A Textbook of Physiognomy: The

Tradition of the Shenxiang quanbian

By

L iv ia K ohn

Kyoto

Chinese physiognomy as the study of the manifestations of man’s fate

in his body is not well known in the West. All that has been generally
accepted are early traces concerning the principles and basic notions
of this divinatory practice, mostly found in philosophical treatises and
biographies of ancient China. Manuals or textbooks of physiognomy,
however, which give concrete and detailed instruction on how to judge
a person’s character and fate from his body are not found before the
Tang. Dated textbooks are only extant from the 10th century onward.
Later there is a great variety of physiognomical manuals, but only a
limited number of these texts survive today.
Hardly any of them have been brought to scholarly attention. It
is the aim of this article to trace the material found in the Shenxiang
quanbian, the most complete and most widely available textbook of
Chinese physiognomy, to earlier, still surviving materials. The study
is not primarily concerned with physiognomy as such; it does not de­
scribe either the theory or practice of body divination. Rather, it
limits itself to an attempted delineation of a tradition of physiognomy
from the earliest manuals of the late Tang to the Shenxiang quanbian
of the early Ming. The major patriarchs and masters of the trade are
described; the main characteristics of the standard type of physiognomy
transmitted to the present day are outlined.
It is found that the physiognomy still practised today can be charac­
terized by three features:

1 . It presupposes a theory on the interaction of body and spirit.

2. It gives detailed rules for analysis of body types and features
according to the theory of the five phases as well as according
Asian Folklore Studies, V o l .45,1986, 227—258
228 I'lV IA K O H N

to animal types.
3. It uses facial complexion analysis, again in accordance with the
five-phases-theory.

Manuals belonging to the main, i.e.，still active, tradition of Chinese

physiognomy inevitably follow this pattern in their description of how
to practise body analysis. The earliest text extant of this type is the
Yuguan zhaoshen ju attributed to Song Qiqiu of the 10th century.
While individual methods of analysis can be traced back much further,
it seems that the streamlined tradition still alive today goes ultimately
back to the Five Dynasties’ period.

T he T ext

The Shenxiang quanbian 示申ネ目全編，‘‘ Complete guide to spirit phy­

siognomy,n is the leading standard textbook of the Chinese traditional
way of reading a person’s health, character, and— most of all— fate, in
his physical appearance. Originally compiled by Yuan Zhongche 袁忠
徹 ，alias Gongda 公達 or Jingsi 靜思，of the early Ming dynasty (1367—
1458)，who was a well-known physiognomist in his days，the book is
now extant in a late Ming edition preserved in the National Central
Library in Taipei, Taiwan. In addition, it has been reprinted in the
great Chinese encyclopedia Gujm tushu pcheng 古 今 圖 書 集 成 （ chapters
631-644). This latter text in turn served as a basis for modern Tai­
wanese pocket-book editions such as the one prepared by Liang Xiang-
run (Liang 1980).
The fact that the Principles of the Shenxianp auanbian are still ap­
plied today is apparent in William Lessa’s work on Chinese physi­
ognomy, or rather “ body divination ” or “ somatomancy， ” as he prefers
to call it (Lessa 1968). An anthropologist, himself without any more
than a basic working knowledge of Chinese, Lessa gained his insights
through the active cooperation of prominent practising physiognomists
of Taiwan, such as Zhang S h i n g u . 1 hereby he made sure that he
dealt with an active practice of the present, not a dead remnant of the
past. His book, Chinese Body Divination, is entirely based on the
Shenxiang auanbian as far as the description of Chinese practices and
concepts goes.
Yet another indication of the present-day importance of the Shen­
xiang auanbian can be found in the fact that the standard Japanese
manual of physiognomy edited by the Tokyo Shrine Administration is
merely an abridged Japanese translation of the old Chinese text (Jingu-
kan 1982).
The text itself divides into twelve chapters in the old Ming edi­
 A T E X T B O O K O F P H Y S IO G N O M Y 229

tion, and into fourteen in the reprint made on the basis of the Tushu
jicheng which follows the latter’s division (chapters 631-644).1 The
order of the material and the material itself, however, are quite the
same. There are only three instances in which illustrations in the
original have been omitted in the reprint:

1 . I n chapter 4 of the Ming edition eight basic human disposi­

tions are diagnosed with pictures of typical faces attached, but
these pictures are not integrated in the reprint edition (Liang
1980: 111).
2. In the same chapter, six standard conditions of life, such as
wealth, nobility, poverty, loneliness, long life, and early death,
are exemplified in text and illustrations, but only the text is
found in the reprint (Liang 1980: 112-115).
3. In the last chapter, there are twenty-four different kinds of
energy analyzed, described, and illustrated, but the illustra­
tions do not appear in the reprint (Liang 1980: 344-345).

Looking now at the old edition, one finds that the first and second
chapters deal mainly with physiognomic principles in general featur­
ing treatises ascribed to such venerable personages as Lii Dongbin,
Guiguzi, Tang Ju, X u Fu, Guan Lu, and even Bodhidharma. There
are also simple expositions on the nature of spirit 神，energy 氣，body 形，
and their interrelation with man and the universe. In addition, about
forty basic physical types are outlined.
Beginning with chapter 3, physiognomic details are analyzed: the
face, the hair, the eyebrows (30 kinds), the eyes (3b Kinds), the nose
(24 kinds), the ears (14 kinds), then the mouth (14 Kinds) together with
details like the lips, teeth, etc. With chapter 4，the whole body is
gradually explored: the back, breasts, hips, legs. Then physical details
such as body movements and activities are interpreted: eating, walking,
standing, sleeping, etc. Basic human dispositions are diagnosed, man
is understood in terms of the theory of the five phases and their respec­
tive geographical directions as well as in terms of the three realms of
heaven, earth, and man.
Chapter 5 sees a return to general expositions which are continued
through chapters 6 and 7. Various methods according to different
masters are given. Chapter 8 discusses the shapes, lines, and signs on
hands and feet, chapter 9 deals with women, children, and with types of
bodies in general(50 kinds). In chapter 10, the meaning of black spots
on different parts of the body is explained along with an exposition on
complexion 色 . This is continued into chapter 1 1 . The work ends
230 L IV IA K O H N

with the diagnosis of diseases by means of complexion analysis and a

description of different types of energy in chapter 12.
This great compendium of theoretical concepts, detailed analysis,
and systematization of Chinese physiognomy can be rightly viewed as
heir to a long tradition. It cites works of all the great masters of phy­
siognomy and uses earlier manuals’ thus summarizing the experience
of ages. The richness gained thereby is what has enabled it to remain
a guide for actual practice down to the present day. Historically speak­
ing, what can we say about this tradition? How much of the material
presented in the Shenxiang quanbian can be traced back how far in
Chinese history? Can we delineate a tradition of physiognomy through
the ages in terms of historical scholarship, i.e., by citing and dating
earlier material still extant? In order to provide the beginning of an
answer to these questions five different aspects shall be considered:

1 . The orthodox tradition— authorship and line of transmission

as claimed in the text itself.
2. Materials from old masters—works cited in the text under
specified names of authors.
3. Comparison to earlier manuals—works used in the text without
any ascription, but found in earlier texts.
4. Forerunners of the tradition— single systems of analysis as
presented in physiognomic texts discovered in Dunhuang.
5. A different system— physiognomy as described in the Yue-
bodong zhongji.

T he O rthodox T r a d it io n

The text itself yields some formal information on its origin. Whereas
we know that it was compiled by Yuan Zhongche, the Ming edition
claims Chen Tuan 陳搏 of the 10th century as its author, with Yuan
Liuzhuang 袁柳莊 as the main transmitter. What do we know about
these men?
Liuzhuang is the alias of Yuan Gong 袁珙 also called Tingyu 廷玉
(1335-1410). He was the father of Yuan Zhongche. Botn fuans were
well known as physiognomists in the early Ming period, both served as
government officials and did numerous prognostications for prominent
personages. Gong even persuaded the monk Yao Guangxiao 姚廣孝
to come out from his monastic seclusion and serve the state. He sup­
posedly recognized Zhu J J i 朱棣 ，later Yongle 永樂 emperor, at an early
age and thereby encouraged him to usurp the throne. Strongly favored
by this emperor, Gong also had a hand in selecting the heir-apparent
(Goodrich, Feng 1976: 1638-1641). After his death in 1410，his son
 A T E X T B O O K O F P H Y S IO G N O M Y 231

Zhongche became the most influential physiognomist at the imperial

court, accompanying the emperor on the expedition against the Mongols
in 1422，where he made some correct predictions about battle situations
and outcomes. Serving under the subsequent emperors, he retired
in 1439 to write and compile several books on physiognomy, including
the Shenxiang quanbian (Goodrich, Feng 1976: 1629-1632).
From this basic information one may rightly conclude that Zhong­
che, the compiler of the Shenxiang quanbian、 learned his skill and re­
ceived the tradition from his father. Yuan Gong, however, does not
seem to have stood in a family tradition. According to legend, a monk
physiognomist named Bieguyai 別古崖 induced him to take up the art
(Goodrich, Feng 1976: 1639).2
Our historical knowledge of Chen Tuan, alias Tunan 圖南 is even
less than that concering the two Yuans. He was born in the latter
half of the 9th century in Henan. Between the years 900 and 930，he
spent much time wandering around famous mountains. During this
period he stayed for a longer while on Mount Wudang where he prac­
tised Taoist techniques of mediation, including breathing exercises
and fasting. In 937 he is found in oichuan where he leaves an in­
scription. It may be assumed that he then resumed his migrations to
settle eventually on Mount Hua in ohensi. Here he spent the rest of
his life— a considerable span, since he supposedly died at an age of 118
in 989.
Between the 940s and his death in 989 he visited the imperial court
twice, once in 956 under Zhou Shizong, and another time in 984 under
Song Taizong. At this latter occasion he was honored with the title
X iyi xiansheng 希夷先生. In addition, a few other reports on Chen
Tuan can be considered within the range of historical authenticity.
He once gave a prophesy for Qian Ruoshui 錢若水，stating that this
high-ranking official would end his career far before the expected time.
He recognized that Zhang Yong 張詠 would have a splendid military
career despite his failure in the official examination. His methods of
prognostication were mainly those of the Ytjtng 易經 and of physi­
ognomy. He passed these methods on Chong Fang 种放，from whom
they were eventually transmitted to such eminent Song dynasty philo­
sophers as Shao Yong 邵雍 and Zhou Dunyi 周 敦 頭 （ Knaul 1981).
From the stories relating Chen Tuan’s prognosticatory activities,
a picture quite similar to that of the two Yuans is drawn. However,
legend has it that his skill in physiognomy was not that of a professional
(fangshi 方士 ), rather it is described as being part of ms oneness with
the Tao. As a Taoist saint he could recognize the flow of energies
before they became mainfest, he could feel the trends of the universe
232 L IV IA K O H N

in his body and act accordingly.

