22 June 2018

The Earliest Dated Heart Sutra Revisited

This material has now been published as: 'Xuanzang’s Relationship to the Heart Sūtra in Light of the Fangshan Stele.' Journal of Chinese Buddhist Studies (2019, 32: 1–30).

Fángshān Stele, rubbing.
He and Xu (2017)
In 2016, a story made its way around the Chinese media (e.g., the Chinese State Administration of Cultural Heritage) that a new discovery had been made of the earliest dated Heart Sutra. A stone stele, inscribed with the Heart Sutra and carrying the date of 661 CE, had been found at Fángshān near Beijing. The story was not picked up in the West. During correspondance with Ji Yun about my review of his article on the Heart Sutra he generously informed me about this inscription and kindly supplied me with a copy of a recent journal article outlining the find (He & Xu 2017) and a book with another transcription (Beijing Library... 1987).

I uncovered some older sources which mention the Fángshān Stele. Firstly, I found that the colophon (containing the date) was transcribed and published in Dàoān and Zhāng (1977). Unfortunately, I cannot get access to this book, except through "snippets" on Google Books. However, I also discovered the text of a pamphlet on Fángshān, which also transcribes the colophon (Lin 1958). And note that Lin 1958 was published in Taipei, Taiwan, not in Communist China and was thus always available to scholars outside the region. The text of Lin (1958) was also used in a pamphlet about the temple on Fángshān, i.e., Yang (2003). Different transcriptions of the colophon disagree on some details. I'm grateful to members of the Omniglot Facebook group and the Chinese Language Stack-exchange for help with deciphering the colophon (though of course any remaining mistakes or infelicities are down to me).

The Fángshān Xīnjīng Stele is of considerable interest because it purports to be carved in 661 CE, three years before the death of Xuánzàng in 664 CE and yet it attributes the translation of the Heart Sutra to him, which as we know is problematic. I have done my best to assemble and evaluate the evidence below.

The text is inscribed on a stone tablet or stele. It's dimensions are unclear, but the ratio of its sides is approximately 2:3 and I would guess at dimensions in the realm of 60 x 90 cm (allowing ca. 3 x 3 cm for each character and some leeway). The surface of the stele seems to be badly damaged so that many characters are obscured. It was broken in half at some point and appears to have been repaired. The lower left corner is missing, obscuring up to nine characters.

The stone tablet now resides at 雲居寺 Yúnjū sì which translates something like Temple Dwelling in the Clouds. The temple is on 房山 Fángshān, which means something like Repository Mountain. Nearby is 石經山 or Stone Sutra Mountain where Buddhist sutras were carved on thousands of stone slabs in an attempt to preserve the entire Buddhist Canon (as described in the 8th Century). Fángshān is about 65km south-west of Beijing.


The title displayed on the stele comes at the end of the text, which is usual for Indic Buddhist texts. The full title of the text is:
The Middle Chinese pronunciation of this can be reconstructed as Banya-baramida-sim-keng. This translates into Sanskrit as Prajñā-pāramitāhṛdaya-sūtra. We can see that the first part—般若波羅蜜多— is an attempt to represent the sounds of the Sanskrit word using Chinese characters, while the last two characters represent whole words.

This transliteration was used by Mokṣala in his translation of the Pañcaviṃśatisāhasrikā-prajñāpāramitā-sūtra dated 291 CE and was also used by Kumārajīva in his translation of the text in 404 CE.

Note that the character 經 is a variant.


The stele attributes the text to Xuánzàng (see detail, right):
三藏法師玄奘奉 詔譯
三藏 Tripiṭaka
法師 Dharma master (Skt dharma-bhāṇaka)
玄奘 Xuánzàng
奉 詔譯 translated with imperial authorisation
Note that there is a full character space between 奉 and 詔. We see the same space in the Beilin Stele. This is added as a mark of respect to the Emperor. The character means:
"An imperial edict. To decree. Appearing in the colophons of translated scriptures, it indicates official authorization at the highest level, indicating the high level of the translatorʼs reputation." (DDB)
Note that this attribution occurs at the beginning of all of Xuánzàng's translations in the Taishō Tripiṭaka. However, it also occurs in his travelogue 大唐西域 (T2087) which was not translated but composed by him. Note also that there are minor variations in some earlier editions suggesting that the wording was not fixed.


