Tibetan Renaissance Seminar > Student Generated Content > Concept Encyclopedia Analyses > Maṇḍala

The Maṇḍala

by Christopher Bell

The Maṇḍala as Cosmogram

The maṇḍala, Sanskrit for "circle," is one of the most powerful religious symbols in Buddhism. In Tibetan the word is kyilkhor (དཀྱིལ་འཁོར་), signifying the center of a circle as well as its diameter. It has its origins in early Vedic rituals and is still recognized for its symbolic wealth in Hinduism. In Buddhism, particularly esoteric Buddhism, the symbolism and significance of the maṇḍala has flourished. The maṇḍala is significant first and foremost because it is the personification of setting. The importance of setting is true with all Tibetan Buddhist deities. These deities are potentially in all places; this ubiquitous nature is symbolized by the maṇḍala, which exists everywhere, in all lands both pure and profane, and ultimately in the minds of the buddhas, which maṇḍalas equally represent.

As its etymology indicates, a maṇḍala is a circle that represents the cosmogram of a buddha or bodhisattva. The central deity can either represent a practitioner’s tutelary deity or anyone from the panoply of Buddhist figures and protectors whom one wishes to propitiate. As a circle, the outer ring of the maṇḍala encompasses a smaller square that represents the palace of the central deity. Four ornate gateways (Skt. toraṇa) mark the entrances of the four cardinal directions. In most maṇḍalas, this main circle and square are elaborated with further concentric layers. Within the palace that is the square, another circle is situated that is then divided into nine sectors. These sectors either repeat the pattern on a smaller scale or house the central deity within the central sector. In the latter case, the surrounding sectors house manifestations of the central deity, representing related Buddhist gods and goddesses. Furthermore, the overall pattern can be contained within a larger square with smaller repeated maṇḍalas at each of the four corners. This is the most popular pattern for the maṇḍala, though there are numerous variations. Within the intermediate spaces there tend to be other buddhas and bodhisattvas or various ornate gardens and animals, and an array of important Buddhist symbols.

David Snellgrove provides the most succinct examination of the maṇḍala, stating that it is a metaphysical representation of the mind and pure realm of the central Buddhist deity. This Buddhist deity belongs to a five-part buddha family, with the key deity at the center, the four other buddhas in the immediate surroundings facing all four directions, and the consorts of these four buddhas residing in the intermediate quadrants between the cardinal directions. Other bodhisattvas representing the retinues of each of these buddhas are depicted further outward. Central to Snellgrove’s interpretation is that the maṇḍala is a sacred space in contrast to and separate from the profane space around it.1 However, Ronald Davidson advances the understanding of the maṇḍala not only by placing its development within the historical context of medieval India but also by stressing its importance as a symbol of ultimate place. Davidson theorizes that the maṇḍala took on greater political significance during the sixth and seventh century in North India as complex feudal relationships developed between various kingships. The fluctuating political alliances at this time were represented by maṇḍalas. The king of a particular dominion stood as the central figure of the maṇḍala. The vassal lords of the surrounding kingships were represented by the immediate kingly figures facing the four directions. These were surrounded by the lesser vassal dominions represented by the broader entourages. This relationship can subsequently change, with previous vassal lords gaining power due to the shifting political climate and thus becoming the central king of the political maṇḍala. This pattern, which Davidson labels "sāmanta-feudalism,"2 is a pragmatic design that ensures that the various kings of the greater domains do not actually come into direct conflict with each other. This system maintains a respectful diplomatic distance, leaving any disputes exclusively within the border regions symbolized by the various entourages.3

The sāmanta-feudalism system is not wholly dissimilar from the "galactic polity" model developed by Tambiah in relationship to Theravādin Buddhist states in Southeast Asia.4 This model was later utilized by Samuel in order to define the decentralized and changing political systems common throughout Tibetan history after the fall of the Yarlung Dynasty.5 Samuel even describes this model as a "maṇḍala-type structure."6 This system also lacks a constant center, though once a temporary center is established, it is recognized as exemplary in relation to neighboring dominions. This shift from central to subordinate status can occur without any drastic change to the identities of the interrelated domains. However, there is one major distinction between the models. While the sāmanta-feudalism model is primarily concerned with land—its acquisition and its tenure under vassal states—the galactic polity model in the context of Tibet is primarily concerned with people. Labor force has been a constant concern of the major political administrations in Tibetan history. Land was abundantly available in Tibet, yet the general population was low by comparison; thus, labor was a greater commodity.7 Davidson does not make this contrast, nor does he seem to be aware of the galactic polity model.

