scrolls, 3 vols. (Rome: Libreria dello Stato, 1949).
ated with the Tibetan Bar-do thos-grol (Book of the Also noteworthy is Jose Arguelles and Miriam Arguelles, Mandala
(Berkeley, Calif.: Shambhala, 1972), which examines figures that may
dead), the so-called tathagatama1J4ala, that of five "tran- be construed as mandalas from a cross-cultural perspective. For the
scendent Buddhas ... , who, together with their female general interpretation of mandalas provided in this text I am largely
counterparts, the Buddha-Dakinls, open up a realm of indebted to Roger Jackson, Carleton College, Northfield, Minnesota.
symbolism rich in interrelationships."5o The sets of 50. Detlef Ingo Lauf, Secret Doctrines of the Tibetan Books of the
related meanings assigned to the various components of Dead, trans. Graham Parkes (Boulder, Colo.: Shambhala, 1977), 64.
51. For a discussion of similar and even earlier ideas from medieval
the mandala are elucidated by the six variations of the Europe, see John E. Murdoch, Antiquity and the Middle Ages, Album
basic diagram. It is noteworthy that at the elemental level of Science (New York: Charles Scribner's Sons, 1984), 360-68.
(view 1), showing water, earth, fire, and air, there is a 52. Lauf, Secret Doctrines, 90-91, esp. fig. 15 (note 50).
can see in published works on Tibetan an are paintings,

they also exist in other forms. Many are produced as
relatively simple black-and-white block prints. I have
already noted that elaborate mandalas are made out of
colored sand for use-and destruction-in religious cere- """"'-1:' ' <ki.
monies. Others, comparably ephemeral, are sculpted by .~
..,.
lamas out of butter mixed with various colored powders
... ~
~~';'.
. ~,~ca'
and worshiped by ordinary people in nightlong cere-
monies. 53 Though most, as previously noted, are essen-
16' .
1.4
",,%" '.
tially two-dimensional, many forms of three-dimensional e.,.J.{-l.~ ~\, ~

,..~j
~)
lOe
mandalas may also be found in monasteries, museums,
and elsewhere. Among these are certain stupas, each part
JlI""'
1
" :i>
+~:
~
of which embodies multiple meanings. For example, the S~

.
.~ -;'',;~~
five vertically arrayed components of a Tibetan stupa cor-
respond, in ascending order, to earth, water, fire, air, and ...",1i:" ....~.. ~
;;f;'"
di.~
ether, and each of those elements has corresponding
't~t;Vt<-
/'
properties such as color, tathiigata (form of a transcen-
,,~lJ,
dant Buddha), and so forth, analogous to those indicated b~t;~~U
in figure 15.13.54
VIEWS FOCUSING ON THE TERRESTRIAL PLANE

1. Sumeru; 2. Piirvavideha; 3. Jambiidvipa; 4. Aparagodaniya; 5. Ultarakuru; 6. Deha;
7. Videha; 8. Camara; 9. Aparacamara; 10. Satha; 11. Ultaramantrina; 12. Kurava;
As I noted in the brief introductory outline of Tibetan 13. Kaurava; 14. Sun; 15. Moon.
cosmography, within the terrestrial plane of the Kamad-

hatu a set of primary and satellite continents are sym- FIG. 15.14. THE GENERAL SCHEME OF THE MERU
metrically arranged around Mount Mem. So too are a MANDALA. This representation of the mandala, with text in
number of additional features, including the sun and the both Tibetan and Chinese, was produced, presumably for aca-
moon, which may be seen in figure 15.14, representing demic purposes, at the Peking Buddhist Institute, probably in
the so-called Meru mandala. 55 The work depicted the late 1930s. The size and media of the original are not known.
A description of the diagram appears in a compilation of minor
appears to have been prepared as part of an academic Tibetan works in a Western format (but bearing no entry regard-
undertaking in Beijing rather than as an object for wor- ing the date or place of publication) that was acquired by Alex
ship. Although some of its components are more Wayman at the Tibetan Press in Dharmsala, Himachal Pradesh,
obviously geographic than others, each has its properly India, in 1970. An analogous diagram appears in figure 15.15.
Some of the more obvious geographical and astronomical fea-
assigned place within the cosmos as a whole. Whether
tures have been identified by the numbers 1-15.
examples exist that show every such feature is not known, After Alex Wayman, The Buddhist Tantras: Light on Indo-
though some include a variety of features not indicated Tibetan Esotericism (New York: Samuel Weiser, 1973), pI. 8
there. (which contains a more complete translation than I have pro-
Depictions of the Meru system are very numerous and vided).
assume a multitude of forms. Many appear as block
prints, some quite simple and others relatively complex. 56 55. See Alex Wayman, The Buddhist Tantras: Light on Indo-Tibetan
Still others are on painted, embroidered, and appliqued Esotericism (New York: Samuel Weiser, 1973), 101-3.
thankas.57 A particularly powerful and expressive assem- 56. Examples of such prints may be found in Nik Douglas, T antric
blage of at least five large cosmographic mandalas, all Charms and Amulets (New York: Dover, 1978), figs. 70-74, with
accompanying text; Peter Gold, Tibetan Re{/ections: Life in a Tibetan
focusing on the terrestrial plain, is to be found in central Refugee Community (London: Wisdom Publications, 1984), 28; and
Bhutan. All are well preserved, but undated, fresco paint- Jackson and Jackson, Tibetan Thangka Painting, 38 (note 42).
ings on the walls of Bhutanese dzongs (castles), three in 57. A very clear example appears in Louis P. Van der Wee, "Rirab
Paro and two in Punakha. Another considerably more Lhunpo and a Tibetan Narrative of Creation," Ethnologische Zeit-
abstract, but nevertheless striking, cosmographic mural schrift, 1976, no. 2, 67-80, illustration on 77. The original work is in
the Teo S"rensen Collection of the Ethnographical Museum of the
is at a nearby dzong in Simtokha. 58 University of Oslo. The cosmography drawn by Giorgi, Alphabetum
Tibetanum (note 5), is also of this type.
58. Four such works from Bhutan are well illustrated, in color, in
53. Antoinette K. Gordon, Tibetan Religious Art (New York: Colum- Sugiura, Ajia no kosumosu + mandara, 39-41 (note 5). The same
bia University Press, 1952),92-93. source also illustrates a painting from nineteenth-century Tibet in a very
54. Detlef Ingo Lauf, Tibetan Sacred Art: The Heritage of Tantra, different style (21). Other sources depicting these and other Bhutanese
trans. Ewald Osers (Berkeley, Calif.: Shambhala, 1976), 139. paintings include Manfred Gerner, Architekturen im Himalaja (Stutt-
scholar) and a tenth-century Tibetan recension of the

originally Sanskrit Kalacakra T antra (text of the wheel
of time), Olschak and Thupten Wangyal discuss them
collectively as a sequentially ordered graphic testament
to the earth's evolution. 59 The earliest phase of the cos-
mogony, depicted on one of the Paro murals, is based
on the second of these texts. The painting gives much
prominence to the elements out of which our world is
aggregated. On it the cosmos is seen from above in the
form of four concentric rings, whose varying colors sym-
bolize the four elements, beginning with the outermost
band, representing air (yellow) and progressing inward
through fire (red), water (light blue), and earth (dark blue).
On this last ring the main and satellite continents are
ranged about Meru in the four cardinal directions, while
inside the ring are eighteen narrow, concentric, brightly
colored rings representing six successive triads of conti-
nent-mountain-ocean, a set of features reminiscent of an
Indian Puranic view. 6o The most central part of the paint-
ing is Meru itself, but a Meru not yet fully formed,
marked by a spiral wind out of which the elements are
condensed. Additionally, the painting includes twelve
narrow intersecting circles, all of the same size but with
centers spaced at thirty-degree intervals with reference to
Meru, that stand for the months of the year and a brick
red ellipse intersecting all the monthly circles that repre-
sents the trajectory of the sun. All of these features, in
turn, overlap the two wide bands symbolizing water and
earth. 61
FIG. 15.15. MERU MANDALA, PAINTED ON A WALL OF

THE PARO DZONG, BHUTAN. This view, of unknown date, gart: Deutsche Verlags-Anstalt, 1987), 46; Chantal Massonaud, "Le
is probably based on Vasubandu's Abhidharmakosa. Like figure Bhoutan," in Les royaumes de ['Himalaya: Histoire et civilisation
15.14, it indicates the terrestrial continents situated in the four (Paris: Imprimerie Nationale, 1982),67-116, esp. 113; Blanche Christine
cardinal compass directions from Mount Meru, each with its Olschak, Ancient Bhutan: A Study on Early Buddhism in the Hima-
distinctive shape and flanked by two smaller tributary conti- layas (Zurich: Swiss Foundation for Alpine Research, 1979), 42; idem,
nents. Meru, the home of the gods, rises from the center of The Dragon Kingdom: Images of Bhutan (Boston: Shambhala, 1988),
the painting and is surrounded by seven ranges of golden moun- front and back inside covers; Olschak and Thupten Wangyal, Mystic
tains (depicted as a set of nested squares). It obscures the north- Art, 108-9 (note 39); and Fran\oise Pommaret-Imaeda and Yoshiro
ern continent, Uttarakuru, whose presence is implied by the Imaeda, Bhutan: A Kingdom of the Eastern Himalayas, trans. Ian
two square tributary continents that are visible. Unlike figure Noble (Boston: Shambhala, 1985), pis. 23 and 26.
15.14, which is oriented with east at the top, this view is oriented 59. Olschak and Thupten Wangyal, Mystic Art, 108-9 (note 39).
toward the north, thus giving greater prominence to the south- 60. Joseph E. Schwartzberg, "Cosmographical Mapping," in The His-
ern continent,jambudvTpa, where humans live. Numerous icons tory of Cartography, ed. J. B. Harley and David Woodward (Chicago:
appear within the various continents, in the intervening oceans, University of Chicago Press, 1987-), vol. 2.1 (1992),332-87, esp. fig.
and on Meru itself. Although these have not been individually 16.2 (text on 336-37) and fig. 16.9 (text on 345).
identified, one may assume they correspond in the main to indi- 61. The foregoing description is based mainly on Pommaret-Imaeda
vidual elements indicated in figure 15.14. and Imaeda, Bhutan, 112 (note 58), and in part on Olschak and Thupten
Size of the original: unknown. From Sugiura Kohei, ed., Aiia Wangyal, Mystic Art, 108-9 (note 39). For a somewhat different inter-
no kosumosu + mandara (The Asian cosmos), catalog of exhi- pretation of the same works, see Massonaud, "Le Bhoutan," 112-14
bition, "Ajia no Uchukan Ten," held at Rafore Myujiamu in (note 58), which also differs from the quoted description in stating that,
November and December 1982 (Tokyo: Kodansha, 1982),41. rather than representing the months of the year, the intersecting circles
©Kodansha Ltd. 1982. actually stand for "douze vents qui sont aurant de chemins" (twelve
winds that are, at the same time, roads) (112); but she does not provide
the basis or significance of this interpretation. A very large thanka in
the Potala palace in Lhasa that appears to incorporate much of the
Although the works at Paro and Punakha are based same content as the Paro mural painting, and probably additional astro-
on two different texts, the Abhidharmakosa (a fourth- nomic information as well, is illustrated, though with minimal explan-
or fifth-century work by Vasubandu, an Indian Buddhist atory text, in Tresors du Tibet, fig. 84 (note 36).
The second and third paintings in the sequence, from

Punakha and Paro, respectively, also appear to be based
on the Kalacakra Tantra. The second shows the pri-
mordial world system just described as it begins to assume
a more differentiated form, with Meru rising out of the
earth and into the clouds and with the abodes of the
gods depicted on its various levels. On the third painting
plants and flowers begin to appear around the sacred
mountain. Both views depict Meru in frontal perspective
in the upper half of the painting and half the earth, in
planimetric perspective, on the lower half. The fourth
and fifth paintings, again from Punakha and from Paro,
are based on the Abhidharmakosa. On the fourth the
world continents are shown in greater detail than pre-
viously, and on the fifth (fig. 15.15), the fully formed
world system, they are even more prominently depicted.
Both paintings combine several perspectives: frontal for
Meru itself (thereby obscuring Uttarakuru, the northern
continent, which lies to its rear), oblique for the platform
out of which Meru rises, and planimetric for the rest of
the system. 62
Apart from representations of the terrestrial plane of
the type just discussed, there are others on which the
surrounding continents are either reduced to insignificant
size or omitted altogether, as in figure 15.16. 63 Other
representations vary widely in appearance and are pre-
sented in a number of published works. 64 There are also
some striking three-dimensional representations of Meru
that appear to conform to the Tibetan Buddhist concep-
tion. A particularly large example, in the Yonghegong
Temple in Beijing, provides a remarkably good match to
FIG. 15.16. MOUNT MERU AND ASSOCIATED FEATURES
the view presented in figure 15.16, though without the ON A TIBETAN TEMPLE BANNER. Painted on a cloth
surrounding continents. 65 Another, also from northern thanka. Date and provenance within Tibet not known. At the
China, is in the form of an elaborate gilt brass mandala base of this painting are what appear to be eight horizontal
with a diameter of 35.1 centimeters and a height of 37.5 rows of mountains (though only seven should properly be
centimeters. In this model the surrounding primary and shown) with intervening ocean. Above these are the four stages
of Meru itself. Within the uppermost stage are the sun and
satellite continents are presented as four sets of three moon, below which appear the Big Dipper and the Pleiades,
respectively. On Meru's summit stands the palace of the gods.
Ranged about Meru are signs for the primary and secondary
continents, Jambiidvipa in the south (bottom center, said to be
62. In this paragraph I have followed Olschak and Thupten Wangyal,
in the shape of the "shoulder blade of a sheep"), Piirvavideha
Mystic Art (note 39).
in the east, Aparagodaniya in the west, and Uttarakuru, rec-
63. This work is illustrated in Wayman, Buddhist Tantras, 104-5,
ognizable only by the two square satellites flanking the moun-
with explanation on 106-9 (note 55); Gordon, Tibetan Religious Art,
tain (the primary continent being hidden behind it), in the nonh.
47, with explanation on 44 and 49 (note 53); and Sugiura, Ajia no
Also shown are the "Seven Jewels of a Universal Monarch,"
kosumosu + mandara, 19 (note 5).
"Eight Buddhist Symbols" (identified with specific ladies), and
64. Two undated views of painted thankas are presented in Van der
Wee, "Rirab Lhunpo" (note 57) (apart from the planimetric view already wish-fulfilling trees.
Size of the original: unknown. Counesy of the Depanment of
indicated therein), one of unknown provenance in Van der Wee's per-
Library Services, American Museum of Natural History, New
sonal collection (fig. 1) and the other, from Ladakh, held by the Volk-
York (neg. no. 319427).
etkundemuseum der Universitat Zorich (fig. 6). The first is also depicted
in Armand Neven, Etudes d'art lamafque et de I'Himalaya (Brussels:
Oyez, 1978), 28 (fig. 12); and the latter in Sugiura, Ajia no kosumosu
+ mandara, 20 (note 5), and Unno, Chizu no shiwa, 192 (fig. 89) (note in which Meru figures only as a subsidiary element (e.g., as being held
20). A very clear oblique Sikkimese version, with an accompanying key, in the hand of the bodhisattva VajrapaQi), on which I will make no
appears in Mark Tatz and Jody Kent, Rebirth: The Tibetan Game of further comment. A very large fresco painting of Meru is reproduced
Liberation (Garden City, N.Y.: Anchor Books, 1977), 38-39. Van der in Moller and Raunig, Der Weg zum Dach der Welt, 293 (nore 34).
Wee, "Rirab Lhunpo" (note 57), also reproduces a number of paintings 65. Illustrated in Gerner, Architekturen im Himalaja, 46 (nore 58).
round, for me it is round. Even if it doesn't exist at

';Ill' C'<t·
"ll",,'C'
all, for me it will always remain round. 67
-----=.1<'"
------...c..
Meru is sacred not only to Tibetan Buddhists, but also
to Bon-pos, as the followers of the pre-Buddhist Bon cult
are called. Over the centuries the two faiths have bor-
rowed so much from one another that it is no longer
possible to untangle the lineage of particular practices
and concepts, including those relating to cosmography.
But it is clear that Bon cosmography, while incorporating
much that is unquestionably within the Buddhist tradi-
tion, also contains much that is idiosyncratic. Until
recently Bon had received little intensive study outside
Tibet itself; but in a pioneering work, mainly of trans-
lation, Snellgrove has set forth the more important por-
tions of Bon doctrine as given in a twelve-volume work,
whose short title is gZi-brjid (The glorious).68 Within that
work, a section titled "The Way of the Shen [a religious
officiant] of the Visual World" relates largely to cos-
mology. The gZi-brjid seems to have attained its present
form in the late fourteenth or early fifteenth century, and
the specific text Snellgrove worked from is believed to
be about four hundred years old. Several maps drawn by
a refugee Bon lama, Tenzin Namdak (bsTan-'dzin rnam-
dag), who collaborated with Snellgrove in preparing the
translation, are included in an appendix. One of these is
the Bon version of the Meru mandala, which in most
important respects conforms to that depicted in figure
15.14. 69 Another diagram from the same work, also relat-
ing to Meru, but more to its immediate precincts, is repro-
FIG. 15.17. BON DRAWING OF THE LHA SUM-CU-SA- duced here as figure 15.17,7° What is left to conjecture
GSUM GYI GZAL-YAS-KHAN (THE PALACES OF THE is the period during which maps of this type began to be
THIRTY-THREE GODS). This pen-and-ink drawing was prob- drawn in Tibet. My surmise, based on the known ante-
ably made in England by a Tibetan refugee, Tenzin Namdak
cedents of other works noted in this and the preceding
(bsTan-'dzin rnam-dag), a learned lama of the Bon cult. It illus-
trates a portion of the text outlining the major doctrines of his section and on the likelihood (to be discussed below)
faith, specifically some verses from "The Way of the Shen of that Bon cartography predates the arrival of Buddhism
the Visual World." In addition to the palaces, the map depicts in Tibet, would be that cosmological portions of the Bon
a sacred bird known as Khyung (the Garuda of Indian mytho-
logy) in a Tree of Paradise at the lower right, a number of parks,
and, it appears, at least four lakes (the stippled rectangular areas),
presumably in the four cardinal directions from the central Pal- 66. Illustrated in Marilyn M. Rhie and Roben A. F. Thurman, Wis-
ace of Victory. At the top of the diagram Mount Meru, becom- dom and Compassion: The Sacred Art of Tibet (San Francisco: Asian
ing ever broader from base to summit, is unmistakable. An Museum, 1991), 382.
Size of the original: unknown. From David L. Snellgrove, ed. 67. Van der Wee, "Rirab Lhunpo," 72 (note 57), my translation from
and trans., The Nine Ways of Bon: Excerpts from "gZi-brjid" the French original.
(London: Oxford University Press, 1967), pI. 21. By permission 68. Snellgrove, Nine Ways of Bon (note 37).
of Oxford University Press. 69. Snellgrove, Nine Ways of Bon, fig. XX (note 37). This and all
the other drawings accompanying Snellgrove's translation relate to items
"which bonpos take for granted as the normal possessions of their high
dignitaries and scholars" (264).
pavilions each. 66 Commenting on these differences, Van 70. For more on this map, see Snellgrove, Nine Ways of Bon, 91
der Wee cites a cosmogonic narrative related by Thubten (Tibetan text on 90), from "The Way of the Shen of the Visual World,"
in relation to fig. XXI (note 37). Since Bon is a little-studied minority
Jigme Norbu, who stated: cult (though Bon practices permeate Tibetan Buddhism, and vice versa)
and since Bon lamas are not numerous outside Tibet, in which country,
If I think of Rirab Lhunpo [one of the many names were it more accessible, older Bon maps could surely be found, it has
for Meru], it doesn't matter whether it be round or been necessary to use a modem drawing here. Given the source, how-
square.... What is important is what my faith signifies ever, I believe it may be considered as adhering rather closely to tra-
for me. Perhaps it is square; but if I believe it to be ditional models.
texts, like those of Tibetan Buddhists, would quickly . Detailed analysis of this and many other Tibetan paint-
~ave given rise to various forms of cartographic expres- Ings reveals what are in effect maps within maps, since
sIon. some of the scenes depicted in the several portions of
the work would individually qualify as maplike views of
townscapes, landscapes, or combinations of the two, usu-
COSMOGRAPHIES FOCUSING LARGELY ON
ally drawn in a characteristically Tibetan oblique per-
THE TEMPORAL DIMENSION
spective. This tendency in Tibetan cosmographic art to
In contrast to the present-oriented and materialistic sec- create a collage of diverse ,elements compounded of views
ular worldview of most inhabitants of industrialized of mythic and nonmythic places is even more evident in
states, that of Tibetan Buddhists is largely concerned with figure 15.25 below.
the hereafter and the rewards or punishments that one A type of Tibetan cosmography, little known or under-
will reap for the cumulative merit earned in past and stood by non-Tibetans, is the tshogs-zhing, or field of
present lives. This attitude regarding sarrzsara (Tibetan assembly (fig. 15.19). Like bhavacakras, such works are
,khor ba, the cycle of transmigration/rebirth) is reflected largely concerned with cosmic time as well as with cosmic
in numerous cosmographic diagrams in which the tem- space. Their principal function is to help worshipers vis-
poral dimension figures prominently, either implicitly or ualize "the objects to which one goes for refuge," here
explicitly or both. Such diagrams form the subject of the conceived as a virtually numberless host of Buddhas and
following section. other spiritually advanced beings (aryas) whose existence
Bhavacakras (wheels of life/existence/becoming) are may be apprehended in both time and space. 75 The image
important cosmographic means for conveying to the presented appears largely in the form of a tree atop which
masses the essentially soteriological belief central to sits a guru (teacher), depicted as either Sakyamuni Buddha
Tibetan Buddhism. Such works, often painted at the or another major religious figure, in this case Tsong-kha-
entrances to temples and monasteries, remind viewers of pa (1357-1419), the founder of the Gelukpa (bGe-Iugs
the realms of existence souls pass through in successive pa, Yellow Hat) monastic order. The literal meaning of
incarnations and warn them of the awful fate awaiting tshogs-zhing, however, is "merit sphere or object," and
those who cling to worldly desires. In keeping with their it may be interpreted unambiguously as referring to a
largely didactic purpose, such diagrams are convention- field, not a tree as some commentators have suggested.
ally simplified depictions of a portion of a complex Further, although the field is presented on the two-dimen-
cosmography relating to the Kamadhatu, or Realm of sional surface of a thanka,
Desire. The bhavacakra is discussed in considerable just as in the visualization of a mandala-what is two-
detail by Waddell. 71 It typically illustrates six subrealms dimensional becomes three dimensional when visu-
within the Kamadhatu, each as a segment of a wheel, alized. Therefore, although ... all the deities below
and by implication includes the Riipadhatu and Aru- the guru appear to be on the front of the tree, they
pyadhatu as well, since the wheel is believed to contain are actually arrayed around the tree. For example, ...
the four world protectors . .. all appear along the
all of sarrzsara. The lowermost segment generally
lower front of the tree, while in a visualization each
includes a multiplicity of hells. Waddell designates it "the
would be found in his appropriate direction.... The
Great Judgment" and shows it as compounded of images guru who is the focus of the tshogs-zhing sits, both
of eight hot hells and eight cold hells, as well as repre-
sentations of the sinner, the weigher (of good and evil
past deeds), the judge, and angels of good and evil. The
71. Waddell, Buddhism of Tibet, 77-122 (note 5).
segments to the left and right of the Great Judgment
72. Waddell, Buddhism of Tibet, illustrations on 102 and 109 (note
represent respectively the world of tantalized spirits (pre- 5).
tas) and the animal world. The uppermost segment, pre- 73. See Eleanor Olson, "The Wheel of Existence," Oriental Art, n.s.,
dictably, represents heaven(s), while the segments to its 9 (1963): 204-9. Olson also provides an illustration of the work as a
left and right signify the human world and the world of whole.
74. See, for example, the work illustrated by Sopa, "Tibetan 'Wheel
titans (asuras). Like the Great Judgment, each of the
of Life,' " 131 (note 41). Other examples of bhavacakras may be seen
other five segments can be subdivided into a number of in numerous museums and in illustrations in many publications, includ-
subrealms. 72 Figure 15.18, showing the upper portion of ing several cited in this chapter.
a bhavacakra, is not identical in its arrangement to the 75. Roger Jackson, "The Tibetan Tshogs Zhing (Field of Assembly):
work described by Waddell, though it is close in spirit. 73 General Notes on Its Function, Structure and Contents," Asian Philo-
sophy 2, no. 2 (1992): 157-72; quotation on 159. Jackson provides a
Alternative views show a five-segment bhavacakra, with
list of fourteen published representations of fields of assembly (170 n.
hell at the bottom, animal and preta realms to its left 3), in addition to the one his own account relates to (157, pI. 1). The
and right, respectively, and, at the top, a realm of both similarities from one view to another in the examples I have seen are
gods and titans to the left and of humans to the right. 74 striking.
FIG. 15.18. A TIBETAN BHAVACAKRA. This picture from nation in particular realms, each symbolized by an animal: a pig
an eighteenth-century (?) thanka, painted on cotton cloth, is for ignorance, a snake for anger, and a cock for desire. On the
representative of many such works depicting the various realms, periphery of the wheel there are symbols denoting each link on
or domains of existence, within the Kamadhatu, the lowermost a twelvefold chain of causation. The entire wheel is in the grip
major component of the vertically structured tripartite universe. of Shinje, a powerful monster, believed to be the wrathful coun-
Within the wheel, the three favorable upper domains are, from terpart of the compassionate AvalokiteSvara, the bodhisattva
left to right, those of asuras (titans or demigods), gods, and regarded as the protector of Tibet. Shinje's spiritual insight is
humans, while the unfavorable domains below, again from left indicated by his third eye.
to right, are those of pretas (hungry ghosts), hells, and animals. Size of the original: ca. 109 X 86 em. By permission of the
Within the circular area at the center of the bhavacakra is a Newark Museum, Newark, N.J. (ace. no. 36.535).
representation of the principal factors responsible for reincar-
FIG. 15.19. THE TIBETAN TSHOGS-ZHING, OR FIELD OF in which power is manifested diachronically. The two spheres
ASSEMBLY. The assemblage of divinities in this diagram repre- converge at the center in the body of a major guru, seen here
sents to Tibetan worshipers a field of merit and a place of as the focus of the two spheres. Maitreya, the future Buddha,
spiritual refuge defined in both temporal and spatial terms. The in his palace in the Tu~ita heaven, at the upper left, and Ami-
bottom half of the diagram, occupied by a hierarchy of eleven tabha, the heavenly Buddha, in his palace in SukhavatI, the
discrete lineages of gurus arrayed on a tree, is conceived syn- western paradise, at the upper right, seemingly preside over the
chronically and seen as representing, in effect, a sphere within entire assembly.
which their power is manifested. The three lineages of deities Size of the original: unknown. By permission of the Gerd-Wolf-
in the upper half are conceived as representing a temporal sphere gang Essen Tibetica Collection, Hamburg.
game board, which may be construed as a map; and the