The date of 661 CE comes from the phrase 顯慶六年二月日造, which occurs at the bottom of the leftmost column on the stele.This is considerably less clear than the attribution.

顯慶 Xiǎnqìng refers to a period of the rule of Emperor 唐高宗 Táng Gāozōng, roughly coinciding with the years 656-661. 唐 was the name of the dynastic lineage, hence Táng Dynasty. Chinese emperors would take special "reign names" (年號) at significant points in their reign. Gāozōng used 14 different names during his time as Emperor (649-683).

The Chinese new year begins on the second new moon after the winter solstice, usually in mid January to mid February. The length of months varied, but generally they were defined from new moon to new moon which was on average 29 ​3260 days long. Thus a typical month might be 29 or 30 days, and this would always leave a few days at the end (12 x 30 = 360 and a year is 365.25 days). [I'm told that this is an over simplification]

However, reign periods did not always change at new year. The Xiǎnqìng period began on 7 February 656 and ended on 4 April 661, to be followed by the 龍朔 Lóngshuò period. Lóngshuò began on the 30th day of the second month (= 5 April), so this stele was made towards the end of Xiǎnqìng, on 13 March, 661.
顯慶 Xiǎnqìng era
六年 6th Year
二月 2nd month = March
八日 8th day = 13
造 made, constructed.
Not all the elements of the first character 顯 are clear, but the second 慶 is clear and there seems little doubt that this is the correct interpretation. This is partly because 慶 is not used in many other names of any other regnal periods and is thus a useful identifier. There is no obvious reign period for which this could be mistaken.

The Text

The rest of the text is presented in 11 columns of 26 characters (or less), most of which are clearly visible and match the text of T251. There are some minor differences, however.

One feature of the text is the substitution of the simplified character 无 for 無 throughout. At first this struck me as odd, but asking around I found that it was actually common, especially in inscriptions where the justification was that it was easier to inscribe. The Wiktionary entry says "First attested in the Warring States period; used interchangeably with 無 until the Tang dynasty." Some of the simplified characters introduced by the PRC government actually have long histories. 

In the dhāraṇī, 帝 is written as 諦 "examine", with the same pronunciation /tei/. This composite character has 言 "speech" as a (vaguely) semantic element and 帝 as a phonetic element. We also saw this substitution in the Beilin Stele. If we explain 无 for 無 as a simplification, then 諦 for 帝 is the opposite, since 諦 is considerably more complex and therefore difficult to carve. However, the so-called two truths are often transliterated as 二諦 and it may that the calligrapher thought this connection too good to pass up. 

The text appears to be signed at the end of the sutra, but I cannot make out the character and none of my sources mentions it. 


The colophon is important because it not only gives us the date of the work, but some details about the donor who paid for the stele to be made. Such items were a fund-raiser for the monastery to help pay for their main project of carving the entire Tripiṭaka into stone (which remained incomplete, but covered thousands of tablets). Indeed, our text is not only the oldest dated Heart Sutra,  it is the oldest dated colophon at Fangshan and thus marks the beginning of a new phase of the project.

By comparing the image of the rubbing from He and Xu (2017) and published transcriptions (which  disagree, are partial, and/or contain errors), I have created a kind of critical edition. The colophon must have had more characters where the corner is broken off (indicated in light grey beow). 
□ = a full character-sized space in the inscription.
What can we find in this? Firstly the inscription was commissioned by 楊社生 Yáng Shèshēng. Unfortunately, he seems not to have made any other mark on history. However, 楊 is a very significant name in Chinese history because the Emperors of Sui were from the 楊 clan; although it is not clear if Yáng Shèshēng was closely related to them, because of his name and rank we can say that he is a member of the aristocracy.