Regarding these various meanings of the maṇḍala, Snellgrove focuses strictly on its religious purpose and Davidson provides a sociopolitical perspective. Davidson considers his research to work contrariwise to Snellgrove’s in one respect. Snellgrove’s religious definition of the maṇḍala functions from an archetypal, metaphysical realm downward to influence the monastic and political institutions of Buddhism. By contrast, Davidson believes that it is rather the structure of North Indian political institutions that provided a ready model to comprehend the metaphysical realms of the buddhas. These buddhas in turn represented the spiritual kings of all existence in contrast with the temporal kings of the various Indian states. Furthermore, Davidson elaborates on the understanding of the maṇḍala in both its religious and political milieus by suggesting the English word "state" as a fitting definition. This combines the term’s meaning as a physical location encompassed by a political institution and as a mental condition.8

The Maṇḍala and Tibetan Deities

To relate this brief survey of the maṇḍala to protector deities, I would like to look more closely at the mythographic nature of the Tibetan landscape. The language and myth of Tibet abounds with the concepts of taming and subjugation. This taming, or disciplining, is most pragmatically that of the initiate who must be disciplined in their tantric practice by their teacher. However, culturally, the concept relates back to the era of Padmasambhava, who tamed the various demonic deities of Tibet as part of the effort to establish Buddhism. Still further back through the curtains of legend, there is the first Buddhist king Songtsen Gampo. This king subdued the giant demoness who represented the entire land of Tibet as a wild and unpredictable environment. Songtsen Gampo, impelled by his two queens, constructed thirteen Buddhist temples throughout Tibet in order to pin down this wild demoness and help firmly establish Buddhism, which was under constant attack by such demonic influences. These temples were constructed in concentric circles radiating out from Lhasa at the center. I accept Robert Miller’s argument for fashioning the diagram of these temples as concentric circles rather than the concentric squares model first volunteered by R. A. Stein.9 The places where these temples were built geographically signified a part of the demoness’s body. Thus, the first concentric circle out from the center consisted of four temples that pinned down the demoness’s right and left shoulders, and right and left hips. The second circle of temples further out pinned her right and left elbows, and right and left knees. The final circle out pinned her right and left palms, and right and left feet. Finally, her heart at the center in Lhasa consisted of a lake that symbolized her life blood. This lake was filled in and the Jokhang temple was built on top. This completed the complex act of subjugating the demoness of the land in order to tame the wilderness and thus properly establish Buddhism.10

I relate this myth because it is clear from its description that a maṇḍala has been symbolically overlaid on the Tibetan landscape in order to subjugate the fierce indigenous deities symbolized wholly by the demoness. The maṇḍala manifests as the ritual act of purifying the land to further Buddhist goals. This, then, can be read as a physical act of redefining the mythographic landscape of Tibet along a Buddhist narrative parallel to the symbolic reenvisioning of the Tibetan past through a Buddhist lens, as discussed in chapter 2. Significantly, such a symbolic reenvisioning through temple construction can also be found in the architecture of Samyé monastery, intentionally structured like a maṇḍala. Thus, the maṇḍala becomes a method of control, superimposing the order of a Buddhist universe on a hostile land. On a smaller scale, the maṇḍala is initially used to establish the abode of a deity before continuing with a ritual program. With the maṇḍalas of tutelary deity, the ultimate ritual goal is to become fully associated with that deity as an expedient means toward enlightenment. In the case of otherworldly deities, the goal is to establish their presence at the site of the ritual by first drawing, in many cases, the maṇḍala of Tamdrin. To give a vivid example, the first chapter of The Warlord’s Tantra with Accompanying Sādhanas [dmag dpon gyi rgyud sgrub thabs dang bcas pa bzhugs so], sets the scene by visually situating the reader within Tamdrin’s maṇḍala: "In the pure realm of the Unsurpassed that reveals magical emanations; in an expanse of a blazing flame heap that represents malicious anger, amid an expanse of turbulent waves that represent lust; within a maṇḍala that is a blazing dark-red triangle, the blessed one, glorious king Tamdrin himself [recites] the mantra to cultivate meditative stabilization, which subjugates the three realms."11 As discussed in the introduction, Tamdrin is a wrathful manifestation of the bodhisattva Avalokiteśvara; as such, he is one of the highest of Buddhist deities. Tamdrin also has the ambiguous position of being closely related to the worldly protector deities whom he must constantly keep in check. It is therefore understandable that one can only ritually access these deities by first propitiating and possibly embodying Tamdrin. The initial act, however, is first to establish the maṇḍala as the abode of these deities and, through this, contact them directly.