rules of play.
The current Tibetan version of the game board consists
of a gridded field of 104 rectangular spaces, in thirteen
rows and eight columns, each of which represents a spe-
cific place in the Buddhist cosmos, whether it be an actual
realm in space (e.g., jambiidvlpa, Meru, Shambhala, or
"The Black Rope and Crushing Hells") or a state of mind
FIG. 15.20. TWO SPACES FROM THE TIBETAN GAME OF or being (e.g., "Wisdom Holder among the Gods of Sense
REBIRTH. These two spaces, 17 (right) and 18 (left), represent Desire").81 To Westerners the distinctions may seem
Jambiidvlpa and Aparagodanlya, the continents lying to the obvious, but to Buddhists they are not. The players'
south and west of Mount Meru. The former is an easily rec-
ognizable shape, said to resemble the shoulder blade of a sheep. object is to advance through the field, according to
The same shape may be seen for Jambiidvlpa in figure 15.15, throws of a die, with the goal of reaching nirvana. These
for example. The shape of Aparagodanlya, however, departs places and states are distributed on the board in an
slightly from its customary circular form. The board as a whole ordered manner, with hells occupying much of the bot-
may be construed as an aggregation of maps composing, in tom two rows and increasingly exalted realms or states
effect, a larger cosmic constellation by which the sacred geo-
graphy of Buddhism may be learned through play. occurring as one moves upward. Each space contains a
From Mark Tatz and Jody Kent, Rebirth: The Tibetan Game symbol indicating its general and more specific nature
of Liberation (Garden City, N.Y.: Anchor Books, 1977). (e.g., all hells are indicated by circles, while individual
hells are differentiated by color or other added graphic
devices), as well as statements specifying the conse-
quences of particular throws of the die that cause one
literally and figuratively, at the intersection of the to be reborn in a higher or lower state. Figure 15.20
diachronic spiritual power of the guru-lineage [in the illustrates two adjacent spaces, those relating to the
sky above] and the synchronic spiritual power of the southern and western continents of the terrestrial realm.
pantheon [arranged on the tree below]. It is into Overall, the scheme constitutes a complex "combination
spheres corresponding to these two types of power of several overlapping systems," which are not apparent
that the tshogs-zhing basically is organized,76 from mere visual inspection of the game board, including,
The place referents in the diagram are numerous. For for example, a path through the realm of sense desire
example, the bodhisattvas Maitreya and Amitabha are within the Mount Meru world system and a path and
seated in their palaces in the Tu~ita heaven and the west- system relating to the various Buddhist heavens. 82 "The
ern paradise, SukhavatI (depicted in the upper-left and
upper-right corners), and the world protectors stand 76. Jackson, "Tibetan Tshogs Zhing," 162 (note 75).
guard over the four cardinal directions at the base of the 77. Jackson, "Tibetan Tshogs Zhing," 164 (note 75).
guru lineage. 77 78. Tatz and Kent, Rebirth, 1 and 19 (note 64).
Concern with the cosmography relating to rebirth is 79. Tatz and Kent, Rebirth, 1 (note 64).
80. F. E. Pargiter, "An Indian Game: Heaven or Hell," Journal of
not merely associated with worship among Tibetans, but the Royal Asiatic Society of Great Britain and Ireland, 1916,539-42.
also finds expression in play. One vehicle by which such Pargiter ascribes the game to the Vaishnavite form of Hinduism but
concern is channeled is the game of "Rebirth" (full cor- offers no probable date for its origin. Despite the alleged Tibetan inven-
rect title, "Determination of the Ascension of Stages"), tion, one cannot rule out the diffusion of some prototype of the game
allegedly invented in the thirteenth century by Kunga from India to Tibet.
81. Tatz and Kent, Rebirth, 62-63 (note 64). A complete folding
Gyaltsen (Kun-dga'-rgyal-mtshan), a scholar of the Sakya game board, drawn by a Tibetan artist, is provided in the end-cover
(Sa-skya) sect. 78 This game, played by Tibetans of all ages, pocket of the book. Photographs of other versions of the game, with
is an important educational device, "inculcating in chil- different numbers and arrangements of spaces, are also provided, includ-
dren the Buddhist map of the world and an understanding ing a Tibetan block print of the original nine-by-nine-square field, a
of the workings of karma."79 Since then, many regional modern Bhutanese version, and a nineteenth-century version; a Korean
version is also described, but not illustrated (11-15). Waddell, Bud-
variants of the game have spread throughout the world dhism of Tibet, 471-73 (note 5), illustrates and discusses a version of
of northern (Mahayana) Buddhism as well as in India, the game that he acquired, probably in the late nineteenth century, with
from whence it was adapted by the English in its present a small playing field of only eight by seven squares. Slusser, Nepal
popular form as "Snakes and Ladders."8o In a remarkably Mandala, vol. 2, pI. 331 (note 13), illustrates a Hindu version of the
thorough and abundantly illustrated study of the game, game from the Vale of Kathmandu. Loden Sherap Dagyab notes that
Tibetan game boards he has seen may be as large as two meters high
Tatz and Kent discuss its origins, development, subse- and one and a half meters wide (Tibetan Religious Art, 2 vols. [Wies-
quent diffusion, and philosophic basis in respect to the baden: Otto Harrassowitz, 1977J, 1:41).
Buddhist belief in karma and rebirth; the nature of the 82. Tatz and Kent, Rebirth, 32-47 (note 64).
various paths may be taken to represent the cultural diver- ingly vague terms as the journey progresses), it is not
sity of humankind.... These may also be interpreted as commonly depicted in paintings, perhaps because it is
differing attitudes toward religion. "83 Though most regarded as attainable only by accomplished yogis. 86
squares, paths, and systems relate to the world of Tibetan Nevertheless,
Buddhism, there are also several paths by which non-
storytellers used to wander Tibet with paintings they
Buddhists-Hindus and Muslims-can rise to the level of would unroll before audiences in order to illustrate
"wisdom holder," and another by which followers of the subjects such as the journey to Shambhala. In one such
indigenous Bon cult of Tibet can attain that same happy performance, witnessed . . . in Lhasa, the storyteller
state, yet far short of nirvana. 84 indicated on his painting how the traveler to Sham-
bhala must climb a stairway to the top of a mountain,
where his body will become light as an insect, enabling
MAPS OF SPECIFIC PORTIONS OF THE COSMOS him to walk to the kingdom on clouds. 87
Apart from features associated with the Mount Meru The Tibetan version of the earlier mentioned Kala-
system, many other parts of the Tibetan spiritual uni- cakra Tantra derives from one brought from India, prob-
verse-heavens, paradises presided over by particular dei- ably in the tenth or eleventh century, though a still earlier
ties, other sacred places, and hells-all find places in the version is traditionally believed to have been brought to
cosmographic art of the region. Here I note only a few India from Shambhala itself. The text describes in great
examples, beginning with an especially popular subject detail that utopian realm, founded by an Indian king,
for religious paintings, the western paradise, SukhavatI Sucandra, in a blessed, mountain-girt land in a location
(literally, the Happy Land), chiefly identified with the revealed to him by the Buddha during his lifetime on
Buddha in the form of Amitabha (Immeasurable Glory). earth (800 B.C. in the Tibetan tradition). Thereafter Sham-
In figure 15.21 SukhavatI is shown on a thanka that forms bhala was ruled by a long line of monarchs, each reigning
the focal object of worship at a Buddhist altar. Detailed for a hundred years and each an incarnation of a partic-
descriptions of this peaceful realm appear in numerous ular bodhisattva. Tibetans continue to believe in the
ancient Buddhist texts, going back perhaps as far as the earthly existence of Shambhala and in the prophecy of
third century A.D. It is said to loom from a dark sea as the Kalacakra T antra that when the world reaches a
a mountain of gleaming copper. From the summit a certain state of moral decay, a savior-king will emerge
bejeweled palace rises into a heaven populated by a host from Shambhala, defeat the forces of evil, and bring
of cloud-mounted gods and goddesses. Those fortunate about a new order of world peace.
enough to be reborn in SukhavatI enjoy continuous fes- The Tibetan recension of the Kalacakra T antra
tivity, marked by music, dancing, bright banners, and
other delights. Surrounding the palace is a zone of silence
marked off by a high embankment built of skulls. Never- 83. Tatz and Kent, Rebirth, 41 (note 64).
theless a winding, narrow road, suggestive of a navel cord, 84. Tatz and Kent, Rebirth, 33 (note 64), present a "Map of the Game
does cross the sea and ascends from the earthly domain Board" that is in fact a deconstruction of the five systems and several
subsystems incorporated within it.
to a golden gate in the wall, thus providing access to the
85. This description was abstracted from a much fuller discussion in
palace. Through appropriate meditation, the worshiper Gerd-Wolfgang Essen and Tsering Tashi Thingo, Die Gotter des Him-
aspires to traverse this difficult spiritual path. The sea that alaya: Buddhistische Kunst Tibets, 2 vols. (Munich: Prestel-Verlag,
must be crossed represents the state of bardo, the period 1989), 1:201-2, including a full-page view of the thanka. The authors
between lives, when the soul, in purgatory, is judged in refer to SukhavatI as the southwestern paradise rather than the western
paradise, but most texts use the latter designation. Other representa-
terms of past good and bad deeds to determine the state
tions, quite different from the one illustrated here, appear in various
to which it will be consigned in its next birth. 85 publications: one that appears to be a woodblock print, with an accom-
Paintings of the northern paradise of Shambhala (fig. panying detailed Tibetan text, is presented by Waddell, Buddhism of
15.22), in contrast to those of SukhavatI, generally Tibet, 140 (note 5); a painted essentially planimetric view in Detlef Ingo
employ an essentially planimetric perspective, thereby Lauf, Verborgene Botschaft tibetische T hangkas / Secret Revelation of
Tibetan Thangkas (Freiburg im Breisgau: Aurum, 1976), 49, with text
emphasizing its wheel-like appearance. Because Sham-
on 48; and a fresco painting three and a half stories high on the chorten
bhala is identified with the deity Kalacakra, a name sig- of Dumtse lhakang at Paro in Bhutan, of what I take to be SukhavatI,
nifying "wheel of time," the depiction of his domain as in Massonaud, "Le Bhoutan," 110 (fig. 39) (note 58).
a mandala (literally, circle) is especially appropriate. 86. An exhaustive and well-illustrated treatment is provided by Edwin
Within the wheel, however, details are shown in an Bernbaum, The Way to Shambhala (Garden City, N.Y.: Anchor Press,
1980). Also useful, though without indigenous maps, is Peter Bishop,
oblique perspective. Curiously, despite the prominence
The Myth of Shangri-La: Tibet, Travel Writing and the Western Cre-
of Shambhala in the popular religion of Tibet and the ation of Sacred Landscape (London: Athlone Press, 1989).
existence of several Tibetan and Sanskrit guidebooks 87. Edwin Bernbaum, "The Hidden Kingdom of Shambhala," N at-
describing the way by which it can be reached (in increas- ural History 92, no. 4 (1983): 54-63, esp. 59.
FIG. 15.21. ALTAR WITH THANKA (SCROLL HANGING) may form focal points in the meditation that is characteristic
SHOWING SUKHAVATI, THE WESTERN PARADISE IN of the worship of Tibetan Buddhists.
THE COSMOGRAPHY OF TIBETAN BUDDHISM. From Size of the original: 58 X 39 cm. By permission of the Gerd-
Tibet, nineteenth century. This painted scroll occupies a central Wolfgang Essen Tibetica Collection, Hamburg.
place in the altar and is one among many sacred images that
FIG. 15.22. SHAMBHALA, THE NORTHERN PARADISE enment amid the depravity and delusion of the surrounding
OF TIBETAN BUDDHISM. From Tibet, late eighteenth cen- world. In this picture mountain ranges not only set off Sham-
tury, a painted cloth thanka. To reach the hidden idyllic king- bhala from the outside world, but also separate the capital from
dom of Shambhala, travelers must undertake an epic journey the eight outer portions of the realm and divide those regions
across deserts and mountains and overcome all manner of inter- from one another. Other paintings show streams, rather than
vening natural obstacles. Those who achieve their goal will mountains, between neighboring outer regions and depict their
encounter a country of beautiful cities and parks divided into shapes with less rigid regularity than in this view.
the shape of a lotus blossom with eight petals; in the center, Size of the original: 65 X 46 cm. By permission of the Gerd-
in a magnificent palace, lives the king, an incarnate bodhisattva. Wolfgang Essen Tibetica Collection, Hamburg.
The lotus here represents the emergence of purity and enlight-
"reflects the influence of various non-Buddhist religions plex and esoteric. A particularly rich assemblage of astro-
that were found in Central Asia ... most notably Nes- logical drawings from a nineteenth-century Tibetan
torian Christianity, Manichaeism, and Islam," suggesting almanac appears in Die Cotter des Himalaya, together
to some Western scholars that there once really was a with a detailed key to its contents. The almanac is of
kingdom of Shambhala somewhere within that general special interest because charts based on the different orig-
region. 88 Helmut Hoffmann actually sought to find it, inal systems appear alongside one another on the same
using the guidebooks in much the same way as Heinrich large page. 94
Schliemann drew on the Iliad and the Odyssey to locate A genre of astrological divination that can be found
Troy, and he concluded that it lay in the Pamir Moun- in related forms in China (its apparent place of origin),
tains to the east of Samarkand. Others have suggested in Tibet, and with considerable modification, in India is
the Tarim basin or the Turfan Depression as likely of the type illustrated in figure 15.23.95 This artifact hap-
regions. 89 Based on the ring of mountains shown on the pens to be in bronze; but more commonly works of this
available cosmographic maps, a case can be made for any type, the making of which was traditionally an important
of those areas, especially if one recognizes that they were source of income for astrologers, would be painted on
once much better watered and therefore more felicitous cloth or paper, especially those kept on family altars. The
environments than they are at present. principal component of diagrams of this type is the body
Within Shambhala, according to followers of the Bon of a tortoise (rather abstractly depicted in this instance)
cult, there exists a "Nine-Stage Swastika Mountain." A or a frog. In the Chinese and Tibetan versions, the body
very elaborate drawing of this mountain, together with of whichever creature is used is divided into nine parts
a lengthy caption in Tibetan, accompanies Snellgrove's in a magic square containing nine smaller squares, in three
translation of Bon doctrines. The much briefer English rows and three columns, each with a number from 1 to
caption simply states that the mountain "representing the
Nine Ways of Bon . .. [is] surrounded by its eight royal
palaces in the country known variously as sTag-gzigs, 88. Bernbaum, "Hidden Kingdom," 56 (note 87).
'Ol-mo-lun-rin, Sambhala, etc.," but it fails to identify 89. Bernbaum, "Hidden Kingdom," 56 (note 87). Among the extant
the other features shown in that intriguing diagram, maps of Shambhala is a striking nineteenth-century Mongolian thanka,
which in some respects resembles Jain renditions of Jam- remarkable in that it shows Shambhala only on the right side of the
painting. This work is held by the Musee Guimet, Paris, and illustrated
budvlpa and surrounding islands. 90
in Rhie and Thurman, Wisdom and Compassion, 378-79 (note 66).
Apart from the examples noted above, there are 90. Snellgrove, Nine Ways of Bon, fig. XXII (note 37); see Schwartz-
numerous additional maps of heavens, hells, and other berg, "Cosmographical Mapping," 367-72 (note 60), for Jain examples.
portions of the Buddhist cosmos. Many form parts of 91. For example, a frighteningly vivid Tibetan depiction of a series
detailed hagiographic compositions (to be examined of hells arranged in eight concentric rings below Mount Meru (pre-
sumably the "Eight Hot Hells" sometimes shown in the lowermost
below), while others focus on a single place or a group
segment of the bhavacakra), painted on a thanka is depicted in Sugiura,
of closely related places. 91 Ajia no kosumosu + mandara, 22-23 (note 5). An exquisitely painted
thanka in the Potala palace in Lhasa shows not only the same set of
concentrically ordered hells, but also a heavenly field above into which
ASTROLOGICAL AND DIVINATORY DIAGRAMS rises a many-layered Meru. This remarkable composition is illustrated
in Tresors du Tibet, 83 (note 36). A comprehensive view of the hot
As previously noted, astrology and other forms of divi- and cold hells is provided in a woodblock print reproduced in Lauf,
nation play an important role in the culture of Greater Secret Doctrines, 132, with a key diagram on 133 and explanatory text
Tibet. 92 Manuals devoted to the subject abound in on 130-31 and 134-37 (note 50). Though it is similar in subject to the
graphic devices that may be consulted to determine out- previously noted work, the two are completely different in appearance,
a fact only partly accounted for by the different media employed.
comes preordained by one's time of birth or by the appar- 92. Rene de Nebesky-Wojkowitz, Oracles and Demons of Tibet: The
ition of various signs believed to be diagnostic guides to Cult and Iconography of the Tibetan Protective Deities (The Hague:
the future. Although much could be written on the sub- Mouton, 1956),291-98 and 455-66, discusses more than a dozen meth-
ject of divination, it lies at the margin of the concerns ods of divination used in Tibet in addition to resort to oracles. Few of
of this volume. Hence I shall do no more than direct these, however, appear to rely heavily, if at all, on graphic aids.
93. Especially useful are Emil Schlagintweit, Buddhism in Tibet, Illus-
attention to a few works in which relevant diagrams trated by Literary Documents and Objects of Religious Worship, with
appear and comment briefly on them, taking note of sev- an Account of the Buddhist Systems Preceding It in India (1863; Lon-
eral genres that relate to Indian cosmography. don: Susil Gupta, 1968), 290-328; and Waddell, Buddhism of Tibet,
The Tibetan system of astrology, an amalgam of 450-74 (note 5).
Indian, Chinese, and indigenous concepts, is discussed 94. Essen and Thingo, Die Gotterdes Himalaya, 2:229-30 and 1:256-
57 (note 85).
and illustrated in a number of nineteenth-century schol- 95. Along with one other bronze example (in a private collection),
arly publications. 93 The diagrams included in almanacs it is illustrated and discussed in Siegbert Hummel, "Kosmische Struk-
consulted by practitioners of astrology can be quite com- turplane der Tibeter," Geographica Helvetica 9 (1964): 34-42.
FIG. 15.23. TIBETAN BRONZE ASTROLOGICAL TABLE. date of birth and the current arrangement of zodiacal constel-
The central component in this assemblage (age not known) is lations-one's future can be foretold. At the top of the diagram
the rather abstract shape of a tortoise, whose head, legs, tail, appear three particularly revered manifestations of the Buddha:
and flanks represent the four cardinal and intermediate direc- Avalokitesvara, Mafijusri, and Vajrapal)i.
tions. On the carapace is a magic square. Other parts of the Size of the original: 35.4 x 28.7 cm. By permission of the Volk-
diagram, primarily with calendrical associations, are also con- erkundemuseum der Universitat Zurich (cat. no. 12664).
sulted, together with its central portion, so that-based on one's
9 so arranged that the sum of the numbers in each column national Gesar epic. Such paintings, which have been pro-
and row adds up to 15. The circular band surrounding duced at least since the fourteenth century, abound in
the magic square contains eight segments, here shown as maplike scenes of places important in the lives of the
the petals of a lotus, representing the eight quarters and personages they honor. Some such scenes are painted
their respective elements. In sequence, north signifies fire, with a great sense of verisimilitude, others are rendered
northeast earth, east iron, southeast heaven, south water, fancifully. The paintings are constructed in many ways,
southwest mountain, west tree, and northwest air. The often with great ingenuity, and reflect schools associated
center (the carapace of the tortoise) stands for the nadir with particular monastic sects. Most place the protagonist
of a vertical cosmic axis. On painted versions of such in the central or most visually prominent portion of the
diagrams, each direction would have its corresponding composition. Others relegate that individual to a marginal
color. It is likely that this system is related to the Indian position, giving greater prominence to a particular place
kurmavibhaga. 96 or set of places within the Buddhist cosmos or to his-
A variant of the type of divination diagram just dis- torically identifiable sacred places. Here I can do no more
cussed, also with apparent Indian analogues, takes the than illustrate a few such paintings and direct attention
form of a wheel, normally with eight radii dividing it into to useful sources in which others can be studied.
eight segments, with the same directional associations Figure 15.24 illustrates one panel of a rather large late
with elements as noted above, and with eight concentric eighteenth-century triptych dedicated to the deity Kala-
rings. 97 Thus there is a total of sixty-four compartments, cakra. The two side panels of this work display most
each representing a combination of two elements. But prominently the realm of Shambhala, the left one as it
the order of elements, while regular in regard to the had been conceived in a bygone era and the right one
sequence of segments, varies within each segment, no two (our illustration) as it was believed to be when the paint-
being identical. Obviously this leads to a more compli- ing was made. Above and below Shambhala on both
cated set of divinatory formulas. Some Tibetan divination panels are tutelary figures and other locales associated
diagrams of this type appear to be even more complicated with it. The central panel reverses this pattern. There
than the basic sixty-four-compartment type just noted, Kalacakra, a giant and fearsome deity, occupies the cen-
containing sixteen or more spokes and more than eight tral position while associated places, largely unidentified,
concentric rings. What I take to be one such diagram is are consigned to marginal positions. Places and person-
painted on the wall of Nechung (gNas-chung) monastery ages depicted in the upper portions of the painting are
near Lhasa, the former seat of the Tibetan state oracle. 98 celestial, while those at lower levels appear to be terres-
What appear to be Indian analogues of the types of
divinatory diagrams just discussed are for the most part
square rather than circular, but they are also composed 96. The account of the Tibetan diagrams is based on Gordon, Tibetan
of a set of directionally identified spokes and, usually, of Religious Art, 24, 27, and 29-30 (note 53); and Schlagintweit, Buddhism
in Tibet, 304-11 (note 93). The figure that Gordon presents is a cloth
nested squares rather than concentric rings. Moreover, thanka and is much more complex than the one illustrated in figure
the two genres differ in that a particular city of reference 15.23. On page 26 she also presents photographs of illustrations, mainly
takes the place of the cosmic nadir at the center of the astrological, from two Tibetan divination manuals. For the Indian kur-
map, and individual cities (seemingly randomly ordered mavibhaga, see Schwartzberg, "Cosmographical Mapping," 337-39
rather than arranged according to their actual azimuths (note 60).
97. Discussed in Schlagintweit, Buddhism in Tibet, 311-13 (note 93).
or distances from the center) take the place of the two- 98. Letter from Toni Huber of the Department of Philosophy and
element combinations of each map segment in the Religious Studies, University of Canterbury, 10 December 1989. Huber,
Tibetan case. Yet the structure and functions of both who enclosed a picture of this painting, was unable to determine the
genres seem to be similar and tied to a set of astrological system for using the diagram, but there is little doubt that it serves an
determinants. The cultural transmission of the underlying astrological purpose. An accompanying diagram of another such wall
painting, seen in the inner courtyard of Zhwa-Iu monastery to the south
organizational principles from India to Tibet, or in the of Shigatse city, shows a circle with sixteen spokes and nine rings. Huber
opposite direction (not excluding the possibility of a writes that his map was designed by the famous fourteenth-century
Chinese link), warrants investigation. 99 Tibetan scholar Buton Rinchen Drub (Bu-ston Rin-chen-grub) and that
"it is intended to correlate the transit of various heavenly bodies with
the Tibetan calendrical cycle (which is a lunar sexagenary one). If you
CARTOGRAPHIC ELEMENTS IN read the inscriptions its sophistication becomes apparent." Photographs
of both paintings were taken by Huber in 1987.
HAGIOGRAPHY AND MYTHOLOGY
99. Two such maps, one square and one circular, are illustrated and
discussed in Schwartzberg, "Cosmographical Mapping," 348-51 (note
Exceedingly common types of Tibetan painting are those
60). One of those and three others (all square) are illustrated and briefly
relating to past, present, and even future lives of the Bud- discussed in Susan Gole, Indian Maps and Plans: From Earliest Times
dha, of bodhisattvas, and of lamas and other saintly mor- to the Advent of European Surveys (New Delhi: Manohar, 1989), 23-
tals, as well as of mythic heroes, especially those of the 24 and 50-53.
FIG. 15.24. PANEL FROM A TIBETAN TRIPTYCH DEDI- ical work on Shambhala. To their right appears the great mon-
CATED TO THE DEITY KALACAKRA. This right panel is astery of Tashilunpo, a chief Gelukpa monastic center, where
the largest of an exceptionally large and remarkably well pre- this painting was probably made, and the surrounding town.
served work dated about 1780. Dominating the work is the Many individual structures therein can be identified. Especially
realm of Shambhala, described in the sacred Kiilacakra T antra, prominent is the tall, windowless "Tower of Silk Paintings"
with Kalapa, the royal capital, at its center. In the heavens above from which giant thankas were unrolled for public exhibition
are, at left, identifiable representations of the four great tutelary on major festive occasions. One wonders if any of these were
deities of the Gelukpa monastic order, one of whom is Kala- maps similar to those displayed in Patan, as shown in figure
cakra, while in the corner to the right are Tsong-kha-pa, the 15.2.
fifteenth-century founder of the order, and two other abbots. Size of this panel: 136 x 86.5 cm. From Armand Neven, Etudes
In the lower left are the tombs of the Gelukpa abbots, including d'art lamafque et de l' Himalaya (Brussels: Oyez, 1978).
that of the Panchen Lama (upper quadrangle), author of a myst-
FIG. 15.25. MORAL LANDSCAPE OF THREE JATAKA most prominent feature depicted. The lower-right quarter of
TALES. This nineteenth-century painting from eastern Tibet the painting relates to Jataka 14, which takes place around Raja-
provides a visual backdrop for the narration of three Jatakas grha, a locality in Magadha, which in the previous story is indi-
(12, 13, and 14), all primarily set in what is now the Indian state cated simply as a grove (here rendered as a small clump of trees)
of Bihar and the adjacent region of Nepal. Jataka 12 occupies under which the Buddha sits. Except for the possibility that the
roughly the upper half of the map. The action begins in the line of mountains, conventionally placed near the top of the
upper left corner and proceeds downward in a zigzag line, but painting, might be intended as the Himalayas, there is no par-
this progression is not rigidly followed. For example, below the ticular geographic logic to this composite moral landscape, but
battle scene involving the Sakya tribe (into which the Buddha the flow of action generally suggests a temporal logic, implying
was born), we see a warrior laying his sword down before the a kind of a route map through time. The ingenious way of
Buddha and rendering homage to him, both depicted near the combining conventionally painted landscape elements in this
left margin; yet the next event relates to a stupa that warrior painting, both as sites for the action depicted and as temporal
built to honor the Buddha, as shown near the upper-right cor- dividers separating one story or domain from the next, is a
ner. Jataka 13 occupies the lower-left quarter of the painting common device in Tibetan religious art.
and relates to events involving the historical personage of Bim- Size of the original: 75 X 56 cm. By permission of the Gerd-
bisara, king of Magadha, in southern Bihar, whose palace is the Wolfgang Essen Tibetica Collection, Hamburg.
trial and are, in part, geographically recognizable. This thought to be an incarnation of the third Panchen Lama.
indicates a tendency to have the organizational structure In brief, after an early life lacking in virtue and a long
of such paintings reflect that of the universe itself. The period of banishment and wandering, Gesar returns pur-
most prominent terrestrial component of figure 15.24 is ified to his native land, leads an army against the evil
the monastery of Tashilunpo (bKra-shis lhun-po), forces of the world, and ultimately becomes a Buddha.
founded by the first Dalai Lama and subsequently the The number of known paintings relating to the Gesar
residence of the Panchen Lama and a major center of epic is not great, but those I have seen are rich in place
the Gelukpa monastic order. Adjoining it is the town of detail that, as with many hagiographic works, combines
Shigatse (gZhis-ka-rtse). Because the third Panchen Lama cosmographic and geographic images. 102 The images
(1737-80) was the author of a mystical doctrinal work, seem to lack the spatiotemporal order noted for figure
The Way of Shambhala, Tashilunpo is closely asociated 15.25, and it therefore is not easy to follow the narrative
with that sacred realm. This monastery is painted in great they depict. 103 .Nevertheless, the scenes are arranged in
detail and with fidelity to its former appearance. Within vivid maplike assemblages that make the epic come alive
it many individual structures, some no longer surviving, to viewers familiar with its content. Bards often convey
can be recognized from late eighteenth-century descrip- the story:
tions by the British ambassador George Bogle (1746-81)
Paintings of these types are used in conjunction with
and lieutenant (later captain) Samuel Turner (1749?- a sung or chanted narrative, the storyteller using a stick
1802). Nearby, as in reality, flows the river Tsangpo, while to point out the scenes on the painting as he goes
to the left are the enclosures of the tombs of the Panchen along. Before circulating in Tibet, this technique of
Lama and other great abbots. 10o illustrated recitation had been employed by monks in
In marked contrast to figure 15.24 is the composite India, China and Japan. As in those countries, the illus-
moral landscape shown in figure 15.25, providing the field trations could be painted in fresco or on portable
for a pictorial accompaniment to the narration of three . scrolls.... The "literary" or narrative character of
successive ]atakas (stories of the past lives of the Buddha). these paintings is [often] emphasized by the presence
The stories in question all took place in or near what is of captions, sometimes quite long, written under each
scene, which serve to identify the figures and episodes
now the Indian state of Bihar and, like the painting itself,
and often reproduce the text of a corresponding man-
combine mythical and historical events in both mythical
ual, with a reference number. 104
and identifiable geographic settings. Though it is impos-
sible to sort out one from the other by visual inspection
of the painting, to the devout believer all are equally real. 100. This description is based on Neven, Etudes d'art lamafque, 45
(note 64). The entire triptych is illustrated and discussed on 40-45,
Unlike much of figure 15.24, the manner of depicting the
while a larger-scale view of Tashilunpo appears on 11. Another, less
features shown in figure 15.25 depends primarily on the detailed view of the monastery appears on a quite different hagiographic
artist's imagination, constrained only by adherence to painting on which the first Dalai Lama occupies the central position.
established iconographic conventions. There is, for exam- This undated work is one of many such paintings illustrated in Hiroki
ple, no way of knowing what the palace of the Magadhan Fujita, Tibetan Buddhist Art (Tokyo: Hakusuisha, 1984), text on 189-
90.
king Bimbisara (d. ca. 490 B.C.) looked like, yet it is
For the descriptions of Bogle and Turner, see Markham, Mission of
depicted in some detail in the lower-left corner of the George Bogle to Tibet (note 35), and Samuel Turner, An Account of
painting. On the whole, the spatial logic of the painting an Embassy to the Court of the Teshoo Lama in Tibet (London, 1800;
is narrative, with places shown in proximity according to reprinted New Delhi: Manjusrl Publishing House, 1971).
the sequence of events transpiring there. This generally 101. The painting is one among numerous hagiographic works
depicted and discussed in Essen and Thingo, Die Gotter des Himalaya,
applies both to the sequence of stories and to the action
1:39-40 (note 85).
within a particular]ataka. But this formula is not rigidly 102. Figure 15.25 is analyzed in Alexander W. Macdonald and Perna
followed; at least one historical place, Rajagrha, appears Tsering, "A Note on Five Tibetan Thail-kas of the Ge-sar Epic," in Die
twice, being first symbolized by a grove of trees under Mongolischen Epen: Bezuge, Sinndeutung und Dberlieferung (Ein Sym-
which the Buddha sits near the left margin of the painting posium), ed. Walther Heissig, Asiatische Forschungen, vol. 68 (Wies-
baden: Otto Harrassowitz, 1979), 150-57. For a brief account of the
and forming the general locale for the whole of ]ataka
origin and nature of the epic, see Rolf A. Stein, Tibetan Civilization,
14. 101 trans.]. E. Stapleton Driver (London: Faber and Faber, 1972), 278-80.
A final illustration (fig. 15.26) relates to the Gesar epic, 103. Rolf A. Stein, "Peintures tibetaines de la vie de Gesar," Arts
known throughout Tibet and Mongolia. Put into its pre- Asiatiques 5 (1958): 243-71, esp. 244. Stein notes that the customary
sent form in the late fourteenth or early fifteenth century, order of presenting scenes on other biographical paintings, for example,
those relating to the Bon saint Milarepa (Mi-Ia-ras-pa), is from the
probably in the Kham region in eastern Tibet, but incor-
bottom up, usually beginning in the lower-left corner. Conceivably this
porating portions of older stories, the epic recounts the provides an intentional means of distinguishing Bon from Tibetan Bud-
career of Gesar (a cognate of Caesar), a warrior-king dhist hagiography.
sometimes associated with Shambhala and sometimes 104. Stein, Tibetan Civilization, 285 (note 102).
GEOGRAPHICAL MAPS
Although the number of known traditional geographic

maps from Greater Tibet is substantial, it appears small
in comparison with the vast body of cosmographic mate-
rials from the same region. This contrast may be attrib-
uted in part to the lesser interest that geographic maps
hold for art historians, students of religion, and other
scholars specializing on the region and in part to the lesser
value Tibetans attach to nonreligious works. Accord-
ingly, relatively little effort has been expended by out-
siders to obtain geographic maps, and relatively little care
seems to have been taken, by either Tibetans or outsiders,
to preserve whatever maps may once have existed. Hence,
while one need not doubt that the disparity between the
number of surviving cosmographic and geographic maps
is real, one cannot draw firm conclusions about the rel-
ative size of the two groups of maps in the past.
Here I shall examine the only two known Tibetan maps
that are believed to be attempts to represent the world,
a not very large number of regional maps, a variety of
pilgrimage guides and other route maps, and, the most
common of all Tibetan geographic maps, those that
depict towns, monasteries, and other specific localities.
WORLD MAPS
Figures 15.27, 15.28, and 15.29 present a modern recen-