Chen Tuan himself was allegedly taught by the famous immortal
Lu Dongbin 呂洞賓 and a rather obscure Taoist named Mayi 麻衣 .
The latter is reported to have given prognostications on the basis of
energy analysis and physiognomy; a work entitled Xinfa 心法 has been
ascribed to him (Knaul 1981:76). Lii Dongbin is a very colorful
figure who has been made into the patron saint of various Taoist prac­
tices and sects. Most of his fame seems to be legendary, since relevant
sources on him dated before the Song are scarce (Baldrian-Hussein

According to the tradition claimed by the Shenxiang quanbian it­

self, the methods of physiognomy set down in this text go ultimately
back to the legendary Lii Dongbin and to the saint Chen Tuan of the
rive Dynasties and early Song. From biographical information on
these early masters of physiognomy the claim of the Shenxianp auanotan
could not be substantiated. Chen Tuan is not officially credited with
a work on physiognomy, even though he is famed for his successful ap­
plication of this art.
Historically speaking, all that can therefore be said so far is that
the text was compiled by the younger of the two Yuans in the early
Ming who considered themselves heirs of the 10th century sage.

M a t e r ia l s From O ld M asters?

Looking now at the various texts contained in the Shenxiang quanbian,

one finds that those attributed to a particular author can be divided
into four groups:

1 . Texts attributed to famous physiognomists before the Tang,

in which cases the attributions are likely to be groundless.
There is no evidence that these texts were really written by
the authors to whom they are ascribed. Rather, the authors
themselves often turn out to be physiognomists only according
to legends. To this group belong texts attributed to Tang
Ju, X u Fu, buan Lu, Bodhidharma, and Guo Linzong.
2. Texts attributed to physiognomists of the Tang, i^ive Dynasties,
and Northern Song. Again, these may be nothing but inven­
tions, but here the possibility of finding circumstantial evidence
for their authenticity— or at least earlier attributions— may not
be excluded. In this group we are dealing mainly with the
three figures mentioned above, i.e., Lii Dongbin, Mayi, and
Chen Tuan. In addition, there are two minor personages,
Xiyu Xiansheng and Hu Seng.
 A T E X T B O O K O F P H Y S IO G N O M Y 233

3. Texts listed under the name of some later, i.e., Southern or

post-Song author who can be positively identified and in
whose case the text might actually be found in another edi­
tion. Here the two Yuans are included as well as Zhang
Xingjian, the author of the Renlun Datongfu.
4. Texts quoted under the name of personages who cannot yet
be identified, such as Luo Zhenren, Yan Shi, Yandian Daoren,
Guiyan Xiansheng, as well as Qiu Tanyue, Taiyi Zhenren, Wu
Xinjian, and Yao Kuazang.

Pre-Tang
Tang Ju 唐擧 was one of the most famous physiognomists under the
Han. She lived in the third century b. c . and is reported to have
made a prediction for Li Tuo，prime minister to the first Chinese em­
peror Qin Shihuang. ‘‘ Within a hundred days you will obtain the
power of the empire ’，{Shiji 79，Qianfu lun 潜夫論 27).3 Tang Ju, who
was originally from Liang, is also mentioned in the “ Feixiang pian ”
非ネ目篇 of the X unzi 荀子 ，an essay critical of ‘‘ the nonsense of physi­
ognomy 55 (3.5). None of her works— if there were any— survive or
are mentioned in the bibliographical section of the Hanshu. Yet the
Shenxiang quanbian quotes two titles in her name:
Xiang shenqi 相神氣 ，“Analyzing spirit and energy” (chapter 1;
Liang 1980: 21-24);
Xuantan shenmiao jue 玄談神妙訣，“ Secret exposition of the won­
ders of spirit ” (chapter 7; Liang 1980: 186).
Xu Fu 許負 has been credited with many short concrete explanatory
sections rather than with theoretical discussions. About tms greatest
physiognomist of the Han two stories have been transmitted. In both
cases X u J^u makes a correct prediction about the future of a noble
person. Zhou Yafu 姑： IMS夫 ，he said, would attain a high rank, yet end
up dying of starvation {Shiji 49; Qianfu lun 27; Lunheng 3; Forke 1907:
308-309). Princess B o , 溥 ，he predicted, would give birth to an em­
peror {Shiji 49; Oianfu lun 27). In the former story concrete indica­
tions are mentioned for the first time in Chinese literature: the sign of
future starvation is found m the “ prependicular lines converging at
the corner of the mouth ” (Forke 1907: 308).
The fact that the first mention of concrete indications is made in
relation to X u Fu may have established a tradition of attributing con­
crete and detailed information to this particular master of old. The
Shenxiang quanbian credits mm with the following sections:
Xiang deqi 相德器，“Analyzing the vessel of virtue” (chapter 1，
234 L IV IA K O H N

Liang 1980: 24-28);

Lun nannii wuguan _ 男 女 玉 “ On the five senses in men and
women ’，(chapter 2，Liang 1980: 58);
Tingsheng pian 聽聲篇 ，“ Hearing and voices ” (chapter 2，Liang
1980: 69);
Xiang kou pian 相ロ篇，“Analyzing the m outh” (chapter 3, Liang
1980: 98);
Xiang chun pian ネ目唇篇，“Analyzing the lip s” (chapter 3, Liang
1980: 100);
Xiang chi pian ネ目齒篇，“Analyzing the teeth ” (chapter 3, Liang
1980: 101);
Xiang shou pian 相手篇，“Analyzing the hands” (chapter 8，Liang
1980: 239);
Xiang zu pian 相足篇，“Analyzing the feet，
’ (chapter 8, Liang 1980:
249).4

Guan Lu 管 路 (209-256) was one of the most famous fate-calculators

and prognosticators of his time. One of a group of diviners at the court
of Cao Cao，he performed many oracles for various high officials. In
his official biography {Sanguo zhi 29) his methods are in two cases related
to physiognomy. In general he relied more on stem-branch calcula­
tion in association with the theory of the five phases, a method closely
linked with his name (Needham 1956: 358). Once when visiting Li
Guo, he recognized the signs of future death through execution in two
of his other guests from the energy between their “ celestial court， ’
(天庭，the area between the eyebrows) and their ears. In another case
he used phySiognomic rules to interpret a dream.
As concerns his works, only one text bearing his name is extant
today, the Guanshi dili zhimeng 管氏地理指蒙，“ Mr. Guan’s geomantic
indicator•， ’ ih c Shenxiang auanbian attributes two works on phisy-
ognomy to him:

Renlun yuarCao fu 人倫淵奥賦，“ The depth of man ” (chapter 7，

Liang 1980: 195-196);
Xiang ying’er 相嬰兒 ，“Analyzing children ’’ (chapter 9，Liang 1980:
264).

ouiguzi 鬼谷子 ，according to standard transmission, was originally

named Wang Xiang 王詡 and lived toward the end of the Zhou dynasty.
He received the Tao directly from the Highest Venerable Lord 太上老
君 and withdrew ot Mount Yunqi 雲氣山 where he managed to retain
extraordinary youthfulness while livin? only on herbs. As his dwelling
was situated in a deep demonic valley he came to be known as the Master
 A T E X T B O O K O F P H Y S IO G N O M Y 235

of the Demon Valley, Guiguzi. He had two famous students, Zhang

Y i 張 義 and Su Q in 蘇秦 .5
Stories concerning Wang Xiang are first found in Shihji 69, but
the name Guiguzi is not mentioned here. On the other hand, there is
a text of this name preserved in the Taoist Canon (DZ 1025), but it
deals mainly with methods of meditation and Taoist concepts of the
mind. According to H u Yinglin’s 胡應麟 Sibu zhengwei 四部正偽，this
text probably goes back to tne works of Znang Yi and du Qin men­
tioned in the bibliographical section of the Honshu. Only a rather in­
direct reference to physiognomy is founa in the text Guiguzi: ‘‘ Dis­
regarding the beauty or ugliness of appearance, one can recognize peo­
ple’s feelings from their features” (DZ 1025; 2，2 8 b ) . 1 he commen­
tator adds, “ One should analyze people’s features to know their feel­
ings."
guiguzi is usually associated with astrology rather than with physi­
ognomy. Some astrological methods have been transmitted under
his name.6 But at least there is some relation according to the tradi­
tional image of Guiguzi to the sciences of fate-calculation. The Shen­
xiang quanotan claims three texts go back to him:

Xiangbian weimang 相辨微芒，“Analyzing and judging the subtle

and the vast ’’ (chapter 1，Liang 1980: 17-19);
Xiang furen ge 相婦人歌，“Analyzing women ’’ (chapter 9, Liang
1980: 253-254);
Yuchang tu 玉掌圖，“ Jade palm chart” (chapter 8). In Liang
1980: 217，as well as in the Japanese edition (Jingukan 1982: 124)，
this chart of the lines and mounts of the palm is found under the
title Bagua shiergong zhi tu 八卦十二宮之圖，“ Chart of the eight
trigrams and twelve palaces,” without the ascription to Guiguzi.
Even more obscure is the relation of Bodhidharma 達磨 ，the semi­
legendary founder of Cmnese Chan 禪 ，to the science of physiognomy.
According to the account generally accepted, Bodhidharma came from
India in about 4フ0 and propagated Buddhism. His most famous ac­
complishment is the meditation of wall-gazing {biguan 壁觀）which he
supposedly practised for nine years without interruption. His life and
death began to be embellished by legend from a very early time (Yam-
polsky 1967: 7-10).
As his legend spread, his venerable name was also cited by other
traditions. In Song dynasty Taoism, for example, several methods
of meditation were transmitted under his name.7 It might therefore
be assumed that physiognomists, too, wanted to profit from the widely-
spread fame of the first Chinese patriarch of Chan. They attributed
236 L IV IA K O H N

the following texts to him in the Shenxiang quanbian:

Wuguan zonglun :1l管總論，“All about the five senses ” (chapter

1，Liang 1980: 46);
Xiang zhushen youqi 相主神もじ，“ Seven ways to analyze mastery
and spirit ” (chapter 2，Liang 1980: 66-67);
Yan you qixtang 眼有七ネ目, “ Seven ways to analyze the eyes ’ ，
(chapter 3, Liang 1980: 85; Japanese Jingukan 1982: 168);
Dongjing lun 動靜論，“Activity and rest” (chapter フ，Liang 1980:
199);
Furen xiang 婦人相 ，“Analyzing women ” (chapter 9，Liang 1980:
251).

