Siku tiyao entry for Shenxian zhuan

Aaron Reich， Kun Li， Ning Luo

(威斯康辛大学麦迪逊分校东亚系,美国威斯康辛麦迪逊,53705；西南交通大学艺术与传播学院,四川成都,610031)

SikutiyaoentryforShenxianzhuan

Aaron Reich， Kun Li， Ning Luo

(威斯康辛大学麦迪逊分校东亚系,美国威斯康辛麦迪逊,53705；西南交通大学艺术与传播学院,四川成都,610031)

BiographiesofDivineImmortals[1]. 10juan.

Collected and Presented by the Liang Huai Salt Administration.

Compiled by Ge Hong葛洪[2](283-343 A.D.) of the Jin Dynasty (265-420 A.D.). This book, according to Ge’s own preface, it is said to have been composed after the completion of the Master Embracing Simplicity Inner Chapters (Baopuzi neipian 抱朴子·内篇)[3], when his disciple Teng Sheng 滕升 asked about the existence of immortals[4]. Recorded in the text are altogether eighty-four individuals[5]. The preface names several hundred individuals that Qin Grand Master Ruan Cang 阮仓[6](?) had recorded, and another seventy-one that Liu Xiang 刘向[7](77-6 B.C.) had compiled. Now we have recopied and collected together the immortals of old as seen in scriptures of immortals, prescriptions of Daoist medicine, the writings of various schools of thought, those spoken of by my late teacher, and those discussed by aged scholars to form ten juan.

The preface further claims that [the entries]Liu Xiang had listed were overly rudimentary, and Ge himself asserts these biographies are improvements over [Liu]Xiang’s[8]. Now in examining its [theShenxianzhuan’s]writings, [we find that]only the two entries for Rong Chenggong 容成公 and Peng Zu 彭祖[9]also appear in [Liu Xiang’s]BiographyofExemplaryImmortals(Liexianzhuan列仙传), and that the remainder all supplement which were not yet included by Liu Xiang[10]. Examples contained in [the SXZ]such as the Yellow Emperor 黄帝 meeting Guang Chengzi 广成子 and Lu Ao’s 卢敖 encountering Ruo Shi 若士 are both of which are parables of Zhuang Zhou 庄周, no different from the likes of Hong Meng 鸿蒙 and Yun Jiang 云将[11], who never actually existed. The king of Huainan 淮南, Liu An 刘安[12](179-122 B.C.), rebellion and suicide, Li Shaojun 李少君[13](2nd century BCE) dying of illness, are both included in theGrandScribe’sRecord(Shiji史记) andtheBookofHan(Hanshu汉书), and indeed there were actually no occurrence of either of them ascending to immortality, [yet]Hong without exception includes [them], [which is]unreasonable farfetched[14]. As for the so-called Xu You 许由 and Chao Fu 巢父, who ingested a yellow elixir left by the stones of Mount Ji 箕山 and are now living in Zhongyue, the central mountain, of modern-day, if these two people remained [alive]in the Jin times, then Ge would have seen them with his own eyes and recorded them, [which is]especially absurd[15].

Nevertheless, the “Biographies of Occult Methods”(“Fangshu zhuan”方术传) inHouHanshuincluded Hu Gong 壶公, Ji Zixun 蓟子训, Liu Gen 刘根, Zuo Ci 左慈, Gan Shi 甘始, and Feng Junda 封君达, all of which largely correspond with [their entries in]theSXZ. Therefore (we) suspect these entries were based on old writings, [as they are not]completely fabricated [by Ge Hong][16]; they were moreover handed down over a long period of time, and later became understood as historical facts. Poets of successive dynasties quoted them continuously so that it is not need to verify the validity of any individual piece.