FIG. 15.26. TIBETAN THANKA FROM A SERIES ON THE sion, a key, and a reference map for what appears to be
GESAR EPIC. This undated, but apparently recent (late nine- the oldest known map of Tibetan provenance. In its orig-
teenth or early twentieth century), painting is one of a set of inal form, the map was undoubtedly associated with the
five, not necessarily complete, relating to the Gesar epic. The
set was acquired by a Swiss citizen in Kalimpong, Sikkim, in Bon faith, which predated Tibet's acceptance of Bud-
1949. The painting depicts numerous events described in the dhism by many centuries. This map, in Tibetan, was pub-
epic, occurring in recognizable places of both heaven and earth. lished in Sgra yi don sdeb snmi gsal sgron me bzugs so
For example, in the upper-right corner six divine personages are (Tibetan Zhang-zhung dictionary), a 1965 edition of a
shown in Lha-Iing, the land of the gods, while the bottom of seventeenth-century collection of Bon texts with Tibetan
the painting represents the (mythic?) land of Ma. The logic by
which the assemblage of scenes is organized, however, is not translations and commentary.106 Zhang-zhung (Shang-
obvious. The four undulating striated bands leading into the shung) is the early language, possibly Indo-European, of
heavens either may be paths by which earthly and celestial char- western Tibet, to which area the Bon faith seems to have
acters in the epic can reach one another's domains or may refer been disseminated from an original source area in Iran. 1O?
to the attainment by great adepts of a "rainbow body" at the
time of death.
Size of the original; ca. 71 x 53 cm. By permission of Alexander 105. The following include an abundance of hagiographical paintings
w. Macdonald. incorporating substantial maplike components: Gordon, Tibetan Reli-
gious Art (note 53); Per Kvaerne, "Peintures tibetaines de la vie de
sTon-pa-g~en-rab," Arts Asiatiques 41 (1986): 36-81, for an important
Bon figure with particularly detailed analyses; Lauf, Verborgene Bot-
schaft (note 85); Liu, Buddhist Art (note 36); Olschak and Thupten
I am not aware of any attempt to study the rich corpus Wangyal, Mystic Art (note 39); Julien L. Tondriau, 20 rouleaux peints
of biographicaljhagiographical paintings from Greater tibetains et nepalais (Brussels: Musees Royaux d'Art et d'Histoire,
Tibet with a view to determining its underlying canons (l964-65?]); Chogyam Trungpa, Visual Dharma: The Buddhist Art of
for representing and ordering geographic and cosmo- Tibet (Berkeley, Calif.: Shambhala, 1975); and Waddell, Buddhism of
Tibet (note 5).
graphic places. Such an undertaking would present a fron-
106. Ni-ma-grags-pa (1616-70), Sgra yi don sdeb snan gsal sgron me
tier for research and promises to offer important insights bzugs so (reprinted Delhi, 1965).
into the cartographic thinking of a highly distinctive cul- 107. For succinct discussions of the origins of Bon and Zhang-zhung,
ture.IOS see Stein, Tibetan Civilization, 35-37 and 230-35 (note 102).
I am not aware of any premodern version of the map or indicates that that still-surviving city was intended.) Other
of any copy held by a museum, library, or other public arguments run into lengthy paragraphs. The conclusion
agency outside Tibet. It is likely, however, that manu- is that the map, which obviously postdates Alexander's
script maps, copied from older sources, are held in conquest of most of the area covered, also predates the
Tibetan monasteries or by refugee Tibetans in India or Roman campaigns against Parthia and, based on numer-
elsewhere. The published Delhi version of the Zhang- ous pieces of internal evidence, can best be ascribed to
zhung map was studied by two Russian scholars, L. N. the second century B.C.
Gumilev, a historical geographer, and B. I. Kuznetsov, a Although Gumilev and Kuznetsov's arguments are
Tibetan phililogist. The following remarks are based informed, carefully reasoned, and (in the absence of con-
largely on the translated account of their thorough analy- trary evidence) plausible, some are less than convincing,
Sis. I08 such as the following:
Gumilev and Kuznetsov's assumption-with which I Ma-thang-bsgral-gling-the transcription of a local
concur-is that the extent of interaction between Tibet place name. Judging from its location in the southern
and lands to its west, especially Persia, was considerably part of the world ocean, south of the Nicobars [the
more extensive in the early historical period than most Nicobar and Andaman Islands are earlier described as
historians recognize and that considerable information places where (human) flesh is eaten], and from its
about distant lands reached Tibetan geographers, at sound, this may be Madagascar. The name Madagas-
either first or second hand, and was incorporated into car, which is not used by its residents, was first
the original precursor of the Zhang-zhung map. A key reported by Marco Polo, and the first description of
to their analysis is their interpretation of what lies at the the island in European geography was given in Periplus
of the Erythrean Sea, i.e., later than the time of our
center of the map, a locale named Bar-po-so-brgyad,
map and without a name. Consequently, this old Mal-
which they identify as Parsogard (Greek Pasargadae), cap-
agasy word must have reached Tibet through India.
ital of the Persian Empire from 550 to 522 B.C., under The Malagasy settled Madagascar about the 3rd cen-
the emperors Cyrus the Great and Cambyses. Within that tury B.C. from Indonesia and thus came to the atten-
central rectangle is a crudely drawn ten-story edifice that, tion of the Indians who were navigating in the Indian
it is said, represents the tomb of Cyrus. Although various Ocean. In other words, the Tibetan cartographer, in
Greek historians left contradictory accounts of that addition to Iranian sources, also relied on Indian
tomb, Aristobulus said that it had the form of a small sources, so that our map is not the product of pla-
tower, and Onesicritus, who accompanied Alexander on giarism, but an original work reflecting the level of
his Persian campaign, stated that it was ten stories high. geographic knowledge in Tibet in the 2d century
B.C. tt2
A Tibetan inscription on the map itself, "swastika hill,
nine stories high," apparently refers to the central figure Apart from the identifications in the Indian Ocean basin
and also notes "crystal columns with inscriptions," "the and eastern Africa, most relate to the Near and Middle
garden of the swastika," "the garden of the wheel," "the East and to Central Asia as far north as the general region
lotus garden," and "the precious garden," almost all of of Lake Balkhash. The westernmost place named is Ionia.
which can be reconciled with Greek descriptions. lo9 Curiously, however, there is no place that falls clearly
Gumilev and Kuznetsov explain the disparity between the within what is now Tibet. This raises the prospect (not
references to a nine- and a ten-story structure by saying suggested by Gumilev and Kuznetsov) that the prototype
that in one instance "the top story was treated sepa- map was brought to Tibet from Iran and then "natural-
rately."110 For all but seven of the sixty-two other places ized" in that area, with considerable fidelity, and trans-
named or annotated on the map (one other is outlined mitted by frequent copying over the subsequent
but not named), Gumilev and Kuznetsov are able to offer centuries. Though some alterations in the original may
a translation or commentary or both; and in more than have occurred, the arguments of Gumilev and Kuznetsov
forty cases they suggest, with varying degrees of persua- suggest that they were not great. This leads to the sup-
siveness, a specific place or broad region with which the
original map feature can be identified. The identifications
108. Gumilev and Kuznetsov, "Two Traditions" (note 21). The two
sometimes overlap and are sometimes repetitious, how- traditions referred to in the title are called "the Irano-Tibetan tradition,"
ever. Some are exceedingly terse and vague while other which is the one that concerns us here, and "the Indian-Tibetan tra-
comments, comparably terse, are quite precise, for exam- dition," which refers to the Buddhist cosmographic tradition. Nothing
ple: "Ne-khri-'bum-thang-the 'Nekribum' plateau"; is said in the article ~bout the latter that will add to what has already
been stated in this volume.
"Grong-khyer-lang-Iing-Jerusalem"; "Ne-seng-dra-ba'i-
109. Gumilev and Kuznetsov, "Two Traditions," 570-71 (note 21).
gromg-khyer-'City of Nesendra,' i.e., Alexandria."lll 110. Gumilev and Kuznetsov, "Two Traditions," 571 (note 21).
(Of course there were many ancient Alexandrias; but the 111. Gumilev and Kuznetsov, "Two Traditions," 575 (note 21).
proximity of this item to a locality identified as Egypt 112. Gumilev and Kuznetsov, "Two Traditions," 574 (note 21).
~~"'l~~~·~t4I.;1~A\~\ '(,'i"'(~\ ~'~'a~i '\~~1
:"~"\\ <\ ~--

1~<:,·~t;·~i,,·\!;qr,,\~~~iir~c;,,\~,,\
..........·:l!\Ill·,·..ill
.,
\\\ \
\\
\\
FIG. 15.27. MODERN RECENSION OF AN ANCIENT
TIBETAN VIEW OF THE WORLD. The information on this
map appears to represent a Tibetan view of the world as of the FIG. 15.28. PARTIAL KEY TO THE CONTENTS OF FIG-
second century B.C. It was made to fit into the form of a mandala URE 15.27. The key numbers are those assigned by Gumilev
centering on the old Persian capital of Parsogard (Pasargadae), and Kuznetsov for their analysis of the map text. Numbers
within which the tomb of Cyrus the Great (d. 529 B.C.), said within shaded fields are keyed to figure 15.29, showing on a
to be ten stories high, provides the focus of the map. No old modern map the ancient regions, cities, and other geographical
version of this map is known to survive. features indicated on the Zhang-zhung map. The analyses of
From Ni-ma-grags-pa (1616-70), Sgra yi don sdeb snan gsal the text for unshaded numbered areas are insufficient to estab-
sgron me biugs so (reprinted Delhi, 1965). lish any equivalent features or do not go beyond the translit-
eration of an unidentifiable name.
After L. N. Gumilev and B. I. Kuznetsov, "Two Traditions of
Ancient Tibetan Canography (Landscape and Ethos, VIII),"
position that the map was accorded a sacred status and Soviet Geography: Review and Translation 11 (1970): 565-79,
thus its integrity had to be maintained. esp. fig. 3 (which contains a more complete translation than I
have provided).
The Zhang-zhung map bears a remarkable resemblance
to another Bon work, which has already been briefly
noted under the heading of cosmographic maps. That
map, identified by Snellgrove as "The Nine-Stage Swas- it should best be regarded as a cosmographic work, as
tika Mountain (representing the Nine Ways of Bon) ... Snellgrove's title suggests, or a geographic map is moot.
in the country known ... as ... Sambhala," was drawn The place of the Ku~aQas on the later map, however, is
(probably in London, and quite possibly from memory) noteworthy in two respects that might relate to the myth
by a refugee Bon lama about 1967. 113 Gumilev and Kuz- of Shambhala. First, the Ku~aQa Empire, a major trading
netsov recognized the unmistakable affinity between the state, was known for its prosperity, which could have
Zhang-zhung map, also said to represent the land of given rise to the notion of Shambhala as an idyllic realm.
Shambhala, and the work brought to light by Snellgrove, Additionally, the original regional hearth of the Ku~aQas
but they observe that the latter omits several of the names in Central Asia was considerably farther north than the
on the Zhang-zhung map and was "compiled somewhat original core area of Persia, and some of the areas under
later and independently of the first version," incorporat- their early control lay in the Tarim basin due north of
ing evidence of the expansion in the region of the Ku~aQa western Tibet. Hence the current Tibetan notion that
dynasty in the first century A.D. 114 The map of the Nine- Shambhala is the northern paradise, although finding no
Stage Swastika Mountain, unfortunately, has yet to be
translated, and the legends on the published version are 113. Snellgrove, Nine Ways of Bon, v and fig. XXII (note 37).
too small to permit an independent translation. Whether 114. Gumilev and Kusnetsov, "Two Traditions," 576 (note 21).
~
70· 80·
40 COUNTR'I
. OFiHE
SAKA
42. COUNTRY OF
THE MASSAGETAE
54. KARAKUM
38. IRANIAN
-17-....,H'I;O'R'"':C;;O:A~\A
BACTAIA 36/37. GRAECO'
'k>. BACiRIA
IS. PARTHIA
24. LAND OF THE

SUAEN CLAN
22.~
SO. GEDROSIA
32. ARABIA
FELIX
Off the map: 48. ANDAMAN & NICOBAA ISLANDS
49. MADAGASCAR
o 250 SOD miles
I iii i I
o 2SO SOD 7SO km
40· SO·
FIG. 15.29. IDENTIFIABLE ANCIENT GEOGRAPHICAL northeastern Africa that are shown on this map, there are two
FEATURES OF THE ZHANG-ZHUNG MAP. The numbers others at more remote locations, the Andaman and Nicobar
indicated on this map are keyed to those of figure 15.28. In Islands and Madagascar, whose inclusion on the Zhang-zhung
addition to the locales in Southwest and Central Asia and in map is suggested by Gumilev and Kuznetsov.
support in the Zhang-zhung map, might be mOre or less topographic information ... was pressed ... into a
appropriate for the one illustrated by Snellgrove, espe- mandala type of framework without much regard for
cially if the Ku~aQa domains came to occupy in the minds actual geography. The map itself ... can hardly be of
of ancient Tibetans the place of Persia as the archetype the epoch represented by the place-names on it since
of Shambhala and if some great and benevolent monarch, the earliest Tibetan inscription dates only from A.D.
767. The religious cosmological view is comparable
such as the Ku~aQa emperor Kani~ka, supplanted Cyrus
with the T-O maps of European culture exemplified
as the archetypal divine monarch. us A question that by that of Cosmas Indicopleustes (fl. A.D. 540).116
remains in comparing the two maps is why the layout
remained virtually identical while the content shifted. But this criticism misses the point. The authenticity of
Very likely, that particular form of mandala acquired- the original map in no way depends on the date when
even earlier than the time of the Zhang-zhung map-a Tibetan inscriptions first became known. The ancient
procrustean sanctity that required fitting the content to
the form and not the opposite. 115. Numerous problems exist in respect to the extent and chro-
nology of the KU~iil)a Empite. These are dealt with in Joseph E.
A. L. Mackay dismisses Gumilev and Kuznetsov's dat- Schwanzberg, ed., A Historical Atlas of South Asia (Chicago: University
ing of the Zhang-zhung map in a single brief paragraph. of Chicago Press, 1978), xxix-xxx, xxxiii-xxxiv, 21, and 174-76.
He observes that the map's 116. Mackay, "Kim Su-hong," 27-38; quotation on 31-32 (note 18).
Zhang-zhung language and later Tibetan were not the for decisions on what places to include and how much
same. But no matter what language the original map may space to assign them. Thus, three of Teramoto's identi-
have been in, there is no reason its script might not have fications are of Indian cities and the rest are of regions
been transliterated many centuries later into Tibetan, or even vast empires. Tibet occupies a more or less central
while retaining some approximation of the original pho- position in the middle panel of the map; see figure 15.31
netic attributes of the original toponyms. for the identification of other parts of the map.120
The second and only other known Tibetan world map It is also difficult to put forward a plausible rationale
was brought to light in 1931 by Teramoto. 117 This enig- for making the map. Paraphrasing Teramoto, Nakamura
matic map (fig. 15.30), with script in Tibetan and Chinese, states:
composed only a small part of a long scroll document
The names on this map, giving a general view of the
relating to Buddhist iconography and was not deciphered communications between India, Persia, Tibet and
before the 1931 study. The original map must have been Central Asia, seem to be based on Chinese and trans-
drawn by the ninth century at latest, since a copy of it lated [into Tibetan] merely to make more easy their
was brought from the Blue Dragon (T'sing-Ioung-tseu) understanding by the Tibetans ... [at a time when]
Temple at Chang'an (modern Xi'an), then the capital of Tibet's national prestige was at its height.... It was
China, to japan by a priest named En-tin who died in therefore most probable during this period that the
A.D. 891. Where the map was originally made is not map was made for political and administrative pur-
known. The japanese copy of the map was kept in the poses.1 21
Onzyozi Temple, and over the following few centuries Although I find this most unconvincing, I can offer no
several additional copies were made, including one in more likely raison d'etre for the work. The map does
1194 by a priest named Zenkaku, attached to Onzyozi not, for example, have any apparent religious purpose.
Temple. Some modern copies were also made, the most Although two of the cities named in India are important
recent in 1890. Zenkaku's copy was lost in the great fire in the early history of Buddhism, the most obvious can-
in Tokyo following the earthquake of 1923, but an album didate for inclusion, Bodh Gaya, seems not to be noted.
of photographs previously made of the entire scroll was The significance of the work, however, seems clearer-
preserved. From 1194 to 1220 there were at least seven assuming that most of Teramoto's identifications of the
unsuccessful attempts to decipher the map. One apparent far-flung regions depicted are valid-in that it proves, as
reason for the failures was that the map's Tibetan text previous maps do not, that Chinese knowledge of the
was mistaken for Sanskrit. In 1893 a copy of the map rest of the Asia was more extensive than what could be
was presented for study to the japanese Bureau for gleaned from the travel accounts of a number of Buddhist
Enquiry into the National Treasures; but the committee pilgrims, among whom Xuanzhuang (602-64) was the
that examined it was "lukewarm about its historical and
geographical value."118 Not until the studies of Tera-
moto, begun in 1921, could anyone claim to have deci- 117. Teramoto Enga, "Waga kokushi to Toban to no Kankei" (The
phered the map. relation between our [Japanese] history and Tibet), Otani Gakuho 12,
The content of the map is sparse, with only twenty- no. 4 (1931): 44-83. It was also briefly discussed in Nakamura, "Old
Chinese World Maps," 19-22 (note 18). Including this map (known
one names identified. As with the Zhang-zhung map, the copies of which exist only in Japan) in an article on Chinese maps
identifications vary in plausibility. The territory covered preserved in Korea relates to its relevance for ascertaining the geo-
appears to be vast, even greater than in the Zhang-zhung graphic knowledge of the Chinese during the Tang dynasty.
map, extending from the Byzantine Empire in the west 118. Nakamura, "Old Chinese World Maps," 20 (note 18). Naka-
to Korea in the east and from Turkic Central Asia in the mura's study of the map is based on the photographs in the album
mentioned above.
northwest to Cambodia (Chen-Ia) in the southeast. Tibet
119. Nakamura, "Old Chinese World Maps," 20 n. 49 (note 18). I
itself is designated once as "Bo" and, implicitly, a second do not know what Chien-pi refers to.
time as "Pa-man" (Eight Barbarian Lands, which were 120. Nakamura, "Old Chinese World Maps," 20 n.49 and fig. 8 (note
generally understood in China to comprise Korea, Chen- 18). The identification of Mangalore, which had little importance as
la, Persia, Tibet, Chien-pi, the land of the Turks, Khitan, far back as the ninth century, strikes me as particularly problematic.
The Tibetan "Mon," which according to Stein "covers all sorts of
and Mo-ho).119 Nakamura does not suggest a principle
aboriginal tribes of the wooded Himalayan Hills ... and is possibly
by which the names are arranged; and in my own exam- related to the word 'Man' used in literary Chinese for all southern
ination of his abstract of the map I could discern none. 'barbarians' " (Tibetan Civilization, 34-35 [note 102]), suggests a more
The map certainly does not have the symmetry of a man- plausible association with the original "Mom."
dala. Neither distance nor direction of places from one 121. Nakamura, "Old Chinese World Maps," 22 n. 51 (note 18).
Nakamura does not say how proficient Teramoto was in Tibetan. If
another is indicated consistently, and it is therefore Teramoto was much more proficient in Chinese than in Tibetan, it
impossible to specify the direction toward which the map would hardly be surprising for him to see more associations on the map
is oriented. Nor does importance appear to be a criterion with essentially Chinese toponyms than with Tibetan ones.
2 3
9
14
16
17 18
o
21
I 1. Pahanyan. Pahan-na? (Fergana); 2. Kin-hin (Kashmir); 3. Tha-ku-sha-si (Tak$llsilaITaxila):

4. Taha Thor-kus (Turkestan); 5. Pu-lin (Byzantine Empire); 6. Sha-he (Sravastf/Set
Mahet); 7. Kuo-kuo (Tokhara); 8. Ya-mah, ? Ghah (Tokhara); 9. Koron (Kurana); 10. Pran
(Polan); 11. Pa-ta-Ior (Bolor. in Gilglt region of modern Kashmir); 12. Thahan (Tashkent);
13. Pa-sin-go (Persia); 14. Ped-she-Ia, ? Pitasela (PflasilalKhuzdar in Baluchistan); 15.
Mom. Mangali, ? Mangale (Mangalore); 16. Koua-kouo (Touen-houang, ? Dunhuang in
modern Gansu); 17. Po (BedlTibet); 18. Kin-ku-ko (Kyrghiz); 19. Hor (Uighur); 20. Than
(China); 21. Pa-man (Eight Barbarian Lands, generally understood to comprise Korea,
Chen-Ia [Cambodia], Persia, Tibet, Chien-pi [unknown], Kitan, and Mo-ho (Manchuria).
Where two names are given within parentheses, the modern one comes second.
FIG. 15.31. KEY TO FIGURE 15.30. The key to the map indi-
cates the spellings of various toponyms, as provided in a 1931
publication by Teramoto. The organization appears rather anar-
chic, and no obvious purpose for the map emerges from study
of it. Although this may be due in part to geographic short-
comings in the original, it is also likely that errors introduced
through repeated copying, together with errors of interpreta-
tion, account for the confusing picture the map presents.
After Hiroshi Nakamura, "Old Chinese World Maps Preserved
by the Koreans," Imago Mundi 4 (1947): 3-22, esp. 21.
FIG. 15.30. FACSIMILE OF SINO-TIBETAN WORLD MAP.
This is one of several facsimiles of the Sino-Tibetan world map.
Photograph counesy of Kazutaka Unno.
FIG. 15.32. NEPALI MAP OF CENTRAL ASIA. In Nepali, reedy shown on the eastern, not the western, shore of that sea.
written in Devanagari script, probably latter half of the nine- Near the left margin the positions of Mashhad, second city
teenth century, painted paper mounted on cloth. Most of the from the top, and "Iran" (Tehran), the city below it, are
details on this map are easily identified with known geographic inverted. European influences appear in the ruled, graduated
features, while a few are mythological. Overall, the level of border, in the naturalistic manner of depicting mountains, and
accuracy is not high. The large body of water at the top of the in the use of color in a not particularly successful attempt to
map, which is oriented toward the west, is a conflation of the signify political jurisdictions.
Aral Sea-into which the largest river on the map, the Amu Size of the original: 76 x 54 em. Private collection, New Delhi.
Darya, actually flows-and the Caspian Sea. The adjacent burn- Photograph courtesy of Susan Gole, London.
ing mountain, which should signify the vicinity of Baku, is incor-
FIG. 15.33. KEY TO NEPALI MAP OF CENTRAL ASIA (FIG. tifications. Names in uppercase letters are inferred political juris-
15.32). Names appearing in neither parentheses nor brackets are dictions. Dashed lines are presumed boundaries between polit-
transliterations from the original text. Names in parentheses are ical jurisdictions shown by color on the original.
modern equivalents. Names in brackets are my inferred iden-
most prominent. Conceivably a part of this knowledge, appear mythological probably have a basis in fact, such
especially that relating to western Asia, was relayed east- as the burning mountain near the top of the map, which
ward via the Bon Zhang-zhung tradition. probably relates to an ancient Hindu fire temple near
Baku that was maintained by priests from Punjab. (Oil
seepages in this area, occasionally ignited, have been
REGIONAL MAPS known for millennia.) Near ten towns on the map there
A Nepali Map of Central Asia is a conventional sign, along with the word garhi (little
fort), suggesting their military importance. What appear
Of the known regional maps from Greater Tibet, the one to be Hindu temples are drawn at several places, possibly
covering by far the greatest area is that illustrated in figure indicating resident communities of Hindu traders. Much
15.32. Although the map is in Nepali and undoubtedly more common, however, are minarets (present in every
was drawn by a Hindu, it is discussed here rather than town shown except for Russian Korali Kadar) signifying
(as in the case of other works of Nepali provenance) with the dominance of Islam in the region. Also near the towns
maps of South Asia because its style is so much more are notes-often derogatory-about the nature of the
akin to Tibetan than to Indian cartography.I22 local population or about matters of commercial interest,
The area covered by the map, including most of Cen- as well as figures giving distances of the towns from either
tral Asia and even extending into southern Russia, is indi- Kabul or Nepal (i.e., Kathmandu), presumably in khos (a
cated by a map key (fig. 15.33) and by its modern unit of about two miles). Roads appear as thin red lines.
counterpart (fig. 15.34) and may be defined by the cities Notes along the edges of the map point the way to several
shown nearest the four corners of the map: Baghdad in places not shown, including China, Mecca, Rum (Tur-
the upper left, Bulgar (modern Saratov) in the upper right, key), and Hinglaj, a place of pilgrimage for both Hindus
Yarkand (in Chinese Turkestan) in the lower right, and and Muslims on the coast of Baluchistan.
Kandahar (in southern Afghanistan) in the lower left. Cur- The degree of detail and the specificity of what is
iously, no part of either Nepal or Tibet is included. The shown tend to be somewhat greater in the eastern portion
orientation is toward the west, but the accuracy of the of the map than farther west, especially in and around
map tends to decrease in that direction. Several promi- Kabul and Kashmir, where Gole was able to identify a
nent physical features appear. The large body of water number of important landmarks. For example, even at
in the upper-left corner appears to be a conflation of the the greatly reduced scale of figure 15.32 one can easily
Aral and Caspian seas, and the large river running into it recognize the famous Shalimar Gardens adjacent to the
is the Amu Darya (Oxus). Several other rivers are indi- latter locale. This is one of a number of reasons to sup-
cated: the Him, probably intended to represent the Syr pose that the artist either had firsthand knowledge of
Darya (]axartes), even though it does not run into the those areas or worked from relatively reliable second-
Amu Darya as shown; and what appear to be the Kabul hand information. Elsewhere, however, egregious errors
and ]helum rivers, running through the city of Kabul and abound, such as the already noted garbled physical geo-
through Kashmir, respectively. Several mountain ranges graphy or the inversion of the relative positions of Iran
are also shown. The one skirting the great bend in the and Mazad (Tehran and Mashhad).
Amu Darya appears to represent the Hindu Kush; but on Although its style is essentially Tibetan, especially the
the whole the representation of orography, including way cities are depicted (compare plate 34 discussed
most of the large mountain chain running along the right below), the map also embodies certain European traits.
edge of the map, appears fanciful. Less prominent, but One of these is the ruled, graduated border. Another,
unquestionably meaningful, are the indications of moun- arguably, is the manner of depicting mountains. Espe-
tains around several cities to show that they are wholly cially significant is the use of color to suggest varying
or partially hemmed in by highlands. Vegetation signs political control over particular areas, for example,
scattered throughout the map, seem to connote the pres- salmon for Iran, yellow for the Russian empire, beige for
ence of orchards or wooded areas but probably are Bukhara, and so forth; but if that is indeed the intention,
largely decorative, especially when not in the immediate the execution is not particularly clear and will not match
vicinity of a city.
Though most of the places shown are fairly well 122. This map was brought to my attention by Susan Gole, who has
known and were important in the nineteenth century had most of the text translated and has studied the map extensively.
when the map was probably drawn, there are also several The text discussion is based mainly on her findings. See Gole, Indian
features that are mythological, such as the caves of the Maps and Plans, 142-43 (a full-page color photograph and two addi-
tional black-and-white enlarged excerpts, along with a brief discussion)
Pandavas, one of the two contending forces in the Indian
(note 99). A much more penetrating analysis, with numerous photo-
epic Mahabharata, which are shown by the six horizon- graphs, is Susan Gole, "A Nepali Map of Central Asia," South Asian
tal oblongs above (west of) Kabul. Some places that Studies 8 (1992): 81-89.
the actual configuration of territorial control at any single

