Did Wu Zetian Name “卍” as “Wanzi”? A Historical Reassessment

Abstract

While scholarly works often attribute the pronunciation of “卍” as “wan” to Empress Wu Zetian in 693, associating it with the meaning “auspicious myriad virtues”, a closer examination of the history of “卍” in Chinese Buddhist translations suggests otherwise. The more accurate transliterations and translations of svastika emerged much later than the term “Wanzi” and had very limited influence. The connection between “卍” and “Wanzi” more likely appeared during the early transmission of Buddhism to China, when people used the accepted cursive form of “萬” to approximate the shape of the svastika symbol. However, as this rationale gradually became obsolete over time, the legend that “Empress Wu Zetian decreed that ‘卍’ be pronounced as ‘wan’” arose during the Song dynasty and has persisted to this day.

1. Introduction

The auspicious symbol svastika (“卍” or “卐”) is frequently depicted in Buddhist scriptures, as well as in various artistic forms such as paintings, sculptures, and decorative patterns, representing one of the “Thirty-Two Marks” or “Eighty Minor Characteristics” of the Buddha.1 In Chinese, this symbol is referred to as “Wanzi 萬字” (“the character of ten thousand”). It can be considered a variant form of the Chinese character “萬”. Phonetic variants of “Wanzi”, such as the Japanese “まんじ” (manji) and the Korean “만자” (manja), have become the recognized names for the svastika symbol in these respective Asian languages.2

Given that the Chinese character “萬” (ten thousand) bears no intrinsic semantic connection to the svastika, numerous dictionaries and scholarly works attempt to explain this association in different ways. These sources frequently reference Fayun’s 法雲 (1088–1158) Fanyi mingyi ji 翻譯名義集 [Collection of Explanations of Translation Terms] (compiled in 1143), which cites the Huayan yinyi 華嚴音義 [The Sound and the Meaning of the Avataṃsaka Sūtra] to suggest that this symbol was first officially recognized as a Chinese character during the second year of the Changshou 長壽 era (693 CE) of Empress Wu Zetian 武則天 (624–705 CE), who established its pronunciation as “wan”:3

(1) … 華嚴音義云：”案卍字，本非是字。大周長壽二年主上權制此文，著於天樞，音之為萬，謂吉祥萬德之所集也。經中上下據漢本總一十七字，同呼為萬，依梵文有二十八相云云。

4

… The Huayan yinyi states: “The character ‘卍’ was not originally a (Chinese) character. In the second year of the Changshou era during the Great Zhou Dynasty, Empress (Wu Zetian) provisionally established it as a (Chinese) character, inscribing it on the Heavenly Pivot and assigning it the pronunciation ‘Wan’ to signify the gathering of auspicious myriad virtues. In the Chinese translations of the scriptures, this symbol is consistently referred to as ‘Wan’, with seventeen instances in total across the texts. According to the Sanskrit text, there are twenty-eight distinct marks” and so on.

The claim that Empress Wu Zetian established the pronunciation of “卍” as “Wan” had gained widespread influence since then, but we should not accept this account without caution. A thorough review of the sources cited by the Fanyi mingyi ji—specifically Huiyuan’s 慧苑 (673–743 CE) Xinyi da fangguang fo huayan jing yinyi 新譯大方廣佛華嚴經音義 [The Sound and the Meaning of the New Translation of the Mahāvaipulya Buddhāvataṃsaka Sūtra] (also known as Huiyuan yinyi 慧苑音義 [Huiyuan’s Dictionary] or Huayan yinyi 華嚴音義, compiled around 732 CE5)—reveals certain discrepancies. In his explanation of the phrase “卐字之形” (the form of the “卐” symbol) as it appears in Śikṣānanda’s 實叉難陀 (652–710 CE) translation of the T279 Da fangguang fo huayan jing 大方廣佛華嚴經 [Mahāvaipulya Buddhāvataṃsaka Sūtra], Huiyuan clarified that the “卍” symbol was originally not a Chinese character. He further indicated that this term in the T279 corresponded to various symbols in the Sanskrit Avataṃsaka Sūtra. When comparing the extant version of Huiyuan yinyi with its citation in the Fanyi mingyi ji, it becomes evident that Fayun’s reference is directly drawn from Huiyuan’s work. This is apparent both from the title “Huayan yinyi” and from the closely aligned phrases, such as 經中上下據漢本總一十七字同呼爲萬 (“in the Chinese translations of the scriptures, this symbol is consistently referred to as ‘Wan’, with seventeen instances in total across the texts”.) which are notably consistent between the two texts. However, one critical discrepancy warrants attention: the passage in Fayun’s text regarding Wu Zetian is entirely absent in Huiyuan’s work:

(2)【卐字之形】今按梵本，卐字乃是德者之相，元非字也。然經中上下據漢本總一十七字同呼爲万，依梵文有二十八相，即八種相中四種相也。謂室利靺瑳、難提迦物多、塞嚩悉底迦、本囊伽咤，又有鉢特忙、斫訖羅、拔折羅等三相。雖於《花藏》《迴向》二品中有，以其可識無謬，故此不列在數。又有盎句奢相，此經總無，故亦不列。其一十七相既非萬字，又非一色之相，今顯異同，謂第八卷有一室利靺瑳相，第九卷有三相：初難提迦物多，次室利靺瑳，後亦室利靺瑳。第二十三有一相，謂塞嚩悉底迦，第二十七有五種相：初室利靺瑳，次塞嚩悉底迦，次難提迦物多，次室利靺瑳，後難提迦物多。第四十八有三相：一塞嚩悉底迦相，二室利靺瑳，三室利靺瑳。第五十七、五十八、六十三、六十五等各有一室利靺瑳相。若謹依梵本，總有二十八相，具顯如《刊定記》說也。6