All written records recordtenjuan, corresponded to the current edition, except in the “Record of Classics”(“Jingji zhi”经籍志) ofSuishu隋书, [where]it is called Ge Hong’sLiexianzhuan, differing in name alone[17].Upon examining theNewandOldBooksofTang’s(Xintangshu新唐书 andJiutangshu旧唐书) respective renditions of Ge Hong’sShenxianzhuan, we know the current version of theRecordofSui,Suizhi隋志 was by chance a mistake; it is not that theShenxianzhuanhad two names.

This edition [in theSKQS]was published by Mao Jin 毛晋[18](1599—1659 A.D.). Pei Songzhi’s 裴松之[19](372-451 A.D.) commentary on the “Biography of Ancient Lord”(“Xianzhu zhuan”先主传) from theRecordofShu,Shuzhi蜀志 quoted from a section on Li Yiji 李意其, a section on Dong Feng 董奉 from the commentary on the “Biography of Shi Xie”(“Shi Xie zhuan”士燮传 and a section on Jie Xiang 介象 from the commentary on the “Biographies of Wu Fang, Liu Dun, Zhao Da”(“Wu Fang Liu Dun Zhao Da zhuan”吴范刘惇赵达传). He considered what Ge Hong recorded came close to deluding the masses, but because Ge’s books and writings were so widely circulated, Pei Songzhi took up and chose several events [from Jie Xiang’s biography]and recorded them at the end of this section[20]. Of these quotations taken from these [Pei Songzhi’s]books, the commentary on theRecordoftheThreeKingdoms(Sanguozhi三国志, completed before 429)[21]is the oldest and corresponded to this edition[in theSKQS]. Then we come to know this is the original edition[22].

TheCollectedBooksofHanandWei(Hanweicongshu汉魏丛书)[23]includes another version[24], whose writings are for the most part identical, but it includes altogether 92 individuals. From the order in which the sections appear, it was compiled by putting together the citations [fromShenxianzhuan]inExtensiveRecordsoftheTaiping[Period](Taipingguangji太平广记)[25]. As there are errors in some of the headings in [Taiping]guangji, and these are also seen in other works, it did not [directly]citeShenxianzhuan. As a result, this edition has quite a few mistakes and omissions[26].Just like the entries for Lu Ao 卢敖 and Ruo Shi 若士, in Li Shan’s 李善 commentary on theSelectionsofLiterature(Wenxuan文选), Jiang Yan’s 江淹ProsePoemsonParting(Biefu别赋), and Bao Zhao’s 鲍照AscenttoHeaven(Shengtianxing升天行), both [Lu Ao and Ruo Shi]are recorded; [and]all [of these sources]name Ge Hong’sSXZ, corresponded to this [theSKQS’s]edition[27]. Because theTaipingguangjidid not cite these entries, theHanweicongshuedition subsequently did not include them, which is adequate enough to prove it [theHanweicongshuedition]is not the complete version.

Original Chinese Text of the Siku tiyao entry

《神仙传》十卷(两淮盐政采进本)

晋葛洪撰。是书据洪自序,盖于《抱朴子·内篇》既成之后,因其弟子滕升问仙人有无而作。所录凡八十四人。序称秦大夫阮仓所记凡数百人,刘向所撰又七十一人。今复抄集古之仙者见于仙经服食方百家之书,先师所说,耆儒所论,以为十卷。

又称刘向所述,殊甚简略,而自谓此传有愈于向。今考其书,惟容成公、彭祖二条与《列仙传》重出,余皆补向所未载。其中如黄帝之见广成子,卢敖之遇若士,皆庄周之寓言,不过鸿蒙云将之类,未尝实有其人。淮南王刘安谋反自杀,李少君病死,具载《史记》、《汉书》,亦实无登仙之事,洪一概登载,未免附会。至谓许由、巢父服箕山石流黃丹,今在中岳中山,若二人晋时尚存,洪目睹而记之者,尤为虚诞。

然《后汉书·方术传》载壶公、蓟子训、刘根、左慈、甘始、封君达诸人,已多与此书相符。疑其亦据旧文,不尽伪撰,又流传既久,遂为故实。历代词人,转相沿用,固不必一一核其真伪也。