historical date. The area I take to be Russia appropriately
envelops most of the conflated Caspian-Aral Sea in the
upper left, but Baghdad's placement within this region
makes no sense. Speculating on the date and purpose of
the map, Gole wrote in her initial effort at interpreting
the map:
It is possible that the map was drawn in the early years

of the 19th century, when attempts were being made
from Nepal to unite the neighbouring countries and
throw out the foreign European invaders, before they
became too powerful. The map may have accom-
panied a travel journal kept by one of the ambassadors
sent to negotiate with the courts of far off countries,
figured in the map. Attempts to learn more about its
provenance in Kathmandu have been unsuccessful. 123
FIG. 15.34. REFERENCE MAP OF CENTRAL ASIA FOR
Although the seriousness of the errors over much of the FIGURES 15.32 AND 15.33. Only relevant place-names and
map argues against its having been made in the course physical features are indicated. Contemporary politi.cal bound-
of, or as a result of, the travels of any ambassador or aries are shown for reference purposes only. No specific referent
for Kazak can be identified. Laghman is included as a regional
other single individual, the idea that a map so elaborate designation. Saratov is off the map to the north.
in execution would have served a diplomatic purpose
strikes me as eminently plausible. In Gole's later and
deeper study of the map, however, she backed away from
that position, which appeared no longer tenable in light of doing this would have been to commission a map,
of her discovery in Russian Korali Kadar (upper right) of with relevant notes on the character of the regions and
a "square building [which] appears to be the fort built peoples in question and on the distances that would have
by the Russians ... in 1855, known as Kazala."124 Assum- to be traversed if their forces were to act in unison. The
ing the correctness of the 1855 date, that would rule out potential key ally in the undertaking would have been
the possibility that the map was made in the period of Afghanistan, which had dealt Britain one of its most
rapid territorial expansion by the Nepali House of humiliating defeats in the Afghan War of 1839-43 and
Gorkha, which was terminated abruptly by the Anglo- which was strategically best situated to strike a new blow
Nepali War of 1814-16. against the raj. This would explain why many of the road
Nevertheless, there are grounds to pursue Gole's orig- distances on the map were to Kabul rather than to Nepal
inal hypothesis in a modified form. Not too long after itself.
its decisive defeat by the British, Nepal, thereafter a Brit- To obtain the desired intelligence the hypothesized
ish protectorate, was witness to the great convulsion of plotters would need to engage a person believed to pos-
the so-called Sepoy Mutiny of 1857-59, which for a brief sess considerable knowledge of the world beyond Nepal
moment in history appeared as if it might signal the end as well as familiarity with mapmaking. The evidence of
of British rule in the subcontinent. Many north Indian the map itself, specifically its combining of the features
princes and landlords then joined forces with the muti- of both British and Tibetan styles of cartography and the
neers, and though Nepal officially remained loyal, the richness of the textual information provided, suggests
temptation among antigovernment factions in Nepal-a that such an individual was indeed commissioned for the
country remarkably prone to political intrigue-to pursue task. That much of what was presented in the map was
an opposite course must have been considerable. To any factually incorrect is beside the point, for there was prob-
such faction, forging a grand alliance of Asian powers ably no one among the conspirators who could challenge
would be an understandable objective. It is even con- the map's authenticity. The mapmaker, of course, need
ceivable that Nepal's close contacts with Tibet inclined not have been sympathetic to the aims of those he was
people toward the idea, embodied in the Gesar epic, that working for or have believed their goals were obtainable.
a savior would come out of the west, or else from the In fact the evidence suggests quite the opposite-that he
mythic realm of Shambhala. In any event, it would have sought, rather, to dissuade the plotters from embarking
seemed reasonable for Nepali conspirators to seek out
whatever intelligence they could obtain that might bear 123. Gole, Indian Maps and Plans, 143 (note 99).
on the feasibility of their grandiose aspirations. One way 124. Gole, "Nepali Map," 82 (note 122).
on a venture doomed to failure and one that might cost tual guides providing topographical information and serv-
him his head were he subsequently to be seen as sup- ing as commentaries on the paintings. As Lauf notes,
porting it. 125 That would explain the numerous, largely "The precision of the details . . . allows us however to
counterfactual, disparaging remarks about the peoples in assume that the artist must have seen these buildings with
the areas portrayed; for example: "The Pathans of Kabul his own eyes."130 Yet the desideratum of verisimilitude
are cowards, it has been determined, they are also deceit- does not prevent the composition from being ordered
ful." "Throughout the kingdom of Kabul, there are "like a mandala of sacred centers of Buddhist doctrine
screws of wood in the swords." "Having determined that in and near Lhasa."131
the mughals of Bokhara are absolute cowards, the wom- The style of depicting landscape on these paintings is
enfolk at least did some brave deeds." "Iranian Muslims said to descend from that of works produced during the
are very arrogant. They take a bath if even the shadow Qianlong period in China (1736-96).132 Taken as a whole,
of a Hindu falls on them."126 the presentation is exalted, being "treated as a vision of
The colors on the map, seemingly used-albeit in- a Pure Land here on earth, which is what Tibetans believe
eptly-to indicate political control, also provide grounds it to be. There are rolling hills, winding streams, jeweled
for supposing that the work dates from the mid-nine- trees, and pastel clouds, on a broad and tilted picture
teenth century, even if they do not enable us to point to plane."133 The manner of composition of all five maps is
a more specific date near 1857. For example, the dis- not greatly different from that of figure 15.32, being
tinctive rose color of the area around Herat suggests that marked in particular by the conventionalized representa-
the mapmaker viewed it as an independent state, which tion in exaggerated size of the major urban places within
in fact it was during 1839-57 or, arguably, until its annex- a walled circular field, and the consequent contraction
ation by Kabul in 1862. Similarly, the beige coloration of the much larger nonsacred spaces between them. Also,
of what seems to be the emirate of Bukhara suggests that because of the seeming reluctance to leave any large
it too was regarded as independent, as was indeed the
case from 1753 until its subjugation by Russia in 1868. 125. For an account of some of the antigovernment conspiracies in
Although these and previously noted arguments are Nepal immediately before the Sepoy Mutiny, including a potential
revolt by Gurung troops within the Nepalese army, see Ikbal Ali Shah,
inconclusive, they are certainly plausible.
Nepal: The Home of the Gods (London: Sampson Low, Marston,
[1938]), 84-97.
Maps Emphasizing Sacred Places in the Regions 126. Gole, "Nepali Map," 86-87 (note 122).
127. The second of the two maps (inv. no. M.G. 21248) is illustrated
around Lhasa and Kathmandu
in Dieux et demons de l'Himalaya, 241, text on 238 (note 33); and in
Vergara and Beguin, Dimore umane, santuari divini, cover illustration
Perhaps the most sumptuous of Tibetan regional maps
with caption on inside cover page (note 27). A full-scale reproduction
available for public display outside Tibet itself are a group of the map proper (61 x 45 em) was available for sale by the Musee
of large and similar works depicting Lhasa and other Guimet in the form of a jigsaw puzzle.
important monastic towns in central Tibet. I know of 128. The Antwerp map (inv. AE 73.25) is extensively discussed in
five such works, though others almost surely exist. Two Detlef Ingo Lauf, Lhasa: De heilige stad van Tibet en haar omgeving
(Antwerp: Etnografisch Museum van de Stad, 1974), in Dutch with
are at the Musee Guimet in Paris: one (plate 34) said to
summaries in French, German, and English. Readers using this generally
be of the eighteenth century (but possibly the nineteenth), exemplary booklet should be advised that the captions therein generally
and a very similar work dating from the nineteenth cen- do not match the numerous photographs next to which they appear.
tury.12? Two are in Belgium: one believed to date from A separately published three-page set of errata corrects this problem.
the second half of the nineteenth century, at the Eth- The Brussels map (Collection Leon Verbert 349) is illustrated with a
brief note in Pia Van der Wee, Louis P. Van der Wee, and Janine
nografisch Museum, Antwerp, and a similar but some-
Schotsmans, Symbolisme de l' art lamafque (Brussels: Musees Royaux
what smaller, undated work in the Musees Royaux d'Art d'Art et d'Histoire, 1988), 32-33; and also in Dieux et demons de
et d'Histoire in Brussels. 128 Finally, the largest of the I'Himalaya, 240 (note 33). It also appears in four museum catalogs
group, thought to be from either the Kham region of listed in the former.
eastern Tibet or Mongolia and to date from either the 129. This map, from the George Crofts Collection of the Royal
Ontario Museum (cat. no. 2193), is published and discussed in Rhie
eighteenth century or the first half of the nineteenth, is
and Thurman, Wisdom and Compassion, 374-75 (note 66). The
in the Royal Ontario Museum in Toronto. 129 This work museum description (personal communication, September 1989) sug-
is so strikingly similar to the one in Brussels that it seems gests an eastern Tibetan provenance, and Rhie and Thurman suggest
safe to assert that one is a copy of the other or that both either that region or, for reasons not stated, Mongolia, which strikes
were copied from some unknown third model. Before me as unlikely; the museum assigns the earlier date and Rhie and Thur-
man the later.
the Chinese takeover in Tibet, paintings of this kind were
130. Lauf, Lhasa, 21 (note 128).
displayed for pilgrims in monasteries and at important 131. Rhie and Thurman, Wisdom and Compassion, 374 (note 66).
points of passage to provide them with a visual guide to 132. Rhie and Thurman, Wisdom and Compassion, 374 (note 66).
the places to be visited. The monasteries also edited tex- 133. Rhie and Thurman, Wisdom and Compassion, 374 (note 66).
empty space, many landscape elements are incorporated of Kathmandu (fig. 15.35) that is conceptually something
into the painting, though some of them may have no of a hybrid between the biographical (actually hagio-
more than a purely decorative function. Of the five maps, graphical) maps and the types of regional maps discussed
those in Antwerp, Brussels, and Toronto are so com- above. This beautiful, richly colored work is painted in
pressed that what is depicted might pass for a single con- a distinctly N ewari style. It is unusual in that it bears a
urbation; yet the area included on those works and on precise dedicatory date, a specific day in the year Newar
the two in Paris would extend over several thousand Samvat 923 (A.D. 1802) and in that among its abundant
square kilometers. The Paris maps show several rivers text is an inscription stating that it was commissioned by
and mountain ranges, none of which I have identified Sri Cikhidi of the Sakya clan, whose genealogy is then
with certainty. These are probably intended to serve provided. This map is the subject of a detailed analysis
largely as regional dividers, much as mountains do on the by Banerjee, who states that "such maps are often per-
hagiographic paintings discussed in the preceding section, sonalized" but does not indicate the grounds for imply-
and it is possible that most of the mountains, if not the ing, as he does, that maps of the type we are considering
rivers, are not meant to represent actual physical features. are a common genre. The map, he says, was not only a
The rivers and mountains shown on the maps in Belgium souvenir of the pilgrimage but in itself "an object of
and Toronto are more sublimated than on those on the reverence."135
maps in Paris and are more naturalistic in appearance. In The content of Cikhidi's map provides a vivid impres-
general, the oblique perspective of the former maps is sion of the domestic architecture of the Vale of Kath-
that of an observer viewing the region from a much lower mandu, of the appearance of a number of its important
position than in the case of the Paris examples. Hence Buddhist and Brahmanical monuments, of the icono-
the format of the Paris maps has a long vertical dimen- graphy of its religious sculpture and mural painting, and
sion, whereas the other three examples (Antwerp, Brus- of the former mode of dress of its Newari inhabitants
sels, and Toronto) greatly compress the vertical (whose religion is itself a Buddhist-Hindu hybrid faith).
dimension. 134 The perspective of the artist is mixed, usually frontal for
The published guide to the Antwerp map identifies all single edifices and oblique for large architectural com-
the major settlements and many individual structures plexes, with individual components of them in a frontal
within them. Although no comparable descriptions of the perspective. As in Tibetan maps, people are shown in
other works are available, comparison with the Antwerp exaggerated size. Almost all the available map space is
example should let one identify most places. For exam- given over to interesting cultural and physical detail. The
ple, on all the maps the identity of the centrally situated road network and drainage pattern are more completely
Potala in Lhasa is unmistakable, and the distinctive and realistically presented than on any of the previous
appearance of the monastery in the lower-left corner of regional maps, though the courses of individual roads
the works at Paris and Antwerp lets one immediately and rivers have been distorted to fit the demands of the
recognize it as Samye (bSam yas) (compare fig. 15.47 composition. The mountain range at the top of the map
below), nearly forty kilometers to the southeast, though is obviously the Himalayas, lying just north of the Vale
in plate 34 it appears to be only a stone's throwaway. of Kathmandu, including a distant range of snowcapped
The relationship between those two settlements suggests peaks and a nearer snow-free crest. But Banerjee errs, I
a general westward orientation to the map; but there does believe, when he states: "A typical pictorial map is ori-
not appear to be a concerted effort by any of the artists ented ... towards the north, in the style of modern maps,
to maintain consistent directional relations among the and is, therefore, scientific in its outlook."136 The north-
places shown. Monasteries that could be intended as ern orientation in this case merely enables a composition
Samye show up near the lower-right corner of the the in which the Himalayas provide an appropriate skyline
Brussels and Toronto examples. I have not established away from which rivers flow toward the bottom of the
all the correspondences among the five works, since it
would be exceedingly difficult using only small-scale pho-
134. The Antwerp map measures 90.5 x 148 em, Brussels 88.5 x 120
tographs, but the task would be worthy of carrying out
em, and Toronto 135.4 x 184.6 em.
from direct inspection of the paintings. Also worthy of 135. N. R. Banerjee, "A Painted Nepalese Paubha in the Collection
investigation is the consistency of orientation with respect of the National Museum, New Delhi," in Buddhist Iconography (New
to the individual centers depicted. My impression is that Delhi: Tibet House, 1989), 154-63 and figs. 27-37; quotations on 155
the directional perspective from which particular towns and 159. An abridged version of this article, but with larger and clearer
photographs, is N. R. Banerjee and O. P. Sharma, "A Note on a Painted
are shown reflects the way they are most likely to be seen
Map of the Kathmandu Valley at the National Museum, New Delhi,"
when approached from Lhasa, without reference to Marg 38, no. 3 [1986]: 77-80. The work is also illustrated and discussed
actual compass bearings. by Gole, Indian Maps and Plans, 68 (note 99).
There is another large scroll map (paubha) of the Vale 136. Banerjee, "Painted Nepalese Paubha," 155 (note 135).
map. A similarly rendered skyline appears clearly at the

top of one of the two regional maps at the Musee Guimet,
and one is also evident, though sometimes more muted,
on the other maps of the Lhasa region (compare plate
34). I am also inclined to discount Banerjee's implication
that the artist might have tried to adhere to any particular
scale, "between half an inch to a mile and an inch to a
mile," in painting the map,137 Finally, I would reject his
suggestion that the artist who painted the map for Cikhidi
might have received training from the British (although
he thought it unlikely), given that comparably detailed
maps were made in Tibet well before the time of signif-
icant British intercourse with Nepal. 138 It seems more
probable that the stylistic traditions of Tibet and Nepal
influenced one another more than either tradition derived
from exposure to British influence, at least until after
Nepal's military defeat by the British in 1814-16.
Tibetan Maps in the Wise Collection
Perhaps the most comprehensive set of relatively large-

scale Tibetan regional maps is the Wise Collection, com-
prising three large folios of maps and drawings held by
the Oriental and India Office Collections, in the British
Library, London. Most of the maps in the collection are
assembled from multiple sheets of European paper
spliced together to form extensive and very detailed
cartographic documents. The constituent map sheets vary
in size, but most approximate forty-eight by sixty cen-
timeters. Figure 15.36, for example, illustrates a portion
of one such map (less than a third of its total surface),
showing all of two sheets and parts of two others that
compose it. Although the maps in the Wise Collection
were at some unknown period the object of considerable
study, as evinced by the hundreds of English annotations
and key numbers they bear, the circumstances of their
compilation and history before their acquisition by the
FIG. 15.35. NEPALI MAP OF THE VALE OF KATH- India Office are shrouded in mystery. The maps and draw-
MANDU. Newari style, text in Newari and Sanskrit, Newar ings, all in black ink and watercolor, are attributed in the
Samvat 923 (A.D. 1802). Painted on cloth in black, red, light undated, anonymous catalog description to "a Tibetan
and dark green, yellow, gold, and white. This map provides a
remarkably vivid and accurate impression of many aspects of artist, probably a lama, who had had contact with Euro-
the cultural landscape of the area around Kathmandu in the peans and developed a semi-European style of draw-
early nineteenth century, especially in respect to places of reli- ing."139 In my view, however, the map style remains
gious significance. The patron who commissioned the work is essentially indigenous. The maps, which could justifiably
shown, along with his family and other members of his retinue, be categorized under the traditional Tibetan genre of sa-
at many pilgrimage sites they jointly visited in that region. The
abundant text of the map aids in the identification of all the khra (picture map), are described as "similar to pilgrim
major places depicted. The orientation is toward the north, maps showing villages and buildings in pictorial form,"
where the snowcapped Himalayas form the skyline. while the "25 drawings of places, monasteries, people
Size of the original: 152 X 82 cm. National Museum, New and ceremonies in Tibet ... appear to have been com-
Delhi (acc. no. 56.1168). From Susan Gole, Indian Maps and
Plans: From Earliest Times to the Advent of European Surveys
(New Delhi: Manohar, 1989), 68.
137. Banerjee, "Painted Nepalese Paubhil," 155 (note 135).
138. Banerjee, "Painted Nepalese Paubha," 159 (note 135).
139. Oriental and India Office Collections, British Library, London,
Add. Or. 3013-43, unpublished catalog description, nine pages, p. 1.
FIG. 15.36. PART OF A LARGE MAP OF A PORTION OF explanatory inscriptions in Tibetan for which-in contrast to
CENTRAL AND EASTERN TIBET. This illustration covers other portions of the colIection-no translation is currently
somewhat less than a third of an irregularly shaped painted map available. The map is believed to have been made by a Tibetan
that forms one of seven large "picture maps" in the Wise Col- who was specialIy trained by a British mentor (possibly Wise,
lection of Tibetan maps. The portion shown in the photograph about whom nothing is known) to carry out mapping and other
comprises all or parts of five pages, out of a total of six, that intelligence gathering sometime between 1844 and 1862.
were pasted together to form the complete map. The principal Size of the detail: ca. 70 x 85 cm. By permission of the Oriental
center shown in this illustration is the famous monastery at and India Office Collections, British Library, London (Add. Or.
Samye, the oldest in Tibet. Included on the map are numerous 3017).
missioned by the writer of the explanatory text."l40 Just