、ju o Linzong 林宗 was an official of the latter Han dynasty, his

proper name was Guo Tai 享K泰 or 1 a i 太 . He was well versed in dif­
ferent walks of life, but one finds he had no specific reputation as a
fate-calculator or physiognomist {Hou Hanshu 98). In the Shenxiang
quanbian, only one text is ascribed to h im :

Xiang wude t>ei wuxing ネN瓦德配Ji:行 ，“Analyzing the five virtues

in correspondence to the live phases ’’ (chapter 1，Liang 1980: 19­
21).

In all, the study of the attributions of texts made by the Shenxiang

quanbian to personages before the 1 ang reveals six famous masters.
Two of them, Tang Ju and X u Fu, are known as physiognomists, two
others are semi-legendary religious masters, i.e.’ Guiguzi and Bodhi­
dharma, one is a famous fate-calculator with occasional recourse to
physiognomy (Guan Lu), and ^u o Linzong, finally, is an official with
no known relation to divinatory practices. Although none of the at­
tributions can be substantiated historically, one gains some idea of
how the tradition imagined its forefathers. In the case of X u Fu, for
example, the credit wnich the Shenxiang quanbian has given him for
concrete instructions on the analysis of individual parts of the body is
true to the image found in other stories about this diviner. All the
material, furthermore, ascribed to Bodhidharma is rather aphoristic
and slightly obscure—just as befits the traditional image of the ven­
erated founder of Chinese Chan. Guiguzi, finally, though traditionally
more famous as an astrologer than as a physiognomist, is credited with
a chart of the human palm where mounts and lines are patterned ac­
cording to various constellations of the stars.
The conclusion that we may draw here is therefore that while the
ascriptions made by the Shenxiang quanbian to personages before the
 A T E X T B O O K O F P H Y S IO G N O M Y 237

'l'ang are not historically verifiable, they are yet very much in line with
the image of the authors found elsewhere in Chinese folklore and reli­
gion. We could not establish an early historical tradition of physi­
ognomy on the basis of this material, but we could gain a feeling for
the identity of the later masters in terms of tradition.

Tang-Northern Song.
Before dealing with the various texts ascribed to the group associated
with Lii Dongbin, Chen Tuan, and Mayi, let me briefly introduce the
two major textbooks of physiognomy which may provide some circum­
stantial evidence for the allegations of the Shenxiang quanbian.
First, there is the Yuguan zhaoshen ju 玉 管 照 神 局 (hereafter ab­
breviated Ymuan)y attributed by all sources to Son? Qiqiu 宋齊邱 of
the Southern Tang. Song was a high official versed in various sciences
of divination, yet the authors of the Siku quanshu zongmu tiyao 四庫全
書總目提要（ hereafter Siku tiyao) thought it doubtful that he wrote the
text himself ( I I I ，2263). The work, which cites Chen Tuan but never
uses his honorific title and might therefore be dated to before 984，was
found in various libraries during the Song.8 The eaition published
in the Siku quanshu goes back to the Yongle dadian 永樂大典. ri his
seems quite identical with the edition available today, i.e.，the one
contained in Lu Xinyuan’s 陸心源 Shiwcm juanlou congshu 十萬卷樓叢書
of about 1880. This edition in turn served for the reprint by Sunny
Books in Taipei, 1982.
As extant today, the Yuguan consists of three chapters. I'he
first (26 Chinese pages; Sunny Books ed.，3-54) mainly gives theoretical
account s . 1 he second (36 pages; Sunny Books ed.，55-126) deals with
details of body analysis, and the third (27 pages; Sunny Books ed.,
127-180) describes the appearance of different types of energy in the
complexion. Although the text itself might not stem directly from
Song Qiqiu of the Southern Tang, it is very probable that it goes back
to the early Song period. Even the critical Siku tiyao authors admit
that it quotes only material from before the Song and was obviously
well known in this period-
More doubts have to be entertained in connection with the other
manual in which materials of the Shenxiang quanbian can be located:
the Taiqing shenjtan 太淸神鑑 in six chapters. Allegedly by Wang Pu
王朴 ’ a high official under Smzong of the Latter Zhou (Jiu Wudai shi
128, Wudaishi 31), this text is not mentioned in any of the bibliogra­
phical sources of the Song. In addition, its introduction claims that
Wang compiled it in a grotto on Mount Linwu 林屋 in Jiangsu, but
detailed study of his life reveals that he never went south of the Yangze
238 L IV IA K O H N

{Siku tiyao I I I ，2264). Similarly any connection of Wang with the

arts of prognostication belongs in the realm of legend. One must there­
fore assume that this text stems from the late Song, if not Yuan, period.
It existed in a Yongle dadian edition, the one accessible to the Siku tiyao
authors. Today two editions can be found: One in Zhang Haipeng’s
張海鵬 Mohai jinhu 墨 海 金 壺 （ 1920 ed.; reprint in the Congshu jicheng
叢書集成) ；another in Qian Xizuo’s 錢熙Jp乍Shoushange congshu 守山閣叢書
(ed. 1844). In addition, the edition of the Siku quanshu might be
available by n o w . 1 he text is usually ranked under “ philosophers”
子部 ，whereas the author is given as ‘‘ anonymous.”
Coming now to the texts ascribed in the Shenxiang quanbian to
personages of the Tang or Northern Song periods, one finds that only
two items are attributed to Lii Dongbin. Firstly, the Xzangfa rumen
相法入門，“ Introduction to physiognomy， ” is found as the first text in
chapter 1 (Liang 1980: 15—17). This text could not be located in any
other physiognomic manual. Secondly, there is the Xiangfu 相賦，
“ Physiognomy’” in chapter 5 (Liang 1980; 156-158)，found also under
the title Lii Dongbin fu 呂洞賓賦 as the first text of the Yuguan. One
might surmise that the claim of a tradition going back to Lu Dongbin,
expressed in both the Shenxiang quanotan and the Yuguan by placing a
work of his at the very beginning, seems to be rather old in physiognomic
circles.
Ascriptions to Mayi, however, appear to be of a later date. There
is no text found under his name in the Yuguan, and in the Shenxiang
quanbian there are only two:

Zalun 雜論 ，“ Miscellanea ” (chapter 7，Liang 1980: 200-201);

Xiangxin 相心，“Analyzing the heart” (chapter 4，Liang 1980:
105-106).9

In the Tushu jicheng edition of the Shenxiang quanbian, a third text is

ascribed to Mayi, the Jinsuofu 金 I I 賦 ，“ Rhapsody of the golden lock ’’
(Liang 1980: 158)，but the text is not cited under his name in the old
Ming edition of the text. On the other hand, it is attributed to Mayi
by the Yuan dynasty commentator on the Renlun datongfu, thus showing
that it was at least extant and associated with Mayi at that time.
More complex is the situation in the case of Chen Tuan, not only
because there are different texts attributed to him，but also because the
Shenxiang quanbian contains one text which quotes his name in the
introduction, and cites him in numerous commentary sections.
The main text by Chen Tuan is the Fengjtan 風鑑 . This is quoted
as Fengjtan ge J風鑑®; arranged in verse, in the Shenxiang quanbian (chap­
 A T E X T B O O K O F P H Y S IO G N O M Y 239

ter 6，Liang 1980: 178-181). It is also found as Chen Tuan xiansheng

fengjian 陳搏先生風鑑 in the Yuguan ( 1 ,4b-9b;10-20) as well as under
the title Shenbi lun 神秘論 in the Taiqing shenjtan (1，3b-8a). In addi­
tion, it is frequently quoted in commentary sections of the Shenxiang
quanbian and other physiognomic manuals.
The two main editions of the Fengjtan, the one in the Yuguan and
the one in the Shenxiang quanbian^ are not identical. The former is in
prose, the latter in verse. But in content they are rather similar, giving
a general survey of the theory and practice of physiognomy. The
Fengjtan as quoted in the Yuguan can roughly be divided into eighteen
sections, most of which are also found in the edition of the Shenxiang
quanbian, though not in the same order. The eighteen sections are:
1 .Definitions of main terms, such as energy, body, essence, and spirit;
2. Physical appearance according to the five phases; 3. Types of bodies
according to animal morphology; 4. The appearance of the ‘‘ wood ’’
type; 5 . impure, i.e., mixed forms; o. Mixtures according to the ‘‘ over­
coming ” of phases; 7. Spirit as deep or shallow; 8. Energy as strong and
weak; 9. Complexion as full or transparent;10. Complexion as young or
o l d ; 1 1. Body and spirit in terms of surplus and insufficiency;12. In­
teraction patterns of the various forces;13. Thin appearance; Signs
of long l i f e ; 15. The importance of mental states;16. Forms of ob­
struction; 17. Women and m e n ;18. Noble and humble.
While the order of topics does not seem very clear in this text, the
three main themes of standard physiognomy are obvious: theoretical
discussions of the principal forces analyzed in physiognomy, typologies
of the human body in terms of the five phases and of animal morphology
and analyses of complexion. However, the text does not give an ex­
haustive survey of physiognomy. Rather，it exemplifies the basic prin­
ciples and main characteristics in discussing specific types, such as the
“ wood type ” with its developments and the type of thin appearance.
Having already been edited in the beginning of the Song in the Yuguan,
this text might actually go back to the teachings of Chen Tuan. More­
over, the fact that it is extensively quoted in all the physiognomic man­
uals since the Song shows the importance of Chen Tuan’s teacmngs in
the modern tradition of physiognomy.
In the Tushu jicheng edition of the Shenxiang quanbiany Chen Tuan
is credited with the Dongxuan jing 洞 玄 經 （ Liang 1980: 257—259). In
the Ming edition (chapter 9), however, this text is cited without any
author’s name under the title Dongxuan png xing you qixtang 洞玄經形有
七丰目，‘‘ Seven types of bodies according to the Dongxuan png,” Under
the title Dongxuan jing zaduan ge 洞玄經雜斷歌 a completely different
text is found in the Yuguan ( 1 ,llb-12b; 24-2b;. It is not ascribed to
240 TJVTA K O H N

anybody.
In addition, the Shenxiang quanbian contains a text on energy and
complexion, Qise lun 氣色論，attributed to Xiyizi 希 夷 子 （ Liang 1980:
32b-326). Chen Tuan might be referred to by this name, as he was
given the title Xiyi xiansheng in 984. However, this text is not found
in any earlier manual. The Yuguan，on the other hand, cites Chen
Tuan as the author of two treatises which are not mentioned in later
texts:

X iu lijin 袖裏金 ，“ Gold up the sleeve ’’ （1 ,9b-lla: 20-23);

Xiang ge 相歌 ，“ Physiognomy ” （ 1，lla b ; 23-24).