[The form of the “卐” symbol]: According to the Sanskrit originals, the “卐” is a mark signifying virtue and is not inherently a character. However, in the Chinese translations of the scriptures, this symbol is consistently referred to as “Wan”, with seventeen instances in total across the texts. According to the Sanskrit text, there are twenty-eight distinct marks, of which four correspond to the eight types of auspicious symbols. These include Śrīvatsa, Nandikāvarta,7 Svastika, and Pūrṇaghaṭa, among others. Additionally, there are three other marks: Padma, Cakra, and Vajra. Although the Huazang [Flower Treasury] and Huixiang [Dedication of Merit] chapters include some of these marks, they are not listed here to avoid confusion as they are clearly identifiable. The Aṅkuśa mark, which does not appear in this sutra, is also omitted. As for the seventeen marks mentioned in the Chinese text, they are neither “Wan” characters nor a single type of mark. The differences are as follows: The eighth volume mentions one Śrīvatsa mark; the ninth volume contains three marks: first, Nandikāvarta; second, Śrīvatsa; and lastly, Śrīvatsa again. The twenty-third volume has one mark, identified as Svastika. The twenty-seventh volume contains five marks: first, Śrīvatsa; second, Svastika; third, Nandikāvarta; fourth, Śrīvatsa; and finally, Nandikāvarta again. The forty-eighth volume mentions three marks: first, a Svastika mark; and then two instances of Śrīvatsa. Volumes fifty-seven, fifty-eight, sixty-three, and sixty-five each contain one Śrīvatsa mark. According to the Sanskrit originals, there are twenty-eight marks in total, as clearly detailed in the Kandinji [Supplementary Record of the Commentary on the Avataṃsaka Sūtra].

The initial question we must address is whether the additional references to Empress Wu Zetian in the Fanyi mingyi ji were introduced by Fayun during his compilation or whether they reflect content that has been lost from the extant Huayan yinyi by Huiyuan. A thorough investigation of all instances where the Fanyi mingyi ji cites the Yinyi works of Xuanying 玄應 (active 645–c. 661/663 CE8), Huilin 慧琳 (737–820 CE), and Huiyuan 慧苑 reveals that Fayun consistently abridged the original content, often summarizing rather than strictly quoting the texts.9 This pattern casts significant doubt on the authenticity of the passage concerning Wu Zetian. It seems only plausible if this narrative did not originate from Huiyuan yinyi, but was instead introduced by Fayun from another source—perhaps an annotation or commentary that was erroneously incorporated into the main text.

Further evidence supporting this hypothesis lies in a subtle wording difference between the two texts. In Huiyuan yinyi, the statements “the ‘卐’ symbol was originally not a Chinese character” and “in the Chinese translations of the scriptures, this symbol is consistently referred to as ‘Wan, ’ with seventeen instances in total” are connected by the conjunction “however” (然), indicating a clear semantic relationship (…元非字也，然經中上下據漢本總一十七字同呼爲萬…). In contrast, Fayun’s version omits this conjunction after inserting the narrative about Wu Zetian, likely to avoid a grammatical inconsistency introduced by the interpolation. This detail suggests that the reference to Wu Zetian was not part of Huiyuan’s original text but rather an addition by Fayun.

More direct evidence can be found in Huiyuan’s commentary on the newly translated Avataṃsaka Sūtra, titled Xu huayan jing lüe shu kandingji 續華嚴略疏刊定記 [Supplementary Record of the Commentary on the Avataṃsaka Sūtra] (referred to as the Kandingji at the end of example 2). In this work, Huiyuan explicitly stated that the designation of the symbol “卍” as the “character of ten thousand” (萬字) arose from an error made by ancient translators who misunderstood the reading and meaning of the Sanskrit term. Notably, he made no mention of any role played by Empress Wu Zetian in this interpretation:

(3) 其相梵名室利靺瑳，此曰吉祥海雲。眾德深廣，利物如雲，海雲即吉祥，吉祥即海雲。古人誤譯“洛剎曩”為“惡剎攞”，遂謂相為字，又以相表利物深廣多故，稱之為萬。理實此中無萬無字，唯云“金剛莊嚴臆室利靺瑳相”，其“萬字”之言都為謬妄爾。

10

This symbol is called Śrīvatsa, which in Chinese means “Auspicious Ocean Cloud”, [because] its numerous virtues are profound and vast, benefiting all beings like clouds. The “ocean cloud” is auspicious, and auspiciousness is the “ocean cloud”. In ancient times, translators mistakenly read lakṣaṇa as akṣara, leading them to refer to this symbol as a “character” (字). Furthermore, since this pattern represents profound and widespread benefits to all beings, they called it “ten thousand” (萬). However, from a logical standpoint, this is neither “ten thousand” nor a “character”. It is simply the “vajra-adorned mark of Śrīvatsa on the chest”. The claim that it represents the “ten thousand character” is entirely mistaken.

Huiyuan was a disciple of Fazang 法藏 (643–712), who was directly involved in the new translation of the Avataṃsaka Sūtra. If the designation of the symbol “卍” as a Chinese character, read as Wan, had indeed originated from an edict by Empress Wu Zetian, it is hard to imagine that he would have rendered such a straightforward and unequivocal conclusion. Moreover, Huiyuan’s direct critic, Chengguan 澄觀 (738–839), explicitly referenced Huiyuan’s commentary (example 3) in his Da fangguang fo huanyan jing shu 大方廣佛華嚴經疏 [Commentary on the Avataṃsaka Sūtra], yet he offered no criticism regarding this particular issue.11