who that writer was is not known. 140. Unpublished catalog description (note 139).
Although the bindings (f. 1890-1900) are lettered 141. Unpublished catalog description (note 139). I would dismiss as
incorrect the handwritten suggestion added to the prefatory note for
Wise, no record of anyone of that name connected
the catalog description that the Wise in question could have been James
with Tibet or adjacent areas can be found in the rec-
F. N. Wise (1834-85), author of Notes on the Races, Castes, and Tribes
ords. The water-marks [1849-56] of the explanatory of Eastern Bengal (London: Harrison, 1883). I have checked the latter's
text as well as internal evidence ... suggest a date for career via his several contributions to the Journal of the Asiatic Society
the drawings of between 1844 and 1862. The notes of Bengal, beginning in 1873, and, in addition to the fact that he would
suggest that "Wise" was acquainted with the western have been too young to be involved with our anonymous lama as early
side of Tibet and had travelled in Ladakh. 141 as 1844, I find no grounds to link him in any way with Tibet.
In light of the British concern with military intelligence animals and plants, are drawn at scales vastly greater than
beyond the frontiers of India at the time the maps were those employed for physiographic features. The actual
made, a period straddling the establishment in 1849 of map size seems to be mainly a function of the amount
a British protectorate over Jammu and Kashmir, of which of detail the artist feels it necessary to depict. But whereas
Ladakh was a part, it is reasonable to suggest that the architectural features are often shown in an oblique per-
"lama" (or lamas) who made the maps in the Wise col- spective in other Tibetan maps, in the Wise maps they
lection was (were) secretly recruited for that task and that are, with relatively few exceptions, shown in frontal per-
the name Wise was simply a pseudonym for some now spective, presumably from the most likely direction of
forgotten British official who saw to his (their) recruit- approach. In respect to map orientation, no general rule
ment and training and who, for diplomatic reasons, could appears to be followed; though each page of a multisheet
not be correctly identified. Much, of course, has been map does appear to have one predominant orientation,
written about the later "pundits" who worked for the the orientation may shift as one goes from one sheet to
Survey of India. They were brilliant individuals selected another, as one can see by contrasting the lower portion
from among the peoples of several faiths living on India's of figure 15.36 with the other visible portions of the map.
borders, who were specially trained to carry out trans- Further, whereas in a purely Tibetan style mountain sum-
frontier intelligence and mapping missions, using assumed mits are usually portrayed as pointing away from an
names and disguised as pilgrims and traders, during the observer on the ground (e.g., in the opposite direction as
period 1863-93. But such missions had precursors as early seen from a road following a river valley), that convention
as the Central Asian expedition of Mir Izzet Ullah in is frequently violated in the Wise maps. Another seeming
1812-13, and it seems likely that the maps in the Wise concession to Western cartographic style is the depiction
Collection were the outcome of a number of such pre- of linear features such as roads and many rivers as nar-
cursor expeditions. 142 Several facts support this line of rower than they would be on most Tibetan maps. The
reasoning. First, the maps were on European paper. Sec- common Tibetan device of having smaller streams appear
ond, they were pieced together in such a way-some from and then disappear behind mountains is largely
pages taking off at right angles from those that had pre- retained, as might be expected from a cartographer
viously been assembled (as in fig. 15.36)-as to suggest unable to follow every stream to its source and trying to
that the mapmaker could not fully anticipate what and map only what he actually saw as he made his way
how much mappable territory lay before him as he made through a topographically complex countryside. Finally,
his way into unknown country. No other Tibetan maps in landscapes largely devoid of forests, lines of trees and
I have seen have this characteristic. Finally, one must village groves are prominently depicted wherever they
consider the map content-for example, the great detail occur.
in which bridges were portrayed, the annotations relative Whether the maps in the Wise Collection were made
to routes and the distances to places beyond the limits by a single artist, as the catalog description suggests, or
of the maps themselves, and so forth. 143 The maps and by two or more individuals is open to question. Although
appended notes also contain much ethnographic and most of the maps in the collection are anything but crude,
other detail that, though of only marginal interest for there are exceptions (items a and b of appendix 15.1),
military purposes, would be highly relevant for the con- and I am inclined to think that at least two individuals
duct of diplomacy.l44 Inclusion of certain more esoteric were at work. But, it is possible that these exceptions are
religious details might be explained on two grounds: first, relatively crude because some circumstance (e.g., the need
as being of personal interest to the map compiler, who,
though working for the British, was probably nonetheless
a devout Buddhist; and second, as a means of seeming 142. A very readable and generally authoritative account is provided
to legitimize an activity that would otherwise raise by Derek Waller, The Pundits: British Exploration of Tibet and Central
unwanted suspicions by Tibetan officials curious to know Asia (Lexington: University Press of Kentucky, 1990), 14 and 22 for
what the mapmakers were up to. Mir Izzet Ullah in particular.
Appendix 15.1 provides a synopsis of some of the 143. Examples of annotations that would have been militarily useful
include the following (all taken from Add. Or. 3014): "Another bridge
important particulars relative to the Wise maps, including across the Indus. The river frozen in winter and passable." "Kera-a
several that do not properly qualify as "regional." To the village; road not passable here in summer when the river is very large."
details provided there, I can add a few more general "Hanle is eleven marches from Le [Leh]."
notes. First, what appears to be most characteristically 144. Examples from Add. Or. 3014 include: "Pituk ... a gunpo
Tibetan in the style of the maps is their attention to [monastery] of the Gelukpa ... sect who came from Tashi-Iunpo."
"Tok, the jaghier [personal estate] of the present raja of Ladak, given
architectural detail. In general, monasteries, chortens (stu- to him by Maharaja Golab Singh, after he conquered the country, ...
pas), palaces, and other major cultural features are ren- Masho, residence of the brother of the Ladak Raja, who is a lama."
dered in conventionalized colors and, along with people, "Liktse, a village & where a tax gatherer of the Ladak govt. stays."
FIG. 15.37. PORTION OF A TIBETAN MAP SHOWING the difficult topography through which the Tsangpo flows in
THE AREA IN THE GREAT BEND OF THE TSANGPO/ eastern Tibet. Particularly striking is the drawing of summits
BRAHMAPUTRA RIVER. The map illustrated here was drawn pointing away from the main axis of travel along the river itself
on non-European paper in black ink and yellow and brown and their apparent convergence within the river bend. As the
watercolor. It probably dates from the late nineteenth or early map text has yet to be translated, a full interpretation of its
twentieth century and was obtained by Heinrich Harrer in content is not yet possible.
Tibet, about 1950. The artist employed many conventional sym- Size of the original: 52 x 171 em. By permission of the Volk-
bols of Tibetan cartography in preparing this vivid depiction of erkundemuseum der Universitiit Zurich (cat. no. 14495).
for security) forced the mapmaker to prepare them with dix is based on a brief firsthand inspection in 1987, study
more haste than was customary. of several published photographs, and notes in a museum
Although a number of published works have drawn catalog relating to an exhibit of Harrer's materials. 146
attention to the maps in the Wise Collection, no one so
far has analyzed them in any detail. 145 Such analysis-not 145. Among the works that cite the Wise Collection are Aziz,
only of the maps, but of the accompanying drawings and "Tibetan Manuscript Maps," in which fig. 1, labeled "folk drawing,"
text-by a scholar with the requisite linguistic skills and is excerpted from one of the Wise maps (which one and the precise
locale depicted are not stated); and idem, "Maps and the Mind," in
knowledge of Tibetan culture and history is sorely
which the same excerpt appears on p. 55, labeled "pilgrim's map" (both
needed and bound to be richly rewarding. in note 12); Marrin Brauen, Feste in Ladakh (Graz: Akademische Druck-
u. Verlagsanstalt, 1980), 14 (fig. 1); and Moller and Raunig, Der Weg
zum Dach der Welt (note 34), in which there are map excerpts showing
Tibetan Maps in the Harrer Collection
the monastery complexes at Shigatse (p. 112), Gyantse (rGyal-rrse) (p.
263), and an unspecified locale (p. 357).
Another noteworthy collection of mainly regional maps
146. The photographs are published in Marrin Brauen, Heinrich Har-
from Tibet is in the Volkerkundemuseum der Universitat rers Impressionen aus Tibet (Innsbruck: Pinguin-Verlag, 1974), pIs. 9
Zurich. These constitute a portion of the Tibetan col- and 110 in color, and fig. 7 (p. 104) in black and white. The book
lection of Heinrich Harrer, acquired by the museum in includes a fifteen-page catalog of the exhibition at the museum that
1972 and still awaiting detailed study. The maps [ have opened in December 1974. The same map excerpt that appears in pI.
9 of Brauen is reproduced on p. 88 of Moller and Raunig, Der Weg
seen are on non-Western paper and rendered in ink and
zum Dach der Welt (note 34), where it is labeled "Landschaft am Knie
various watercolors. One of the maps in the collection des Tsangpo-Flusses," which would only barely place it, as Brauen does,
is illustrated in figure 15.37. Appendix 15.2 provides in the Tibetan province of Kham. Another excellent photograph, which
details of those maps and some of the others. The appen- appears on stylistic grounds to be remarkably similar to most in the
The age of the maps is uncertain, but I would judge as Huber pointed out, although "one end of the map
that none predates the late nineteenth century or post- does indeed depict some [my emphasis] of the south-
dates the first quarter of the twentieth. Since Harrer left western section of the .long Tsa-ri rong-skor pilgrimage
Tibet early in 1951, after a sojourn of seven years, that route, ... the majority of it covers areas well to the west
year represents an absolute terminus ad quem for the
entire collection. 147 How and why Harrer acquired these
specific maps is at present not known. Since several of
Harrer Collection, appears in Tichy, Himalaya, 133 (note 43). This very
the works in his collection appear to relate to a military
detailed map, labeled "Alte tibetische Landkarte," is not described in
campaign, however, as is evident from the depiction of the work cited and is not attributed in any way to Harrer. Rather, it
weapons, military encampments, a field with a few is stated that the photograph came from Ella Goldschmidt of Vienna
corpses, and what appear to be battle lines, we may sup- (who may well have obtained it from or through Harrer). My attempts
pose that he hoped to use at least some of the maps to to establish contact with her proved futile.
147. Harrer, an Austrian mountaineer, was interned in India during
put forward a Tibetan view of the several Anglo-Tibetan
World War II and in 1944 escaped from detention in the Indian town
skirmishes resulting from the British Younghusband of Dehra Dun, subsequently making his way into Tibet. After a long
expedition to Lhasa in 1904. 148 and difficult journey to Lhasa, he was allowed to reside in that city,
Although it is virtually certain, on stylistic grounds, that and in time he became a trusted employee of the Tibetan government
the Harrer maps were made by several different hands, and a tutor of the young Dalai Lama. He did not leave Tibet until
March 1951, some months after the Chinese occupation of that country.
most show broad similarities, and all use conventions of
He describes his Tibetan sojourn in Seven Years in Tibet, trans. Richard
the type depicted in figure 15.9 to show landscape ele- Graves (London: Rupert Hart-Davis, 1953), but he does not mention
ments, especially various types of mountains. Other con- in that work that Tibetan maps were among the substantial body of
ventions for depicting religious edifices and settlement Tibetan materials he took with him to India.
also tend to be broadly similar and comparable to those 148. I am, however, aware of no attempt by Harrer to use the maps
in the way suggested. Of greater concern to Harrer was the cause of
used in the earlier maps of the Wise Collection.
Tibetan independence. To have called attention to Tibet's past humil-
A singular exception, however, is one uncataloged iation by Britain, a nation generally inclined to support Tibet's interests
map, a portion of which-judging from the house types vis-a-vis China, would have to be a diversionary exercise that would in
and rope suspension bridges shown on it-relates to an no way further the aims of Tibet in the mid-twentieth century. Harrer
area of transition from Tibetan culture to the tribal cul- makes no mention of the maps in his work Return to Tibet, trans.
Ewald Osers (New York: Schocken Books, 1984).
ture of what is now the northeast Indian state of Arun-
149. Escaping twice from Tibetan captivity, Kintup managed to carry
achal Pradesh. That area was of considerable interest to on with the mission (1880-84) after the defalcation of his Chinese
Anglo-Tibetan and Anglo-Chinese diplomacy during the master, who sold him into slavery in May 1881. His account of the
period leading up the Simla Conference of 1914, when routes he explored was largely discounted, and Kintup was all but
Britain and Tibet agreed on the McMahon Line as their forgotten until 1914, when his achievements were finally recognized.
The stories of Kintup, Morshead and Bailey, and others concerned with
common border in that region, a border which China
making known the region in question are recounted in Waller, Pundits,
stoutly rejects. The area was first surveyed for Britain by 214-47 (note 142). It is conceivable that information obtained from
Henry T. Morshead and Frederick M. Bailey in 1912- Kintup while he was a captive in Tibet was in some way related to the
13, but it was visited in 1884 by Kintup C'KP"), the illit- decision by the Tibetans themselves to prepare the map and possibly
erate Sikkimese servant of a Chinese "lama" employed others that may yet come to light. It is also conceivable, though less
likely, that the map was made using information the Tibetans extracted
by the Survey of India to determine whether the Tsangpo
from Kintup, whose powers of recollection were said to be extraor-
and the Brahmaputra were the same river, as had been dinary. In either event, Harrer's being entrusted with the map when he
suspected. 149 left Tibet for India in 1951 may have related to some now unfathomable
Tibetan diplomatic purpose.
150. Although I have had an opportunity to study the map at first
Other Regional Maps hand, my discussion of it is based on work by Huber, who in con-
junction with a Tibetan colleague, Tashi Tsering, resident in Dharmsala,
A small section of a large and elaborate cotton scroll India, has translated all of the abundant text on the map and prepared
map is illustrated in figure 15.38. This map has been stud- glosses, some of them rather lengthy, keyed to most of its 140 inscrip-
ied in great detail by Huber. 1so It was given to Hugh tions; Toni Huber, "A Tibetan Map of IHo-Kha in the South-eastern
Himalayan Borderlands of Tibet," Imago Mundi 44 (1992): 9-23. Hub-
Richardson in Lhasa about 1944, when he was the repre- er's analysis is by far the most extensive and thorough scholarly com-
sentative of the British government in Tibet. The work, mentary on a Tibetan map I have seen. His kindness in making this
which Richardson described as "well used" at the time work available to me in draft form is acknowledged with gratitude.
he acquired it, is not dated, but it is ascribed by the British 151. Aziz, "Tibetan Manuscript Maps," 29, with a partial illustration
on page 30, and idem, "Maps and the Mind," 54-55, with a different
Museum to the nineteenth or twentieth century and des-
illustration overlapping those two pages (both in note 12). A small
ignated a "pilgrim map." The map was noted and partially portion of the map has been printed in color on the dust jacket of Aziz
reproduced in several publications by Aziz, who describes and Kapstein, Soundings in Tibetan Civilization (note 28), but this will
it as a pilgrimage map of Tsa-ri (rTsa-ri) rong skor.t s1 But, probably have been removed by most libraries.
FIG. 15.38. SMALL PORTION OF A LARGE TIBETAN MAP record of physical, administrative, ethnographic, and other
OF THE TSA-RI REGION AND AN ADJACENT PORTION needed intelligence relating to an area of conflicting political
OF ARUNACHAL PRADESH, INDIA. Painted on a cotton interests in the early part of this century. The map is exceedingly
scroll, this undated Tibetan map shows an area along the upper rich in place-specific information, as is evident not only from
Subansiri River straddling Tibet's border with the present Indian its visual detail, but also from its numerous inscriptions.
state of Arunachal Pradesh. Long thought to be a pilgrimage Size of the entire original: 63.5 x 351 em. By permission of the
map, the work appears instead to have been prepared, very likely British Museum, London (1986.5-26.01).
at the behest of the Tibetan government, as a cartographic
of Tsa-ri proper."152 It is also clear from the map text 4. "Crossing the Gla-'khol pass from bKra-shis monastery it is
that its author had little or no concern with religion; but one day's journey as far as IHun-rtse (rDzong) [castle)."
rather, the map was related to travel in general, questions 35. "A forty-live-step ladder [clearly shown on map) to the upper
valley."
of political and fiscal administration, and topographic and 49. "Peaks on the eastern bDag-ri range."
ethnographic intelligence. IS3 53. "The precipice which is the extremity of the (bDag-ri [refer-
ring to another note]) mountains."
152. Huber, "Tibetan Map," 9 (note 150). Aziz appears also to have 55. "From this resthouse above the Bya-chu [a river) it is live
erred in stating that Richardson acquired the map "before World War days' journey as far as the nine passes and nine valleys."
II" and in placing Tsa-ri in northern Tibet and in a valley in a "legendary 58. "All (this area is) the continuous frontier of the country
area accessible only by passage through dangerous lands inhabited by with the lDing-klo [a tribal group of Arunachal Pradesh) in those
brigands and wild forest people," rather than near Tibet's southern (places) behind, to the south of here."
border ("Maps and the Mind," 55 [note 12]). Although the area covered 67. "The narrow footpath to Klung."
is indeed difficult to reach, estimates of the number of pilgrims from 75. "There is only Klo(-pa) [tribal) country behind the passes
all parts of Tibet making the Tsa-ri pilgrimage (held every twelve years) on the entire southern Rank of these snow mountains (i.e., the
range from as few as 10,000 to 15,000 to as many as 100,000 (Huber, Grand Himalayan range)."
"Tibetan Map," 17 [note 150]). From Huber, "Tibetan Map," 4-18, passim (note 150).
153. Consider, for example, the following translated map inscriptions Those who wish to compare the content of the map with that of
(using Huber's numbering system): the Tibetan guidebook discussing a number of the features the map
The map content is largely conveyed by inscriptions begun in the year 1648 and repeatedly painted over during
and, even more, by its highly varied graphic elements/ the following three centuries, are said to adorn "every
signs, using both shape and color to convey meaning. It hall, every chapel, and every corridor," often being drawn
is thus a particularly good example of the sa-khra (picture with a "vertical view and diversified perspective" sug-
map) genre. Several different signs are used to show gesting maplike images. 156 Further, it is reasonable to
mountains, some suggesting glacial terrain, others karstic suppose that what was true of the Potala was also true,
highlands, and still others lower, rounded summits. As though obviously to a lesser degree, for other major mon-
in many other Tibetan maps the summits point away from asteries. Regrettably, however, relatively few Tibetan
the observer's route in a valley, in this case of the Suban- monasteries have survived the cultural vandalism and offi-
siri River, that forms the main axis of the map. Although cially sanctioned destruction that Tibet has suffered since
it might at first appear that the mapmaker had little con- 1950, especially during the Chinese Cultural Revolution
cern for what layaway from this axis, the notes on the of 1966-69. 157
map-many of them referring to the number of days' Mural and scroll maps are also known to exist in Bud-
march to other locales and the passes to be crossed to dhist monasteries outside Tibet. For example, a seven-
get there-prove otherwise. Other map notes point out teenth-century map of Bhutan may be seen on a wall of
areas of special interest such as grassy flats and plains. the Punakha dzong (castle). And as noted in the intro-
Many regions and areas occupied by particular tribes and duction, long cloth map scrolls (paubhas) are periodically
clans are explicitly identified. Dozens of settlements-not put on public display in the courtyards of various Nepali
all of them identifiable-are depicted, as are more than monasteries. No inventory of mural and scroll maps has
a dozen monasteries, a comparable number of estates, yet been attempted, but it is highly probable that more
smaller numbers of temples and stupas, several forts, and will come to light in the near future. It is not always easy
at least one trading post. Footpaths are frequently labeled to classify the few known examples as regional, route,
as such, and at least one "cattle path" is noted. or locality maps. One such painting appears in figure 15.2
and an excerpt from another, an enormous work of the
The western portions of the map ... show the barren, eighteenth or nineteenth century, in figure 15.46.
rolling landscape of the Tibetan plateau marked reg-
Although the latter could be regarded as a regional map,
ularly by small villages (in general, only 1 or 2 buildings
it seems preferable to discuss it below under locality
are used to represent a village), religious establish-
maps.
ments and plots of land under cultivation. Moving into
the eastern portions the settlements become less fre-
quent and the Tibetan style dwellings begin to give PILGRIMAGE AND OTHER ROUTE MAPS
way to the thatched huts used by the pre-literate Arun-
achal tribals who visited the upper Subansiri valleys Travel and visibility in Tibet and neighboring areas are
on the Tibetan border in order to trade. The scene largely constrained by difficult mountain topography
painted here is marked by dramatic cliffs, waterfalls, marked by long and deep orographic troughs (such as
and a variety of vegetation ... including bamboo, those along the upper Indus, Sutlej, and Tsangpo rivers).
flowering herbs and different tree species, all reflecting Consequently, what might be intended as maps of major
the high rainfall that this area is well known for. 154
regions may take the form of rather elongated strips (as
do several in the Wise Collection), making the distinction
Huber does not suggest a date for the map or a respon-
sible agency, but I believe it is similar in date, origin, and
purpose to the uncataloged map in the Harrer Collection
discussed above. The circumstances surrounding the depicts may consult Elena Rossi Filibeck, "A Guide-Book to Tsa-ri,"
in Reflections on Tibetan Culture: Essays in Memory of Turrell V.
drawing of the Harrer map would have applied equally
Wylie, ed. Lawrence Epstein and Richard F. Sherburne (Lewiston, N.Y.:
for the map obtained by Harrer's contemporary, Rich- Edwin Mellen Press, 1990), 1-10.
ardson, since both works refer to a troubled and polit- 154. Huber, "Tibetan Map," 10 (note 150).
ically sensitive frontier region. 155. Letters from Todd Lewis to me dated 1 September 1987, 12
We can only speculate on how many and what sorts February 1988, and 28 March 1988.
156. The Potala Palace of Tibet, compo Cultural Relics Administra-
of additional regional maps survive in Tibet. Todd Lewis, tion, Tibet Autonomous Region (Shanghai: People's Art Publishing
an anthropologist specializing in the Himalayan region House, 1982), 52. This work contains numerous photographs with
and Tibet, reports having seen "very extensive frescoes glimpses of mural paintings, but none are good enough to permit a
that depict major regions of Tibet" in the Potala and meaningful assessment of their cartographic quality.
Norbulinka (Nor-bu-ling-ka, Summer Palace) in Lhasa, 157. Since the opening of Tibet to foreign tourists beginning in 1982,
the Chinese authorities have restored portions of some of the more
but he was not able to obtain photographs of such important monasteries. What effect, if any, this may have had on the
works. I55 The Potala is vast, and its treasures have yet to state of their cartographic murals is not known. Photography within
be made fully known beyond Tibet. Its mural paintings, the monasteries is largely forbidden.
FIG. 15.39. MAP PORTION OF A TIBETAN SCROLL presiding deities, occupy distinctive fields of color, and are iden-
SHOWING PILGRIMAGE PLACES IN THE VALE OF tified by inscriptions in the Newari script. Only a few of the
KATHMANDU. This map occupies the upper half of a large identified tfrthas shown, however, correspond clearly to cur-
fragment of a two-part Nepali cloth scroll completed in Newar rently frequented sites. The rivers on the map reflect the actual
Samvat 755 (A.D. 1634). Referred to as a "banner painting," the network of the Vale (but with the topological distortion
entire work was executed in opaque watercolors (various greens, required by the scroll format) and group those sites by subre-
blues, salmon, red, black, white, and gilt) on cotton. The lower gions. We may assume that this work, like other banner paint-
register of the work (not shown) is a narrative painting relating ings, was hung our for periodic display in the courtyard of some
to a traditional Buddhist history of the deity Svayambhiinath. Nepali vihara (monastery), and it is likely that exposure to the
The portion shown here is designated a ttrtha mahatmya, which elements is the reason only a fragment of the original remains.
signifies that its function was to glorify ttrthas (pilgrimage Size of the entire original: 39 x 130.2 em. By permission of the
places), in this case in the Vale of Kathmandu, and guide pilgrims Cleveland Museum of Art, gift of Mrs. Albert S. Ingalls (54.788).
to them. The ttrthas are symbolized by icons representing their
between regional and route maps more arbitrary and lasa) and other places of pilgrimage in its vicinity. The
problematic than in other parts of the world. Neverthe- work in question is a bas-relief bronze tablet that includes
less, I have selected here a number of works whose prin- a rather detailed representation of the physically visible,
cipal objectives appear to be to illustrate routes or places as well as mythological, features on and in the vicinity
with relatively little regard for the nature of the terrain of Kailas CTi-se to Tibetans). This is taken by Buddhists,
away from those routes. Some of these have the elon- Bon-pos, Hindus, and Jains to be the earthly manifes-
gated shape characteristic of many route maps, while tation of the sacred Mount Meru, the axis of the uni-
others do not. verse. 160 Despite its absence of text, the most important
Oldest among the known works of the latter type is a features shown on the tablet (see fig. 15.41) may be readily
fragment of a Newari scroll painting (fig. 15.39) that illus- recognized on medium- and large-scale maps of the
trates an ancient text, the Svayambhu Purii1Ja. 158 In its region it depicts, in particular Mount Kailas, which rises
original form, the entire scroll probably was of a type 6,700 meters above sea level; the lesser mountain, Dorje
similar to the one illustrated in figure 15.2. The fragment Tijung to its west (the tablet being oriented with the
shown here comprises two horizontal components, a viewer looking to the north); and Lakes Manasarowar
lower narrative register of scenes and an upper register
that has been designated as a trrtha mahiitmya (i.e., a
158. This map is discussed at length in Slusser, "Serpents, Sages, and
record of "eminent places of pilgrimage"). The map
Sorcerers" (note 13).
includes fifteen or so named pilgrimage places in the Vale 159. The entire fragment measures 39 by 130.2 centimeters, of which
of Kathmandu, of which most can be identified icono- the upper half is the map portion. Slusser, "Serpents, Sages, and Sor-
graphically, and a few additional scenes of worship. cerers," 68 (note 13), provides a key diagram on which the several map
These are regionally grouped into elongated fields (shapes elements are noted and, wherever possible, identified. Each is further
discussed in the accompanying text. The work is also illustrated in
necessitated by the scroll format), set off from one
Slusser, Nepal Mandala, vol. 2, fig. 569 (note 13). A photograph of a
another by the course of the Bagmati River and several more modern, though undated, painting, somewhat similar to the work
identifiable tributaries. 159 In its use of rivers as dividers, discussed by Slusser and relating to the same region-but with a dis-
this scroll map is similar in appearance to many cos- tinctly more Indian appearance-was given to me in Brussels in 1987
mographic scenes in the lengthy scroll presentations of by Armand Neven, a historian of South Asian art. At the time the work
was in the London gallery of Jean-Claude Ciancimino, but I do not
the Thai Buddhist Trai phum, discussed in chapter 17 of
know its present location. As in the earlier work, the various tirthas
this volume (compare fig. 17.23). This raises the possi- are grouped into fields by a network of rivers and are recognizable both
bility of the existence of an intermediate Indian form, by the images of their tutelary deities and by names written on the
which might have spread to both Nepal and Siam from painting itself. A line of mountain peaks near the top of the painting
a sacred site such as Bodh Gaya, frequented by pilgrims presumably represents the Himalayas.
160. See the expert analysis by Eva Stoll, "Ti-se, der heilige Berg in
from both countries. Tibet;' Geographica Helvetica 21 (1966): 162-67. The tablet is also
Totally different in form from the tirtha mahiitmya, illustrated and more brieRy discussed, in the broad context of Buddhist
though similar in function, is an undated Tibetan work beliefs about Mount Sumeru, by Van der Wee, "Rirab Lhunpo," 71-
(fig. 15.40) that relates to the sacred Mount Kailas (Kai- 72 and fig. 7 (note 57).
FIG. 15.40. TIBETAN BRONZE TABLET SHOWING particular note are the pilgrims circumambulating the mountain
PLACES OF PILGRIMAGE IN THE AREA OF MOUNT TI- clockwise.
SE (KAlLAS). This undated bas-relief tablet shows not only Size of the original: 34.2 x 28 em. By permission of the Yolk-
many places and monasteries that pilgrims are enjoined to visit, erkundemuseum der Universitiit Zurich (cat. no. 12665).
but also unreachable features that are purely mythological. Of
orienting pilgrims toward realizing the several objects of

their journey.
Included in Stoll's article are photographs of several
paintings of Ti-se and its surroundings. These are among
a number of pilgrimage maps that are remarkably similar
in content to the Zurich tablet: (a) a work by a Tibetan
lama, Nav-Kushok, taken from a book on the region by
the Indian explorer Swami Pranavananda, who spent
years in the vicinity of Kailas; (b) a wall painting, rendered
in a naive folk style, found in a Nepali village near the
Tibetan border; (c) a thanka painted by Tulku Tsewang,
remarkably similar to the earlier-mentioned wall map,
found in a monastery in Dolpo, Nepal; and (d) a painting
on paper of unknown date and provenance, but with
accompanying Tibetan text, that is rather different in style
and composition from all of the foregoing, though gen-
erally similar in content. 161
Well to the southeast of Kailas and Manasarowar is
another sacred mountain-lake dyad, the peak of Gosain-
than (Shisha Pangma) and the lake of Gosainkund, with
respect to which one pilgrimage map is known to exist.
This work, painted on cloth and measuring 125 by 79
centimeters, is held by the Department of Oriental Anti-
quities, British Museum, London (registration number
1928.0707.1). Found in Kathmandu, the map is believed
1. Ti-se (Mount Kailas); 2. Mount Dorie Tijung; 3. other unspecified neighboring to date from the early nineteenth century. Although I
mountains; 4. pilgrim route around Ti-se; 5. peak signifying the Tsering-Changa group; know it only from a black-and-white photograph, I
6. Lake Manasarowar; 7. Langag, Lake of the Demons; 8 and 9. Tarchen-C'u and
Lha-C'u, tributaries to Langag; 10 to 14. five gompas (monasteries) of which four lie in would judge that it is the product of a Shaivite Hindu
the cardinal directions around Ti·se (10. Tint'ipu, north; 11. C'oku, west; 13. Zupral, artist, despite some apparent influences from Tibetan
east; 14. Gyantrag, south; and the fifth, 12. Silung); 15 to 17. Lakes Kurkyal C'ungo,
Ding Ts' o. and T' uki Zingbo; 18. lotus bud in Manasarowar; 19. traces of the struggle models, notably in its embossed border, similar to those
between Saint Mila Repa and the Bon-po; 20. Mila Repa's hand and footprints; 21. Mila found on many Buddhist thankas, and its composition,
Repa on Ti-se; 22. Mila Repa circumambulating Ti-se; 23. flagpole; 24. the Buddha's
footprint; 25. pilgrims of whom the one in the high hat is a Bon-po. in some ways reminiscent of the cosmography of Suk-
havati depicted in figure 15.21. The map is richly detailed
FIG,15.41. KEYTO FEATURES SHOWN ON FIGURE 15.40, and seems to show many Shaivite temples on either side
After Eva Stoll, "Ti-se, cler heilige Berg in Tibet," Geographica
Helvetica 21 (1966): 162-67, esp, fig. 1.
161. Only works a and b are illustrated by Stoll, "Ti-se" (note 160).
Work a also appears in Swami Pranavananda, Kailas-Manasr(War (Cal-
(Ma pham) and Langag (La-ng) to the south. Some idea cutta: S. P. League, 1949), fig. 93, with numbers keyed to an accom-
of the scale of the area shown may be obtained from the panying list of forty-four specific features it depicts, largely duplicating
knowledge that circumambulation of Manasarowar, the those of the bronze tablet in Zurich. Pranavananda provides even more
detailed explanations of these features (pp. to, 14, and 123) than does
larger lake, normally takes several days, the circuit
Stoll and appears to be the source Stoll largely relied on. Stoll's source
extending for seventy kilometers, or roughly one hundred for work b is Olschak and Thupten Wangyal, Mystic Art, 6 (note 39),
kilometers if one visits all eight of the surrounding mon- which presents the work at a larger scale. The map was discovered by
asteries that are also depicted on the tablet. A shorter the Swiss geologist Augusto Gansser en route to Mount Kailas, on the
circuit is that of the mountain itself, but that the more wall of a small monastery in Tinkar, southwest of the Tibetan town
of Taklakot and near the junction of the borders of India, Nepal, and
arduous trip is also enjoined on the devout is made evi-
Tibet. Both it and work c are illustrated in Blanche Christine Olschak,
dent from the clearly marked path on the tablet along Augusto Gansser, and Andreas Gruschke, Himalayas (New York: Facts
which two pilgrims and a monk are shown making their on File, 1987), 192 and 81 respectively. Work d is illustrated in Louis
clockwise circumambulation. Regrettably, the published P. Van der Wee, "A 'Cloister-City'-Tanka," Journal of the Indian
accounts of the tablet say nothing about the circumstan- Society of Oriental Art, n.s., 4 (1971-72): 108-20; relevant note on 113
and illustration in fig. 5. The thanka, in the Rijksmuseum voor Volk-
ces of its acquisition, or about its date or provenance.
enkunde, Leiden, is incompletely and somewhat misleadingly identified
What is reasonably certain, however, is that the work in the relevant museum note that Van der Wee cites. I have referred
conforms closely to published Tibetan pilgrimage guides here only to works I have seen photographs of; others, however, are
for the region, and that it would have served well in known to exist.
FIG. 15.42. TIBETAN MAP OF THE COMMERCIAL route depicted begins at the Potala, shown at the top of the
ROUTE FROM LHASA TO ASSAM. This seemingly naive map map, continues through Lhasa itself, and thence through Samye.
makes use of numerous conventional signs in Tibetan carto- The second half of the map (not shown) continues through
graphy, evident from the annotations along the margins. It is Tawang, now in India, and south to a point near the plains of
the northern half of the map taken from Hermann von Schlag- Assam. The scale is notably larger in the better-known northern
intweit-Sakonhinski, Adolphe von Schlagintweit, and Robert part of the map than in its southern portion.
von Schlagintweit, Results of a Scientific Mission to India and Size of this portion: ca. 30 x 20 cm. By permission of the
High Asia (1861). The map embodies far more specific detail Bodleian Library, Oxford (Maps 206a.2).
than the uninitiated non-Tibetan viewer might suppose. The
of the Trisuli River along the pilgrimage route to The route this map deals with runs generally north-
Gosainkund and beyond to Gosainthan, as well as lines south at right angles to the predominant grain of the
of pilgrims wending their difficult way toward those high terrain. Hence the mountains do not point away from
Himalayan destinations whose sanctity for Nepali Hindus the observer following a route in a river valley, as they
rivals that of the more renowned region around Kailas. conspicuously do so on several other maps we have exam-
Despite the recent inroads of modern technology and ined. Rather, they point away from the southerly locale
Western education in the Himalayan region, it appears of Narigun, from which the author's dominant view is
that pilgrimage maps embodying elements of traditional toward the north. As with most Tibetan maps, the scale
cartography are still being made. Snellgrove illustrates is not consistent from one part of the map to another;
one such map, taken from a modern printed Tibetan in this case the scale for the northern portion is greater
pilgrimage guide to the holy places in the region of Thak- than for the south. However, that nearly half of the ver-
kola (Thag)-now included in the Nepali region of Mus- tical dimension of the northern half is taken up by the
tang, just south of the Tibetan border-and he also Potala, the most prominent feature shown, and the city
provides a key to its contents, drawn on a modern geo- of Lhasa would in itself account for most of this differ-
graphic base. 162 The guide the map is taken from is one ence.
of many such works that have been produced in the areas
of Tibetan Buddhism over a period of centuries. A sys- PLANS OF TOWNS, MONASTERIES, AND
tematic search through such guidebooks may well OTHER SMALL LOCALITIES
uncover many more pilgrimage maps.
The final route map that I shall consider is among the Apart from cosmographic maps, the most common type
earliest works of Tibetan cartography to be brought to of cartographic artifacts from Greater Tibet are those
the attention of Western scholars. Drawn on a scroll for that relate to specific localities, especially major religious
Hermann von Schlagintweit-Sakunlunski in the monas- centers or particular monasteries or other religious edif-
tery town of Narigun in January 1856, it was published ices within those centers. Many of these depictions, as
in a black-and-white facsimile, at two-thirds the original we have seen, are appended to much larger, essentially
size, in the atlas volume that formed part of the massive hagiographical works, being designed to indicate the
report of the scientific expedition of the Bavarian Schlag- sanctified places associated with a particular religious fig-
intweit brothers to India and "High Asia."163 The map ure, whether mortal or divine. Here I shall consider only
illustrates a route followed by Bhotia traders between paintings in which a particular locality forms the domi-
Lhasa and Assam, via Samye and Tawang. Figure 15.42 nant focus of the composition. Only a small sample of
reproduces its northern (upper) half. A note on the map, such maps-chosen to demonstrate the diversity of their
printed in English, indicates that it was drawn by Davang modes of representation-will be individually discussed.
Dorje, a Bhotia chief, with the assistance of some lamas. Brief notes relating to the remainder of those known to
His signature appears in a panel near the bottom of the me-probably a reasonably representative sample of a
map, a rare instance of attributing authorship on a much larger corpus-are provided in appendix 15.3. 164
Tibetan map. All the original map text is in Tibetan, but Possibly the most intriguing, best-studied, and oldest
along both sides of the facsimile appear notes in English among locality maps is one focusing on Svayambhunath
keyed to its contents: thirty-two, along the left margin, stupa, on the outskirts of Kathmandu (fig. 15.43).165
relating to "inhabited places and stations," and twenty-
three on the right referring to mountains, rivers, and other 162. David L. Snellgrove, trans. and ed., "Places of Pilgrimage in
Thag (Thakkhola)," Kailash 7 (1979): 72-132; Tibetan text 133-70,
physical and biological features.
maps on 72 and 132. The date of the text the map relates to may be
Notwithstanding the circumstances attending its prep- 1607, 1667, or 1727, depending on the specific sixty-year Tibetan cal-
aration, the map is in an unadulterated Tibetan style, endric cycle the stated date refers to. The map itself, however, has
employing many common conventions. For example, clearly been made to conform broadly to recent geographic knowledge
map annotations relating to clouds indicate not only that of the region depicted.
163. Schlagintweit-Sakunlunski, Schlagintweit, and Schlagintweit,
those behind the Potala are the "usual ornaments for the
Results of a Scientific Mission, vol. 4, map 3 (note 10).
top of Buddhist drawings," but also that others "are used 164. Appendix 15.3 includes a few maps that I have seen personally
in connection with high peaks." Elsewhere, in relation and a much larger number that I know only through photographs, either
to the lone yak depicted not far from the lower-right published-with or without explanatory text-or sent to me by various
corner of the illustration, the note reads, "Symbol of sources. Obviously any attempt at analysis based solely on small-scale
photographs has serious limitations. The citations of published works
numerous herds of wild Yaks, in these regions very fre-
in the final column of the appendix are limited to those I have personally
quent." Thus, just as a single house, on this and other examined.
Tibetan maps, may signify a village, one yak may indicate 165. The first attempt at scholarly analysis of this map was that of
an entire herd or herds. Theodore Riccardi, Jr., "Some Preliminary Remarks on a Newari Paint-
FIG. 15.43. NEWARI MAP OF THE PRECINCTS OF SVAY- tain rim of the Vale, shown in a highly stylized manner in the
AMBHONATH STUPA AND THE VALE OF KATH- arch that largely envelops the stupa and also at the bottom of
MANDU. This unusual painted cloth map was prepared in con- the map, and the Bagmati River system. The placement of spe-
nection with the rededication of the ancient Svayambhiinath cific towns and religious sites within this system illustrates the
stupa in A.D. 1565. The work has two relatively discrete com- basic orderliness of the presentation. The abundant text on the
ponents: an upper portion, in which the stupa and the structures map, in both Newari and Sanskrit, enables the positive iden-
in its immediate vicinity are depicted at a very large scale against tification of most of the features shown.
a field of cinnabar, and a lower portion in which several dozen Size of the original: 101 X 85 em. Photograph by Mary Shepherd
places in the Vale of Kathmandu are shown within a field of Slusser. By permission of Princeton University Press.
muted gray. Prominent features of the map include the moun-
.. N
1. Vidyiisvari; 2. Sobhii Bhagavat'; 3. Pulchok; 4. Patan; 5. Yarigal (southern