Another mention of Chen Tuan is made in the Shenyi fu 神異賦

(chapter 5，Liang 1980: 125-133), which figures as the first as well as
the longest of the theoretical treatises contained in the central part of
the Shenxiang quanbian. Here Chen Tuan is described as receiving the
physiognomic tradition from Mayi “ in mid-winter sitting around the
earth stove in Mayi’s grotto on Mount Hua. , . . There were no words
used in the transmission of the teacnmg. Mayi gave it to him in se­
crecy.” As the commentator explains, “ he used glimming sticks of
wood and wrote characters into the ashes of the fire ” (Liang 1980: 125).
The text Shenyi fu then consists of these very instructions first given
with so much obscuring effort by Mayi. The attribution of the text
to Chen Tuan represents another milestone on his way to becoming
the founding saint of modern standard physiognomy.10
With the three great masters of the 10th century, Lii Dongbin,
Mayi, and Chen Tuan, our picture of the tradition of the Shenxiang
quanotan has moved into a somewhat more concrete framework. The
Yuguan of the early Song has helped to date several ascriptions to the
10th century, and it now seems probable that the standard modern
tradition of physiognomy does indeed go back to Chen Tuan whose
main physiognomic worK Fengjian is quoted again and again in the
textbooks since the Song.

Southern and Post-Song.

This section will deal with two kinds of texts: The Renlun Datongfu
by Zhang Xingjian of the Jin and its commentary of the Yuan, on the
one hand, and the texts ascribed to the two Yuans of the Ming, on the
other.
The Renlun datongfu 人倫大統賦 is found in entirety but without
the commentary in chapter 6 of the Shenxiang quanbian (Liang, 1980:
lo 丄ー178). It is also found in three congshu collections, in each case
 A T E X T B O O K O F P H Y S IO G N O M Y 241

accompanied by the commentary.11 The author, Zhang Xingjian 張行

簡 ，alias Jingfu 敬南 ，was an official in the Ministry of Rites under the
Jin. He graduated to the civil service in 1179 and was well known for
his proficiency in various divinatory sciences, such as astrology and
numerological speculation. He left behind several works, notablv one
compendium on ritual matters, but there is no mention in his biography
of any work on physiognomy {Jinshu 106). On the other hand, the
Renlun datongfu is mentioned in Huang Yuji’s 黃虞稷 Oianqing tang
shumu 千頃堂書目. Here, however, no author’s name is given. The
author and title are connected only from the Yongle dadian edition on­
ward {Siku tiyao III, 2265).
The extensive commentary to the Renlun datongfu stems from an
official of the Yuan named Xuc Yannian 辟延年- . It is dated to the year
1313，which supports the claim that the text goes back to the Jin (Siku
tiyao I I I ，2265). The commentary with its precise dating is useful in
judging the existence and popularity of physiognomic texts under the
Yuan.
In content the Renlun Datongfu, a short treatise of no more than
2000—3000 characters, follows the standard pattern of modern physi­
ognomy. i-irst it gives a general exposition on the interrelatedness of
body and spirit, then it recounts detailed methods of analysis of the
human body based on the system of the five phases, tinally，it discus­
ses colors and complexion.
As the text stands therefore directly in the line of tradition leading
from Chen Tuan’s Fengjian to the Yuguan, on to the Taiqing shenjtan，
and later to the Shenxiang quanbian, its commentary also quotes the
same texts and authorities found in the other textbooks of physiognomy.
About 70% of the texts and authorities quoted by Xue Yannian are
the same as those found in the Shenxiang auanbian.12
As concerns the texts left behind by the two Yuans, one notices
first that the name of the latter Yuan, Zhongche, identified by historical
sources as the actual complier of the Shenxiang quanbian, is not men­
tioned in the text itself. On the other hand, several texts are ascribed
to his father, Yuan Gong, usually called Liuzhuang, who is also quoted
occasionally in commentary sections.
First there is the Shiren ge 識人歌 ，“ Song on knowing man， ” in
chapter 6 of the Shenxiang quanbian (Liang 1980: 181—183). But we
know nothing of this text other than its title. Yuan Gong’s only extant
work today is a collection of poetry, the Liuzhuang xiansheng shtji 招Pffi先
生詩集，wherein no reference to physiognomy is made.
Another text ascribed to Yuan Gong is the Zalun 雜論 ，“ Miscel­
lanea/' in three sections found in chapter 6 of the Shenxiang quanbian
242 L IV IA K O H N

(Liang 1980: 199-200). This, surprisingly enough, is identical with

the Zalun contained in the Taiqing shenjtan (2，la-2a). Even though
the commentator claims that this Zalun is already found in the Yuguany
this supposition could not be substantiated, at least for the edition of
the Yuguan extant today. Nevertheless, the ascription to Yuan Gong
must be regarded as fictional.
A third text ascribed to Yuan Gong is the Renxiang fu 人象賦，“ On
the image of m an， ” found in chapter フof the Shenxiang quanbian (Liang,
1980: 188-192). This is probably identical with the Renxiang fu by
Yuan Zhongche, a text appended to his collection of popular stories on
physiognomy, the Jingu shijtan 今古識鑑. This text, although con-
tainea in the Siku Catalogue、was not available to me (see, however,
Goodrich, Feng 1976: 1632).
To summarize, after the heyday of physiognomic systematization
in the 10th century, one finds the tradition continued in the Renlun
datongfu of the Jin, its commentary of the Yuan, and through the two
Yuans of the early Ming, to culminate in their grand physiognomic
collection, the Shenxiang quanbian. All these texts belong to the stand­
ard modern tradition of physiognomy and share the same organiza­
tional pattern.

Unidentified Authors and Their Texts.

The following authors to whom certain materials contained in the
Shenxiang quanbian were ascribed could not be further identified:

ouiyan xiansheng 鬼 眠 先 生 X ia n g fa 相法
chapter 6 (Liang 1980: 18+-186)
Luo Zhenren 羅 眞 X ia n g fu 相賦
人
chapter 7 (Liang 1980: 198-199)
Qiu Tanyue 秋 潭 月 L m n ilr e n 論女人
chapter 9 (Liang 1980: 251—253)
Shuo furen g e 説婦人歌
chapter 9 (Liang 1980: 254—255)
Qise g e 氣色歌
chapter 10 (Liang 1980: 306-308)
Taiyi Zhenren 太 乙 眞 人 S hu書
chapter 6 (Liang 1980: 182)
W u Xinjian 吳心鑑 Tongyuan f u 通元賦
chapter 6 (Liang 1980: 193-194)
Yandian Daoren 廉 電 道 人 Sketiyan j i n g 神眼經
chapter 5 (Liang 1980: 154—156)
Yang Shi 楊氏 Lun s h e n q i論神氣
 A T E X T B O O K O F P H Y S IO G N O M Y 243

chapter 7 (Liang 1980: 199)

Yao K u a z a n g 姚括蒼 Yuman j u e 玉管訣
chapter 6 (Liang 1980: 182-185)

In order to complete this survey of texts given with a specific title

and/or attributed to a certain authority in the Shenxiang quanbian, a list
of all the materials quoted in the commentary to the first five texts of
the Shenxiang quanbian (chapter 1 ; Liang 1980: l^-z8), presumably
by the editor Yuan Zhongche, is given here.

T exts and P ersonages Q UOTED IN THE COM M EN TARY TO Shenxiang

Quanbian C hapter 1 .(Nunnber of times in parantheses)

Texts
Bijue 秘訣 (45)
Boyue 伯樂 ( 1 ) (also quoted by Xue Yannian)
Dingjue 定訣 a)
Fengjian 風鑑 (16) (Chen Tuan’s manual)
Hunyi jing 混儀經 (i)
Linfeng ji 麟鳳記 (i)
Lingtal jing 靈臺經 a ) (also quoted by Xue Yannian; not in-
dentical with D Z 288)
Mingdu jing 冥度經 ( 2)
Niixin jmg 女心鏡 ( 1)
Pinnii jinjing 貧女金鏡 (1)
Qingjian 淸鑑 (2 ) (also quoted by Xue Yannian)
Renling shujing 人靈樞經 (1)
Renlun fu 人倫賦 ( 1 ) (Renlun datongfu)
Shengfan lun 聖凡論 ( 1)
Shenji 神機 ( 3)
Shenjie 神解 ⑴ (also quoted by Xue Yannian)
Tongxian lu 通仙銶 (2) (also quoted by Xue, and in the Taiq-
ing shenjtan)
Tongyuan fu 通元賦 ⑴ (by Wu Xinjian)
Wuxing (ge) 無形 (歌） ( 2)
⑴ (A Xinjing 心鏡 is quoted by Xue
玉玉元元易

Xinjing
管篪談神經

心經
Yannian)
1 ljing
銶訣

Yuanshen lu
Yuantan
Yuchi
Yuguanjue (by Yao Kuazang)
244 L IV IA K O H N

Zhouhou jing 肘 後 經 3)
（
Zhudan jing 燭 膽 經 1) (full text in Liang 1980: 196-197)
（