Another piece of evidence can be found in the slightly later work Yiqiejing yinyi 一切經音義 [The Sound and the Meaning of All Scriptures] by Huilin 慧琳 (also known as Huilin yinyi 慧琳音義 [Huilin’s dictionary], compiled before 808 CE12), which incorporates Huiyuan’s annotations on T279 and briefly references Huiyuan’s explanation in his own annotation on the term “卍字之文” in Volume 20 of the T310 Da baoji jing大寶積經 [Mahāratnakūṭa Sūtra]. In his annotations, Huilin not only omitted any mention of Wu Zetian’s alleged decree that “卍” be recognized as a Chinese character “Wan”, but he also explicitly refuted this claim, further claiming that referring to “卍” as “Wan” was erroneous:

(4) 卍字之文：梵云“室哩(二合)末蹉(倉何反)”。唐云吉祥相也。有云萬字者，謬說也。《花嚴經》第八卷中具說此相等亦非是字也，乃是如來身上數處有此吉祥之文、大福德之相。

13

The symbol “卍”: In Sanskrit, it is pronounced ‘Śrīvatsa’. In the Tang language, it is referred to as an auspicious sign. Some have claimed that it is the character “Wan”, but this is an incorrect statement. (In Huayan yinyi’s annotation of) the eighth volume of the Avataṃsaka Sūtra, it is clarified that the symbol is not a character, but rather an auspicious mark found in various places on the body of the Tathāgata, signifying great merit and virtue.

The claim that Wu Zetian decreed the pronunciation of “卍” as “Wan” is largely absent from Tang dynasty historical records and seems to have originated predominantly from Song dynasty sources,14 casting doubt on its credibility. Thus, the precise timing and underlying reasons for the designation of “卍” as “Wanzi” requires further rigorous scrutiny and analysis.

2. The Initial Naming of “卍” in Chinese Translations

From the chronological perspective of Chinese Buddhist translations, the use of the character “萬” (or “万”) to represent the “卍” symbol on the Buddha’s body appears quite early. This practice can be traced back to the translations of Zhi Qian 支謙 (fl. 222–252 CE). The following are some examples from relatively early translations:

(5) 披褻相太子，見有三十二相：軀體金色，頂有肉髻，其髮紺青……毛右旋，一一孔一毛生，皮毛細軟，不受塵水，胸有万字。

15

(Asita) uncovered the prince’s inner garment to examine him and observed that he possessed the thirty-two major marks: his body was of a golden hue, a fleshy protuberance adorned the top of his head, his hair was deep blue, … each hair curled to the right, with a single hair emerging from each pore, his skin and hair were fine and soft, impervious to dust and water, and a svastika symbol was present on his chest.

(6) 時阿夷以偈答王言：“今觀太子身……頰車如師子，四牙萬字現……是以眉間毫，白淨如明珠。”

16

At this moment, Asita responded to the king with a verse: “Now, observing the prince’s form… his jaw is like that of a lion, (with large, white) four teeth and the svastika symbol appearing (on his body)17… therefore, the tuft of white hair between his eyebrows is pure and radiant, like a luminous pearl”.

(7) 是時，父王慇懃再三，重問相師：“汝等更觀太子三十二相，斯名何等？”時諸相師即披太子衣，說三十二相：“……十六、胸有萬字……是爲三十二相。”

18

At that time, the king earnestly inquired of the physiognomists again: “Look once more at the thirty-two marks of the prince, what are they?” The physiognomists then unveiled the prince’s robe and revealed the thirty-two marks: “…sixteenth, the svastika symbol on the chest… these are the thirty-two marks”.

Aside from the Buddhist translations, the earliest reference in Chinese literature to the svastika symbol on the chest of a Buddha image appears in the Gaoseng zhuan 高僧傳 [Biographies of Eminent Monks] completed in 519 CE. In this text, the svastika is also referred to as “Wanzi”:

(8) 初僧護所創，鑿龕過淺，乃鏟入五丈，更施頂髻，及身相克成，瑩磨將畢。夜中忽當萬字處，色赤而隆起。今像胸萬字處，猶不施金鎛，而赤色在焉。像以天監十二年春就功，至十五年春竟。

19

Initially, Senghu carved a shallow niche, but later deepened it to five zhang, adding a topknot and completing the bodily features. As the polishing was about to be finished, suddenly, in the middle of the night, at the site of the svastika symbol, the area turned red and raised. Even today, the svastika symbol on the chest of the statue remains unadorned with gold but retains its reddish color. The statue was completed in the twelfth year of the Tianjian era and finished in the fifteenth year.

When used as a decorative motif on objects, the symbol “卍” is also referred to as “Wanzi”:

(9) 彼畫鉢中作蒱桃蔓蓮華像，佛言：“不應爾。”彼鉢中作萬字，佛言：“不應爾。”彼畫鉢作己名字，佛言：“不應爾。”

20

The monks painted on the bowl an image of a grape vine and lotus flower, to which the Buddha said, “This is inappropriate”. They then painted the svastika symbol, to which the Buddha again said, “This is inappropriate”. They painted their own names, and the Buddha said, “This is inappropriate”.

While pre-Tang Chinese texts frequently refer to the svastika as “Wanzi”, it is important to consider the possibility that these references may have been altered in subsequent periods. Since most extant versions of ancient Chinese literature were copied or printed long after their original composition—particularly given that many editions of the Buddhist canon were produced during the Song and Yuan dynasties—it raises the question of whether the designation of the svastika as “Wanzi” might have been a post-Tang modification. There are numerous textual variants in the examples cited earlier. For instance, in example (5) from the Taizi ruiying benqi jing 太子瑞應本起經 [The Sutra on the Auspicious Signs and Previous Lives of the Prince] by Zhi Qian, the Qisha edition (磧砂藏), Song edition (思溪藏), and Imperial Household edition (宮內廳本) use “卍”, while the Dunhuang manuscript (Dunyan 187) and Japanese manuscripts from Kongō-ji 金剛寺 and Shichi-ji 七寺 use “萬”. In this context, Dharmarakṣa’s translation of the T186 Puyao jing 普曜經 [Lalitavistara] is largely consistent in expression with its parallel text, the earlier Taizi ruien benqi jing translated by Zhi Qian,21 and its Korean edition (高麗藏) and Jin edition (趙城金藏) use “卐”, while the Song, Qisha, and Imperial Household editions use “卍”. However, the Fangshan stone inscriptions (房山石經) and the Jinglü yixiang 經律異相 [Compendium of Divergent Aspects of the Sūtras and Vinayas]22 cite it as “萬”.