诸家著录皆作十卷,与今本合,惟《隋书·经籍志》称为葛洪《列仙传》,其名独异。考新、旧唐书并作葛洪《神仙传》,知今本《隋志》殆承上《列仙传》赞之文,偶然误刊,非书有二名也。

此本为毛晋所刊。考裴松之《蜀志·先主传》注,引李意其一条,《吴志·士燮传》注引董奉一条,《吴范刘惇赵达传》注引介象一条,并称葛洪所记,近为惑众,其书文颇行世,故撮举数事,载之篇末。是征引此书,以《三国志》注为最古。然悉与此本相合,知为原帙。

《汉魏丛书》别载一本,其文大略相同,而所载凡九十二人。核其篇第,盖从《太平广记》所引钞合而成。《广记》标题,间有舛误,亦有与他书复见,即不引《神仙传》者,故其本颇有讹漏。即如卢敖若士一条,李善注《文选》江淹《别赋》鲍照《升天行》,凡两引之,俱称葛洪《神仙传》,与此本合。因《太平广记》未引此条,《汉魏丛书》本遂不载之,足以证其非完本矣。

注释：

[1]Scholars have translated the title in a number of ways. Robert Campany, one of the leading authorities on theShenxianzhuan, renders it asTraditionsofDivineTranscendents(2002), which is also used throughout theEncyclopediaofTaoism(2008). In the present translation of theSikutiyaoentry, we have opted forBiographiesofDivineImmortals, following Stephen Durrant’s entry in theIndianaCompanion(1986). There is no complete version of theBiographiesofDivineImmortals,Shenxianzhuan神仙传 (hereafterSXZ) in the Daoist Canon. Instead, most scholars refer to theLongweibishu龙威祕书 edition (1794, reconstructed largely fromTaipingguangji太平广记 or to the Ming version included in theSikuquanshu四库全书 (1782). See Benjamin Penny (2008),pp.887-888. Peter Bumbacher argues that the text was written before 317/8. See Bumbacher (2000),p.732.

[2]Various dates ranging from 253 and 363 have been indicated for Ge Hong in the past, but most scholars agree on those marked here,pp.283-343. Ge’s style name is Zhi Chuan 稚川, and one of his epithets was the Master Embracing Simplicity, Bao Puzi 抱朴子, which doubled as the title of one of his most well-known work. He heralds from Danyang Gourong 丹阳句容, modern day Jiangsu Province. His histories derive from several texts, includingJinshu晋书,Luofuji罗浮记,Jinzhongxingshu晋中兴书,Daoxuezhuan道学传 andMasterEmbracingSimplicityOuterChapters,BaopuziWaipian抱朴子·外篇. To posterity Ge is known as an encyclopedia writer in the Jin, reportedly writing more than seventy books, though most of them have been lost. Scholars believe Ge was over 30 years old by the time he wrote theSXZ. See Fabrizio Pregadio (2008),pp.442-443 and Chen (2008),p.318.

[3]Ge Hong’s most famous work, theBaopuzi is divided into Inner Chapters mainly devoted to descriptions and comments of religious practices, and Outer Chapters concerning “discourses of the literati”(rushuo儒说). See Pregadio (2008),p.215.

[4]This section of theSikutiyaoentry derives from the introduction toSXZ, which reveals the relation between it andBaopuzineipian. Chen Shangjun asserts that Ge wrote the work in response to Confucian scholars who had criticized Daoism, hoping that it would serve as evidence from the existence of immortal beings. See Chen (2008),p.318.

[5]Liang Su 梁肃 (753-793A.D.), a Buddhist scholar of the Tang, reports in his “Shenxian zhuan lun”神仙传论 (“On the Shenxian zhuan”;QuanTangwen[Zhonghua shuju reprint of the 1814 edition,p.519.10a-11a]that theSXZoriginally had 190 biographies while modern versions have only ninety or so. See Penny (2008),p.887.