Unspecified texts

\
11

1/ )/
ge 歌
子

^
—

\ X- - - - - - - - - - - /
jing

s
經

O
宗

11
yu 語
子

Personages
南

Baige Daozhe 白閣道者 (3) (acc. to F ozxl tongjt 43/396 identical

子 德陳成管鬼敦胡孔來李柳呂麻孟裴宋太王謝

with Mayi)
和圖和輅箭林曾子和靖莊尚衣子行齊沖朔靈

Baohezi (2)
Chen Tunan (13) (Chen Tuan)
Chenghezi (5) (also quoted by Xue fanman)
Guan Lu ⑴
G ui Jian (2 )
Guo Linzong ⑴
Hu Seng (2 )
Kongzi (1)
Laihezi (2 )
L i Jing (1)
Liuzhuang ⑶ (once cited as Yuan Liuzhuang)
Lii Shang (2)
( 2 ) (Chen’s teacher)
Mengzi ⑴
儉岳子運

Pei Xingjian (2)

Song Qiqiu ⑴ (author of the Yuguan)
1 aizhongzi ( 2)
Wang Shuo ( 1)
Xie Lingyun ( 2 ) (fifth c. poet; also cited by Xue Yan-
man)
Zhuzi 朱子 （ 1)

C o m p a r is o n to E a r l ie r M anuals

So far our concern has been limited to texts that were ascribed to an
author in the Shenxiang quanbian. By comparison to earlier manuals
of physiognomy several texts could be found to date back as early as the
beginning of the Song dynasty. The three texts used for comparison
were the Yuguan zhaoshen ju of the early Song, the Taiqing shenjtan of
the late Song, and the Renlun datongfu of the Jin with its commentary
of the Yuan dynasty. In this section, material of the Shenxiang quan-
 A T E X T B O O K O F P H Y S IO G N O M Y 245

bian without ascription to any author which can yet be located in any
of these earlier texts will be discussed.
Upon closer inspection, two sections of the Yuguan have been in­
tegrated into the Shenxiang quanbian. First, the whole chapter on the
lines and signs of the palm (2，17a-35a; 87-123) is found in chapter 8 of
the Shenxiang quanbian (Liang 1980: 221-238). This section consists
of altogether seventy-two pictures of hands with lines and signs, in
each case accompanied by an explanatory poem of four lines of seven
characters. There are mainly two differences between the Yuguan
and the Shenxiang quanbian editions of this section. Firstly, the Yuguan
gives a short prose description of the signs and lines which is placed
above the picture. This is not found in the Shenxiang quanbian. Sec­
ondly, the order of the seventy-two pictures is the same in the Yuguan
and in the Ming edition of the Shenxiang quanbian，but it differs from
the order given in the Tushu jicheng reprint and therefore in Liang’s
edition of the text. The summarizing translation of Lessa (1968: 99一
112) follows the latter version, its content, however, goes back to the
Yuguan.
Second, another long section has been taken over from the Yuguan.
This is the part on different types of bodies according to animal mor­
phology (3，3a-15a: 131-155) which appears in chapter 9 of the Shen­
xiang quanbian (Liang 1980: 26o-z75). Similar differences as found
above apply here. The Yuguan edition is more detailed in its prose
description, but the poems are found m both texts e q u a l l y . 1 he order
of the material differs. Yet in this case, a more decisive difference is
to be discovered in that the Yuguan has only thirty-one types of bodies
in comparison to fifty-one types listed in the Shenxiang quanbian. The
types not mentioned in the earlier edition are the dragon and the phoe­
nix, and there are only about half as many bird types in the earlier text.
Also, many domestic animals, such as the pig ana the cat, do not figure
in the Yuguan, and finally, there are fewer sea creatures. Only two
are added in the older edition, the “ dark warrior ” and the “ sea turtle.”
In summary, it can be said that the types of hands and kinds of
bodies found in the Shenxiang ouanbian go back to the Yuguan. The
system applied in these two texts is the same, but differs from the meth­
od set forth in the Taiqing shenjian. Here we do not find any list of
lines and signs of the palm at all; the hana is only discussed theoretically.
Only seven types of bodies are eiven in the Taiqing shenjtan, i.e.，
crane，phoenix, turtle, rhinoceros, tiger, lion，and dragon. Each type
is, however, subdivided according to categories like “ straight ’’ 正 ,
“ small” 小 ，“ simple ’’ 單 ，“ sick ” 病，etc. {Taiqing shenjtan 4, 3b-7b).
In many other instances, however, the Shenxiang quanbian owes
246 L IV IA K O H N

much to the systematization of the Taiqing shenjtany especially in the

general expositions on single parts of the body. Moreover，the sec­
tions on energy and complexion stem from this late Song dynasty man­
ual. A full list of comparison is given below:

Shenxiang quanbian T exts F ound Taiqing Shenjian

in the

Shenxiang
Shenjian chjp Title chjLiang 1980
2/la-2a Zalun 雜 論 （ Miscellanea) 7/199-200
2ハ2b Sidu 四 讀 (4 Streams) 2/58
2/13a Wuxing xiangsheng ge 五行相生歌 4ハ1フ
(Song of five elements producing)
2ハ3b Wuxing xiangke g e 五行ネ目克歌 4/117
(Song of five elements overcoming)
2/13b Wuxing bihe xiangying 五行比和相應 4/11フ
(Five elements in harmony and correspondence)
2/13b-14a Sixuetang l u n 四學堂論 1/41
(On the four halls of learning)
2ハ2b Wuguan 五 管 (five senses) 2/4フ-57
(w. comm)
3/4a Lun s h e n y o u y u 論神有餘 1/42
(On surplus of spirit)
3/4b Lun s h e n b u z u 論神不足 1/43
(On deficiency of spirit)
3ハ3b Qingse c h u m o 青色出沒 11/329
(Coming and going of green complexion)
3/13b-14b Qingse jixiong g e 青色吉凶歌 11/329-330
(Fortune according to green complexion)
3ハ4b Huangse c h u m o 黃色出沒 11/335
(Coming and going of yellow complexion)
3/14b-15b Huangse jixiong ge 黃色吉凶歌 11/335-336
(Fortune according to yellow complexion)
3ハ5b Chise c h u m o 赤色出沒 11/339
(Coming and going of red complexion)
3/15b-16a Chise jixiong g e 赤色吉凶歌 11/339
(Fortune according to red complexion)
3/16ab Baise c h u m o 白色出沒 11/331
(Coming and going of white complexion)
3/16b-17a Baise jixiong g e 白色吉凶歌 11/331
(Fortune according to white complexion)
3ハフa Heise c h u m o 黒色出沒 11/332
(Coming and going of black complexion)
 A T E X T B O O K O F P H Y S IO G N O M Y 247

ぬ
—
1
3 /

7 i
Heise jixiong g e 黒色吉凶歌 11/333
(Fortune according to black complexic>n)
3

8
Zise c h u m o 紫色出沒 11/337
a
i
.
—
_
— .

(Coming and going of purple complexion)

i
t

Zise jixiong g e 紫色吉凶歌 11/337

3 / / /-

a
1
-

(Fortune according to purple complexion)

3

Lun x in g b u z u 論形不足 1/43

4 .

(On deficiency of form)

— / /4 —/

Lun x in g y o u y u 論形有餘 1/42

3
b

(On surplus of form)

Lun sheng 論 聲 (On the voice) 2/68
' -
7
.

Wuyin 五 音 (The five tones) 2/68

—4 .4

b
9

)b Xingbu ィ了部（ On walking) 4/107

b
.

11
a
— .4.

-l
b l a b

Lun zuo 論 坐 (On sitting) 4/107-108

9
b

(quoted un der youyun 又云）

1
— 4

11

Lun e 論 跃 (On sleeping) 4/108

o
../ .

Lun s h i y i n 論食飲 4/108

1 1

(On eating and drinking)

— 4.

b
/ /4/.

Lun ebu 論 額 部 （ On the forehead) 10/278

Zhentou 枕 頭 (The back of the head) 10/285
a
-— 5

(^similar)
2b
./ / / / / / /
5 4b

Meibu 眉 部 （ On the eyebrows) 3/81 (similar)

5 55 5
6

Yaobu 腰 部 (On the hips) 4/103-104

Lun shou 論 手 (On the hands) 8/216
A
8

ab Xiang changwen ネ目掌文 8/239-240

9
5

(Analysis of palm patterns) (from youyun 又云 onward)

S/9b-10a Shoubeiwen 手背文 8/240
(Patterns on the back of the hands)
4
8
8

5/10ab Guabu 瓜 部 （ The nails)

8

4
8

10b- Lun zu 論 足 （ On the feet)

1

6/la Heizhibu 黒 疲 部 （ Black spots)

2
9

6/7a Niiren jiushan 女人九善 262

(Nine good points in women)
7b Niiren j i u e 女人九惡 ,62
(Nine bad points in women)

The wealth of material of the Shenxianp auanbian, which is found

in ealier manuals, and especially in the Yupuan of the early Song, to a
certain extent substantiates the ciaim of the physiognomic tradition
that it ultimately goes back to the 10th century.
248 L IV IA K O H N