However, if we were to rely on the account in the Fanyi mingyi ji and argue that the term “Wanzi” was not used to refer to “卍” before the Tang dynasty—suggesting that all variants of “萬” are post-Tang corrections of “卍”—such a claim would be highly suspicious. There is substantial evidence that “卍” was referred to as “Wanzi” well before the Tang dynasty and that this term carried an inherent connection to the number ten thousand in the mind of Chinese readers. This is exemplified by Emperor Jianwen of Liang 梁簡文帝, Xiao Gang 蕭綱, in his inscription “Wujun shixiang bei” 吳郡石像碑 [The Stone Statue Stele of Wu Commandery]:

(10) 千輪足起，萬字胸書。

23

A thousand-spoked wheel on the feet, a svastika symbol inscribed on the chest.

It describes the image of a Buddha statue with the “卍” symbol on the chest and a thousand-spoked wheel mark on the feet. The inscription mentions “the fourth year of Zhongdatong 中大通” and “Prince Linghou of Linru 臨汝靈侯” (the posthumous title of Xiao Yuanyou 蕭淵猷), suggesting that the composition was written after 533 CE and before Xiao Gang’s death in 551 CE. This paired phrase was later adapted by Yancong 彥琮 (557–610 CE) in his Tong ji lun 通極論 [Treatise on Reaching the Ultimate]: 開萬字於胸前，躡千輪於足下 “Displaying the svastika on his chest and treading on the thousand-spoked wheel beneath the feet”.24 Although this widely admired passage is sometimes cited with “萬” rendered as “卍” in its many iterations,25 the use of “Wanzi” (萬字) and “Qianlun” (千輪) as a parallel couplet in this prose clearly indicates that the author referred to the svastika symbol on the Buddha’s chest as “Wanzi” and believed it to be, at least in a literal sense, related to the number ten thousand.

This practice of linking the svastika symbol with the number ten thousand can also be observed in certain Chinese Buddhist translations:

(11) 下生於人間，而得大人相，胸字有萬數。以此相好故，無有諸疾病，若在家出家，常得受快樂。若獲剎利種，得王四天下；若出家學道，得成無上尊，純受上妙樂。

26

Descending to the human world, he attains the great man’s marks, with the svastika symbol on his chest. Because of these auspicious marks, he is free from all ailments. Whether a householder or a monastic, he always experiences happiness. If born as a kshatriya, he rules over the four continents; if he leaves home to pursue the Way, he attains supreme honor and enjoys the highest pleasures.

As previously discussed, the svastika on the Buddha’s chest was originally a symbol of auspiciousness and was not inherently connected to numerical values. However, in example (11), to conform to the five-character line structure of Chinese verse, the character “數” was added after “萬”, clearly indicating that the svastika on the Buddha’s chest was understood in relation to the number ten thousand. This interpretation did not originate in India but rather reflects a deeply ingrained Chinese cultural perspective that did not hinder the understanding of the translation.

Example (11) is from the T579 Youpoyi jingxing famen jing 優婆夷淨行法門經 [Upāsikā’s Pure Practices Dharma Gate Sūtra]. It is listed in extant Buddhist canons as having an unknown translator, with a note that “Sengyou’s catalog records it as an ‘alternative translation from Liang’ appended to the Northern Liang catalog (僧祐錄云安公涼土異經附北涼錄)”. This identification likely stems from the fact that the Chu sanzang ji ji includes a text called the Jingxing jing 淨行經 in its Xin ji An Gong Liangtu Yijing Lu 新集安公凉土異經錄 [A revised catalog of alternative Buddhist texts from the Liang region, based on Master Dao’an’s collection] Hayashiya (1941, pp. 997–1001) suggested that the language of T579 was influenced by the translations of Kumārajīva (344–413 CE) and that it was unlikely to be the same text as the Jingxing jing from Dao’an’s 道安 (312–385 CE) time. Nonetheless, Sui dynasty catalogs, including the Zhongjing mulu 眾經目錄 [Catalogue of Scriptures] by Fajing 法經27 (compiled in 594 CE), Zhongjing mulu 眾經目錄 [Catalogue of Scriptures] by Yancong 彥琮28 (compiled in 602 CE), and the Lidai sanbao ji 歷代三寶紀 [Record of the Three Jewels Through the Ages] by Fei Changfang 費長房29 (compiled in 597 CE), all list the Youpoyi jingxing jing as a two-scroll text with an unknown translator. Even if this text postdates Dao’an and Sengyou, it was at least translated before the Sui dynasty.

The svastika symbol on the Buddha’s body was referred to as “Wanzi” in Chinese as early as the Three Kingdoms period, suggesting that this rendering was adopted from the earliest instances of the symbol’s appearance in Chinese Buddhist translations. In many pre-Tang sources, the svastika was commonly referred to as “Wanzi”, leading to its association with the number ten thousand. This connection had already become deeply ingrained in the popular consciousness well before the reign of Empress Wu Zetian.