[6]The dates for Ruan Cang’s life remain unknown, but it is reported that he wrote a book that does not survive namedLiexiantu列仙图, which recorded more than seven hundred immortals, ranging from Huang Di 黄帝, Yao 尧, Shun 舜, Yu 禹, and others, up to the Qin period. In Eastern Han, The Emperor of Zhang 章帝 (58-88A.D.) granted it to Liu Cang 刘仓(c.29-83A.D.), but this version lost. See Chen (2008),p.319.

[7]Liu Xiang, style name is Zi Zheng 子政, was a famous writer in the Han Dynasty. He wrote theLiexianzhuan, which reportedly had originally recorded seventy-two immortals, though now it lists only seventy. See Chen (2008),p.319.

[8]Durrant notes that Ge’s entries are indeed much more detailed than those of theLiexianzhi, wherein the longest biographies are rarely more than two thousand words, and some consist of only one or two lines of text. See Durrant (1986),p.678 and Chen (2008),p.320.

[9]Ge Hong expanded the story of Peng Zu from the seventy-eight characters of theLiexianzhuanto more than 1500 in theSXZ. See Chen (2008),p.320.

[10]The additional immortals can be divided into two types. The first kind includes those who lived before the Western Han that Liu Xiang did not list, such as Guang Chengzi 广成子, Ruo Shi 若士 and Bai Shisheng 白石生. The second type includes individuals who lived in the periods between Liu Xiang and Ge Hong, such as Zhang Daoling 张道陵, Zuo Ci 左慈, Ge Xuan 葛玄.

[11]Hong Meng and Yun Jiang were completely fictitious characters. They appeared in aZhuangzi庄子 chapter entitled “Zaiyou在宥”, wherein Yun Jiang travels to the east and meets Hong Meng while passing by the Fu Yao Tree. See Chen (2008),p.319.

[12]Liu An’s suicide was recorded inShijiandHanshu. TheSXZstates that Liu An had inclinations toward Daoism and consumes an elixir of immortality to rise to Heaven. See Chen (2008),p.321.

[13]Li Shaojun was among the many Masters of Methods,fangshi方士 invited to the court of Emperor Wu of Han, Han Wu Di 汉武帝, who had a strong interest in proto-Daoist longevity techniques. Several texts, for exampleShijixiaowubenji史记·孝武本纪,Shijifengchanshu史记·封禅书, andHanshujiaosizhi汉书·郊祀志, record that he died of illness; however, it is recorded that Empire Wu maintained that Li Shaojun ascended to Heaven, and it is this latter account that appears inSXZ. See Chen (2009),p.320.

[14]The persons ofSXZcan be divided into two types: fictional and historical. Huang Di and Guang Chengzi are examples of the first type, while Liu An and Li Shaojun are examples of the second.

[15]The stories of Xu You and Chao Fu are not actually recorded inSXZ, only their names appear in the entry of Wei Shuqin 卫叔卿, where it states that he played game with the two of them. Because Wei Shuqin lived in the period of Empire Wu of Han, and Xu You and Chao Fu were contemporary with Yao 尧, Chen Shangjun concludes that readers should be skeptical towards the content ofSXZ. See Chen (2008),p.321.

[16]Due to similarities between theFangshuliezhuanandSXZ, Chen Shangjun and other scholars generally conclude that Ge Hong probably functioned more as an editor than an author, drawing together stories from a variety of sources. Ge Hong states in his introduction that he learned of these stories via external sources, which corroborates this interpretation of Ge Hong’s editorial role in the production of the text. Chen suspects that it shares similar root sources with those Ge used to compile theBaopuzi. See Chen (2008),p.322.

[17]The attribution of theSXZto Ge Hong remains consistent from the Sui period onward; however, the reference to the text in the bibliographical treatise ofSuishuis the most problematic. It reads: “Liexianzhuan,tenjuan, compiled by Ge Hong.”SeeSuishu隋书 (Beijing: zhonghua shuju, 1973), 33:979. See also Penny, “Text and Authorship,”p.174.

