བོན་ · བོན་པོ་ · ཞང་ཞུང་Bon · Bonpo · Zhangzhung

这三个词属于不同层面。先把词义分开，才可能把信仰史、政治史与语言史重新接合。

བོན་：从“呼唤”到宗教名

Géza Uray 的文献考证把 bon 的可证古义限定为“请求、献供、低声诵念祷文咒语、呼唤与召唤”。Helmut Hoffmann 因而推测，它最初可能指以咒语召请神灵的行为。敦煌文书中的 bon 尚不是今日完整的教义制度；R. A. Stein 与 Samten Karmay 都要求区分古代仪式信仰与 10—11 世纪以后成形的雍仲本教。[1][2]

ཞང་ཞུང་：一个非藏语地名

Christopher Beckwith 认为 zhang zhung 没有藏语内部词源，藏文记录的是一个非藏语固有名；汉文“羊同”的中古音与其构拟相合。638 年《新唐书》所记“吐蕃率羊同共击吐谷浑”提供了政治史坐标。然而 Bettina Zeisler 质疑把“羊同”与“象雄”的疆域完全重合；严谨说法应是：汉文“羊同”通常被比定为象雄，但边界与对应范围仍有争论。[3][4]

本教宗教纪年与现代学术断代并不属于同一种证据体系，不能互相替代。

象雄不是一块边界固定、可直接套上现代地图的领土。传统地理、唐代史籍、藏文叙事与考古记录必须分层阅读。

三部象雄与圣地地理

本教传统把象雄分为外、里、中三部：外象雄从吉尔吉特、拉达克一带延伸至藏北；里象雄与 Tagzig 相连；中象雄地望未定。沃尔摩隆仁被描述为中心灵境，九重雍仲山常比定为冈仁波齐。这是宗教宇宙地理，不能直接等同政治疆域。

物质证据把讨论拉回地面

John Vincent Bellezza 自 1994 年起系统记录上部西藏的堡垒、巨石、墓葬与岩画，主张约公元前 5 世纪至公元 7 世纪存在复杂的灌溉、采矿、建筑与贸易网络，但 7 世纪前尚未发现当地文字系统或真正城市。[5]

卡尔东遗址调查显示规整居住区、公共／祭祀建筑、防御墙与地下通道；多方测年提示公元 2—5 世纪最活跃。附近故如甲木墓地在 2012—2014 年发掘 11 座墓葬，出土“王侯”铭文锦、黄金覆面与早期茶叶实物，说明阿里与中原、新疆及更广贸易网络相连。它能证明社会复杂度，却不能自动证明墓主就是某位象雄王。[5][6]

传播区

制度化本教后来分布于卫藏、康、安多以及尼泊尔多尔波和木斯塘、印度多兰吉与拉达克等地。Karmay 与长野泰彦主持的寺院调查被机构转引为西藏自治区约 300 座本教寺院，但这是机构统计而非人口普查。[7] 不丹与云南德钦的具体本教社区资料仍待核，不能凭地理邻近推定。

传统叙事建立神圣时间，文献学则追问文本何时出现、经谁整理、如何传播。二者可以同时陈述，但标签不能丢。

顿巴辛饶的十二事业

本教传统称顿巴辛饶出生于 Tagzig 中心的沃尔摩隆仁，完成弘法、婚配、降魔、出家与涅槃等十二事业。重要母题包括：以青稞面塑代替动物祭祀；一箭穿山开辟“箭道”；经藏焚毁后从灰烬显现五勇士字母并化出 368 字母。这些是宗教传记中的文明化、失落与复得叙事，不是考古事件。

三大传记与伏藏

《经集》（mDo’dus）与《孜美》（gZer mig）通常被视为 10—11 世纪伏藏；《光荣经》（mDo gzi brjid）由 14 世纪鲁登宁波口传而出。1017 年沈钦鲁嘎掘藏，是雍仲本教复兴的关键节点。《洞中宝藏》（mDzod phug）是唯一较长的象雄语—藏语双语典籍，也是判断象雄语真实存在的核心材料。[8]