Line 1. 雍州櫟陽縣遊騎將軍守左衛淥城府左果毅都尉楊社生

Yáng was from 雍州 Yōng Zhōu or Yong Province in which the Tang capital, Chang'an (長安) was located (modern day Xian). More specifically, he was from 櫟陽縣 Yueyang county.* Yueyang was a temporary capital of the Han (200-205 BCE). It is now in the Yanliang District (阎良区) about 50 km from Xian.
* note that the usual Mandarin pronunciation of 櫟 is lì, but the name of the County is definitely Yue, probably based on the pronunciation of 樂 yuè.
Yang was a military officer. With help from Charles Hucker's (1985) Dictionary of Official Titles in Imperial China we can determined that he held () the prestige title of General of Mobile Cavalry (游騎將軍), but served as Courageous Commander" (果毅都尉) of the left () in the guard of the left (左衛) in the garrison of 淥城 Lùchéng. The early Tang military was divided into 12 armies, each comprised of a number of garrisons (~ 650 in total). Each garrison had an overall commander and two "courageous commanders" (果毅都尉), left and right. The "courageous" part related to the way of referring to different garrisons.

The place name 淥城府 or Lùchéng Garrison seems to correspond to modern day 涿州为 [涿州為] or Zhuōzhōu Province which is about 30km south-east of Fángshān. This explains why the stele was found in Yúnjū Temple rather than somewhere closer to Chang'an.

Line 2: 父楊 母段息懷慶玄嗣玄器玄貞女大娘二娘隸利巫山

Line two is all summed by Yang (2003) as 其全家 "his family". It begins with his Mother 母 Duan段. Duan would be her family name. Next is his wife 妻 Hu 扈; his sons (息): 懷慶, 玄嗣, 玄黎*, and 玄貞; his daughters (女) 大娘 and 二娘 (i.e., first daughter, second daughter); and finally someone named 利巫山 who is a servant or dependent (隸). Perhaps a "ward" given that he is included with the family. The person missing from all this is his father. Since the tablet has columns of 26 characters, there are potentially three characters missing from the end of each colophon column. We can conjecture that the end of line one included the word father (父) and his name, which was presumably also 楊.
* 玄器 is an alternative reading of 玄黎.
† in this context we might expect 太 rather than 大.
The traditional Chinese system of names is relatively complex. They have a family (originally a clan) name, in this case 楊. Then they have a given name (名) which may be given by the head of the family rather than the parents and only used in the family. Women only used their family and given names. Boys might have an infant name (乳名) used up to adulthood. At adulthood men get a 字 or "courtesy name" which is the name they use in everyday life, though intimates may also call them by a nickname (號). At ordination monks take a Dharma name (法號). It's possible that the younger sons became monks and that their names with the common element 玄 (which they share with Xuanzang) reflect this. Other names, such as a nom de plume, or posthumous names were also common. Emperors often took a new name when they took the throne.

Three characters are missing at the end of this line.

Line 3: 家眷屬緣此功德齊成正覺顯慶六年二月八日造經

The third line asks that family (家) members (眷屬) be caused (緣) by this merit (此功德) to attain awakening (成正覺) together (齊).

The date 顯慶六年二月日 the sutra was made 造經 we have already discussed.


For the first time we have physical evidence linking the Heart Sutra to Xuánzàng during his lifetime and naming him as translator (譯). However, we need to be cautious. What this tells us is precisely that those involved in the production of the inscription believed that Xuánzàng had translated the sutra. Xuánzàng is mentioned in this inscription, but he wasn't involved in it.

I asked Dr Jeffrey Kotyk if he could shed any light on the chronology from the traditional histories. In the 《釋氏通鑑》, a Buddhist history of China up to ca 960 CE, we find a single mention of Xuánzàng for year 5 of Xianqing 顯慶 (kindly translated for me by JK, but with some slight modifications of my own):
「三月。西明寺靜之禪師遷逝。甞鼻患肉塞。百方無驗。有僧令誦般若心經萬遍。恰至五千。肉鈴便落(本傳)○奘法師。於玉華譯般(若經)○」(CBETA, X76, no. 1516, p. 88b2-4)
"3rd lunar month. Chanshi Jingzhi of Ximing-si passed away. He once suffered from blocked nasal passages. Hundreds of remedies were ineffective. There was a monk who had him recite the Prajñā-Heart Sutra ten-thousand times. At exactly five-thousand [recitations], his [nasal] flesh tinkled [like a bell]. (original biography). Master Xuánzàng at Yuhua translated the Pra(jñā Sūtra)."
Words in square brackets are added to help make sense of the translation. The words in parentheses are notes from the CBETA edition.
From the 《大唐大慈恩寺三藏法師傳》, a biography of Xuánzàng composed by慧立 Huìlì, edited and published by 彥悰 Yàncóng in about 688 CE (T 2053), we know that Xuánzàng moved to Yuhua late in the 4th year of Xianqing (659), and started translating the Mahāprajñāpāramitā (i.e., T220) at the beginning of the 5th year (660). See below for more on dates. What this passage suggests is that the Heart Sutra predates the translation of the Mahāprajñāpāramitā.