The Maṇḍala in The Blue Annals

In The Blue Annals, the maṇḍala has a very prominent presence, underlying every major discussion of religious practice. This is not difficult to understand, since maṇḍala rites are an integral part of initiation into tantric lineages. There are innumerable instances throughout The Blue Annals where important religious figures compose and transmit maṇḍala ritual texts (dkyil chog), practice such rituals, and experience intense spiritual visions of maṇḍalas or even transform into maṇḍalic fields. The context for such rites and experiences concern many notable tantric lineages and the gods with their retinues that reside in such complex maṇḍalas. These lineages include the Guhyasamāja (gsang ba 'dus pa), the Hevajra (kye rdo rje), the Cakrasaṁvara ('khor lo bde mchog), the Kālacakra (dus kyi 'khor lo), the Māyājāla (sgyu 'phrul drwa ba), and several other minor traditions. Indeed, a tantric tradition cannot flourish without including a maṇḍalic transmission. A vivid example of this is when prominent figures of the Khön ('khon) clan, the forebearers of the Sakya (sa skya) sect, adopted the Path and Fruit (lam 'bras) system. The Path and Fruit did not come equipt with an explicit maṇḍala system, wich is so necessary for initiation and generation stage (bskyed rim; Skt. utpattikrama) yoga; to remedy this, the Path and Fruit was allowed to be associated with either the Hevajra or Cakrasaṁvara systems.12 No doubt, this was a necessary step for the lineage—now complete with either of two rich maṇḍalic systems—to become so successful.

As mediums of exchange, there are a number of instances in The Blue Annals where maṇḍalas are indicative of more underlying conflicts, usually between different tantric traditions. Two instances stand out as both didactic and informative. First, after composing the Vajrāvali (rdo rje phreng ba), the 12th-century Indian master Abhayākaragupta set about propagating it and its maṇḍala system irrespective of different yogic systems. This caused consternation for some who felt the practice should be divided along different tantric and ritual lines according to a textual collection of sādhanas also propagated by Abhayākaragupta, the Ocean of Sadhānas (sgrub thabs rgya mtsho). These disputing voices felt that several teachers, sadhānas, and maṇḍala systems were necessary for proper propitiations:

Now, the ācārya Abhaya composed a maṇḍala rite belonging to the Sampaṇakrama yoga, basing himself on the Tantric text (Guhyasamāja) which said: 'One should know the three classes of Yoga, that of blessing, that of imagination (yongs brtags), and that of the complete manifestation of form.' There were many who used to think: 'If one would not propitiate according to the methods of different sādhanas, preached by various teachers, and (expounded) in different Tantras, the propitiations would not be complete.' If so, a fully enlightened (Buddha) cannot propitiate even a single maṇḍala, because he is free of constructive thought. Can you avoid this contradiction! Such people should be initiated according to different rites, and not according to the Vajrivali belonging to the system of Abhaya. Similarly also one should not speak slightingly of those who bestowed initiations according to the Ocean of Sadhānas (sgrub thabs rgya mtsho), transmitted through Abhayākaragupta, Puṇyakaragupta (dge ba'i 'byung gnas sbas pa), the siddha glong zhabs, the paṇḍita Kīrticandra, and the lo tsa ba grags pa rgyal mtshan, because one is unable to establish whether one is fit or not to enter into this great mystic sphere (dkyil 'khor chen po). (Roerich 1996, p. 1048)

As illustrated above, the resolution was to leave such people to their divisions, which were necessary for some, while leaving the more holistic system of the Vajrāvali to more responsible adepts. This is a common approach to resolution in Tibetan textual doxography, where systems become divided along practices to suit the needs of particular religious constituencies.