本教九乘与宁玛九乘在递进结构上高度对应。Snellgrove 的《The Nine Ways of Bon》使其进入西方学术。结构相似是佛教影响论的重要证据，但具体文本与修法的影响方向必须逐案判断，不能一句“抄袭”或“完全本土”结束讨论。[9]

甘珠尔、丹珠尔与大圆满

本教甘珠尔分经、般若、续、宝藏四部；丹珠尔收律仪、注疏、密续、仪轨、护法、占卜与禅修。Kvaerne 推测甘珠尔约在 1450 年前后编成。象雄耳传、阿赤与仰则隆钦构成本教大圆满三大传承；现存耳传文本的实际编纂年代多在 10—13 世纪。

象雄语确实存在，但留存材料极少；“存在”不等于我们已经能完整复原它。

Dan Martin 的方法论警告尤其重要：后期象雄—藏文词表可能混入现代藏西方言，许多词的“象雄属性”在证实前都应保持怀疑。迄今也没有获得学界认可的出土象雄文字实物；传统所说玛钦文等五种文字与“李迷夏创制象雄文”，仍缺乏考古旁证。[3][8]

最好的解释并不是在“纯粹本土”与“佛教复制品”之间二选一，而是追踪观念、仪式与制度在不同年代的重组。

本土传统，还是佛教影响下形成？

Geoffrey Samuel 认为本教本质上是藏传佛教变体，却保留真实古老成分；David Snellgrove 称其为佛教的一种异端形式；John Powers 强调制度化宗教是在佛教影响下发展；Christopher Beckwith 更激进地否认帝国时期存在非佛教本教。Karmay、Kvaerne 的框架更能容纳复杂性：王朝时期有 bon 信仰与祭司，10—11 世纪出现受佛教深刻影响的制度化宗教，同时民间仪式传统继续存在。

为什么不宜称“萨满教”

早期西方著作常用“萨满教＋万物有灵”概括本教。Achard 2022 年专文批评这种做法：禳解、医卜与神谕并不能抹去本教的经论、戒律、辩经、寺院和大圆满哲学。今日学界趋向于承认藏地各传统都有某些“萨满型”实践，但反对把本教整体标签化。[10]

• 误解：《大唐西域记》记载象雄。 核查：玄奘未进入吐蕃／象雄，书中没有“象雄国”条目。
• 误解：本教是中国官方“第五大教派”。 核查：这是流亡藏人社群的宗教地位表达；未找到中国官方同名认定。
• 误解：“雍仲本教”已列入国家级非遗。 核查：未见项目证据；藏族天文历算是非遗，但不等于本教专属。
• 误解：卡尔东已经被证明就是穹隆银城。 核查：广泛比定，直接文字证据不足。
• 误解：象雄拥有“一万八千年文明、二十万军队”。 核查：无考古支持。

本教研究不是单一学派：文献学、语言学、考古与人类学彼此校正，才可能把传统叙事转化为可检验的问题。

20 世纪流亡重建使本教从地域宗教转为跨国传统，也改变了教法呈现方式。

1969 年起，印度多兰吉重建梅日寺；1978 年建立辩经学院，1986 年首届格西毕业。1987 年雍增丹增诺布在加德满都重建催丹诺布策。Tenzin Wangyal 1992 年在美国创办 Ligmincha，以居士、在线、多语种与梦瑜伽等路径传播；法国 Shenten Dargye Ling 则保留更完整的传统学修轨道，2005 年获法国宗教团体承认。[11][12]

这形成两条互补路线：一条将大圆满、梦瑜伽、五元素等转译为现代身心实践；另一条复制寺院课程、闭关与师徒传承。全球化扩大可及性，也提出新的研究问题：当仪式脱离地域、语言与寺院共同体之后，什么仍然构成本教身份？

The terms belong to different analytical levels. Keeping them apart is the first step toward reconnecting religious, political, and linguistic history.