3. Efforts to Restore the “Correct Translation”

Starting with Kumārajīva’s 鳩摩羅什 (344–413 CE) translations, the svastika symbol on the Buddha’s body began to be labeled by new terms. In earlier translations, the svastika had consistently been called “Wanzi” (萬字), with textual variants such as “卍字” or “卐字”, and no alternative names were recorded. However, Kumārajīva introduced new terms for the symbol, referring to it as “de zi 德字” (virtue character) or “de xiang 德相” (virtue mark), a practice that was later adopted by Xuanzang 玄奘 (602–664 CE) in his translations as well:

(12) 是時世尊從三昧安詳而起，以天眼觀視世界，舉身微笑，從足下千輻相輪中放六百萬億光明，足十指、兩踝、兩𨄔、兩膝、兩髀、腰、脊、腹、脅、背、臍、心胸德字、肩、臂、手十指、項、口四十齒、鼻兩孔、兩眼、兩耳、白毫相、肉髻，各各放六百萬億光明。

30

At that time, the World-Honored One arose calmly from samādhi and, using his divine eye, gazed upon the world. He smiled subtly, emitting six hundred million billion rays of light from the thousand-spoked wheels on his feet. Each of his ten toes, two ankles, two calves, two knees, two thighs, waist, spine, abdomen, ribs, back, navel, the “virtue character” on his chest, shoulders, arms, ten fingers, neck, mouth with forty teeth, two nostrils, two eyes, two ears, the white tuft between his eyebrows, and the cranial protuberance each emitted six hundred million billion rays of light.

(13) 八十者手足有德相。

31

The eightieth (minor mark), his hands and feet bear the “virtue mark”.

(14) 世尊手足及胸臆前俱有吉祥喜旋德相，文同綺畫，色類朱丹，是第八十。

32

The World-Honored One’s hands, feet, and chest bear auspicious, joyous turning “virtue marks”, patterned like brocade and colored like vermilion; this is the eightieth [minor mark].

The above examples (12) through (14) depict various auspicious signs on the Buddha’s body. In the Sanskrit Pañcaviṃśatisāhasrikā Prajñāpāramitā [The Perfection of Wisdom in 25,000 Lines],33 the term “德字” corresponds to śrīvatsa,34 while “德相” and “吉祥喜旋德相” correspond to śrīvatsa-svastika-nandy-āvarta,35 all referring to the “卍” symbol located on the Buddha’s chest or limbs. Both “德字” (virtue character) and “德相” (virtue mark) are semantic translations, with “德” (virtue) conveying meanings of “auspiciousness” and “blessing”.36 Translating the auspicious symbol “卍”—particularly its Sanskrit term svastika—as “德” is semantically more appropriate. In contrast, the association between the numeral “萬” and svastika is less evident in terms of meaning. Consequently, the Yiqiejing yinyi by Huilin dismisses the term “萬字” as an “incorrect statement”, likely reflecting this semantic consideration.

It is intriguing that, despite Kumārajīva’s newly coined translations often replacing earlier renditions by translators like An Shigao, Zhi Qian, and Dharmarakṣa and becoming the standard for later translations,37 the translation of “卍” as “德” (virtue) did not gain widespread acceptance. Even though Xuanzang, another highly influential translator, also adopted Kumārajīva’s terminology, this particular translation did not take hold. This suggests that the practice of reading the symbol “卍” as “Wan” was likely already well established before Kumārajīva’s time.

Even in Buddhist translations and Chinese writings postdating Kumārajīva, the “卍” symbol continued to be frequently referred to as “Wanzi”. There are also examples where translators combined the terms “德” and “萬” in their translations, possibly as an attempt to reconcile the old and new translation methods:

(15) 自有衆生樂觀如來胸德字萬印38相，三摩尼光相者。

39

Some beings find joy in gazing upon the “virtue character/Wan mark”, on the chest of the Tathāgata, from which radiates the light of three maṇi jewels.

(16) 云何觀如來頸相、缺瓫骨滿相、臆德字相、萬字印相？……如是圍繞諸光畫中，是名佛頸出圓光相、胸德字文、萬字印中，缺瓫滿相、腋下珠相。

40

How should one observe the Tathāgata’s neck mark, the complete supraclavicular mark, the “virtue/Wan character” mark on the chest? … Thus, surrounded by various lights and patterns, these are known as the Buddha’s neck-radiating mark, the “virtue/Wan character” on the chest, and the complete supraclavicular mark, with the pearl mark under the arm.

(17) 爾時世尊智慧觀察現在大衆，非肉眼觀，如師子王奮迅視眄，呵呵大笑，頂上肉髻放無量光，肩脅腰髀胸卐41德處及諸毛孔，皆放一切無量光明，如空中虹、如日千光，如劫盡時大火熾然猛炎之相。

42

At that time, the World-Honored One observed the present assembly with wisdom, not with physical eyes, but with the gaze of a lion king aroused in vigor, and laughed. From the topknot on his head, he emitted countless rays of light, and from his shoulders, ribs, waist, thighs, chest, where the “卐” virtue symbol is, and from all his pores, he emitted boundless light, like a rainbow in the sky, like the thousand rays of the sun, like the blazing fire at the end of a kalpa.

Additionally, some Buddhist translations from the Tang and Song dynasties began to include phonetic translations of Śrīvatsa or Svastika:

(18) 三者臆前有室利婆瑳像。

43

Third, the “Śrīvatsa” symbol on the chest.