F orerunners of the T r a d it io n

If the assumption that the modern tradition of physiognomy began in

the 10th century is correct, there should be some evidence of different
types of body analysis before this time, i.e., types which were either
compatible with the later standard system and therefore integrated into
it, or types entirely different from modern physiognomy so that no
mention of them has been made after the 10th century. Physiognomic
manuals before the 10th century can indeed be found. Two main
sources can be distinguished, the material discovered in Dunhuang
and a text entitled Yuebodong zhongji 月波洞中記. It turns out that
the Dunhuang texts can be considered forerunners of the later tradi­
tion in that the systems presented therein are similar, n not identical,
with later methods, whereas the theories and patterns of the human
body set forth in the Yuebodong zhongji are not compatible with the
later tradition. Let us first look at the forerunners.
Physiognomic manuals discovered at Dunhuang can be divided
into three groups:

a. those dealing mainly with the complexion of the face syste­

matized according to colors and five-phases-theory;
b. overall surveys of the human body and the meaning of individ­
ual features;
c. texts on the significance of moles or black spots 黒子 on dif­
ferent parts of the body.13

a) P. 3589 (in part) and P. 3390. The latter text has been studied
in great detail by Hou Cning-lang (1979). Dated to the first half of the
10th century (around the year 950), this text already contains the stand­
ard nomenclature used for the different parts of the face in the later
tradition. In addition, it systematizes the various colors of types of
complexion as they appear on different parts of the face and gives their
respective significance in terms of ensuing good or bad luck. The
terminology as well as the system of colors can be traced back to sources
on Chinese medicine (IIou 1979: 63), the latter is also found in a Taoist
scripture on exorcism.14 The main colors used here are yellow 黃，red
赤，white 白，black 黒，blue or green 青，i.e., the colors of the five phases.
In addition, more specified tints are given, such as “ peach-blossom， ’
桃花，“ blue-black” 青黒 ，“ dried flesh” 乾脯，etc. (Hou 1979: 63).
The most auspicious color is yellow, whereas blue-green, red, white,
and most of all, black indicate unfortunate occurrences (Hou 1979: 64).
The terminology used for the different parts of the face— thirty
terms have been translated and illustrated by Hou (1979: 68 and 69)—
 A T E X T B O O K O F P H Y S IO G N O M Y 249

can certainly be found in the later tradition. Here, however, the terms
and their respective facial points have increased considerably.
The division of colors is also present in the Shenxiang quanbian.
However, no specified tints like “ peach-blossom， ’ can be found. On
the other hand, ‘‘ purple ” 紫 has been added to the basic five colors
of the phases, a color which is most auspicious, as it corresponds to yel­
low on a higher level (Porkert 1961:443). Notwithstanding these
minor differences, P. 3390 can be considered a direct forerunner of
modern standard physiognomic treatises.
b) P. 2572, P. 2797, and P. 3589. Together with S. 3395 and
S. 5969, these texts are different versions of a text entilted Xiatwshu
ネ目書，“ Physiognomy,” and ascribed to X u Fu. It gives a survey of the
human body from top to bottom, explanning the significance ox its
different features (Hou 1979: 5フ) . The order of the survey is basically
the same as found in the Shenxiang auanbian and other later manuals.
As an example, the list of contents of P. 2572 is given below.

Section Title ri ranslation

3 Xiang wuguan bu 相五官部 Tne nve senses
4 Xiang liufu bu 相六府部 The six departments
5 Xiang mian bu 相面部 The face
6 Xiang ra bu 相髪部 The hair
7 Xiang e bu 相額部 The forehead
8 Xiang mei bu 相眉部 1 he eyebrows
9 Xian? van bu 相眼部 1 he eyes
10 Xiang bi bu 相鼻部 The nose
11 Xiang erlei bu 相耳類部 fypes of ears
12 Xiang bi renzhong bu 相鼻人中部 Ih e nose as the human
center
13 Xiang chun bu 相唇部 The lips
14 Xiang kou bu 相ロ部 The mouth
15 Xiang chi bu 相齒部 The teeth
16 Xiang sheng bu 相聲部 The voice
17 Xiang she bu 相舌部 The tongue
18 Xiang yihan bu 相頤頷部 The chin
19 Xiang yuzhen ding 相玉枕頂 The back and top of the
head
20 Xiang bei xiongyi 相背肯臆 Back and breast
21 Xiang xiefu 相脅腹 Ribs and stomach
22 Xiang nai duzi 相妳肚腈 Women’s breasts, abdo-
men, and private
parts
250 L IV IA K O H N

23 Xiang yuchui daiqi 相玉薹袋器 Men’s penis and testi­

cles
24 Xiang pi? 相脾 (Abdomen?、
25 Xiang jiaohuai 相脚踝 Legs and ankles
26 Xiang xing bu 相行歩 Walking
27 Xiang bishou 相臂手 Arms and hands
28 Xiang haomao 相毫毛 Fine body hair
29 Xiang renmian bu santing The three facial sec­
相人面部三亭 tions
30 Xiang nanzi 相男子 Male children
31 Xiang niiren jiue 相女人九惡 Nine bad points in wo­
men
32 Xiang ewen 相額文 Lines on the forehead
33 Xiang shou changwen 相手掌文 Lines on the palms
34 Xiang jiaozu xiawen 相脚足下文 Lines on the soles of the
feet
35 Xiang renmian seqi 相人面色氣 Facial complexion

The major differences to the later tradition are obvious. Firstly,

not only those parts or the human body generally visible to the public
are analyzed. Mention is also made of “ women’s breasts, abdomen,
and private parts ” as well as of “ men’s penis and testicles.” In P.
3589, illustrations of naked male figures are included. Secondly, the
number of lines and signs identified on the soles of the feet according
to P. 2572 and P. 2797 is significantly higher than that found in the
Yuguan, which in turn gives about three times as many signs as are
noted in the Shenxiang quanbian. It seems that the later tradition was
limited much more strictly to parts of the body generally visible and
as a rule excluded private parts and the soles of the feet from its analysis.
Earlier relations of physiognomy and the medical sciences might be
speculated on in this context.
More than that, these two major differences between the Dunhuang
texts and physiognomic manuals since the Song mirror important
changes in Chinese culture. While ‘‘ the Chinese of the Tang period
did not object to women showing their bare throats and bosoms, during
the Song and after, bosom and neck were covered by the upper rims of
the robe’ and then by the nigh, tight-fitting collar of an under-jacket ’ ，
(van Lruiik 19ol: 18o入 The fact that less and less emphasis is placed
on the lines on the soles of the feet, moreover, might be related to the
introduction of the custom of binding women’s feet, dated to the first
half of the tenth century. “All literary and archeological evidence
points to the custom having begun . . . in the intervals of ca_ fifty years
 A T E X T B O O K O F P H Y S IO G N O M Y 251

that elapsed between the Tang and Song dynasties ’’ (van Gulik 1961:
216). In due course, “ women’s feet came to be considered the most
intimate part of her body ” and were thus “ strictly taboo ’，(van Gulik
19ol:218). The emergence of a modern standard tradition of physi­
ognomy in thus to be understood in the context of the development of
new Chinese customs and ethical feelings which have persisted into the
twentieth century.

A D i f f e r e n t S y s t e m : t h e Yuebodong Zhongji
rhe Yuebodong zhongji 月波洞中言己 is of rather obscure origins. The
preface is dated to the year chiwu 赤鳥 20，the alleged transmitter of
the text is Zhang Zhongyuan 張中遠 of the Wu 吳 . Although there is
a reign period called chiwu in the Wu kingdom of the Sanguo era, this
only lasted fourteen years (238-251). It is therefore probable that
this reign title is mythological. According to the Sandong zhunang
三洞珠囊，chiwu is a reign period or a kalpa in the era of the Earthly
August One 地 皇 (DZ 1139; 8，29a). Rather than giving information
on the actual date of the text, the reign title cntwu is therefore an indica­
tion for its high standing according to Taoist mythology.
The preface of the Yuebodong zhongji claims that the teachings of
the text stand in the Zhong-Lii tradition, i.e., the tradition which claims
Zhongli Quan 鐘離權 and Lii Dongbin as its patriarchs (Baldrian-Hus­
sein 1984: 23). This claim is not very likely to go back to before the
Song and thus throws some doubts on the antiquity of the preface, if
not the text itself.
Another lineage outlined in the preface goes back to the Highest
Venerable Lord himself. In one of ms revelatory phases he supposedly
descended into the Moon-Wave Grotto 月波洞 on Mount Taibai 太白
in Shensi and transmitted this text, duly called the “ Record of the
Moon-Wave-urotto， ” to a recluse not specifically named. This latter
account is also mentioned in Song dynasty catalogues, such as Zheng
Jiao’s 鄭樵 Tongzhi yiwen hie 通志藝文略 and Chao Gongwu’s 晁公武
Dushu zhi 讀書志 . Here the name of the recluse is given as Ren Xiaoyao
任逍遙 of the Tang, but— as almost everything regarding the origins of
this text— this man cannot be further identified.
What is certain about the Yuebodong zhongij is that a text of this
title in one scroll and nine sections existed under the Song. This
makes— — as the Siku tiyao authors note (III, 2263)— — a pre-Song dating
probable. Today the text is found in two editions.15 It is twenty-
seven Chinese pages in length.
The Yuebodong zhongji is a complete guide to physiognomic analy­
sis. In several instances the information it provides resembles the
252 L IV IA K O H N

methods of the Yuguan and the Shenxiang quanbian, but in most cases,
the system of the Yuebodong zhongji differs considerably from the kind
of body analysis practised since the tenth century.
Four major resemblances can be noted:
1 . The five sacred mountains and four streams of the human
face (forehead —north; left cheek bone— — east; right cheek bone—west;
nose— center; chin^south; and ears~Jiang 江 ；nose— Ji 濟 ；eyes— Wei
渭 ; mouth— He 河）are the same in the Yuebodong zhongji and the Shen­
xiang quanbian, with the qualification that the last two items are rev­
ersed in the later text i.e., mouth—
— Wei and eyes— He (Yuebo 4b; Shen­
xiang chapter 2; Liang 1980: 58).
2. j"ive-phases-theory is found in the relation seen by the Yue­
bodong zhongji between the heart and the phase fire, on the one hand,
and the eyes and the phase wood, on the other. Inasmuch as wood
produces fire, the eyes reveal the heart (oa;.
3. The five basic colors of human complexion are mentioned
with the same interpretations as given in P. 3390 and in the later man­
uals (7ab).
4. Among the different bone shapes to be detected in the ‘ jade
pillow,” i.e. the occipital bone, of which the Yuebodong zhongji (8a)
gives eighteen, sixteen are also found in the Shenxiang quanotan (chapter
10; Liang 1980: 287-288; Lessa 1968: 86—88). The latter text gives a
total number of 35 different shapes. The two items mentioned in the
Yuebodong zhongji’ but not in the Shenxiang quanbian, are the “ A weight)
character pillow ” and the “ 十 (ten) character pillow.” Among the
items mentioned in both texts, in some cases the Shenxiang quanbian
gives a more fanciful name, such as e.g., “ three frontier pillow ” for
‘‘ 三 (three) character pillow ， ” or ‘‘ three star pillow ” for ‘‘ 品 (rank)
character pillow.”
Whereas in the four instances given above, the Yuebodong zhongji
can be considered as another forerunner of the later standard tradition,
in all other instances it represents a different system of physiognomy.16
First, the basic pattern of analysis is given in terms of nine (la-
3a). The human head is mapped out as if seen from above with the
jade pillow in the north, the nose in the south, and the two ears as east
and west. The other directions are defined by the cheekbones (south­
east and south-west) and the temples or “ halls of longevity ’’ (north­
east and north-west). The center would accordingly be the top of
the head.
Next, nine different bones are given as the mainstays of analysis:
“ cheek bone” 顴 ，“ post house” 驛，“ general” 將軍，i.e., the nose,
“ sun horn ” H 角 above the left eye, “ moon horn ’，j ]角 above the right
 A T E X T B O O K O F P H Y S IO G N O M Y 253

eye, “ dragon palace” 龍宮 ，“ rhinoceros horn” 伏犀 ，“ sea turtle”