བོན་: from an act of calling to the name of a religion

Géza Uray’s philological study restricted the securely attested early meanings of bon to asking, giving or offering, murmuring prayers and charms, and invoking or summoning. Helmut Hoffmann therefore conjectured that it once referred to calling gods through magical formulae. The bon encountered in imperial-period Dunhuang documents is not yet the fully institutional religion known today. Both R. A. Stein and Samten Karmay insist on distinguishing early rites and beliefs from the Yungdrung Bön that crystallized after the tenth century. [1][2]

ཞང་ཞུང་: a non-Tibetan place name

Christopher Beckwith argues that Zhangzhung has no internal Tibetan etymology and transcribes a non-Tibetan proper name. His reconstruction corresponds closely to the Middle Chinese pronunciation of “Yangtong,” the name found in Tang records. In 638, the New Book of Tang says Tibet led Yangtong in an attack on Tuyuhun. Bettina Zeisler, however, cautions against assuming that the territories denoted by Yangtong and Zhangzhung overlapped perfectly. The careful formulation is: Yangtong is generally identified with Zhangzhung, but the exact geographical equivalence remains debated. [3][4]

Bön religious chronology and modern historical dating are not competing measurements of the same evidence; they belong to different evidentiary systems.

Zhangzhung was not a territory with fixed borders ready to be transferred to a modern map. Sacred geography, Tang records, Tibetan narratives, and excavated material must be read on separate levels.

The three Zhangzhungs and the sacred center

Bön tradition divides Zhangzhung into outer, inner, and middle regions. Outer Zhangzhung stretches across Gilgit and Ladakh toward the northern plateau; Inner Zhangzhung is linked with Tagzig; the location of Middle Zhangzhung is unresolved. Olmo Lung Ring is the sacred center, and its nine-tiered Yungdrung mountain is often identified with Mount Kailash. This is religious cosmography, not a surveyed imperial boundary.

Material evidence

Since 1994, John Vincent Bellezza has documented forts, megaliths, cemeteries, and rock art across Upper Tibet. He argues that irrigation, mining, monumental building, and long-distance exchange supported socially complex communities from approximately the fifth century BCE through the seventh century CE, while noting that no local writing system or true city is yet known before the seventh century. [5]

Survey and limited excavation at Khardong revealed planned residential, public or ritual, defensive, and subterranean structures; independent dating programs indicate peak activity in the second to fifth centuries. At nearby Gurugyam, eleven graves excavated between 2012 and 2014 yielded a brocade bearing the characters “king and lords,” a gold face covering, and very early physical remains of tea. These finds demonstrate status and long-range connections, but they do not identify any occupant as a named Zhangzhung king. [5][6]

Later distribution

Institutional Bön eventually extended across Ü-Tsang, Kham, and Amdo and into Dolpo and Mustang in Nepal, Dolanji and Ladakh in India, and global diaspora centers. A monastery survey edited by Karmay and Yasuhiko Nagano is cited by Bön institutions for roughly 300 monasteries in the Tibet Autonomous Region; this is an institutional count, not a population census. [7] Specific claims about Bön communities in Bhutan and Deqin still require field verification.

Religious narrative creates sacred time; philology asks when a text appears, who transmitted it, and how it changed. Both can be reported, but their labels cannot be discarded.

The Twelve Deeds

Bön tradition locates Tönpa Shenrab’s life in Tagzig and Olmo Lung Ring and narrates twelve deeds—birth, teaching, marriage, subjugation, ordination, and nirvana among them. Memorable motifs include substituting barley-dough animals for blood sacrifice, shooting an arrow through a mountain to open a passage, and the emergence of five heroic letters and 368 letters from the ashes of a burned canon. These are sacred biographical accounts of reform, loss, and recovery—not archaeological events.