Second, in the process of questioning masters, it is a common motif for practitioners to initially be confused by the cryptic responses of such masters before coming to the epiphany of realization. With the following example, not only is the primacy of the maṇḍalified worldview of the Māyājāla questioned, but there is an added conflict of disciples desiring to change teachers, knowing that the result could be unpleasant from the social perspective of respect:

Having deputed the four, the four went into the presence of lha rje [zur chung shes rab grags pa; 1014-1074]. At first skyo ston shAk yes inquired: 'lha rje chen po lags! Do you hold in high esteem the method of meditation of the school of the Great Perfection (rdzogs pa chen po) only?' The latter replied: 'ls my mind attached to any particular object?' – 'Are you not concentrating your mind on the Great Perfection?' lha rje chen po said: 'Why should I be disturbed?' The interrogator feeling powerless, remained speechless. After that glang ston shAk bzangs inquired: 'Are you not maintaining that all visual objects exist as the sphere (maṇḍala) of gods and goddesses, as stated in the system of gsang ba sgyu 'phrul?' lha rje then replied: 'Who will deny the validity of the proof of direct sense perception of the visual objects as independent material bodies?' – 'Well then, are you not maintaining this?' – 'Who will be able to contradict it, as it had been deducted from many teachings, the Sūtras and Tantras, in order to remove the wrong illusion characteristic of the living beings regarding the independent existence of visual object?' Again this remained without answer. ln this manner the four put questions (to him) in turn, but he answered them by remaining silent. Then they said: 'Nowhere did we find such a kalyāṇa-mitra possessed of the incontrovertible understanding of the doctrine of the Mahāyāna, endowed with the understanding of emancipation and science! If we suddenly become his followers, our teacher will be displeased.' Therefore they made a solemn promise that next year they were to leave their teacher and honor zur chung ba. Next year they came to lha rje (zur chung), and these became the 'Four Pillars.' (Roerich 1996, p. 119-120)

No doubt such concerns regarding the nature of the maṇḍala in notable tantric systems is a common source of conflict in the evolution of tantric systems.

One final example is not so much indicative of conflict overtly but rather of ambiguous categorization. While most instances of maṇḍalas in The Blue Annals can be categorized under one of three major axioms, text, practice, and experience, there are certain instances where an event can encompass all three:

On one occasion he [glang lung pa; 1123-1193] read through the large commentary on the rtsa ltung (rtsa ba'i ltung ba'i rgya cher 'grel pa), and by the power of his faith (in that book), he saw for six days his own body as a cakra maṇḍala. (Roerich 1996, p. 298)

Here, we have a religious practitioner engaging with a text, no doubt involving himself in related practices, and as a result he has a visionary and transformative experience. This indicates that texts (and not just ritual texts) and bodies alike are potential vehicles for the unfolding transformative power of maṇḍalas. See the Appendix below for a chronological enumeration of the appearances of the maṇḍala in The Blue Annals, organized along the three axioms given.

Let us recall that maṇḍalas are cosmic representations of land, which, as discussed above, very commonly come to imbue the land of Tibet on multiple levels. Therefore, another potential source of conflict is with land acquisition and disputation, as well as hegemonic shift, events that fill Tibetan history. Since there are numerous maṇḍalas, numerous tantric systems utilizing these maṇḍalas, and numerous "clans" of gods populating these maṇḍalas—many of whom have strong ties to specific localities in Tibet—the use of maṇḍalas in Tibetan religious history vividly illustrates a constant negotiation and renegotiation of land, tantric systems, religious histories, sacred paradigms, political control, and population. All of these elements come together in particularly powerful confluences during the key moments of Tibetan history, such as the consolidation of the Tibetan empire in the 7th century, the re-appropriation of Tibet under the rule of the Sakya in the 13th century, and the reunification of Tibet under the 5th Dalai Lama in the 17th century. In The Blue Annals and indeed all Tibetan religious literature, maṇḍalas are an integral mythographic element of initiation, transformation, soteriology, exchange, negotion, and acquisition, all of which evolve along several fluctuating textual paradigms and lineages of practice. If there is any single concept that completely permeates the Tibetan religious, cultural, and historical landscape, the maṇḍala is it.


Chronological References to Maṇḍalas in The Blue Annals

This chronology is not exhaustive but indicates the most prominent examples in The Blue Annals (Roerich 1996) in which the maṇḍala appears. This chronology is divided along the three contexts in which maṇḍalas are most prevalently discussed: textual composition or transmission, rites and practices, and visionary or transformative experiences. There is one unique instance where the Tibetan word for maṇḍala, "dkyil 'khor" is listed as part of a name, and this is listed last.