巨驚，and “ dragon horn ’’ 角 . Although not in all cases the position
of these bones is positively established, they all refer to the skull as
seen from above. It is therefore probable that the first two form a
pair, i.e., that the “ post house ” refers to a cheek bone. Furthermore
one might speculate that the “ dragon palace ” and the “ rhinoceros
horn ” have corresponding positions, possibly indicating the bones of
the temples. The two final items must then refer to some bones toward
the top of the head.
These nine bones are then set into correspondence with nine d s v -
chological and physiological characteristics of man, to be judged ac­
cording to the elevation and strength of the skull bones. Spirit 招}神，
judged by the cheek bone, is evaluated in terms of clarity. The soul
魂魄 ，judged by the post house, is ranked according to nobility. Body
stature 形貌 ，parallel to the nose, should be erect. Complexion 氣色，
evaluated by the sun horn，is measured in terms of radiance. Body
movements f/j止 ，related to the moon horn, are judged according to
their calmness. Man’s activity 行藏 ，in correspondence to his dragon
palace bone, should be proper. His glance 膽視 ，as identified in his
rhinoceros horn, should be full of clarity. Furthermore, perception
才智，as judged by the sea turtle bone, is evaluated according to swift­
ness. And demeanor,德行，finally, judged by the dragon horn, should
be law-abiding (3ab).
After a lacunae between pages 8b and 10b, the text thirdly describes
eight limits 八限 . These are the ears, forehead, eyebrows, eyes, nose,
upper and lower lips, and the chin. The term “ limit ” refers to their
function as limiting a certain period of human life. The theory behind
this notion holds that the spirit gradually moves from the top of the
human face to its lower parts; therefore the later years of life can be
analyzed through the chin, the earlier in the forehead (25b). ri he nose
in the center of the face is the human center 人中 and represents the
most productive years of man’s life. The eight limits are as follows:
E a rs :15 years; forehead:10 years; eyebrows: 4 years; eyes: 6 years;
n o s e :10 years; upper l i p : 10 years; lower l i p : 10 years; chin: ca. 5
years (10b-15a).
This makes a total average lifespan oi /0 years, with the most
productive years between the ages of 35 and 45 signified by the nose.
The chin here corresponds only to the last five years of life, but in older
texs one finds the chin as an indicator of man’s posterity (Zuozhuan’
Wen 1 ) . 1 hough therefore the system of the Yuebodong zhongji is not
compatible with the later tradition and cannot be found in any of the
later manuals, its basic concepts go back very far in Cmnese history.
254 L IV IA K O H N

The most outstanding characteristic of the Yuebodong zhongji is

its close relation to Taoism. Not only is nine the sacred number of
the Taoists, but the nine palaces 九宮 of the human head play an im­
portant role in Taoist meditation, where they are visualized inwardly
rather than being outwardly visible (Robinet 1979: 195). As noted
above, the text claims that the Highest Venerable Lord himself descend­
ed to reveal it to mankind in a mythological age. More than that, al­
most all sections of the text are summarized in poems entitled Laojun
ge 老君歌 ，“ Song of the Venerable Lord.” These verses are unique to
this text and could not be identified with any of the numerous poems
in the later manuals. Another point in which the systems of the Yue­
bodong zhongji resembles Taoist theories is the connection it postulates
between the mind and the eyes, a notion elaborated in Taoist mystical
texts.17 Furthermore the general theory underlying the system of the
Yuebodong zhongji with its claim that wholeness 全 of the body 形 causes
wholeness of the blood 血 which in turn ensures wholeness of the energy
氣 ，and finally of the spirit f 申(25a) closely resembles the Taoist under­
standing of the mystical process.18

C o n c l u s io n

Following the tradition of the Shenxiang quanbian from the beginning

of the Ming dynasty to the earliest extant manuals of Chinese physi­
ognomy, it has become clear that there is a standardized type of physi­
ognomical analysis. This type is still being practised today, as can be
seen from the popularity of the Shenxiang quantum in Taiwan and
Japan. The texts contained in this compendium can be traced to
several earlier physiognomic treatises, especially to the Renlun datongfu
of the Jin, the Taiqing shenjian of the late Song, and the Yuguan zhaoshen
ju of the 10th century. The date of this latter text gives the beginnings
of modern physiognomy. By tracing the material of the Shenxiang
quanbian back to the tenth century, the claim by its authors that their
tradition goes back to the saint Chen Tuan could be substantiated to a
certain degree. It is within the range of historical probability that the
text Fengjtan^ which is already contained in the Yuguan zhaoshen ju
and which can be considered the first example of the later standardized
type, does indeed go back to Chen Tuan.
In contrast to the established tradition from the tenth century
onward, earlier materials tend to represent only parts of later integrated
methods. The manuals found in Dunhuang can generally be con­
sidered forerunners of the later tradition, but some of them are still
written in a quite different cultural atmosphere, i.e., under less restric­
tions regarding physical exposure. The Yuebodong zhongjiy finally,
 A T E X T B O O K O F P H Y S IO G N O M Y 255

with its explicit connection to the Taoist tradition gives systems en­
tirely different from those that make up the standard tradition.
In conclusion one may not only say that standard modern physi­
ognomy indeed goes back to the tenth century and is rightfully con­
nected with the name Chen Tuan, also but that physiognomy, before
being established as a standard tradition, was transmitted under the
auspices of Taoism. As is physiognomy itself, Taoism is seriously
concerned with the human body and closely related to the various
sciences of divination. However, when in the tenth century physi­
ognomy began to stand on its own feet, its interrelatedness with the
Taoist tradition became less obvious.19 Thus modern physiognomy
claims a unique position of its own.

NOTES

1 . References to the Shenxiang quanbian w ill always give the chapter num ber of
the old M in g edition together w ith the page num ber of the 1980 reprint by Liang.
To make cross-references easier— also w ith the Tushu jicheng edition— a list is given
below:
chapter Tushu chapter Liang 1980: pages,
1 631 15-45
2 632-633 m iddle 46-78
3 633 m iddle 一 634 latter half 78-103
4 634 first h a lf ~ 635 103-125
5 636+ two sections of o37 125-158
6 637-638 158-187
フ 639 188-215
8 640 216-251
9 641 251-277
10 642 278-308
11 643 309-333
12 644 334-356
2. Another very famous physiognomist w ith the name of Y uan is found in the
records of the T ang dynasty (Tangshu 95; J i u Tangshu 6 1 ; Taiping guangji 2 2 1 ; Tushu-
jicheng 647). H is name is Y uan Tiangang 袁天綱 and he was originally from Sichuan.
H e occupied several official posts during the seventh century and made numerous
prognostications for high officials. Accoding to standard legend, he even recognized
the later Empress W u when she was just an infant.
3. A ll the various chapters in W ang F u ，
s Qtanfidun dealing w ith divination have
been translated by Anne Behnke in her doctoral thesis at the University of M ichigan,
A n n Arbor.
4. In addition, there are two sections attributed to X u F u in the M in g edition,
b ut not in the Tushu jicheng reprint:
X ia n g renzhong pian 相人中篇 ，‘ ‘Analyzing the hum an center ” (chapter 3; Liang
1980: 95);
X ia n g xing pian 相行篇，“ Analysis of walking ” (chapter 4 ，Liang 1980: 107).
Furthermore there are two texts ascribed to X u F u in the later edition, but not in the
256 L IV IA K O H N