Biographies and treasure texts

The mDo ’dus and gZer mig are generally treated as tenth- or eleventh-century treasure revelations. The mDo gzi brjid emerged through the fourteenth-century oral transmission of Loden Nyingpo. The 1017 revelation by Shenchen Luga marks a key moment in Bön revival. The mDzod phug, the only extended Zhangzhung–Tibetan bilingual work, is central evidence for the historical existence of the language. [8]

The progression strongly resembles the Nyingma Nine Vehicles. Snellgrove’s The Nine Ways of Bon brought the system into Western scholarship. Structural resemblance is major evidence in arguments for Buddhist influence, yet direction and degree of borrowing must be established text by text rather than settled by slogans such as “copy” or “purely indigenous.” [9]

Canon and Dzogchen

The Bön Kanjur is arranged into sutra, perfection of wisdom, tantra, and treasury divisions; the Katen or Tengyur contains ethical, ritual, tantric, protective, divinatory, and contemplative commentaries. Kvaerne has proposed that the Kanjur was compiled around 1450. The Zhangzhung Oral Transmission, A-khrid, and Yangrtse Longchen are the three major Bön Dzogchen systems; extant oral-transmission texts were substantially compiled between the tenth and thirteenth centuries.

Zhangzhung was real, but its surviving corpus is tiny. Existence does not mean complete recoverability.

Martin’s methodological warning is crucial: later Zhangzhung–Tibetan vocabularies may contain modern western Tibetan dialect words misidentified as ancient survivals. Every item’s “Zhangzhung-ness” must be tested. No excavated inscription in a distinct Zhangzhung script has received scholarly acceptance; traditions of five scripts such as Smar chen, and claims that King Ligmi created a script, remain uncorroborated. [3][8]

The most productive model does not force a choice between “purely indigenous” and “a Buddhist copy.” It tracks how rites, doctrines, and institutions were reorganized over time.

Indigenous survival or Buddhist formation?

Geoffrey Samuel calls Bön essentially a form of Tibetan Buddhism that retains genuinely old elements. David Snellgrove described it as a heterodox form of Buddhism; John Powers stresses its development as a self-conscious system under Buddhist influence. Christopher Beckwith takes the more radical minority position that there is no evidence for a non-Buddhist Bön in the imperial period. Karmay and Kvaerne better accommodate the layered evidence: imperial bon beliefs and priests, a deeply Buddhist-influenced institutional religion emerging in the tenth and eleventh centuries, and continuing local ritual practices can all be historically real.

Why “shamanism” is inadequate

Older Western works often reduced Bön to “shamanism plus animism.” Achard’s 2022 critique shows why this fails: rites of divination, healing, and spirit mediation do not erase canon, ordination, dialectics, monasteries, or Dzogchen philosophy. Current scholarship may acknowledge “shamanic-type” practices across Tibetan religions while rejecting “shamanism” as a label for the whole Bön tradition. [10]

• Myth: Xuanzang’s Great Tang Records describes Zhangzhung. Finding: Xuanzang did not enter Tibet or Zhangzhung, and the book has no Zhangzhung entry.
• Myth: Khardong has been proven to be Kyunglung. Finding: the identification is influential but lacks direct epigraphic proof.
• Myth: Zhangzhung had an 18,000-year civilization and an army of 200,000. Finding: no archaeological evidence supports these figures.
• Myth: Yungdrung Bön is listed as a Chinese national intangible heritage item. Finding: no such listing was located; Tibetan calendrical calculation is listed, but is not exclusively Bön.

Bön studies is not a single school. Philology, linguistics, archaeology, and anthropology correct one another and turn inherited stories into testable questions.

Exile reconstruction turned a regional religion into a transnational tradition and changed the media through which it presents itself.

Menri was re-established at Dolanji, India, from 1969; a dialectics college opened in 1978, and the first geshes graduated in 1986. Yongdzin Tenzin Namdak re-established Triten Norbutse in Kathmandu in 1987. Tenzin Wangyal founded Ligmincha in the United States in 1992, emphasizing lay, online, multilingual forms and practices such as dream yoga. Shenten Dargye Ling in France preserves a more traditional study-and-retreat track and received French recognition as a religious congregation in 2005. [11][12]

The result is a dual pathway: one translates Dzogchen, dream yoga, and the five elements into modern contemplative language; the other recreates monastic curricula, retreats, and teacher–student transmission. Globalization expands access while raising a new research question: when practices leave their regional, linguistic, and monastic setting, which elements continue to define Bön identity?