Kathmandu); 6. Yal]1bu (northern Kathmandu); 7. Thal]1 Bah.; 8. Bodhniith; 9. Deopatan;
10. Kumbhesvara (?); 11. Theeva; 12. Sanagaon; 13. Harasiddhi; 14. Pharping (?); 15.
Thimi; 16. Lubhu; 17. Bhaktapur; 18. SOrya Viniiyaka; 19. an unnamed viMra.
FIG. 15.44. KEY TO PART OF FIGURE 15.43. The numbered

sites are those that bear identifiable labels on figure 15.43. With
two exceptions (10 and 19), these are named and similarly num-
bered on figure 15.45.
After Mary Shepherd Slusser, "The Cultural Aspects of Newar
Painting," in Heritage of the Kathmandu Valley: Proceedings o 1 2 3 4 miles
of an International Conference in Lubeck,june 1985, ed. Niels I '. Ii' II
Gutschow and Axel Michaels (Saint Augustin: VGH Wissen- o 2 4 6km
schaftsverlag, 1987), 13-27, esp. fig. 2.

FIG. 15.45. REFERENCE MAP FOR FIGURE 15.43. This is a
modern map of the Vale of Kathmandu. Key numbers on this
map are identical to those of figure 15.44. Sites 5 and 6 represent
Although the area of this map includes most of the Vale southern and northern Kathmandu respectively. Sites 10 and 19
of Kathmandu and is similar to that of the regional map on figure 15.44 cannot be located on a modern map.
illustrated in figure 15.35, I have chosen to consider it After Mary Shepherd Slusser, "The Cultural Aspects of Newar
here as a locality map because of the extraordinary Painting," in Heritage of the Kathmandu Valley: Proceedings
of an International Conference in Lubeck,june 1985, ed. Niels
emphasis given to a single dominant religious complex.
Gutschow and Axel Michaels (Saint Augustin: VGH Wissen-
The date of the map is Newar Samvat 685 (A.D. 1565), schaftsverlag, 1987), 13-27, esp. fig. 4.
when it was, according to its inscription, commissioned
in a monastery in the Nepali town of Patan in connection
with the reconsecration of the nearby renowned ancient
stupa that is its central object. Long forgotten and seri- runs the Bagmati River, eXltmg the Vale to the left
ously deteriorated over the following four centuries, the (south). Over much of the map, captions in Newari, vary-
map resurfaced in 1967 and passed into the hands of a ing in legibility, identify many of the places depicted.
private collector, who saw to its restoration. Between the places so marked, we can identify the major
The painting is arranged in two sections, the upper components of the drainage system by interpolation. Fig-
one devoted to the glory of Svayambhii, the lower to ure 15.44 shows the major features of the map apart from
the towns and sacred sites-Buddhist and Hindu-that the stupa itself and the buildings in its immediate vicinity;
bathe in its refulgence. In both sections the arrange- these may then be compared with the corresponding fea-
ment is by no means haphazard but represents on the tures on a modern map as presented in figure 15.45. 167
one hand an exact site plan of the Svayambhii com- Also from Patan is another map relating to the city
pound and on the other a correct map of the Kath-
mandu Valley. Indeed, the painting ... serves as a
ing of Svayambhilnath," Journal of the American Oriental Society 93
pilgrim's guide to the sacred geography of the Kath-
(1973): 335-40. A more complete study is that of Slusser, "Cultural
mandu Valley. In short, it is a map.166
Aspects," and Slusser, Nepal Mandala, vol. 1,299, and vol. 2, fig. 495
Arching above the stupa and to both sides is a highly (both in note 13); and Pratapaditya Pal, The Arts of Nepal, pt. 2, Paint-
ing (Leiden: E. J. Brill, 1978), 80-81, 152, and fig. 108.
stylized representation of the mountain wall that forms
166. Slusser, "Cultural Aspects," 20 (note 13).
the western, northern, and southern rims of the Vale, a 167. Slusser, "Cultural Aspects," figs. 7 and 8 (note 13); Slusser's fig.
pattern one sees also at the bottom of the map, which 6 is a modem plan of the Svayambhilnath compound to which one can
represents the Vale's eastern margin. Between the two also relate the features in the appropriate portion of the original map.
FIG. 15.46. NEPALI SCROLL MAP OF THE VALE OF tion to be readily identifiable. Despite the restrictive format of
KATHMANDU. This is a small portion of an enormous cloth the painting, there is an unmistakable topographical logic to
banner painting that is periodically displayed in the courtyard the organization of the features, including various meandering
of the Kwa Bahal monastery in the town of Patan in the Vale streams, rolling hills, and-marking the horizon-cloud-capped
of Kathmandu. It depicts numerous sacred places and several mountains. In many locales flora, fauna, and people enliven the
towns of the region, but it has not been analyzed sufficiently work.
to determine the full extent of the area it relates to. Many of Size of the entire original: ca. 90 x 1,800 cm. Photograph cour-
the features depicted have been labeled, presumably in Newari, tesy of Mary Shepherd Slusser.
while other features are distinctive enough in their representa-
and neighboring places in the Vale of Kathmandu. This from a particular point, or does some other principle
work, believed to date from the mid-nineteenth century, guide its composition? These questions also apply to the
appears on an enormous painted cloth scroll that is hung other scroll map from Patan noted in appendix 15.3.
from time to time in the courtyard of the Kwa Bahal In striking contrast to the Nepali map focusing on Svay-
monastery. A small part of the map is illustrated in figure ambhiinath are a number of Tibetan representatations of
15.46. Neither Slusser nor Pal, who discuss it in several the renowned Tibetan monastery of Samye, founded in
works, provides an estimate of its overall dimensions, but A.D. 775 and restored several times since. Of these, the
based on color slides and black-and-white prints of the best known is undoubtedly the one reproduced in figure
entire painting, I estimate its length as approximately 15.47, which has been dated as early as the sixteenth
eighteen meters and its height as about ninety centime- century but more likely dates from the seventeenth or
ters. 168 On this map eighteenth. Samye is of particular interest because it is
all the principal sacred places and a number of towns
the first and oldest of all Tibetan monasteries and was
are graphically illustrated, some further identified with also consciously designed, largely following the model
written labels. They occupy a landscape of cloud- of the monastery of Odantapura, in what is now the
capped mountains, rolling hills, and meandering Indian state of Bihar, to symbolize the Buddhist universe.
streams enlivened with diverse flora, fauna, and peo- The central square temple represents Mount Mem.
ple, the latter largely occupied with religious affairs. . .. Its three stories are each in a different architectural
Given the limitations of the long narrow format, the
shrines are so well organized topographically that,
168. This work is illustrated in several works by Slusser: Nepal Man-
labeled or not, most can be easily identified. 169 dala, vol. 2, fig. 98; "Pilgrim's Guide from Nepal," 30-31; and "Cultural
Regrettably, this is the extent of Slusser's description. Aspects," 26 (all in note 13); and also by Pal, Arts of Nepal, fig. 164,
plus note on 132 (note 165). I am indebted to Todd Lewis for his
One might also wish to know how far beyond the urban sending me sets of black-and-white photographs and color slides cov-
core of the Vale the map extends and also the point of ering the entire painting.
view of the artist. Is the map a panoramic view, as if seen 169. Slusser, "Pilgrim's Guide from Nepal," 30 (note 13).
FIG. 15.47. TIBETAN MAP OF SAMYE MONASTERY. This a model of the universe as conceived by Tibetan Buddhists.
exquisitely painted thanka faithfully depicts the eclectic nature Size of the original: 53 x 38 em. By permission of the Newark
of its central temple, which is compounded of Tibetan, Indian, Museum, Newark, N.]., Shelton Tibetan Collection (ace. no.
and Chinese styles, one for each of its three stories, and makes 20.271).
evident the attempt to re-create, within the complex as a whole,
FIG. 15.48. MAP OF DREPUNG MONASTERY. This stylis- was presumably affiliated), but also numerous features of the
tical1y distinctive mid-eighteenth century thanka of the place natural landscape. The seeming disregard for perspective-note
where the Dalai Lama resided before the construction of the in particular the vertical bar representing a street in the lower
Potala is believed to have been made by a pilgrim from the left portion of the city-is deceptive; general1y speaking, the
Amdo region of northeastern Tibet. Though not evident at the higher the feature within the image, the greater its distance from
scale of the photograph, the painting abounds in descriptive the observer. An inscription on the map indicates that its sanctity
text, which has enabled it to be studied more intensively, per- is such that individuals apprehending it, by sight or touch, wil1
haps, than any other Tibetan locality map. The text identifies be rewarded with a speedy attainment of liberation (nirvana).
not only scores of individualistical1y portrayed buildings within Size of the original: 115 x 68 em. By permission of the Musees
the city, most of them serving religious functions (including a Royaux d'Art et d'Histoire, Brussels (Col1ection Leon Verbert
number associated with a Tantric col1ege with which the pilgrim 350).
FIG. 15.49. MAP DEPICTING A PROCESSION OF MONKS the Potala are intended to be villages in the vicinity of Lhasa
ENCIRCLING THE POTALA IN LHASA. This large map, of or sacred centers in their own right. Possible Sinic elements in
uncertain date and provenance, appears to combine Tibetan and the map include the rings of trees that surround these peripheral
Chinese stylistic elements and may well have been produced by areas and the wave pattern in the stream running along its lower
a follower of Tibetan Buddhism in China. It was acquired in edge.
Beijing in 1931. The work has yet to receive careful study, and Size of the original: unknown. ©The National Museum of Eth-
it not certain whether the small areas of settlement surrounding nography, Stockholm (H.2882). Photograph by Bo Gabrielsson.
style: Tibetan, Indian, and Chinese. Four temples .to ings of Samye of which I have seen photographs (all noted
the north, south, east and west represent the four in appendix 15.3), though none of these is quite as
worlds [i.e., dvrpas or island continents]; smaller tem- detailed as the Newark example.
ples represent the [tributary] islands.... Two further About eight kilometers west of Lhasa and some seventy
temples symbolize the sun and the moon. 170 kilometers northwest of Samye lies Drepung ('Bras-
Most of the elements just noted can be readily dis- spung), another important monastery town (formerly
cerned on the painting. For example, the three cusp- inhabited by 7,000 to 10,000 monks); a painted map is
shaped structures in the foreground presumably represent illustrated in figure 15.48. The contrast between the maps
the continent of Jambiidvlpa and its two island tributar- of Drepung and Samye is so striking that one w,ould
ies, while the temples in the upper-right and upper-left hardly suppose that both fell within the same broad
corners probably signify the sun and moon. Additional Tibetan Buddhist tradition. Although it is not apparent
noteworthy features are the two zigzag walls enclosing at the scale of our photograph, the Drepung map is
the compound, said to represent the ring of mountains
surrounding the terrestrial plane of the universe. I71 The 170. Catalogue of the Tibetan Collection and Other Lamaist Mate-
fidelity of the map is noted by Reynolds: "Except for the rial in the Newark Museum,S vols. (Newark, N.].: Newark Museum,
1951-71),3:65.
fanciful hill and stream landscape, this painting faithfully
171. Trungpa, Visual Dharma, 130 (note 105).
represents the main elements of Sam-ye as it looks in 172. Valrae Reynolds, Tibet: A Lost World, exhibition catalog for
photographs taken in the mid-20th century."172 A similar the Newark Museum Collection of Tibetan Art and Ethnography (New
concern for fidelity seems to infuse the three other paint- York: American Federation of Arts, 1978), 118.
replete with text identifying more than one hundred is contrary to what one would actually find in the pre-
buildings, natural features, icons, and other elements. vailingly bleak environment of the Tibetan Plateau and
Hence the map has lent itself to what is perhaps the most would support the hypothesis of a non-Tibetan origin.
thorough scholarly analysis ever made of any Tibetan What appears to be a more Chinese than Tibetan feature
locality map.173 Based on both inscriptional and stylistic of the map is the wave pattern within the stream that
considerations, it has been suggested that the map was runs along the bottom edge.
made in the mid-eighteenth century, probably by a pil- Shigatse, the locale of Tashilunpo, the monastic seat
grim from the region of Amdo in northeastern Tibet. Of of the Panchen Lama and the second largest city of Tibet,
particular concern to the artist were the institutions asso- is the subject of the final illustration, figure 15.50. This
ciated with a Tantric college in Drepung, and notes on seemingly unfinished and undated painting on canvas is
them are especially numerous. A particularly important on display at the American Museum of Natural History,
legend, a "formula of consecration" inscribed just below New York. It is of interest for a number of reasons. First,
the town's lower wall, reads: "Through the merit of [i.e., it effectively combines a remarkable diversity of per-
conveyed by] the perfect realization of this image of the spectives: planimetric, oblique, and frontal. Second, the
glorious monastery of Drepung, well known in the three features portrayed are rendered with great fidelity to what
worlds, one will obtain speedy liberation by seeing it, by actually existed in Shigatse at the time the painting
understanding it through touching it, and by remember- appears to relate to, roughly the third quarter of the
ing it."174 The composition of the painting has a decep- nineteenth century. Third, it is one of at least two maps
tively naive appearance. Although there appears to be a that appear to have been copied from the whole or a
general lack of concern for perspective, the distance of part of some unknown prototype. The upper-left portion
buildings from the hypothetical observer increases with of the painting, depicting Tashilunpo, bears a resem-
their apparent altitude on the prominence that dominates blance to another work (item kk in appendix 15.3) in a
the composition, which in fact may not be particularly private collection, which is believed to be based on the
prominent. The buildings are drawn in considerable same prototype. 176 Finally, there exists for this map a
detail, mostly painted white, while temples, abbots' res- very detailed key (though not as complete as for the map
idences, and other religious edifices are surmounted by of Drepung), made to accompany a faithful copy of it
an ocher band. reproduced in a work by Sarat Chandra Das, a celebrated
The view of Lhasa illustrated in figure 15.49 resembles Bengali student of the religion, geography, ethnography,
the maps of Samye and Drepung in its strong focus on a and history of Tibet, who between 1879 and 1883 made
major monastic complex, and it also resembles that of several extensive journeys to that country and to China
Svayambhiinath in placing its dominant object within a proper as an agent of the British Indian government. 177
much more inclusive spatial and religious context. This
painting has not previously been published, so far as I
am aware, and I know it only from several photographs 173. The most extensive discussion, together with a detailed map
and the tantalizingly brief descriptive note provided to key, is that of Philippe van Heurck, "Description de la than-ka repre-
me by the museum, which reads, "Big painting depicting sentant Ie monastere de Drepung," Bulletin des Musees Royaux d'Art
et d'Histoire 57, no. 2 (1986): 5-29. Briefer notes with illustrations
a procession encircling the Potala in Lhasa-acquired in appear in Dieux et demons de ['Himalaya, 58, 237-38 (note 33); Lauf,
Peking 1931."175 Neither the provenance nor the date of Tibetan Sacred Art, 12, 16 (color photograph) (note 54); and Vergara
the painting is known. The style appears more or less and Beguin, Dimore umane, santuari divini, 80 (note 27), on which
Tibetan, but the work could be by other followers of striking similarities in the compositions of the Drepung and Samye paint-
Tibetan Buddhism, possibly resident in China's capital. ings are demonstrated, notwithstanding their huge stylistic differences.
174. Heurck, "Description de la than-ka," 22 (note 173).
The map does not appear to have any inscriptions. A 175. Letter from Claes Hallgren, 23 October 1989.
characteristic it shares with several other paintings I have 176. The privately owned map of Tashilunpo is illustrated and
noted (see especially plate 34) is that it shows Lhasa sur- described in Van der Wee, " 'Cloister-City'-Tanka," 109, 114-20, and
rounded by a cluster of smaller centers, none of which figs. 1 and 7 (note 161). It appears at first glance to be only a fragment
I can identify positively. In this case, however, the striking of a larger work covering the same area as figure 15.50, but Van der
Wee observed, on removing the map from its frame, that the right
degree to which those lesser centers are subordinate to border, which would have been painted were that the case, was a blank
the Potala justifies classifying this work as essentially a area in which the artist tried his colors before working on the rest of
locality map. It is not clear whether the map's peripheral the composition. Thus that map could not have been the one from
settlements are meant to be distinct from Lhasa and at which figure 15.50, or any other showing the whole of Shigatse, was
a distance from it or merely quarters of what one might copied.
177. Sarat Chandra Das, Journey to Lhasa and Central Tibet (1902;
consider "greater Lhasa." On other maps the lesser cen- reprinted New Delhi: Mafijusri Publishing House, 1970), Shigatse map
ters are generally shown surrounded by walls, but here facing 45. The caption for figure 15.50 is based primarily on Das's keyed
they seem to be ringed by circles of broadleaf trees. This notations fo~ many of the major features depicted. Das's work includes,
FIG. 15.50. TIBETAN PAINTING OF THE TOWN OF SHI- at the lower right; a "camp for exercise," the nearly square
GATSE, INCLUDING THE TASHILUNPO MONASTIC feature to the left of the nobleman's compound; another low-
COMPLEX. This seemingly incomplete yet remarkably detailed lying dzong, marked by a wall with battlements, a bit farther
painting on canvas appears to have been copied from a larger to the left; the temple of the war god, Gesar, directly below
original. Although this map is undated, the work it is based on the dzong; a market area, in the empty space above the dzong;
appears to have been painted in the latter half of the nineteenth an important stupa (not in the usual form), between the market
century. Where that original may be is not known; but it is area and Shigatse dzong; and the summer palace of the Grand
possible that a painted cloth map, restricted to the area of (Panchen?) Lama, lower left. Among the more conspicuous fea-
Tashilunpo (shown here at the left), now in a private collection, tures in Tashilunpo itself are the mausoleums of the first four
may be a fragment of that work. The map is marked by a Panchen Lamas, marked by their Chinese-style roofs and shown
remarkably wide range of perspectives, generally looking north: in a straight line upper left in Tashilunpo (but not the mau-
planimetric, oblique (from various angles and occasionally soleum of the fifth Panchen Lama, who died in 1882); a tall
divergent), and frontal. The artist seemingly chose whichever stone structure (said to be more than thirty-five meters high)
type of view would best enable him to convey a visual impres- above them and to the right, from which great banners were
sion of particular types of features. Despite this inconsistency, hung at certain festivals every year; and "the Park of Happi-
or perhaps because of it, structures are rendered with great ness," the walled garden attached to the monastery at the lower
fidelity, and dozens of them have been positively identified. right, where the Grand Lama's parents reside.
Among the main components of the map, in addition to Tash- Size of the original: ca. 66 x 94 cm. Courtesy of the Department
ilunpo, are Shigatse dzong (fort), in the top center; the town of Library Services, American Museum of Natural History, New
of Shigatse proper, in a descending arc to the right of the dzong; York (cat. no. 70.2-187; neg. no. 335058).
the compound of an important nobleman, shown by the oblong
No record exists of how the museum obtained the in addition to the copy of the map of Shigatse, two illustrations that
Shigatse map, but it is not unlikely that it was a bequest appear to be faithful copies of Tibetan locality maps of the Potala and
the Jo-khang palaces and provides a few notes, in English, on their
from Das's friend and fellow student of Tibet, W. W. major features (facing 154 and 160). I have not been able to determine
Rockhill (1854-1914), an American diplomat, who may the originals ftom which they were taken and have not noted the works
have received it from Das himself. In his 1899 introduc- in appendix 15.3.
tion to Das's account of his Tibetan travels, Rockhill to individual maps and provided valuable insights into
acknowledges his "lasting debt of gratitude for the val- the mental processes underlying them, none has tried to
uable information which he gave me while in Peking." assess the entire corpus. Only Aziz and Slusser have
Rockhill, who edited Das's travel narrative, worked for sought to make generalizations based on the study of a
the Smithsonian Institution for several years; that he may number of maps, and in both cases that number was
also have had a connection with the American Museum smal1. 181 We still await a comprehensive study by a
of Natural History in New York City is suggested by the scholar with the requisite linguistic skills, cultural know-
fact that he was residing in Block Island, New York, when ledge, and sensitivity. What is provided in this chapter
he wrote his introduction. 178 As to the date of the should be considered no more than an overview of the
unknown original, one may safely infer that it was painted many types of maps that call for further investigation.
between 1854 and 1882, the years when the fourth and The most important point to be made about indige-
fifth Panchen Lamas died, since the painting prominently nous cartography from Greater Tibet is that its impetus
shows, in a single line, the mausoleums of the first four was overwhelmingly religious. While exceedingly inform-
such dignitaries but fails to show the fifth. 179 This would ative maps (e.g., figs. 15.32 and 15.38) for essentially sec-
accord with the period of Das's visits to Shigatse. ular purposes were certainly made (though few
Though not illustrated here, one additional locality premodern examples survive), the vast majority of the
map worthy of brief consideration is a mural painting of maps that have come to light served some religious end.
Chang'an, the capital of China under the Tang dynasty Among these were cosmographies-often including a
(618-907). It is one of a large number of paintings of temporal dimension-of widely varying complexity,
localities occupying the corridors of the Potala in Lhasa. intended to aid in religious education and meditation;
That undated map is illustrated in a Chinese work on maps forming components of hagiographical and bio-
Buddhist art from Tibet, but unfortunately it appears graphical compositions, used to promote the veneration
without any explanatory text apart from the caption, of important religious figures; maps painted as adjuncts
"This is an interesting depiction of the famous grid layout to illuminated religious texts, such as the Jataka stories;
of the Tang capital."180 The painting's combination of tirtha mahatmyas, glorifying places of pilgrimage; other
an overall planimetric frame for the city and the set of types of pilgrimage guides; and cartographic records of
vertical elevations for the buildings within its sixteen pilgrimages successfully undertaken (e.g., fig. 15.35).
compartments and horizontal perspectives to depict the Additionally, though they are only cursorily considered
landscape above and below provides a starker duality of in this work, there are innumerable religiously sanctioned
perspectives than in any other of the Tibetan maps I have astrological charts and other graphic aids (e.g., fig. 15.23)
considered. Obviously the painting is no more than an that still guide Tibetans in their day-to-day lives and are
idealized and relatively abstract view of what Chang'an regarded as indispensable resources in making major deci-
was supposed to look like, and it lacks the sense of fidel- SIons.
ity, or at least personal familiarity with the subject, that Although for organizational purposes a basic distinc-
characterizes most other Tibetan locality maps. It war- tion has been made in this history between cosmographic
rants notice, however, in that its manner of execution and geographic mapping, one must note that such a dis-
reinforces the proposition that Tibetan painters were not tinction might appear arbitrary to most people from the
locked into a uniform artistic style as they prepared the region that concerns us. The road to Shambhala, for
maps I have sought to describe in this chapter. example, wherever that land may be, certainly began, in
the minds of the faithful, in locales that some of them
CONCLUSIONS experienced at first hand, as is evident from the scroll
maps showing the route that were once carried about by
For Greater Tibet, no less than for India and Southeast itinerant Tibetan storytellers. Conversely, pilgrimage
Asia, the standard histories of cartography to this point maps, such as those of the sacred region around Mount
lead one to suppose there was virtually no indigenous Kailas, included numerous wholly mythological places.
tradition of cartography. As this history has demon- The scale of maps from Greater Tibet varies enor-
strated, such a conclusion is clearly untenable for all three mously. Leaving aside cosmographies, the range is from
regions. The cartographic legacy of Greater Tibet in par-
ticular displays a richness, variety, and vigor that are 178. Rockhill's introduction to Das, journey to Lhasa, xv (note 177).
remarkable given the region's meager population and its 179. Van der Wee, " 'Cloister-City'-Tanka," 115-16 (note 161).
relative isolation, both physical and political, from the 180. Liu, Buddhist Art, 187 (note 36).
181. Aziz, "Tibetan Manuscript Maps," Tibetan Frontier Families,
rest of the world. and "Maps and the Mind" (all in note 12); Slusser, "Serpents, Sages,
Although several scholars, mainly art historians and and Sorcerors," Nepal Mandala, "Pilgrim's Guide from Nepal," and
anthropologists, have written exemplary articles relating "Cultural Aspects" (all in note 13).
what were probably intended to be world maps (e.g., figs. also trees, houses, and other features. A more common
15.27 and 15.30) to plans of an individual edifice. Within practice, however, is to adopt an oblique perspective, as
this broad spectrum, maps of individual localities appear if from a perch in space. Even more common is the use
to be most common (see appendix 15.3). Furthermore, of multiple perspectives (e.g., fig. 15.50), showing some
some maps of considerably larger regions, especially of features, such as buildings, mountains, and trees, from a
the area around Lhasa, are drawn with so much compres- horizontal perspective (in frontal elevation); others from
sion of sacrally "empty" space between major religious one or more oblique perspectives (oblique frontal also
centers that the uninitiated observer gets the impression being very common for buildings); and still others, such
that a single, though rather expansive, locality is being as lakes and large compounds, from a vertical (plani-
depicted. metric) perspective. The use of a divergent perspective
Within individual maps uniformity of scale seldom (e.g., fig. 15.10), the very opposite of that conventionally
appears to be a desideratum. The use of varying scales used in Western drawing, is also common. On maps
for different map elements seems to be a general feature showing a number of towns, the perspective chosen for
of Tibetan cartography. Thus it is not uncommon to each would presumably be the one that normally obtains
show human beings and occasionally animals at a much for travelers approaching it along the most traveled
larger scale than the buildings they move among. What routes. Thus Lhasa would generally be shown as if one
rules govern such decisions by mapmakers is not entirely were looking toward the north, while on the same set
clear, but there are several probable reasons for depicting of maps another town, say Shigatse, might be seen as if
certain map elements at an exceptionally large scale. First, one were looking at it toward the south.
the need to draw features large enough so the artist can The foregoing generalizations on perspective apply as
show all their essential attributes (e.g., within a city, the much to cosmographic as to geographic maps. In the
main gates, towers, important religious edifices, and so former, when-as is often the case-the emphasis is on
forth or, within a building, the number of stories and the the vertical dimension of the universe, the obvious choice
general nature of the construction). Second, the desire to is to employ a frontal perspective such that the dominant
underscore the religious or political importance of par- orientation of the map is at a right angle to the horizontal
ticular places. Third, the need to make visible small but plane, which is generally assumed to be the norm in mod-
significant features that might otherwise be overlooked ern cartography. Similarly, in showing the cities or celes-
(frequently to aid a teacher or storyteller using a map in tial palaces in which specific divinities are believed to
a didactic way). Finally, the need to highlight the exist- reside, there is usually one widely accepted correct view
ence of a common, but physically small, element in a based on specific religious texts or precedents drawn from
particular locale by depicting only one or a few such earlier cosmographies. But in composite depictions (e.g.,
elements at a large scale, assuming that the map reader fig. 15.15), multiple perspectives are also common.
will recognize it/them as signifying a general class. Con- If there is any general principle at work here, it is that
versely, scale compression might have been regarded as cartographers consciously choose the perspective or com-
appropriate when none of the foregoing considerations bination of points of view that will best allow commu-
applied. nication with their intended audience. They feel in no
There appears to be no general rule with respect to way hamstrung by a scientistic need for consistency in
the "correct" orientation of maps in Greater Tibet. It is how they accomplish that purpose, just as they feel no
not especially significant that a number of maps from obligation to maintain consistency of scale. Moreover,
Nepal appear to be oriented toward the north. Since it mapmakers seem to give their audience credit for being
is a common convention for mapmakers within the able to make the mental adjustments necessary to derive
region to show a line of mountain summits at the top of the intended meaning from the map, even when the task
a map, the choice of the Himalayas for the purpose in is relatively difficult-for example, in seeing the "field of
Nepal appears to be the most appropriate; on the other assembly" (fig. 15.19) as constituting, in essence, two tan-
hand, the frequent use of other real or aesthetically con- gent spheres rather than a mere two-dimensional repre-
trived crests in Tibet proper, without particular regard to sentation of a tree.
their direction vis-a-vis the map viewer, suggests that the The rules that govern the composition of maps in
Nepali use of north is merely fortuitous. Greater Tibet appear to be far from uniform, and it is
It is common for mapmakers of Greater Tibet to orient not at all clear, especially outside the realm of cosmo-
features so that they point away from the map reader. graphic maps, in what ways and to what extent general
This is particularly true of route maps, especially when principles apply. For cosmographies, of course, especially
the routes depicted lie within long river valleys. In such for mandalas (see figs. 15.3 and 15.8), rules tend to be
cases mountain peaks typically point away from the valley relatively rigid, and learning them is an important aspect
bottom, in opposite directions on its two flanks, as might of the training of many monks. Also, in Tibetan painting
in general, there are well-developed canons of compo- square) to designate that place. Tibetan Buddhists, Bon-
sition and principles of aesthetics that guide the artist. In pos, and Nepali Hindus are likely to have no difficulty
geographic maps, on the other hand, individual carto- in making the necessary associations, and what laymen
graphers appear to have enjoyed considerable personal might fail to see, their preceptors will make clear to them;
discretion, judging from the idiosyncratic nature of much but followers of alien faiths will encounter much diffi-
of the available corpus. This question calls for further culty in interpreting, or even recognizing, the signs that
study. It is certainly reasonable to assume that part of many maps employ. Even more difficult for the unini-
the difficulty in interpreting the Zhang-zhung world map tiated will be the interpretation of the many religious
(fig. 15.27) derives from the fact that its author felt con- icons that are metaphoric or allegorical-for example, the
strained by the necessity of placing all the places shown narrow, winding road, suggestive of a navel cord, that
in such a way as to constitute a mandala. But what of leads across the black sea of barda (an intermediate state
more modern maps? It would be interesting, for example, of existence that prefigures one's next rebirth) to Suk-
to compare all the regional maps of Lhasa and its environs havatl, the "happy land" of the west (shown in fig. 15.21).
(i.e., plate 34 and maps analogous to it) to determine how To this point we have been considering what is shown
far the selection, placement, and directional relationships on maps from Greater Tibet and how. But it is also nec-
of and among the various places depicted were kept more essary to ask what is omitted. Silences are often pregnant
or less uniform and were guided by aesthetic, hierarchic, with meaning, yet easy to overlook. One wonders, for
and other nongeographic considerations. Unfortunately, example, if Tibetan Buddhist maps made a point of omit-
such an exercise is virtually impossible to undertake using ting holy places sacred only to Bon-pos and vice versa.
only small photographs and would require firsthand study Similarly, within the former tradition, did maps by fol-
in the various museums that hold these works. Further, lowers of the dominant Gelukpa (Yellow Hat) sect fail
on the same set of maps, as well as on many others, one to show establishments associated primarily with other
might wish to ascertain whether the mountains and rivers sects? I have noted that on primarily secular maps, even
that set off one part of the map from another had any those believed to have been drawn primarily for intelli-
function other than to define sections of the map for the gence purposes at the behest of the British, places of
reader to scan in turn. That is, do those features on the religious importance figured prominently. But was the
map even exist in nature, rather than being mere aesthetic opposite true? What place, if any, did sites with little
or didactic devices used by the cartographer? And if they religious importance find on explicitly religious maps?
do exist, how much liberty may the cartographer take in The roots of cartography in Greater Tibet extend far
literally bending them to a particular purpose, as back in time and probably first took hold outside the
obviously happened in the depiction of rivers on the map region itself. Among the earliest surviving objects from
focused on Svayambhiinath (e.g., fig. 15.43)? Likewise, in Tibet that we may designate as maps are works such as
the case of purely cosmographic maps, despite the exist- the fragment, possibly of the ninth century, of a depiction
ence of widely shared views, one should not rule out the of the Dunhuang region of Gansu (fig. 15.7) and two
role of an individual's faith in determining the image, as detailed models of the Mahabodhi Temple thought to
evinced by the passage describing the diversity of ways have been carried away from Bodh Gaya early in the
Mount Meru is depicted (p. 624). thirteenth century (noted on p. 612). Even earlier, and
That there are widely used graphic conventions in certainly not later than the tenth century, some sort of
Tibetan mapping is beyond question. Many of these con- model must have been taken to Tibet from India to
ventions may be seen in figure 15.9. The signs used there enable the construction, dated 971, of Samye monastery
apply not only to geographic maps but also to many in the form of the one at Odantapura in what is now
cosmographies and are commonly employed in works of Bihar. Also borrowed from India were cosmographic
art that have no specific cartographic quality. Hence their schemata such as the bhavacakra, or wheel of life, which
meanings are, on the whole, self-evident and easy for would have been transmitted at a still earlier date. The
laymen and even non-Tibetans to comprehend. Other diffusion of cosmological ideas back and forth between
conventions relate to color; for example, red is used to Tibet and China also appears to be indisputable, and one
indicate religious buildings and white to show ordinary may reasonably suppose that with them went cosmo-
residences or to identify specific cosmic elements. A cat- graphic artifacts. Turning again toward the west, if we
egory of sign that calls for special mention is the religious accept the views of Gumilev and Kuznetsov relative to
icon. Such signs are of particular importance in cosmo- the Zhang-zhung world map, we would have to believe
graphic maps but are used in others as well. The icon (be that substantial geographic knowledge from Persia and
it a particular deity, a bejeweled palace, or a particular lands even farther west made its way to Tibet long before
type of tree) is often taken to represent a specific place the establishment of Buddhism in that region and was
and may be used with no other sign (such as a circle or incorporated in the earliest of all Tibetan maps. Other
seeming relationships between Tibetan and Western map- these styles were different in respect to cartography than
ping relate to cosmography. How, for example, do we in respect to painting in general needs to be investigated.
explain the remarkable similarity between the basic To what extent, one should ask, were they a product of
schema of the tathagatamarj4ala (fig. 15.13) and that of competition among various faiths (Hindu vs. Buddhist in
many Western cosmographic schemata? At a much more Nepal, Buddhist vs. Bon in Tibet) or of religious orders
recent date, what accounts for the similarity in the style within the community of monks? What did they owe to
of the Nepali map of western Asia (fig. 15.32), obviously the personal influence of particularly charismatic lamas
the work of a Hindu, and various regional maps showing or of scholarly pilgrims? What role was played by military
the Buddhist holy cities centered on Lhasa (of which plate conquest, from which not even Tibet was immune? How
34 is an excellent example)? Can we establish a connec- was mapping affected by the introduction of paper? What
tion, as was suggested, between the kindred Tibetan other factors were at work?
beliefs in a blessed land to the west, SukhavatI, and in The areas of provenance of known Tibetan maps sug-
the virtuous warrior-king, the protagonist of the Gesar gest that only a few centers account for a very large
epic, who comes out of Shambhala to rid the world of proportion of the total corpus: central Tibet (especially
evil forces (p. 637) and the notion that this Nepali map Lhasa), the Vale of Kathmandu, and arguably Ladakh. It
may have been commissioned, perhaps during the time is not clear, however, whether other centers also existed,
of the Indian Mutiny of 1857-59, with a view to forging since the interests of the outside world, especially of the
a grand alliance to drive the British out of Asia? British, focused on only a handful of places in Greater
Although the specific events, routes of transmission, Tibet. In particular, the portions of the Tibetan cultural
and particular ideas and artifacts transmitted in all direc- sphere within the Chinese provinces of Qinghai and
tions to and from Greater Tibet are at present only dimly Sichuan are poorly represented among the artifacts I have
discernible, it seems clear that the cartography of Greater considered, even though the number of ethnic Tibetans
Tibet, especially in its formative phase, owes much to living in those areas is considerably greater than in Tibet
foreign cultural influences. As historical research proper (the Autonomous Region of Tibet). It is therefore
advances, it should shed fresh light on the relevant dif- doubtful that my sample of artifacts is representative.
fusion processes. This is clearly a topic on which new With respect to the few known centers of mapmaking,
thought and research are needed. In this regard the intri- one is struck by, and must try to explain, the profound
guing, though partially tendentious, arguments of Gum- diversity among surviving maps. This is especially note-
ilev and Kuznetsov and of Teramoto (as put forward by worthy in the case of the Vale of Kathmandu, long the
Nakamura) with respect to the two supposed Tibetan hearth of the particularly vibrant Newari culture. Com-
world maps (pp. 639-43) call for reexamination and inde- pare, for example, the works illustrated in figures 15.35,
pendent confirmation or refutation. 15.39, 15.43, and 15.46, to which might be added other
Beyond Greater Tibet, particularly in Mongolia, there strikingly different maps that I have not illustrated. Of
are other areas where virtually the same form of Bud- course the works in question were not contemporaneous;
dhism is practiced and for which cosmographic and other but it is doubtful if that factor alone explains the differ-
cartographic artifacts are known to exist. I have not had ences.
an opportunity to study those works in detail. Many of Finally, one cannot ignore the possible influence of the
the topographic maps in Mongolian script have been physical environment on the development of carto-
studies by Walther Heissig, especially for their toponymy, graphic sensibility and on the propensity to use and
but the corpus has not been analyzed from the viewpoint understand maps. Over most of Greater Tibet, there exist
of the history of cartography. A brief introduction, with high vantage points from which largely barren expanses
basic bibliographic references and representative illustra- of land stretch before the observer, who would see them
tions, is found in appendix 15.4 and in figures 15.51 and through the clear mountain air as if they were living maps.
15.52. Can viewing such sweeping vistas and, in the case of
Within Greater Tibet itself there was a progressive dif- habitual travelers, comparing one such view with many
ferentiation of ideas relative to the style and content of others possibly be without effect in the development of
cosmographic and geographic maps. Again, research is mapping skills? Might not habituation to such views
needed to establish the periods, routes, and nature of largely account for the popularity of high oblique per-
transmissions and to discover when and where significant spectives in Tibetan mapping? Further cognitive studies
endogenous developments occurred. It is probably too such as those carried out by Fisher and others of the
early to state whether identifiable "schools" of carto- mapmaking skills of Nepali schoolchildren (pp. 618-19),
graphy arose, but distinctive regional styles, identifiable and analogous studies of Tibetans and control groups
by art historians (e.g., that of the map of Drepung, illus- including adult populations, appear warranted.
trated in fig. 15.48), certainly came into being. Whether To conclude, I quote an observation by Slusser that,
though written about Nepali maps, has much more gen- is a venerable science."182 In my judgment that suggestion
eral applicability within the region that concerns us here: is likely to prove valid for most, if not all, of Greater
"We know little about the Newar penchant for map mak- Tibet. Surely the time has come for historians of carto-
ing, perhaps because we have not thought about it." After graphy to take a fresh and much more penetrating look
noting some of the known, though admittedly remote, at the remarkable evolution of mapping in that fascinat-
antecedents of Nepali mapping going as far back as an ing remote corner of the world.
Indian map sent to China in A.D. 648, Slusser suggests
that "it may well be that in Nepal as well, map making 182. Slusser, "Cultural Aspects," 26 (note 13).
APPENDIX 15.1 TIBETAN MAPS IN THE WISE COLLECTION,