(19) 足現千輻輪、莎悉帝迦相。

44

The thousand-spoked wheel and the svastika marks appear on his feet

Some scholars argue that the reading of “卍” as “Wan” is related to phonetic translation.45 A commonly cited argument is based on Zanning’s 贊寧 theory of the “six rules of translation (六例)” in Song gaoseng zhuan 宋高僧傳 [Biographies of Eminent Monks of the Song Dynasty]:

(20) 謂譯字譯音爲一例……初則四句：一譯字不譯音，即陀羅尼是；二譯音不譯字，如佛胸前卍字是；三音字俱譯，即諸經律中純華言是；四音字俱不譯，如經題上 二字是。

46

The first aspect of translation concerns whether to “translate”47 the script or the sound. This aspect can be summarized in four statements: 1. “Translate” the script but not the sound, as in the case of dhāraṇī. 2. “Translate” the sound but not the script, as in the case of the “卍” symbol on the Buddha’s chest. 3. “Translate” both the sound and the script, as in the case of the purely Chinese expressions found in various sutras and vinayas. 4. “Translate” neither the sound nor the script, as in the case of the characters48 found in the titles of some scriptures.

Zanning’s theory is often cited as one of the specific reasons why Wu Zetian designated the 卍 symbol as “萬字”, with the pronunciation of “萬” being (partially) similar to the Sanskrit word svastika.49 However, upon closer examination, there are several issues with how Zanning’s theory should be understood. The renowned scholar Yinke Chen (2001, p. 283) suggested that there may be a textual inversion in the phrases “譯字不譯音 (translate the script but not the sound)” and “譯音不譯字 (translate the sound but not the script)”, as the dhāraṇī mantras in Buddhist scriptures are typically transliterated phonetically rather than semantically.50 Even if Yinke Chen’s conjecture is correct, the connection between the concept of “translating the script but not the sound” and the “卍” symbol remains quite challenging to understand.

The key to addressing this issue lies in understanding that the concept of “譯” in the Song gaoseng zhuan differs from the contemporary notion of “翻譯” or “translation”. To comprehend Zanning’s notion of “譯”, it is crucial to consider his statement: 譯之言易也，謂以所有易所無也 “Translation implies change; it means substituting what is available for what is lacking”.51 Therefore, his theory of “譯字 (translating the script)” or “譯音 (translating the sound)” should not be viewed as a case of textual inversion. Instead, it should be interpreted as “altering the script while preserving the pronunciation when translating dhāraṇī” and “retaining the symbol while changing the pronunciation when translating the ‘卍’ symbol on the Buddha’s chest”. This explanation actually indicates that the “卍” (svastika) and its Chinese pronunciation “wan” have no phonetic connection.

When considering the historical development of translating the “卍” symbol in Chinese Buddhist texts, it becomes evident that the full phonetic transliterations derived from the Sanskrit names śrīvatsa or svastika appeared relatively late, and their pronunciation differs significantly from the Chinese “Wan”,52 and the transliterations (such as the previously mentioned “室利婆瑳” and “莎悉帝迦”) do not use the character “萬”. Therefore, “wan” should not be considered as an abbreviated transliteration of these terms.

4. The Rationale Behind the Naming of “卍” as “Wanzi” in Chinese

As discussed earlier, the Sanskrit names for the “卍” symbol, such as śrīvatsa and svastika, whether in their phonetic transliterations (“室利婆瑳” and “莎悉帝迦”) or semantic translations (“德字” and “德相”), appeared relatively late. Moreover, there is no evident connection between these terms and the Chinese word “萬” in terms of pronunciation or meaning. This raises another significant question: why was the symbol “卍” regarded as a “character” in Chinese? In Huiyuan’s commentary, he attempted to attribute this issue to early translators’ errors (see example 3 above). While it is true that early translations were often based on Middle Indic languages rather than standard Sanskrit, leading to occasional confusion or mistakes,53 the overwhelming and widespread designation of “卍” as a “character” (rather than sporadic or alternative instances) is difficult to attribute solely to the confusion between lakṣaṇa and akṣara. Furthermore, early translators consistently rendered lakṣaṇa correctly as “xiang 相” (“sign, mark, characteristic”),54 and there is no verifiable evidence of cases where it was misunderstood as akṣara.

The reason the auspicious symbol “卍” came to be known as “Wanzi” likely stems neither from phonetic nor semantic translation but rather from the use of the character “万”, a popular form of the numeral “萬” at the time, to visually mimic the shape of “卐”. The character “万”, which is now the simplified form of “萬”, has a long history in the development of Chinese script, appearing as early as the Oracle Bone Script from the Shang dynasty (c. 1600–1046 BCE). However, at that time, “万” was merely an ancient form of the character “丏”. Due to the phonetic similarity between “丏” and “萬” in Old Chinese,55 “万” began to be used as a phonetic loan character for “萬” as early as the Warring States period (475–221 BCE).56 In the Han dynasty, the character “万”, representing the numeral, also gave rise to forms like “”.57 These variants of “万” and “” as popular forms of “萬” continued to be used after the Han dynasty and appear in Dunhuang manuscripts as well as in printed editions of the Chinese Buddhist canon from the Song dynasty onwards.58 

Table 1. Early Forms and Usages of the Character “万”.

Shang Dynasty
	The form of the character “万” in the Shi group small-script oracle bones (Inscription No. 20824). (Zongkun Li 2012, p. 1219)	The form of the character “万” in the anonymous group oracle bones (Inscription No. Tun 0062). (Zongkun Li 2012, p. 1219)	The form of the character “万” in the Huayuanzhuang East oracle bones (Inscription No. Hua 226). (Zongkun Li 2012, p. 1219)
Warring States period
	A copper seal from the Yan state, engraved with the characters “万千”. (Zhang 2021, p. 136)	A copper seal from the Yan state, engraved with the characters “万”. (Zhang 2021, p. 137)	The knife-shaped coins of the State of Yan have the combined character “八万” (eighty thousand) on the reverse side. (Wu 2006, p. 247)	The knife-shaped coins of the State of Qi bear the character “万” on the reverse side. (Wu 2006, p. 149)
Han dynasty
	The character “万” in the Juyan bamboo slips, found in the phrase 凡七万五千四百廿九(“a total of 75,429”) (Slip No. 261·27A, 261·13A). (Y. Li 2014, p. 934)	The character “万” in the phrase 奉錢万二千 (“salary of 12,000”) from the Juyan newly discovered bamboo slips (Slip No. E.P.S4.T1: 14B). (Bai 2014, p. 976)	The character “万” in the phrase万年 (“ten thousand years”) from the Jianshui Jinguan bamboo slips (Slip No. 73EJT9: 136). (Ren 2014, p. 334)	The character “万” in the phrase 万六百三十亖 (a) (“10,634”) from the Jianshui Jinguan bamboo slips (Slip No. F3: 259).
Dunhuang Manuscript
	The character “万” in 万錢 (“ten thousand coins”) (b)	The character “万” in 孤山高万仞 (“The solitary mountain rises ten thousand ren”) (c)

