. 11
( 15)


Ancient China. (Cambridge, Mass.: Harvard University Press, 1983), p. 232, ¬g.
61 ˜f.™ In the same work (p. 280), however, we also ¬nd an identical graph, repro-
duced in a chart of ¬‚ags from Hayashi Minao™s article “Chugoku senshin jidai
no hata,” whose assigned meaning is a hand that is holding a standard.
Chang, Art, Myth and Ritual, pp. 91“92.


was the one initially charged with the duty of keeping track of hits at these
contests.4 Archaeologists and specialists in early Chinese history generally
agree that there must originally have been a class of people who special-
ized in the engraving of oracle bones and in the practice of writing; thus
the ancient shih could have been an “engraver” or, simply, “one who could
The existence of a category of shih functionaries as “makers of books”
is shown by the inscription on a recently discovered p™an basin,6 where it
is recorded that the vessel was made by order of a shih “named Ch™iang
whose lineage specialized in tso ts™e, the making of ˜bamboo books.™ ”7 This
points again to the historian as someone engaged above all in the “craft of
writing,” an activity that had to be performed at different levels of the
bureaucracy and was intimately connected with political and ritual func-
tions. Highly esteemed because of the magical and ritual powers attributed
to the written word in early China, this occupation later acquired special
relevance because of the moral, political, and ideological implications inher-
ent in preserving the past. The writings of the ¬rst historians, therefore,
were used to assist the ruler in the performance of sacri¬cial rites; the div-
ination records these functionaries inscribed on animal scapulas and turtle
shells represent instances of their recording speci¬c events.8
By the time of the Chou dynasty the shih had acquired important duties
that included assisting with astronomical and astrological affairs, especially
through the selection of auspicious and inauspicious days for the perfor-
mance of particular duties and rituals; accompanying the ruler to sacri¬ces,
on military expeditions, or to diplomatic meetings; and attending archery
contests. Historians under the Chou were also invested with the authority

Burton Watson, Ssu-ma Ch™ien, Grand Historian of China (New York: Colum-
bia University Press, 1958), p. 70. For a different hypothesis, see Shirakawa
Shizuka, “Shaku shi,” in id., Kokotsubungaku ronshu (Kyoto: Hoyu shoten,
1955), 1: 1“66; and Leon Vandermeersch, Wangdao: ou, La voie royale:
recherches sur l™esprit des institutions de la Chine archaique, 2 vols. (Paris: Ecole
Fran§aise d™Extrême-Orient, 1977“80).
According to K. C. Chang, many historians believe that the “earliest historiog-
raphers were also religious, possibly shamanistic ¬gures.” Yet characters that
speci¬cally indicate shamans, such as shih and wu, were already used at a very
early stage and exclude the possibility of an identity of functions between the his-
torian and the priest. Therefore, the thesis of an originally religious (possibly
shamanistic) role of the shih, albeit highly attractive and full of interesting impli-
cations, can be accepted only as a hypothesis.
T™ang Lan, “Lüeh-lun Hsi-Chou wei shih chia-tsu chiao-ts™ang t™ung-ch™i te
chung-yao yi-yi,” Wen-wu, 1978.3: 14, 19“24.
Chang, Art, Myth, and Ritual, p. 91.
C. S. Gardner, Chinese Traditional Historiography (Cambridge, Mass.: Harvard
University Press, 1938), pp. 7“9.


to judge the morality of actions by the ruler.9 The writing of annals was a
later development that probably arose when historians had to keep a record
of the of¬cial activities of rulers for future reference.10 According to the
Confucian tradition, the wang of the Spring and Autumn period was always
¬‚anked by a “right” and a “left” shih, with the duty of recording, respec-
tively, the “deeds” and the “words” of the ruler.11 In this respect the posi-
tion of the shih was comparable to that of the medieval chronicler of the
West or perhaps of the “logographer” of ancient Greece.
In the Warring States period, the Confucian tradition ascribed a moral
value to the work of the historian by charging the past with the preeminent
quality of being the repository of human experience.12 Thus historians were
invested with a moral authority derived not only from their knowledge of
the past but also from their institutional role as interpreters, or judges, of
the past.13 As the functions of the shih came to involve rituals, such as the
selection of auspicious days or the interpretation of planetary movements,
the de¬nition of the profession of the shih came to mean not only someone
who could write but also someone who was engaged in the acquisition and
control of an ever more complex and esoteric body of knowledge. Astron-
omy, the calendar, and the recording of human events all fell within the
realm of historical knowledge and were glued together by the universal
belief in the co-terminal existence, close relationship, and mutual in¬‚uence
of the human and heavenly worlds.14

History Writing during the Early Han

The dual function of the shih as recorder of both heavenly and human events
was institutionalized during the Former Han dynasty (206 b.c.“a.d. 9),

H. G. Creel, Shen Pu-hai. A Chinese Political Philosopher of the Fourth Century
B.C. (Chicago: University of Chicago Press, 1974), p. 85
B. Watson, Ssu-ma Ch™ien, p. 70.
Otto Franke, “Der Ursprung der chinesischen Geschichtschreibung,” Sitzungs-
berichte der prüßischen Akademie der Wissenschaften 23 (1925): 276“309.
A. F. Wright, “On the Uses of Generalization in the Study of Chinese History,”
in Generalizations in the Writing of History, ed. L. Gottschalk (Chicago:
University of Chicago Press, 1963), p. 37; Creel, Shen Pu-hai, p. 85.
On the Confucian position on history, see Roger Ames, The Art of Rulership: A
Study in Ancient Chinese Political Thought (Honolulu: University of Hawaii
Press, 1983), pp. 1“6. But other schools of thought also exhibited similar con-
cerns. For instance, Mo Tzu said that “the sources of our knowledge live in what
is written on bamboo and silk, what is engraved on metal and stone, and what
is cut on vessels to be handed down to posterity” (Chang, Art, Myth and Ritual,
p. 89).
Benjamin Schwartz, The World of Thought in Ancient China, (Cambridge, Mass.:
Harvard University Press, 1985), p. 353.