ORIENTAL AND INDIA OFFICE COLLECTIONS, BRITISH LIBRARY, LONDON
Number and Approximate
Volume and Arrangement of Dimensionsa
Catalog Number Sheets (h x w) (cm) Area Covered Remarks
a. Vol. 1, Three, in a line 62 x 153 Lhasa and environs, More a locality than a regional map.
Add. Or. 3013 especially the Potala and Relatively crudely rendered.
Jo-khang
b. Vol. 1, Four, in a line; 70 x 283 Ladakh, from Leh to Abundant annotations are keyed to
Add. Or. 3014 one end sheet at Hanle, about 200 km to numbers on the map. Even more crudely
a right angle to the southeast rendered than item a.
others
c. Vol. 1, Seven, in a line; 74 x 335 From west of Demchok Seen as if the imaginary viewer is looking
Add. Or. 3015 one irregular in Ladakh, past Mount south (actually south-southwest). A
end sheet at a Kailas and Lake prominent river on the map, labeled
right angle to Manasarowar to Pekut "Makchakabad," flowing out of Lake
others Lake in Tsang Province Manasarowar and into the Tsangpo,
of central Tibet, i.e., cannot be identified on modern survey
from approximately 79° maps. Parts of five sheets left· blank.
to 86°E
d. Vol. 1, Seven, in a line 48 x 300 Central Tibet, middle Sheets 6 and 7 of the map are said to
Add. Or. 3016 sheet includes relate to "Namtsho?" the location of
Tashilunpo monastery in which has not been determined. More
Shigatse (ca. 89°E) than 150 key numbers on map, but no
accompanying text.
e. Vol. 2, Six, some joined 160 x 185 Central and eastern Includes numerous Tibetan captions, but
Add. Or. 3017(a) to others at right Tibet, including Samye without accompanying explanatory text.
angles and Chunggye Partially illustrated in figure 15.36. See
also item f.
f. Vol. 2, One, affixed to 48 X 62 Chunggye Scale much larger than that of item e, to
Add. Or. 3017(b) right of item e which this is affixed. Fifty-seven edifices,
taken to be the tombs of Tibetan kings,
are named in Tibetan. Text to match
Arabic key numbers is not available.
g. Vol. 2, Five, in a line 33 x 194 Zanskar valley in Ladakh Accompanied by abundant explanatory
Add. Or. 3018 and adjacent areas text
aBecause most of the maps do not form perfect rectangles, their actual areas may be significantly smaller than the maximum vertical and
horizontal dimensions cited suggest.
675
APPENDIX 15.2 TIBETAN MAPS IN THE HARRER COLLECTION,

VOLKERKUNDEMUSEUM DER UNIVERSITAT ZURICH
Reference
Approximate in Brauen,
Catalog Dimensions Impressionen
Number (h X w) (cm) Medium Area Depicted Description/Remarks aus Tibeta
a. 14481 87 X 97 Black ink; red Southeastern Area of relatively dense settlement. 104 and
and gray wash Tibet Seems to relate to a military catalog, p. 3
campaign, probably the
Younghusband expedition of 1904.
b. 14482, 52 X 67 Black, red, blue, Presumably Undoubtedly related to a military
side A and yellow similar to item a campaign; seems to be by the same
watercolor artist as item a
c. 14482, 52 X 67 Black ink A single locality, Very large-scale depiction.
side B presumably in Compounds shown planimetrically,
southeastern but individual houses, of several
Tibet types, shown in frontal elevation.
Much detail relating to religious
structures. Association with item b
not clear.
d. 14485 32 X 95 Painted Not known Not seen by me. Described as catalog, p. 3
cCLandkarte.:"
e. 14488 31 X 190 Painted Southern Tibet As for item d catalog, p. 3
f. 14493 32 X 94 Painted Not known As for item d catalog, p. 3
g. 14495 52 X 171 Black ink; yellow Area of eastern See figure 15.37 and text pI. 110
and brown bend of
watercolor Tsangpo/
Brahmaputra
h. ? 52 X 62 Same as for item Kham Province Style similar to item b but richer in pI. 9
b in eastern Tibet detail, especially in regard to
vegetation. Appears also to have a
military purpose. One building flies
British flag. Ferries prominently
shown.
i. ? 62 X 95 (L- Black ink; Presumably on or Bears note: "Restauriert Winter
shaped, two brown, yellow, near Tibet's 1982"
sheets of green, and mauve border with
paper) wash Arunachal
Pradesh in India
. ~
26 X 46 Black ink and A small locality Style similar to item b. Monastery
J••
red watercolor shown in oblique perspective, with
several hamlets.
k. ? 32 X 106 Black ink; blue, One main river Filed in same drawer as items b
red, and yellow valley (Chumbi?) and c. Also seems to relate to
watercolor with several military campaign. Flags flying in
tributaries several crudely drawn settlements.
aMartin Brauen, Heinrich Harrers Impressionen aus Tibet (lnnsbruck: Pinguin-Verlag, 1974), including a fifteen-page catalog of the exhibition
at the museum that opened in December 1974.
676
APPENDIX 15.3
Dimensions
Localitya Place Where Map Is Held Provenance and Date (h x w) (cm) Language IT ext b
a. Chang'an, China Potala, Lhasa Lhasa; date unknown Not known None evident
b. Chokpuri Hill, Potala, Lhasa Lhasa; date unknown Not known None evident
Lhasa
c. Chokpuri Hill, Potala, Lhasa Lhasa; date unknown Not known None evident
Lhasa
d. Drepung, near Musees Royaux d'Art et Probably around Amdo in 115 x 68 Numerous
Lhasa d'Histoire, Brussels, northeastern Tibet; inscriptions in
Collection Leon Verbert 350 eighteenth century cursive Tibetan
e. Dza-rung, Nepal (?) Not known Gi-Iong Nor-bu (artist); 1975 Not known None evident
fig. Gang-gar, Ding- Private collection of Barbara Pan-tan P'a-Ia (artist), former Not known None evident
ri, Tibet Nimri Aziz? inhabitant of town; ca. 1975
h. Gorkha, Nepal Museum, Bhaktapur, Nepal Nepal (Newari style); early Not known None
nineteenth century
i. Gyantse, Tibet Palchor monastery, Gyantse, Gyantse, Tibet, date Not known, but None evident
Tibet unknown quite large
j. J0- khang Palace, Potala, Lhasa Potala, Lhasa; date unknown None evident
Lhasa
k. Lhasa Not known Probably Darjeeling, India; None
late nineteenth century
1. Lhasa Gerd-Wolfgang Essen Tibet; ca. 1900 93 x 160 None
Tibetica Collection,
Hamburg
aChokpuri Hill, Drepung, Jo-khang, and the Potala are all within the pung," Bulletin des Musees Royaux d'Art et d'Histoire 57, no. 2 (1986):
orbit of Lhasa; Tashilunpo is similarly within the orbit of Shigatse. 5-29; Lumir Jisl, Tibetan Art, trans. Ilse Gottheiner (London: Spring
bThe note "none evident" signifies that text cannot be discerned on Books, 1957); Detlef lngo Lauf, Tibetan Sacred Art: The Heritage of
available photographs and is not mentioned in relevant references. Tantra, trans. Ewald Osers (Berkeley, Calif.: Shambhala, 1976); Lobsang
cThe citations in this column are: Barbara Nimri Aziz, Tibetan Fron- P. Lhalungpa, Tibet, the Sacred Realm: Photographs, 1880-1950, exhi-
tier Families: Reflections of Three Generations from D'ing-ri (Durham, bition catalog, Philadelphia Museum of Art, 20 March-22 May 1983
N.C.: Carolina Academic Press, 1978); Catalogue of the Tibetan Col- ([Millerton, N.Y.]: Aperture, 1983); Li Jicheng, The Realm of Tibetan
lection and Other Lamaist Material in the Newark Museum, 5 vols. Buddhism (New Delhi: UBS Publishers' Distributors, 1986); Liu
(Newark, N.J.: Newark Museum, 1951-71), vol. 3; Sarat Chandra Das, Lizhong, Buddhist Art of the Tibetan Plateau, ed. and trans. Ralph
Indian Pandits in the Land of Snow (1893; reprinted Delhi: Delhi Kiggell (Hong Kong: Joint Publishing, 1988); Alexander W. Macdonald
Printers Prakashan, 1978); Gerd-Wolfgang Essen and Tsering Tashi and Anne Vergati Stahl, N ewar Art: Nepalese Art during the Malla
Thingo, Die Gotter des Himalaya: Buddhistische Kunst Tibets,·2 vols. Period (Warminster, Eng.: Aris and Phillips, 1979); Claudius C. Moller
(Munich: Prestel-Verlag, 1989); Werner Forman and Bedrich Forman, and Walter Raunig, eds., Der Weg zum Dach der Welt (lnnsbruck:
Art of Far Lands, ed. Lubor Hajek, trans. W. Cungh and H. Watney Pinguin-Verlag[1982]); Ngapo NgawangJigmei et aI., Tibet (New York:
([London]: Spring Books, [1958?]); Manfred Gerner, Architekturen im McGraw-Hill, 1981); Blanche Christine Olschak, Augusto Gansser, and
Himala;a (Stuttgart: Deutsche Verlags-Anstalt, 1987); Philippe van Andreas Gruschke, Himalayas (New York: Facts on File, 1987);
Heurck, "Description de la than-ka representant Ie monastere de Dre- Blanche Christine Olschak and Geshe Thupten Wangyal, Mystic Art of
677
GREATER TIBETAN LOCALITY MAPS
Medium Description Where Published c
Mural painting See text, p. 670 Liu, Buddhist Art, 187 (fig. 343)
Mural painting Various unidentified buildings in oblique frontal Li, Realm of Tibetan Buddhism, 166-67
perspective ranged about hill; rugged hills in (mislabeled)
background
Mural painting As for item b, but gentler terrain in background Liu, Buddhist Art, 187 (fig. 344)
Painted thanka See text, pp. 666-68, and figure 15.48 Heurck, "Description de la than-ka," 5-29; Lauf,
Tibetan Sacred Art, pI. 5; Lhalungpa, Tibet, 22;
Tibetische Kunst, fig. 96 and pp. 72-73; Van der
Wee, Van der Wee, and Schotsmans, Symbolisme,
34-35; Vergara and Beguin, Dimare umane, 80
Ink and paint on Town, within sight of Mount Everest; oblique Aziz, Tibetan Frontier Families, unnumbered plate
paper perspective; background naturalistically rendered between 96 and 97
Painted on paper Oblique perspective of mountainside town Aziz, Tibetan Frontier Families, unnumbered plate
between 96 and 97
Wall painting Palace and neighboring buildings set in middle Vergati, "Les royaumes de la vallee de
ground in circuit of trees; forests to rear and Katmandou," 202-3
cultivated fields and river in foreground, each
marked by characteristic animals drawn larger than
many buildings; very diverse treatment of
vegetation
Mural painting Several clusters of temples and monasteries in Liu, Buddhist Art, fig. 302
oblique frontal perspective, separated by areas of
relatively verdant terrain
Mural painting Festival scene with large gathering of monks; Liu, Buddhist Art, 184 and fig. 339
mixed oblique and frontal perspectives
Ink on paper Multiple, partially divergent perspectives Waddell, Buddhism of Tibet, facing 287
Painted thanka Painted as guide map for pilgrims; varying oblique Essen and Thingo, Die Gotter des Himalaya,
with brocade perspectives; much open terrain between city 1:245-47 and 2:221-22
border proper and Potala
Ancient Tibet (1973; Boston: Shambhala, 1987); Valrae Reynolds, Tibet: Himalaya (Vienna: Anton Schroll, 1968); Chogyam Trungpa, Visual
A Lost World, exhibition catalog for the Newark Museum Collection Dharma: The Buddhist Art ofTibet (Berkeley, Calif.: Shambhala, 1975);
of Tibetan Art and Ethnography (New York: American Federation of Louis P. Van der Wee, "A 'Cloister-City'-Tanka," Journal of the
Arts, 1978); Theodore Riccardi, Jr., "Some Preliminary Remarks on a Indian Society of Oriental Art, n.s., 4 (1971-72): 108-20; Pia Van der
Newari Painting of Svayambhonath," Journal of the American Oriental Wee, Louis P. Van der Wee, and Janine Schotsmans, Symbolisme de
Society 93 (1973): 335-40; Mary Shepherd Slusser, Nepal Mandala: A I'art lamafque (Brussels: Musees Royaux d'Art et d'Histoire, 1988);
Cultural Study of the Kathmandu Valley, 2 vols. (Princeton: Princeton Paola Mortari Vergara and Gilles Beguin, eds., Dimore umane, santuari
University Press, 1982); idem, "On a Sixteenth-Century Pictorial Pil- divini: Origini, sviluppo e diffusione delfarchitettura tibetana /
grim's Guide from Nepal," Archives of Asian Art 38 (1985): 6-36; idem, Demeures des hommes, sanctuaires des dieux: Sources, developpement
"The Cultural Aspects of Newar Painting," in Heritage of the Kath- et rayonnement de f architecture tibetaine (Rome: Universita di Roma
mandu Valley: Proceedings of an International Conference in Lubeck, "La Sapienza," 1987); Anne Vergati, "Les royaumes de la vallee de
June 1985, ed. Niels Gutschow and Axel Michaels (Saint Augustin: Katmandou," in Les royaume de I' Himalaya: H istoire et civilisation
VGH Wissenschaftsverlag, 1987), 13-27; David L. Snellgrove and Hugh (Paris: Imprimerie Nationale, 1982), 164-208; Laurence Austine Wad-
Richardson, A Cultural History of Tibet (London: Weidenfeld and dell, The Buddhism of Tibet, or Lamaism, with Its Mystic Cults, Sym-
Nicolson, 1968); Tibetische Kunst: Katalog zu Ausstellung, 8.-30. bolism, and Mythology, and Its Relation to Indian Buddhism, 2d ed.
Miirz, 1969, Helmshaus, Zurich, 17 Apr. bis 11 Mai 1969, Gesell- (Cambridge: W. Heffer, 1935 [first published 1895]); Siddiq Wahid,
schaftshaus zu Schutzen, Luzern (Bern: TIBETA, 1969); Herbert Tichy, Ladakh: Between Earth and Sky (New York: Norton, 1981).
678
APPENDIX 15.3
Dimensions
m. Lhasa Private collection of David Tibet (?); probably twentieth None evident
Tremayne, London century
n. Lhasa and National Museum of Provenance and date Very large None evident
environs Ethnography, Stockholm, unknown
cat. no. H 2882
o. Menri monastery, Oriental and India Office Menri (?), Tibet; ca. 1900 None
Tsang Province, Collections, British Library,
Tibet London
p. Nepal ITibet Ashmolean Museum, Oxford Tibet or Nepal None evident
border? Unknown University
stupa and
surrounding holy
places
q. Patan, Nepal Guita-bahil monastery, Patan, Nepal; seventeenth to Very large, more Newari
Patan, Nepal eighteenth century than 10 m long
r. Patan and environs Kwa Bahal monastery, Patan, Patan, Nepal; ca. 1850 Est. 90 x 1,800 Newari
in Vale of Nepal
Kathmandu, Nepal
s. Potala, Lhasa Private collection of Blanche Tibetan monastery in Devanagari script
Christine Olschak northern Nepal; seventeenth
century (?)
t. Potala, Lhasa Private collection Tibet; seventeenth-eighteenth 81 x 56 None evident
century
u. Potala, Lhasa Probably Darjeeling, India; in Tibet; probably late None evident
late nineteenth century nineteenth century
v. Potala, Lhasa Potala, Lhasa Potala, Lhasa; date unknown Height ca. 3 m; Abundant text,
width not presumably
known Tibetan
w. Potala, Lhasa Potala, Lhasa Potala, Lhasa; date unknown As for item t Tibetan
x. Potala, Lhasa Potala, Lhasa Potala, Lhasa; date unknown As for item t
y. Potala, Lhasa Potala, Lhasa Potala, Lhasa; date unknown Very large None evident
z. Potala, Lhasa Potala, Lhasa Potala, Lhasa; date unknown None
aa. Potala, Lhasa Potala, Lhasa Potala, Lhasa; date unknown None evident
(future site of)
bb. Samye, Tibet Newark Museum, Newark, Kham region, eastern Tibet, 53 x 38 Tibetan text in
N.J.; acc. no. 20.271 seventeenth-eighteenth red and gold
century
cc. Samye, Tibet Samye (?), Tibet Samye (?), Tibet; date Cursive Tibetan
unknown
679
(continued)
Painted on paper Oblique perspective, terrain presented in relatively