^(a) The character “四” (four) written as “亖” is a characteristic of the Xinmang bamboo slips, and this form appeared after the 3rd year of Shijian’guo 始建國 (11 CE). See J. Li (1989). ^(b) P.2524 Yu Dui語對. ^(c) S.126 Taizi Rushan Xiudao Zan 太子入山修道贊.

below illustrates the early forms and usages of “万”:

The earliest extant Chinese statues featuring the “svastika” symbol can be traced back to the Northern Wei dynasty (see Jin 2016, pp. 190–91), revealing that the “svastika” symbol on the Buddha’s chest can either be the counterclockwise “卍” (Figure 1) or the clockwise “卐” (Figure 2). The latter resembles the popular forms of the Chinese characters “万” and “” (as shown in the red-highlighted illustration: ).

The “卐” symbol in Buddhism was originally an auspicious motif rather than a character, but its designation as “Wanzi” (literally “character Wan”) in Chinese, rather than later translations like “Virtue Mark” (德相) or “Śrīvatsa Image” (室利婆瑳像), suggests that the symbol was named based on its resemblance to the shape of a Chinese character. This approach is similar to how modern Chinese uses terms like “十字架” (cross), “八字眉” (arched eyebrows), or “丁字路口” (T-junction). The practice of using Chinese characters to depict shapes was common in medieval times, with literature often using the phrase “…字” to express this concept:

(21) 鱉目凹陷者，及厭下有王字形者，不可食之。

59

Turtles with sunken eyes or with a “Wang (王)”-shaped mark on their undersides should not be eaten.

(22) 世云堯眉八采，不然也，直兩眉頭甚竪，似八字耳。

60

It is said that Emperor Yao’s eyebrows had eight different colours; however, this is not true. His eyebrows were simply very vertical, resembling the character “Ba (八)”.

(23) 君不見西陵田，縱橫十字成陌阡。

61

Have you not seen the fields of Xiling, crisscrossed into quadrants shaped like a character “Shi (十)”?

Although the form of the character “万” and the complete pattern of “卐” differ in certain details, the former serves as a rough approximation of the latter. In the process of abstracting and representing natural patterns through written characters, such degrees of distortion are often unavoidable and, once conventionalized, tend to be widely accepted. This can be seen in examples 21 to 23, as well as in analogous expressions in other languages, such as the terms “A-line skirt” and “C-clamp” in English. Another supporting piece of evidence is that even during the period when 卐 was widely recognized as a Chinese character pronounced “wan”, various alternative forms of the symbol continued to appear in literature, such as , , ,62 , , , , 63, ,64 65. This suggests that even in the post-Tang Dynasty periods, the precise details of the “卐” symbol remained unclear to many Chinese, leading them to partially distort its shape according to their habits of writing Chinese characters. Given that the reading of “卐” as “wan” certainly emerged before the Tang Dynasty, it is quite likely that early translators or believers chose a relatively similar, though imperfect, Chinese character to approximate its form.

The practice of referring to the svastika symbol on the Buddha’s body as “万字” may not only be attributed to its resemblance to the Chinese character but could also be connected to an ancient Chinese belief that the appearance of characters on a person or animal often signified divine will. This belief has existed since pre-Qin times. In the Zuo zhuan 左傳, for instance, it is recorded that when Zhongzi 仲子 (?–721 BCE), the daughter of Duke Wu of Song 宋武公 (?–748 BCE), was born, she had characters on her hand reading “to become the wife of the Duke of Lu (為魯夫人)”. Consequently, the people of Song followed the divine mandate and married her to Duke Hui of Lu as a secondary wife.66 Similarly, Tang Shu 唐叔 (reigned 1042–? BCE), the first ruler of the Jin state, was born with the character “虞” (Yu) on his palm, and thus he was named Yu.67 Likewise, Ji You 季友 (?–644 BCE), a minister in the state of Lu, was named because he was born with the character “友” (You) on his hand.68 The Shiji 史記also recounts an instance where a divine tortoise naturally manifested a long inscription on its shell: 甲子重光，得我者匹夫爲人君，有土正，諸侯得我爲帝王 (“On the day of the Jiazi, a solar halo appeared. Whoever gains me on this day, though a commoner, will become a ruler; possessing the mandate of the earth, those who gain me among the nobility will become kings”.). Although we cannot conclusively verify through existing literature that the svastika on the Buddha’s body was called “Wanzi” due to this same idea, it ultimately became associated with a similar concept: by the Song dynasty, “萬字” came to be understood literally as “the gathering of all auspicious virtues (ten thousand virtues)”.