The phrase 三藏法師玄奘奉 詔譯 appears at the beginning of T220, Xuánzàng's translation of the collected Prajñāpāramitā sūtras. By contrast, Huìlì and Yàncóng state that the work was translated due to a request from the "people"
東國重於《般若》,前代雖翻,不能周備,眾人更請委翻 (T 2053.275c.17-19)
In the Eastern Country the Mahāprajñāpāramitā Sūtra was highly esteemed. Although it had been translated into Chinese during a previous dynasty, the translation was incomplete, so the people [眾人] requested that the Master kindly translate it anew. (Li 1995: 327)
From what I can make out, such translations were presented to the Emperor after completion and then received the imperial seal of approval.

The 玉華宮 Yuhua Gong, or Palace of Jade Flowers, is the place where Xuánzàng's translation team worked on T220. It is about 100 km north of Changan, well away from the distractions of life in the capital (and quite far from where Yang lived, also). According to Huìlì and Yàncóng, Xuánzàng moved out to Yuhua in 顯慶四年十月 or November 659 (T2053.275c). Yaowang Mountain, about halfway between Chang'an and Yuhua also has a collection of stone sutras.

The date of 顯慶六年二月日 for the Fángshān stele is interesting because it's in the middle of the period during which Xuánzàng and his team of translators were translating the collection of sixteen Prajñāpāramitā sūtras known as the 大般若波羅蜜多經 or Mahāprajñāpāramitā Sūtra (T220). This took about four years and occurred between: 顯慶五年正月一日 and 龍朔三年十月二十日. The table below shows the key events and the dates in the traditional Chinese and Gregorian calendars.

EventChinese dateGregorian
顯慶 begins (i.7)656 Feb 7
Move to Yuhua顯慶四年十月 659 Nov
T220 Trans begins顯慶五年正月一日660 Feb 16
Fángshān stele顯慶六年二月八日661 March 13

龍朔 begins (ii.30)661 Apr 4
T220 Trans ends 龍朔三年十月二十日663 Nov 15

麟德 begins (i.1)664 Feb 2
Xuánzàng dies麟德一年二月五日664 March 7

As we can see, the stele purports to be from a time a little over a year into the translation of the Prajñāpāramitā Sūtra collection. And at a time when Xuánzàng had retreated from public life in the capital to a mountain retreat 100 km away. If we take this at face value, then Xuánzàng must have "translated" the Heart Sutra attributed to him (T251) before he started this magnum opus. I use scare quotes on "translated" because it is clear from other evidence that he did not translate it. Note that Xuánzàng died within a few months of completing the Mahāprajñāpāramitā Sūtra translation.

All of the circumstantial evidence points away from Xuánzàng's being involved in translating it (see Nattier 1992: 189ff for a survey of the evidence).
  1. The Heart Sutra is a 抄經 (chāo jīng) or "sutra extract" rather than a translation.
  2. The extraction was from Kumārajīva's Chinese translation of the Large Sutra (T223), with other parts inspired by the same text, and a dhāraṇī from elsewhere. It now clearly predates the completion of T220.
  3. Like a lot of English (so-called) "translations" the text attributed to Xuánzàng (T251) is an edited version of an existing text (T250). Two lines were removed and the characters for two names and one technical term (skandha) were changed.
  4. All the terms changed were introduced by Xuánzàng, but were seldom taken up by later translators.
  5. No text translated by Xuánzàng ever replaced one translated by Kumārajīva in popular Chinese Buddhism - Kumārajīva's texts are still in use today.
  6. Xuánzàng's biography mentions him being given the text, not translating it.
  7. Xuánzàng's own travelogue doesn't mention the Heart Sutra at all.
  8. The Heart Sutra does not appear in T220, Xuánzàng's collection of Prajñāpāramitā sutras translated from Sanskrit (though we have reason to believe he already possessed a version). No other Prajñāpāramitā text translated by Xuánzàng occurs outside of T220.
So, Xuánzàng was, at best, an editor of the text. And such edits as occurred were relatively minor. In a forthcoming essay I will show that the character 譯 does not always mean "translate" but can mean precisely "edit". In any case, the resulting text, or one very like it, was attributed to Xuánzàng three years before he died (early in 664 CE) by someone who lived several hundred kilometers away.