Composed or Transmitted Ritual Texts: pp. 107, 172, 356, 358, 370, 371, 385-386, 389-390, 394, 663, 1046.

Rites and Practices: pp. 109, 118, 246, 256, 250-251, 287, 770, 1048.

Visions and Transformative Experiences: pp. 104, 119-120, 128, 250, 298, 369.

As Part of a Name: p. 805.


Blondeau, Anne-Marie and Yonten Gyatso. 2003. "Lhasa, Legend and History." In Lhasa in the Seventeenth Century: The Capital of the Dalai Lamas. Françoise Pommaret, ed. Leiden: Brill, pp. 15-38.

Davidson, Ronald. 2002. Indian Esoteric Buddhism: A Social History of the Tantric Movement. New York: Columbia University Press.

—. 2005. Tibetan Renaissance: Tantric Buddhism in the Rebirth of Tibetan Culture. New York: Columbia University Press.

Gyatso, Janet. 1987. "Down with the Demoness: Reflections on a Feminine Ground in Tibet." In The Tibet Journal 12(4), pp. 38-53.

Marko, Ann. 2003. "Civilising Woman the Demon: A Tibetan Myth of State." In History of Tibet, vol. 1. Alex McKay, ed. London: RoutledgeCurzon, pp. 322-335.

Miller, Robert. 2003. "‘The Supine Demoness’ (Srin mo) and the Consolidation of Empire." In History of Tibet, vol. 1. Alex McKay, ed. London: RoutledgeCurzon, pp. 336-353.

MPG, abbr. Ngari Paṇchen Padma Wangyel (Mnga’ ris Paṇ chen Padma dbang rgyal; 1487-1542). 19th century. The Warlord’s Tantra with Accompanying Sādhanas (dmag dpon gyi rgyud sgrub thabs dang bcas pa bzhugs so). In The Great Treasury of Termas (rin chen gter mdzod chen mo), vol. 62. Jamgön Kongtrül Lodrö Tayé (’Jam mgon Kong sprul blo gros mtha’ yas; 1813-1899), pp. 299-332.

Roerich, George. [1949] 1996. The Blue Annals. Delhi: Motilal Banarsidass Publishers.

Samuel, Geoffrey. 1993. Civilized Shamans. Washington: Smithsonian Institution Press.

Snellgrove, David. 2002. Indo-Tibetan Buddhism: Indian Buddhists and Their Tibetan Successors. Boston: Shambhala Publications.

Stein, R. A. 1972. Tibetan Civilization. Stanford: Stanford University Press.

Tambiah, Stanley. 1976. World Conqueror and World Renouncer: A Study of Buddhism and Polity in Thailand against a Historical Background. Cambridge: Cambridge University Press.


1 Snellgrove 2002, pp. 198-213.
2 The term "sāmanta-feudalism," as Davidson duly notes, was actually coined by Brajadulal Chattopadhyaya. This is a distinct term that holds the initial definition of feudalism as signifying a specific political arrangement that entails land tenure and authority without the necessary ideological baggage that has come to surround the term. Adding the Sanskrit word sāmanta, meaning "neighbor," adds a complex Indian understanding signified by the maṇḍala structure.
3 See Davidson 2002, pp. 131-144.
4 See Tambiah 1976, pp. 102-131.
5 See Samuel 1993, pp. 61-63.
6 Samuel 1993, p. 62.
7 See Samuel 1993, p. 62.
8 See Davidson 2002, p.131-132.
9 See Miller 2003 and Stein 1972, p. 39.
10 For further references on the "supine demoness," see Gyatso 1987, who provides a survey of this myth in multiple contexts, and Miller 2003, who discusses this legend in the context of the consolidation of the Tibetan empire. For a feminist interpretation, see Marko 2003. See also Blondeau and Gyatso 2003.
11 MPG, p. 300.2: ’og min cho ’phrul bstan pa’i zhing khams na / zhe sdang gdug pa me dpung ’bar ba'i klong / rakta chags pa’i rba klong ’khrugs pa’i dkyil / dmar nag gru gsum ’bar ba’i dkyil ’khor na / bcom ldan ’das dpal rta mgrin rgyal po nyid / khams gsum dbang du sdud pa’i ting nge ’dzin bsgoms par sngags so.
12 See Davidson 2005, p. 191.