[18]This is the version included in theSikuquanshu. While theSXZwas lost during the Southern Song, three editions—each in 10juanand each attributing authorship to Ge Hong—appear at the end of the Ming within a period of seven years. The third of these was printed during thechongzhenperiod (1628-1644A.D.) by Mao Jin in hisJiguge汲古阁. For more on Mao Jin, see Campany (2002),p.385. The originalSXZhad tenjuan, which is confirmed by a number of external sources. Although some material seems to have been lost during the Song, theSXZstill divides into tenjuan. See Chen (2008),p.323.

[19]Pei Songzhi was a scholar of the Former Song 前宋 who doubted the veracity of the biographies contained in theSXZbut cited them nonetheless in his writings. See Durrant (1986).

[20]This sentence of our translation is adapted from Benjamin Penny (1996),p.166.

[21]Pei Songzhi completed hisSanguozhizhuin 429, approximately 100 years after theSXZappeared; it is the oldest book to cite theSXZ. See Chen (2008),p.322.

[22]Chen Shangjun notes his discrepancy with the editors ofSikuquanshu, stating that Mao Jin’s version could not be the same as the original book due to textual variants. He notes that the numbers of immortals differ across versions.SikuQuanshu’shas eighty-four individuals, while Liang Shu’s 梁肃 (759-793A.D.)Shenxianzhuanlun神仙传论 is said to include 190. In the introduction toXianyuanbianzhu仙苑编珠, compiled by Wang Songnian 王松年 (fl. the Five Dynasties period), he states theSXZincluded seventy-three individuals. TheDaozang道藏 compiled in the Ming Dynasty did not include theSXZ, and theDaozangquejingmulu道藏阙经目录 states that theSXZhad been lost, facts which attest to its scarcity during the Ming dynasty. See Chen (2008),pp.323-324.

[23]Bumbacher notes thatSikutiyaospeaks simply of the “Han Wei congshu,” yet thecongshubearing this title and having been published during the Ming Wanli period did not contain theSXZ. He suggests that either theGuangHanWeicongshu, published in 1592, or theZengdingHanWeicongshu, published during the Qingganlongperiod (1736-1795A.D.) must be meant, believing the former being the more likely one. See Bumbacher (2000), 800, note 296.

[24]There are two version ofSXZstill extant: one in theHanweiCongshu, and the version published by Mao Jin, which is the one included in theSKQS. The primary difference between the two versions is the numbers of individuals each records. TheHanweicongshurecords ninety-two individuals, while theSikuQuanshuversion only includes eighty-four, with only seventy-seven being the same across both versions. Among these seventy-seven are further variants in content. See Chen (2008),p.325.

[25]TheTaipingguangjiis a large collection of supernatural events throughout ancient history. One of the so-called “four large books” (Songsidashu宋四大书) of the Northern Song 北宋 (960-1126A.D.), it was compiled on imperial order by Li Fang 李昉, Hu Meng 扈蒙, Li Mu 李穆 and other members of the National University (taixue 太学) during the years 977 and 978. See Cheng (1995).

[26]We have adapted part of this section from Penny’s translation. See Penny (1996),p.183.

[27]Li Shan’s commentary onWenxuan文选 cites theSXZa total of fifteen times. Jiang Yan’sBiefu, Bao Zhao’sShengtianxing, and also Jiang Yan’sGuonongnongyouxianall cite the stories of Lu Ao and Ruo Shi, which also appear in the Mao Jin version. The version ofHanweiCongshudoes not include these entries, which proves thatHanweiCongshudoes not include the entire version of theSXZ. Chen Shangjun indicates that Mao Jin’s version could be a fragment of a Song Dynasty edition. Yu Jiaxi 余嘉锡(1883-1955A.D.) argues that there were two editions, but that Mao Jin was more comprehensive and careful with his editing. See Chen (2008),p.326.