Even if the story about the blocked nasal passages is not exactly historically accurate, it probably does reflect the use to which the Heart Sutra was put in the 10th Century, when the commentary was composed. And this is confirmed by other sources. While a handful of scholars studied and interpreted the text as a document of Buddhist ideas, the majority of Buddhists, then and now, see it in magical terms, in which understanding the text is secondary, if it has any importance at all.

The Fángshān Stele can now claim to be the oldest dated Heart Sutra. It forces us to review the relationship between Xuánzàng and the Heart Sutra, though I do not think that we can take the attribution to him at face value. Since the Heart Sutra Xuánzàng had was almost certainly already in Chinese, we cannot say that he translated it. It is possible, even likely, that he edited it for publication. If he did so, it was most likely before he embarked upon his translation of the Mahāprajñāpāramitā Sūtra. And he probably did not include the Heart Sutra in this collection, because it was already in Chinese. If the biography of Huìlì and Yàncóng can be believed, then Xuánzàng treated the text as a locally produced (magically efficacious) dhāraṇī, not as an authentic Indian sutra. However, the commentaries of Kuījī and Woncheuk (which I have written about before) clearly do treat the text as having an Indian origin and as being a text about ideas rather than simply apophatic magic.


Chinese Canonical texts from the CBETA Reader, except where stated.

北京圖書館金石組, 中國佛教圖書文物館石經組編 (1987) ‘房山石經題記匯編’. 书目文献出版社 : 新華書店北京發行所發行, 1987. = The Beijing Library Metal and Stone Group and The Chinese Buddhist books and Cultural Relics Museum Stone Sutra Group. (1987). Classified Compilation of Headings and Records of the Stone Scriptures on Mt. Fang, Beijing: Bibliographic Literature Publishing House and Xinhua Bookstore.

道安 and 張曼濤. (1977)「大藏經硏究彙編」(2 Vols.) 台北: 大乘文化出版社. = Dàoān and Zhāng Màntāo. (1977) Collection of Tripiṭaka Research. (2 Vols.). Taipei: Mahāyana Culture Press.

賀銘, 續小玉, “早期《心經》的版本”,房山石經博物館/房山石經與雲居寺文化研究中心,編輯,《石經研究》,第一輯,頁12-28. 北京:北京燕山出版社,2017年。= He Ming, Xu Xiaoyu. (2017) “the Early recessions of Heart Sutra”, in Fángshān Stone Sutras Museum & Research Center of Fángshān Stone Sutras and Yunju Temple, ed., Stone Sutras Studies, Vol,1, pp.12-28. Beijing: Beijing Yanshan Chubanshe.

Hucker, Charles O. (1985). Dictionary of Official Titles in Imperial China. Stanford University Press.

Li Rongxi (1995) A Biography of the Tripiṭaka Master of the Great Ci'en Monastery of the Great Tang. Numata Centre for Buddhist Translation and Research.

林元白。(1958)「唐代房山石经刻造概况」現代佛學 , 3。 一九五八年。= Lin Yuanbai. 'A General Survey of Fángshān Stone Sutras from the Tang Dynasty. Modern Buddhist Studies, 3, 1958. www.baohuasi.org/qikan/xdfx/5803-011A.htm. Cached copy.

杨亦武. (2003) 云居寺. 华文出版社. = Yáng Yìwǔ (2003). Yún jū Temple. Huawen Publishing House.

Additional Links

Chinese news story: http://www.sach.gov.cn/art/2016/9/27/art_723_133778.html

Video on Fángshān showing caves and stone steles with carved sutras. http://www.ikgf.uni-erlangen.de/videos/china-academic-visit-2013/the-video-on-Fángshān.shtml
Related Posts with Thumbnails