naturalistic style
Painted on cloth See text, p. 668, and figure 15.49
Paint and ink on Shows a major monastery of the Bon religion; Moller and Raunig, Der Weg zum Dach der Welt,
paper buildings in frontal and terrain in oblique 373, 375
perspective, considerable attention to vegetation
Temple banner Stupa at a very large scale and surrounding holy Snellgrove and Richardson, Cultural History,
painted on cloth places at much smaller scales, all in frontal appended photographic plate
perspective within mountainous terrain presumably
near Nepal/Tibet border; work illustrates a
pilgrimage
Painted cloth Detailed townscape in panoramic form, oblique Slusser, Nepal Mandala, vol. 2, fig. 97; idem,
scroll perspective. See also figure 15.2. "Pilgrim's Guide," 30
Painted cloth See text, pp. 663-64, and figure 15.46 Slusser, Nepal Mandala, vol. 2, fig. 98; idem,
scroll "Pilgrim's Guide," 31 and 32; idem, "Cultural
Aspects," fig. 11
Painted thanka Does not show buildings constructed in reign of Olschak and Thupten Wangyul, Mystic Art, 78
fifth Dalai Lama (1617-82); oblique perspective
Paint and ink on Oblique perspective, structures very Forman and Forman, Art of Far Lands, 204
cloth individualistically portrayed
Ink on paper Multiple, partially divergent perspectives Waddell, Buddhism of Tibet, 40
Mural painting Shows Potala being extended; oblique frontal Li, Realm of Tibetan Buddhism, 166; Jisl, Tibetan
perspective Art, 29 and fig. 33
Mural painting As for item t; further stage in construction Jisl, Tibetan Art, 29 and fig. 34
Mural painting Shows a near terminal phase of construction; Gerner, Architekturen im Himalaja, 98
frontal perspective
Mural painting Festive scene showing unfurling of giant thanka Liu, Buddhist Art, 184 and fig. 338
on south wall; oblique frontal perspective
Mural painting Shows Potala before expansion; mixed frontal and Liu, Buddhist Art, 187 and fig. 342
oblique perspectives
Mural painting Shows original temple and fortifications where Jisl, Tibetan Art, fig. 32
Potala was to be; oblique perspective
Painted and gilt See text and figure 15.47 Catalog of the Tibetan Collection, 3:64 and 91;
thanka on Reynolds, Tibet, 118 and 126; Trungpa, Visual
Chinese silk Dharma, 37; Vergara and Beguin, Dimore umane,
79-80
Fresco painting Extraordinary composition; monastery complex Ngapo et aI., Tibet, 247-49
shown in varying oblique perspectives, within
circular wall, set against variegated landscape
depicted at much smaller scale, also in mixed
perspectives
680
APPENDIX 15.3
Dimensions
dd. Samye, Tibet Samye monastery Samye, Tibet; date unknown None evident
ee. Samye, Tibet Main temple in Gangtok, Sikkim, date unknown None evident
Sikkim
ff. Sherpa village, Not known Kalden, a twentieth-century None
Nepal Sherpa artist
gg. Shigatse, Tibet American Museum of Tibet; late nineteenth or ca. 66 x 94 None evident
Natural History, Hall of twentieth century
Asian Peoples
hh. Svayambhilnath, Collection of A. Peter Patan, Nepal; 1565 101 x 85 Newari and
Nepal Burleigh Sanskrit
ii. Svayambhilnath, Private collection, Paris Central Nepal; late 90 x 70 None evident
Nepal seventeenth century
jj. Svayambhiinath, Private collection, Brussels Nepal; late eighteenth 82 x 55 Devanagari
Nepal century (?) script, language
not known
kk. Tashilunpo, Shigatse, Tibet, owner not Shigatse, Tibet; late None evident
Shigatse, Tibet known eighteenth century
11. Tashilunpo, Private collection Central Tibet; between 1854 None evident
Shigatse, Tibet and 1882
mm. Tashilunpo, Library at Alchi monastery, Ladakh (?); date unknown None evident
Shigatse, Tibet Ladakh
nne Tashilunpo, Traktok monastery, Ladakh Traktok monastery, date None evident
Shigatse, Tibet unknown
00. Unknown city Somewhere in Tibet? Tibet? Very large Considerable
text, presumably
Tibetan
pp. Unknown National Museum of Northern Tibet or Qinghai 138 x 186 None evident
monastery complex Ethnography, Stockholm (?); date unknown
681
(continued)
Fresco painting Oblique view of monastic complex in its original Gerner, Architekturen im Himalaja, 53
form
Fresco painting Approximation of Western one-point perspective Olschak, Gansser, and Gruschke, Himalayas, 175
Painted, on paper Village painted against background of vividly Tichy, Himalaya, frontispiece
(?) depicted Mount Khumbila. People and animals
shown larger than houses. Oblique frontal
perspective.
Paint and ink on See text, pp. 668-70, and figure 15.50
paper
Painted on cloth See text, pp. 661-63, and figure 15.43 Riccardi, "Preliminary Remarks"; Slusser, Nepal
Mandala, vol. 2, pI. 495; idem, "Pilgrim's Guide";
idem, "Cultural Aspects," 20-27
Painted on cloth Frontal perspective for stupa, oblique perspective Macdonald and Vergati Stahl, Newar Art,
for its surroundings frontispiece
Painted on cloth Frontal perspective for stupa, oblique perspective
for some nearby structures
Painted thanka Mixed frontal and oblique perspectives; terrain Liu, Buddhist Art, 228 (fig. 435)
presented in relatively naturalistic style
Painted on cloth, Oriented toward north, many structures can be Van der Wee, "Tanka," passim
formerly in silk individually identified; wall in foreground,
frame mountains in background; oblique frontal
perspective
Painted thanka Oblique frontal perspective Wahid, Ladakh, 88-89
Painted, on paper Essentially a frontal perspective Gerner, Architekturen im Himalaja, 85

(?)
Appears to be Very detailed representation of a townscape with Das, Indian Pandits, frontispiece
mural painting many large edifices (monasteries?) and places of
assembly in a rather verdant setting. Mixed
oblique and frontal perspectives. Conventional
crest of mountains in background.
Painted on cloth Monastery complex mountain girt on three sides;
foreground shows laymen with horses, camels,
wagons, and tents; varying oblique perspectives
APPENDIX 15.4 to prevent border disputes and to limit the groups' movements.
MONGOLIAN CARTOGRAPHY By restricting pastoral nomadism, it fundamentally changed the
G. HENRIK HERB
1. "Mongolian" in this context is clearly not restricted to the present
Although the Mongolian people of the Central Asian steppes Mongolian state, since this political unit does not encompass all Mon-
have had enormous historical influence on China, the Middle golian people. Large numbers of Mongols inhabit Inner Mongolia in
East, and even Europe, an authoritative account of a Mongolian China, Chinese Turkestan (now Xinjiang Uygur Autonomous Region),
cartographic tradition is still lacking. Indigenous sources for the parts of Manchuria, and the Buryat-Mongol lands of Siberia.
reconstruction of this tradition either have not survived or have 2. This does not mean there was no indigenous Mongolian carto-
yet to come to light. 1 The only Mongolian geographical maps graphic tradition. It is possible that surviving maps might not have come
that have been preserved were made after the middle of the to light and that others might have been drawn directly on the ground,
or on materials such as animal hides and wooden boards, as well as
eighteenth century-that is, after close contact with well-estab-
paper or cloth, that did not survive the rigors of time.
lished foreign cartographic traditions. 2 As a result, it is extremely
3. The maps are preserved at the University Library, Uppsala. They
difficult to isolate indigenous Mongolian elements from the
are reproduced in John F. Baddeley, Russia, Mongolia, China; Being
existing maps. Apart from geographical maps, there is the ques- Some Record of the Relations between Them from the Beginning of
tion of cosmographical artifacts. It would be surprising if the the XVllth Century to the Death of the Tsar Alexei Mikhailovich, A.D.
traditional use of such works among Tibetan Buddhists were 1602-1676, 2 vols. (London, 1919; reprinted New York: B. Franklin,
not mirrored to some extent by their Mongol coreligionists. 1964), and Nicholas Poppe, "Renat's Kalmuck Maps," Imago Mundi
However, I have not had the opportunity to investigate that 12 (1955): 157-59.
Issue. 4. The Oirat script was introduced in 1648 by Zaya Pandita, a Bud-
Even the two maps customarily cited as the oldest Mongolian dhist monk and scholar; see Poppe, "Kalmuck Maps," 157-58 (note
3).
artifacts-which have been termed Renat 1 and Renat 2 in the
5. There is considerable uncertainty about the authors of the maps.
literature-thus appear to have been made with little Mongolian
Heissig stated that Renat 1 was drawn by the Kalmyk ruler Galdan
contribution. 3 These maps were brought to Europe in 1734 by
Tseren (1727-1845), whereas Poppe mentioned that the map was given
a Swedish officer, Johan Gustav Renat, after a seventeen-year to Renat by Galdan Tseren. In the case of Renat 2, Heissig believed
captivity by the West Mongolian Kalmyks. Both Renat 1, which that the map was made by Oirats (Oloten), who made improvements
covers the area from about 70° to 90° east longitude (the west- to a Chinese original, whereas Poppe claimed that Renat had obtained
ern terminus is Samarkand), and Renat 2, which covers an even the map from Chinese troops who had attacked the Kalmyks. See
larger region stretching from 79° to 107° east longitude (the Walther Heissig, "Uber Mongolische Landkarten," Monumenta Serica
eastern terminus is Ulan Bator), are oriented to the south and 9 (1944): 123-73, esp. 124 and 127, and Poppe, "Kalmuck Maps," 157
use Oirat, a modified Mongolian script. 4 Beyond the script, there (note 3).
is little that indicates a Mongolian tradition; they are believed 6. The holdings in Ulan Bator are mentioned in Walther Heissig, ed.,
Mongolische Ortsnamen, 3 vols. (Wiesbaden: Franz Steiner, 1966-81),
to be copies of Chinese originals. 5
1:VIII, and see the Mongolian reference cited there. The maps in Berlin,
The remaining Mongolian maps-about 565 artifacts are
which were collected by Herrmann Consten and Walther Heissig, are
known-differ considerably from the Renat maps and form a cataloged in Walther Heissig, Mongolische Handschriften, Block-
unified corpus. They are regional manuscript maps that depict drucke, Landkarten (Wiesbaden: Franz Steiner, 1961). Twenty-six of
Mongolian tribal and pasture territories. Their most striking the maps were reproduced in 1966 (2 in color and 24 in black and
features, which are clearly visible in most of them, are anno- white), and 118 of them were reproduced in 1978 (black-and-white
tations extending beyond the boundaries in different directions facsimiles); see Heissig's Mongolische Ortsnamen, vols. 1 and 2. The
(see figs. 15.51 and 15.52). A few of the maps date from the maps in this collection were made between 1890 and 1920; most date
middle of the eighteenth century, but most were made between from 1907 (Heissig, Mongolische H andschriften, 338; this source also
the middle of the nineteenth century and the early twentieth contains a list of the maps in Tenri, 493-94, with references).
century. 7. The map in Copenhagen, MS. Mong. 562, is of the Setsen Khan
district (Khalkha), 1913. Heissig mentioned that he saw a map depicting
The maps are held in several depositories: the largest collec-
the Khalkha district in the collection of L. Kotwicz in Krakow (Mon-
tion, 335 maps, is part of the Mongolian State Archive in Ulan golische Ortsnamen, 2:XII [note 6]). Heissig has mentioned that there
Bator, followed by the Staatsbibliothek zu Berlin-Preussischer are three maps in private collections in addition to the 182 maps kept
Kulturbesitz, with 182 maps and the Tenri Central Library in at the Staatsbibliothek zu Berlin, referring to his 1944 article for further
Tenri, Japan, with 44. 6 Additional maps are preserved at the information. However, only two maps could be identified from Heis-
Royal Library in Copenhagen, the Archive of the Polish Acad- sig's sources. (1) Map of the Khangghin Banner of the Ordos region
emy of Sciences in Krakow, and in private collections. 7 from 1909, owned by the German professor Walter Fuchs (described,
The German scholar Walther Heissig-who was the first to including a list of 189 toponyms, and reproduced in "Uber Mongolische
describe these maps in detail and on whose work this brief Landkarten," 126, 136-47, pI. XIII [note 5]). A similar map is part of
the Berlin collection (Hs. Or. 108); see Mongolische Handschriften,
account is based-believes that the incentive for their produc-
337 (note 6). (2) Map of the Dzungghar Banner of the Ordos region,
tion came from the outside. s He links their production to the
undated, which is owned by Heissig (described, including a list of 210
division of Mongols into "banners" that took place in 1649, toponyms, and reproduced in "Uber Mongolische Landkarten," 126,
shortly after submission to Manchu rule. The creation of ban- 160-70, and pI. XV). A similar map is part of the Berlin collection (Hs.
ners-that is, the delimitation of enduring territorial boundaries Or. 836); see Mongolische Handschriften, 337).
for the pasture areas of Mongolian tribal units-was intended 8. Heissig, "Uber Mongolische Landkarten," 128 (note 5).
.
.(
..•
FIG. 15.51. MAP OF THE rOMBOSURUN BANNER IN that traverse the banner from west to east, leading to Kiiriyen
THE SETSEN KHAN DISTRICT (KHALKHA). Mongolian (Urga).
manuscript map from 1907, depicting sixty-nine boundary Size of the original: 96.3 X 64.3 cm. By permission of the Staats-
markers (oboya) extending beyond the pasture territory. In addi- bibliothek zu Berlin-Preussischer Kulturbesitz, Orientabteilung
tion to a variety of toponyms, it also shows six caravan routes (Hs. Or. 101).
FIG. 15.52. MAP OF THE TUSALArCI GONG BANNER IN and mountain ridges, sand dunes, the monastery Yeriigel-i
THE SETSEN KHAN DISTRICT (KHALKHA). This Mon- dayayaysan sumee, and the site of a religious school as well as
golian manuscript map from 1913 depicts a tribal pasture ter- a homestead. Seven annotations in the margins define the dis-
ritory. Thirty-five annotations extending radially from the tances between the border markers.
boundary ipentify the directional position of boundary markers, Size of the original: 49.5 x 75 em. By permission of the Oriental
called oboya. It also depicts the names of fifty-five mountains Department, Royal Library, Copenhagen (Mong. 562).
tribal social structures. The banner boundary markers, called another country." Also requested were the "mountains, rock
oboya, were not permanent. They consisted mostly of piles of formations and ruins, temples and monasteries, bridges, ravines
stones, sand, or earth, and the imperial government ordered and mountain passes that exist in each banner; moreover, the
that they be checked and repaired periodically.9 products of the region and the submitted taxes, persons of rank
The production of maps is not mentioned in the context of and name, laws and customs, the number of settlements, and
the first Manchu orders to fix banner boundaries, but it is doc- the size of the families."l1
umented for later years. A Mongolian treatise from 1802 men- A printed map of the Dolonor (Dolon Nor) region, which
tions that maps were made in 1686 at the order of an imperial has been dated to about 1911, also provides corroborative evi-
envoy to document newly established banner boundaries. A
letter from the colonial office in Beijing to the ruler of the Tumet 9. Documents indicate that yearly checks were requested in 1832,
Banner in Koke Khota in the summer of 1690, written in Mon- 1855, and 1889; see Heissig, "Uber Mongolische Landkarten," 130-31
golian, asked for detailed geographical descriptions about the (note 5).
individual Mongolian banners. to 10. Heissig, "Uber Mongolische Landkarten," 128-30 (note 5), and
The information requested, which was to be used in preparing idem, Mongolische Ortsnamen, I:VII (note 6). I have followed Heissig's
spellings for Mongolian place-names throughout this appendix; Heissig
the Da Qing yitong zhi (Comprehensive gazetteer of the Great
also provides maps of some of these areas.
Qing realm, completed 1746), included "the pasture areas of 11. From the German in Heissig's "Uber Mongolische Landkarten,"
each banner and also the name of the banner and its territory; 128-30 (note 5), where he used the word aufschreiben for the recording
the extent of each banner territory in miles in all cardinal direc- of this geographical information, which could also mean preparing a
tions; and also the numbers of paths and, furthermore, from list. However, he later used the word aufzeichnen, which implies draw-
which direction the banner borders meet with the borders of ing a map (Mongolische Ortsnamen, I:VII-VIII [note 6j).
dence. Its title states that it was based on a manuscript map alphabetizing, and cross-referencing the place-names on many
submitted at the request of the imperial government in 1742. 12 of the maps in the Staatsbibliothek zu Berlin. This resulted in
Imperial involvement in the production of maps of Mon- the most comprehensive and authentic collection of Mongolian
golian banners is documented for yet another case. Mostaert place-names to date and provides a solid basis for etymological
reports that on 16 December 1739, a meeting took place that research. 22
included the imperial emissary Liou Bayar, the head of the Mon- Although the scarcity and late date of the extant sources
golian confederation in the Dolonor region, and the rulers of makes it difficult to trace the origins of Mongolian cartography,
the seven Dolonor banners. 13 The purpose of the meeting was further synthesis is still feasible and desirable. A first step would
to delimit the boundaries for the seven banners because a dis- be a translation from Mongolian and a critical examination of
pute had arisen. A map was drawn, and each of the seven rulers the work by Goncigdorz mentioned by Heissig. Goncigdorz's
affixed his seal to state his approval. The map was sent to the findings appear to be very promising, particularly his claim that
colonial office in Beijing. 14 the maps of the Khitans were precursors to Mongolian maps.
Heissig claims that these individual requests by the imperial In order to free the study of Mongolian cartography from its
government for maps of Mongolian territories were routine linguistic isolation, and to prevent the introduction. of nation-
procedure: all Mongolian districts, confederations, and banners alist bias, it will be necessary to enlist the help of scholars from
were required to submit maps every ten years to the colonial different disciplines and with different linguistic competence.
office in Beijing, where they were registered and translated. 1s
The maps in the collection of the Staatsbibliothek zu Berlin 12. The scripts on the map were Chinese and Mongolian (Heissig,
give testimony to five stages in the process from production to "Uber Mongolische Landkarten," 130-31 [note 5]). Note that Heissig
use by the colonial office: (1) initial Mongolian drafts; (2) com- never consulted the original printed map, but had only seen a repro-
pleted Mongolian maps, not colored; (3) finished Mongolian duction of the printed map in a journal.
maps; (4) finished maps with small glued-on pieces of paper 13. See Mostaert's introduction in Sayang Secen (Ssanang Ssetsen),
Erdeni-yin TobCi: Mongolian Chronicle, 4 vols., ed. Antoine Mostaert
bearing Chinese transcriptions and translations; (5) maps with
(Cambridge: Harvard University Press, 1956), 1:85-86.
Chinese lettering that were based on Mongolian originals. 16
14. The 1739 map has not been preserved, but Mostaert showed that
Despite foreign incentives for the production of these maps, it was the archetype of a 1740 map he illustrated. Mostaert reproduced
Mongolian scholars claim they are part of a Mongolian tradi- a 1903 photo of the original 1740 map as well as a 1927 copy; Sayang
tion. According to Heissig, B. Goncigdorz traced the maps and Secen (Ssanang Ssetsen), Erdeni-yin Tobci, pI. 1 and additional plate
their construction to the Khitans, whose cartographic activities (note 13). Joseph Kler, "A propos de cartographie mongole," Bulletin
apparently have been documented for as early as 1179. 17 Chag- de la Societe Royale Beige de Geographie 24, pts. 1 and 2 (1956): 26-
darsurung (Shagdarsurung in other transcriptions) pointed out 51, also reproduced the map. However, as Mostaert pointed out, Kler
that the maps of Mongolian banners and districts were based reproduced not the original map, as he claimed, but the 1927 copy.
on what he called the "compass rose system of Mongolian According to Mostaert, the 1740 map is preserved at the "ia men de
la banniere d'Otoy" (126 and additional note at the end of the book).
cartography," which used twenty-four directional points to
15. Heissig, Mongolische Handschriften, 338 (note 6).
define the locations of the boundary markers, or oboya. I8 These
16. Heissig, Mongolische Handschriften, 338 (note 6). In addition to
directional points were made up of the twelve elements of the works conforming to these five stages, Heissig described two maps that
Chinese zodiac (which included the four cardinal directions), in were unique: Hs. Or. 253, which he considered "so modern in design
combination with eight colors and with four points related to that it is certainly purely Chinese," and Hs. Or. 33, which has Manchu
elements of divination. I9 In addition, he identified a set of inter- script (p. 337).
mediary points that enabled a further division into forty-eight 17. Heissig, Mongolische Ortsnamen, 2:XII (note 6). I have been
directions. To support his argument he gave a translation of the unable to locate the Goncigdorz work cited by Heissig.
Mongolian lettering on a map of the Ceringwangduyibabu- 18. Ts. Chagdarsurung, "La connaissance geographique et la carte
des Mongols," Studia Mongolica, vol. 3 (2) (1975): 345-70, e~p. 347.
dorJi Banner in the Setsen Khan district, Khalkha. On this map,
19. The twelve elements are, clockwise: rat (north), bull, tiger, hare
thirty-two annotations extended vertically from each boundary
(east), dragon, snake, horse (south), sheep, monkey, rooster (west), dog,
marker beyond the boundary line, similar to the representation and pig; and the eight colors are two shades each for black, blue, red,
in figure 15.51. Each of these annotations identified the location and white. See Chagdarsurung, "La connaissance geographique," 347-
of that marker by giving reference to the directional points and 50 (note 18), and Heissig, Mongolische Ortsnamen, 2:XII-XIII (note 6).
also its distance to the next boundary marker clockwise. 20 20. Chagdarsurung, "La connaissance geographique," 355-58 (note
Apart from these tentative findings by Mongolian scholars, 18); Heissig notes that the descriptions of the 32 boundary markers
little attention has been given in the general literature to the correspond exactly with Hs. Or. 146 in the Berlin collection (Mon-
precursors and designs of regional Mongolian pasture area golische Ortsnamen, 2:xv and pI. 65 [note 6]).
maps. Most efforts have been directed at their geographical 21. Heissig, Mongolische Ortsnamen, 1:IX (note 6).
22. Magadborin Haltod collected 13,644 Mongolian place-names
content, because the maps are unique resources for place-name
from the 182 maps in the Staatsbibliothek zu Berlin-Preussischer Kul-
research. Virtually all authors who discuss them include long
turbesitz, and they were published as an alphabetical list in volume 1
lists of toponyms in their works. As comparisons between maps of Mongolische Ortsnamen (note 6). Volume 2 contains facsimiles of
from different time periods have shown, the place-names men- 118 of the 182 maps. In volume 3 the place-names have been cross-
tioned in the maps have changed little over the last two to three referenced by S. Rasidondug, H.-R. Kampfe, and V. Veit. Other lists
hundred years. 21 To make this resource more accessible, of place-names were often based on inaccurate Chinese translations;
Walther Heissig initiated an ambitious project of collecting, Heissig, Mongolische Ortsnamen, l:IX-XI (note 6).

Maps of Greater Tibet

Uploaded by

Copyright:

Available Formats

Maps of Greater Tibet

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

Maps of Greater Tibet

Uploaded by

Copyright:

Available Formats

15 · Maps of Greater Tibet

INTRODUCTION Tibetan maps are known to exist or are likely to be dis-

o 100 200 300 km

FIG, 15.1. REFERENCE MAP FOR GREATER TIBET.

lives of Tibetan Buddhists cannot be overemphasized. tains no reference to an indigenous work. 17

because of their appeal to tourists. Moreover, in surviving

the centuries, as have route maps intended especially for

sPa-gro Byang Tshang-pa:

here follow Aris.)

Not only were the layout and external appearance of

sources, sometimes as a consequence of Tibetan conquest

FIG. 15.10. THE POTALA OF LHASA. This detail from a nine-

FIG. 15.11. INDIGENOUS MAP OF DING-RI VALLEY,

can see in published works on Tibetan an are paintings,

tially two-dimensional, many forms of three-dimensional e.,.J.{-l.~ ~\, ~

of which embodies multiple meanings. For example, the S~

VIEWS FOCUSING ON THE TERRESTRIAL PLANE

cosmography, within the terrestrial plane of the Kamad-

scholar) and a tenth-century Tibetan recension of the

FIG. 15.15. MERU MANDALA, PAINTED ON A WALL OF

The second and third paintings in the sequence, from

round, for me it is round. Even if it doesn't exist at

game board, which may be construed as a map; and the

Although the number of known traditional geographic

Figures 15.27, 15.28, and 15.29 present a modern recen-

:"~"\\ <\ ~--

24. LAND OF THE

I 1. Pahanyan. Pahan-na? (Fergana); 2. Kin-hin (Kashmir); 3. Tha-ku-sha-si (Tak$llsilaITaxila):

the actual configuration of territorial control at any single

It is possible that the map was drawn in the early years

map. A similarly rendered skyline appears clearly at the

Tibetan Maps in the Wise Collection

Perhaps the most comprehensive set of relatively large-

missioned by the writer of the explanatory text."l40 Just

orienting pilgrims toward realizing the several objects of

1. Vidyiisvari; 2. Sobhii Bhagavat'; 3. Pulchok; 4. Patan; 5. Yarigal (southern

FIG. 15.44. KEY TO PART OF FIGURE 15.43. The numbered

schaftsverlag, 1987), 13-27, esp. fig. 2.

APPENDIX 15.1 TIBETAN MAPS IN THE WISE COLLECTION,

APPENDIX 15.2 TIBETAN MAPS IN THE HARRER COLLECTION,

GREATER TIBETAN LOCALITY MAPS

Medium Description Where Published c

z. Potala, Lhasa Potala, Lhasa Potala, Lhasa; date unknown None

Medium Description Where Published c

Painted on paper Oblique perspective, terrain presented in relatively

Medium Description Where Published c

Painted, on paper Essentially a frontal perspective Gerner, Architekturen im Himalaja, 85

You might also like