As for the records in texts such as the Fanyi mingyi ji since the Song dynasty, which claim that Wu Zetian first used the “卍” symbol to represent the character “萬” on the “Heavenly Pivot”,69 this may reflect a post hoc interpretation that emerged after the original connection between “卐” and “万” had been forgotten. The “Heavenly Pivot”, a commemorative bronze pillar completed in the second year of the Yanzai 延載 era (695 CE),70 bore the inscription 大周萬國頌德天樞71 “The Heavenly Pivot Commemorating the Praise of Virtue by All Nations of the Great Zhou”. If the account recorded in the Fanyi mingyi ji has some degree of truth, we might speculate that the “萬” character on the Heavenly Pivot could have been written as “卍”.72 This suggests that Wu Zetian may accept the already established equivalence between “卍” and the Chinese character “萬” (see Shi 1984), rather than being the origin of the practice of pronouncing “卍” as “Wan” in Chinese. Moreover, the first batch of characters promulgated by Empress Wu Zetian in 689 CE included a new form for “月” (moon), which was written as “”. As previously discussed, the representation of “卍” within this new character underwent a transformation due to the ambiguous recollection of this symbol by Chinese writers, combined with the habitual practices of Chinese calligraphy. This resulted in the emergence of alternative forms such as “”, “” and “囝” (see Zhang et al. 2022, p. 2569). Empress Wu Zetian’s decision to utilize the “卐” in creating the new character for “moon” was likely inspired by the idea that the svastika on the Buddha’s chest could emit light. Additionally, it is visually connected to the new character for “日” (sun), which was written as “”. However, this new form for “月” has no relation to the pronunciation (wàn) or meaning (ten thousand) of “萬”.

It is also worth noticing why the Fanyi mingyi ji includes the claim, purportedly sourced from the Huayan yinyi, that Wu Zetian was the first to pronounce “卍” as “Wan”. This likely stems from a combination of two historical factors: Wu Zetian’s enthusiasm for creating new forms of Chinese characters and her deep reverence for the Avataṃsaka Sūtra. Wu Zetian sent envoys to Khotan to seek out the Sanskrit original and, in 695 CE, organized the retranslation of the eighty-volume Xinyi da fangguang fo huayan jing 新譯大方廣佛華嚴經 (T279) by Śikṣānanda 實叉難陀, Bodhiruci 菩提流志, and Yijing 義淨. The new characters introduced during Wu Zetian’s reign, along with the newly translated T279, spread to neighboring countries, and ancient manuscripts of the T279 from the Nara and Silla periods in Japan and Korea frequently feature these new characters.73 The Japanese monk Myōgaku 明覺, in his work Shittan Yōketsu 悉曇要訣 [Essentials of Siddham] (compiled circa 1101 CE), also addressed the script issues found in the Huayan jing of his time:

(24) 彼經中多用古文，星字作○形、日字作形、地字作埊等是也。故萬字古文作卐形歟？但天竺卐形本非字也，乃是印形，其形自當漢土萬字古文，故云萬字歟？如十字羯磨，羯磨形似十字，故云十字羯磨，或直云十字印，非以十字即爲羯磨，十字是此土字形，非西土字，然三藏傳漢土時，以此土十字令知羯磨形歟？

That scripture often uses ancient scripts, such as a “○” for the star, a “” for the sun, and “埊” for the earth. Thus, the ancient script for “萬” might have been “卐”? But in India, the “卐” symbol is not a character; it is a symbol. Its form likely resembles the ancient Chinese script for “萬”, which is why it is called “Wanzi”. Just as a cross-shaped karman resembles the Chinese character “十”, so it is called as “Shizi jiemo”, or simply “Shizi yin”. This does not mean that the Chinese character “十” is the karman itself; “十” is a Chinese character, not an Indian one. However, when the Tripiṭaka was transmitted to China, the shape of Chinese character “十” was used to represent the karman form”.

Although the characters for “star” and “sun” cited by Myōgaku from the Huayan jing are derived from the new script introduced during Wu Zetian’s reign and not from ancient Chinese characters, and although the symbol “卐” is likewise not an ancient form of the character “萬”, it is noteworthy that Myōgaku was among the first to discern a visual connection between the Buddhist symbol “卐” and the Chinese character “萬 (万)”. 74 This observation was indeed insightful, yet it is unfortunate that this perspective has not garnered the scholarly attention it merits. Similarly, the modern Japanese scholar Morohashi Tetsuji correctly observed the connection between the character “万” and the svastika (Morohashi [1955] 1999, p. 104). However, due to the incomplete availability of materials such as oracle bone inscriptions at the time of his writing, Morohashi mistakenly reversed the order of their development and the causal relationship between the two. He incorrectly posited that the vulgar character “万” was a derivative form of the svastika and, like most lexicographers, attributed the naming of the svastika to Empress Wu Zetian.

5. Conclusions

The well-accepted view, found in dictionaries and scholarly works, that the Buddhist “卍” symbol was referred to as “Wanzi” in Chinese, by the decree of Empress Wu Zetian during the construction of the Heavenly Pivot, is not reliable. The connection between “卍” and the Chinese character “萬” likely originated from the visual resemblance of the symbol “卐” to the popular forms of the numeral “萬” in its variants “万” and “”. This translation appeared early and was widely used in the Chinese translations of Buddhist scriptures, and it is probable that this interpretation was established and widely accepted when Buddhism was first introduced to China. As a result, even though influential translators like Kumārajīva and Xuanzang translated the Sanskrit term svastika as “德相” (virtue mark), they could not change the public’s habit of reading “卐” as “Wanzi”. This practice, through the spread of Chinese Buddhism, eventually influenced surrounding regions.

Over time, as the original rationale for the connection between “卐” and “万” was gradually forgotten, and given that Empress Wu Zetian was both an ardent patron of Buddhism and used the “卐” symbol in creating new characters during her reign, the legend arose that “Empress Wu Zetian designated this symbol as a Chinese character, inscribed it on the Heavenly Pivot, and assigned it the pronunciation as ‘Wan,’ which signified the gathering of auspicious myriad virtues”. This story was recorded in works such as the Fanyi mingyi ji and has continued to influence interpretations to this day.