M in g original:
X ia n g er pian 相耳 篇 ，“ Analyzing the ears ” (chapter 3; Liang 1980: 95);
Xiang she pian 相舌篇，“ Analyzing the tongue ’，(chapter 3，Liang 1980:102).
5. There are several biographies of G uig uzi found in the Taoist Canon (abbreviated
D Z and cited after Schipper 1975):
L ishi zhenxian tidao tongjian 歷jit ； 眞丨ill 體道通鑑 ( D Z 296) 6, 6 n ;
錄異記（ D Z 5 9 1 ) 1 ,la ;
X u an p in lu 玄 品 錄 （D Z 7 8 1 )1，8b;
Scm dong qunxian lu 三 洞 群 仙 錄 （ D Z 1 2 4 8 )14，8b;
Xiaoyao xujing 逍 遙 虛 經 （ D Z 1 4 6 5 )1，12a.
6. T he role of G u ig u zi as an astrologer becomes clearer when one remembers that
G u ig u is the name of a star in the N orth Culm en 北極 ，traditionally the place where the
demons meet (Shtjt 117). In addition, G u ig u during the H an was seen as closely re­
lated to G uiy uju 免臾區 ，a minister of H uangdi, alias D ahong 大 鴻 {Shiji 28), D ahong
in turn is found as the chief astrologer of H uang di {Hanshu 20, 30) also under the name
of H ongm eng 鴻 蒙 ，‘ ‘ Great Concealment ” {Zhuangzi 11； . In this latter role he plays
the part of a rather mysterious minister of H uangdi in a dialogue w ith “ Cloud C h ie f，
”
Y un jia n g 雲將^ the master of rain. G u ig u is therefore the name of a star, the minister
of astrology under H uan g di, and the mysterious grand-master of this science. A
Taoist recluse, living presumably in a place called G u ig u w ould therefore be easily as­
sociated w ith astrology and fate-calculation in H an dynasty folklore.
7. These are one called ‘ ‘ w all-gazing” and another ranked as “ embryonic re­
spiration.^ See Daoshu 道 樞 ( D Z 1 0 1 7 )14，8b; Y unji qiqian 雲 窟 七 籤 (D Z 1032) 59，
14b; Chu zhensheng taishen yongjue 諸 眞 聖 胎 神 用 訣 （
D Z 826) 5b-6a).
8. I t is m entioned in the bibliographical section of the Songshi 宋 .出，in Chen
Zhensun’s 陳振孫 Shumu ■书 ，as well as in W u Renchen’s 吳任ぼ Shiguo chunqui
十 l阐春秋 . Th e inform ation given on num ber of chapters, length, and organization of
the text does not correspond in all these sources. The Shiguo chunqiu’ for instance,
relates that the text consisted of ten chapters, and was on the whole divided into a form ­
er, yang and a latter, yin part. T he yang section supposedly gave the analysis of con­
crete outer signs of the hum an body, while the yin part described all those hum an fea­
tures w hich are not at once obvious to the eye (Siku tiyao I I I ，2264).
9. T his latter text m ight be related to a text called X in fa 心、
法 traditionally as­
sociated w ith M ayi. T his association is first founa in L ii D o n g b in ’s Taoist biography
( D Z 305) and in the Fozu tongji 佛祖通計 of 1250. The text X in fa is today found as
M a y i daozhe zhengyi xinfa 麻衣道者正易心法 in a J in d a i bishu 津逮秘書 edition. H o w ­
ever, as H u Y in g lin says in his Sibu zhengwei, it deals m ainly w ith inner alcnemy and
Y ijin g philosophy and probably stems from a later date.
10. There are two m inor personages claimed as authors of texts in the Shenxiang
quanbian who can be dated w ith the help of the Yuguan. First, the Master of the
Western Peak, X iy u xiansheng 西嶽先生 ，w ith a text called X ia ng fa 相 法 ，“ Physiog-
rto m y ，
，(chapter 6，L ian g 1980: 186-187; Yuguan l/14b-15a; pages 30-31). Again,
the Barbarian M o n k, H u Seng 胡曾 ，w ith a text called X iangjue 相 訣 ，“ Physiognomic
secrets ” (chapter 6, L ian g 1980:197-198). However, the text attributed to him in
the Yuguan (l/15a-16b: 31-34) is entirely different. It bears the title H u Seng lun
yuguan xiangshu zongyao jue 胡曾論玉管相書總、 要訣 ，“ H u Seng on general and particu­
lar secrets of the physiognomic m anual Yuguan，” which does not sound as if it was part
of the original Yuguan of the early Song. O u r inform ation on H u Seng therefore re­
mains scarce.
11, Shiwan juanlou congshu 十萬卷樓叢書 by L u X in yuan 陸心源 around 1880.
 A T E X T B O O K O F P H Y S IO G N O M Y 257

Ling long shanguan congshu 玲擺山齊叢 :碁，alias Yiyatart^ congshu 益雅常叢宵 by an

anonymous author of the Q in g; 1889.
Baibu congshu 百部叢書 chapter 591.
12. Texts quoted by X u e Y annian of the Y uan :
Boyue jin g 伯 樂 經 (also found in the commentary to the Shenxiang quanbian); Dong-
wei yujian 洞 微 玉 鑑 ；Dongyuan jin g 洞 源 經 ；G uijian 顧 鑑 ；力nWw baoyin 金 書 寶 印 (also
quoted in Taiqing shenjtan)\ Jinsuo jm g 金 鎖 經 (ascribed to M ayi); Qunpin guangjian
群 品 廣 鑑 ；L ingtai bijue 靈 臺 秘 訣 (quoted also in the commentary to the Shenxiang
quanbian; a IJn g ta ijin g 靈臺經 is found in D Z 288 and 289, b ut this text is an astrologi­
cal treatise); Qingjian ネ街鑑(quoted in the commentary to the Shenxiang quanbian) ;
Sanfu xinshu 三 輔 新 書 ； Shenjie 神 解 (quoted in the commentary to the Shenxiang
quanbian) ; Shiren lun 識人論 ，Tongxian jin g 通 仙 經 (also quoted by the Yuguan) ; X injing
ン已、
鏡 (as X in jin g ン
こ、 糸沿m the commentary to the Shenxiang quanotan) ; Yuebodong zhongji
月波洞中言己(cf. below ); Yuguan zhaoshen lun 玉 管 照 神 論 (i.e., the Yuguan’ also cited
as Yuguan zhaoren lun 玉管照人論 ）；Zaoshen lun 造 神 論 ；
Names of authorities quoted: Chen X u n 陳順；C h en g he zi 厂及和子；o u a n h u 资輅；
G u ig u zi 鬼 谷 子 ；G uiy an 鬼 眼 ；Song Q iq iu 宋 齊 邱 （
author of the Yuguan) ; Tang Ju
唐 擧 ；X ie L in gy un 謝 靈 運 . Except for C hen X u n , these personages are also cited in
the Shettxiarw auanbian.
13. A m ong the various D un hu ang manuscripts dealing with physiognomy the fol-
】
ow ing were not available to me. S. 3395 and S. 5969, both variant versions of the
Xiangshu discussed under b) below ; P. 2829, P. 3492 V° 1，and S. 5976 mentioned in
H o u 1979: ^7. These three latter texts are different versions of a work on the signifi­
cance of moles or black spots, a type of body analysis which figures prominently in the
Shenxiang quanbian (chapter 10, Liang 1980: 290-295) as well as in other physiognomic
manuals after the tenth century.
14. T he Zhengyifawen xiuzhen yaozhi J|::.一‘ 法文條眞要旨 of the sixth century ( D Z
1270; 13b-14a) gives various colors signifying a good complexion. Variants are m e n ­
tioned of the five basic colors: red, white, yellow, black, and blue (Strickmann 1985).
15. L i Tiaoyuan 李調元 ，H a n h a i 函 海 （
eighteenth century) and L iu W anying 劉晚染
Shugu congchao 述 古 叢 鈔 （
1885).
16. It can be considered a forerunner of the Taiqing shenjtan in its systematization
of body analysis according to anim al morphology. It gives five basics types, i.e., m o n­
key, ox, elephant, bear, and tiger, w ith the specifications of <c straig h t，
，，“ sim ple,” etc.
(23b-24b).
17. There are m ainly W u Y u n ’s 吳箱 X in m u lun 心 目 論 （D Z 1038), as 'veil as
sections of Sima Chengzhen’s 司馬承酿 Tianyinzi 天 隱 子 （ D Z 1026; sect. 5) and his
Znowanglun 坐 忘 論 （
D Z 1036; 4b).
18. For the mystical process in Taoism see Zaowanglun ふ
； 論 (D Z 1 0 3 6 ),1 5a
and Dingguan jin g 定 觀 經 （
D Z 400) ，6b-フb. G radual refinement here begins w ith the
body being purified to energy, energy to spirit, and spirit to the Tao itself.
19. There is only one work on physiognomy in the Taoist Canon today, the Lingxin
jin g z h i 靈倍經 n ( D Z 1425; variant D Z 1481). Here most emphasis is placed on the
significance of the eyes, ears and the m in d which are called the <c three lights ’’
of m an. The close relation between the eyes and the m in d is noted. A new element
is added to types of physiognomic analysis studied so far in that facial symptoms are
linked with certain periods of tim e (Kalinowski 1985).
258 L IV IA K O H N

R E F E R E N C E S C IT E D

B a l d r ia n -H u ssein , Farzeen
1984 Procedes secrets du joyau magiqae. Paris: Les D eux Oceans.
1986 L ii T ung-pin in northern Sung literature, Cahiers d ’Extreme Asie 2/3, 135-171•
F o r k e , Alfred
1907 Lun-heng. Leipzig: Harrassowitz.
G o o d r ic h , L .C . and F e n g , C.
1976 Dictionary of M in g biography 1368-1644. N ew York: C olum bia University
Press.
G u l ik , Robert van
1961 Sexual life in ancient China. Leiden: E. J. Brill.
H o u Ching-lang
1979 Physiognomie d ’apr さs le teint sous la dynastie des Tang [Physiognomy
according to facial complexion under the Tang], Contributions aux etudes de
Touert-houang，55—フ0. Paris, Geneve.
J i n g u k a n 神宮館
1982 Ninso no hana shi 人 相 の 話 （O n physiognomy). Tokyo: Jingukan (adopted
from the teaching of Takashima D onzo 高島呑象 ).
K a l in o w s k i , M arc
1985 L a litt^rature divinatoire dans ]e D aozans [Texts dealing with divination in
the Daozang]. Paper presented at the Quatrieme Colloque Pluridisciplinaire
Franco-Japonais, Paris, October 1985.
K n a u l , Livia
1981 Leben und Legende des C li’en T ’uan [Life and legend o f C h ’en T ’uan"l.
Frankfurt: Peter Lang.
L essa, W illia m
1968 Chinese body divination. Los Angeles: U nited W o rld Press.
L ia n g X ia ng run 梁湘潤
1980 Shenxiang quanbian 神相全編 . Tai pei : Hungye.
N e e d h a m , Joseph
1956 Science and civilization in China I I . Cambridge: Cambridge University
Press.
P o r k e rt , M anfred
1961 U ntersuchungen einiger philosopmsch-wissenschaftlicher orundbegriffe und
-beziehungen im Chinesischen [Examination of some philosophical scientific
terms and relations in Chinese], Zeitschrift der deutschen Morgenlandischen
Gesellschaft 110.2，422-452.
R o b in e t , Isabelle
1979 Meditation taoiste. Paris: Dervy Livres.
S c h ip p e r , Kristofer M .
1975 Concordance du Tao-tsang. P aris: £cole Franqaise d ’Extreme-Orient.
S t r ic k m a n n , M ichel
1985 Ensigellation: A Buddho-Taoist technique of exorcism. Paper presented at
the Quatrieme しolloque Pluridisciplinaire Franco-Japonais, Paris, October
1985
Y a m p o l s k i , P h ilip B.
1967 The Platform S utra of the Sixth Patriarch. N ew York: C olum bia U niver­
sity Press.