when the shih became a high-ranking of¬cer in the central administration
of the state, who specialized also in astrological matters.15 The duties of the
T™ai-shih-ling (usually translated as Grand Historian or Prefect Grand
Astrologer) included drawing up the annual calendar, memorializing about
the monthly calendar on the ¬rst day of each month, ¬nding auspicious days
for state rituals, and keeping a record of portents and omens. He was also
supposed to supervise the tests given for the appointment of the Masters of
Documents, that is, the imperial secretaries, who were required to know a
large number of characters and different styles of writing.16
After inheriting the position of Grand Historian from his father, Ssu-ma
Ch™ien did not limit himself to data collecting and memorializing on aus-
picious days. He also allegedly continued, privately, the labor of his father,
Ssu-ma T™an, in compiling a history that was not simply a collection of doc-
uments, but had a worldview, was politically “engaged,” and did not refrain
from interpretation and moral judgments. The project for a universal
history of China down to the Han, usually attributed to the son in its mature
formulation, was grandiose, but the search for the reasons Ssu-ma Ch™ien
undertook it have so far failed to yield a satisfactory explanation.
Ssu-ma Ch™ien™s work and thought defy easy characterization, and
attempts to see him as the exponent of a given school of thought “ Confu-
cianism, Taoism, or something else “ have not been particularly successful
in interpreting the genesis of the Shih chi. A theory that has had consider-
able currency has pointed out the proximity of Ssu-ma Ch™ien to the
so-called Huang-Lao thought. Ssu-ma Ch™ien™s intellectual position has been
seen as a synthesis of ideas drawn from the Huang-Lao tradition and the
cosmological elements of Han Confucianism, in particular those linked to
the yin-yang theory.17 But the Grand Historian also criticized popular Taoist
political ideals that ran counter to the notion of the centralized state, a posi-
tion possibly attributable to his upbringing, which presumably stressed the
“Confucian” ethos for civil service (his father having been, like him, a
Grand Historian, from the beginning of his career he always remained close
to the government and the palace).18 The presence of elements belonging to
different philosophical traditions and the syncretic tendencies detectable in
the Shih chi are supported by Ssu-ma Ch™ien himself, who declared, in his

Paul Pelliot, “L™©dition collective des oeuvres de Wang Kouo-wei,” T™oung pao
26 (1929): 118.
Hans Bielenstein, The Bureaucracy of Han Times (Cambridge: Cambridge
University Press, 1980), p. 19.
Lai Hsin-hsia, “Ts™ung Shih Chi k™an Ssu-ma Ch™ien te cheng-chih ssu-hsiang,”
Wen shih che 1981.2: 56.
This is the case, for instance, of the Taoist belief in “small states and poor people”;
see Wu Ju-yu, “Ssu-ma Ch™ien te ju tao ssu- hsiang pien-hsi,”in Jen-wen tsa-chih
1984.3: 87.


famous letter to Jen An, that he wanted “to form a single school of thought”
(ch™eng yi chia chih yen).19 Arguably, it is in his endeavor to unify and
resolve the many variances and lacunae of an as yet not uni¬ed historical
past within a coherent whole that the author found a major raison d™être
for his monumental work.
To gain an understanding of Han historiography, the relationship
between history writing before Ssu-ma Ch™ien and the Shih chi has emerged
as an obvious but nonetheless engaging issue. Both Édouard Chavannes and
Burton Watson placed Ssu-ma Ch™ien™s work within the so-called Confu-
cian historiographical tradition, of which the Shih chi represented a more
advanced evolutionary stage. According to Chavannes, Ssu-ma Ch™ien™s
main advance was his organization of already existing material within a
historical context; therefore, Ssu-ma Ch™ien was a compilateur rather than
an original thinker.20 This was possible, according to the French sinologist,
owing to a general ¬‚ourishing of literary activity “ a renaissance des lettres
“ during the reigns of Wen-ti and Wu-ti.21 This interpretation leads us to
two conclusions. First, Ssu-ma Ch™ien could not be considered the “father”
of Chinese historiography because his own conception of history had
been generated by the Confucian one. Second, based on contextual simi-
larities, a genetic relationship is assumed to have existed between the Shih
chi and works such as the Spring and Autumn Annals and the Tso-chuan,
for which the Han historian is regarded as the conscious interpreter and
Seeking to justify the radical differences between the Ch™un Ch™iu and
the Shih chi in conception, structure, and philosophy of history, Piet van
der Loon proposed an evolutionary scheme according to which Chinese
historiography developed in three stages. Before Confucius there was a
“ritual” historiography; then, with and after Confucius, there was a shift
toward a “moralizing” use of history; and ¬nally, with Ssu-ma Ch™ien, there

This statement has been the object of close scrutiny by numerous exegetes of the
Shih chi. For instance, see Wu Chung-kuang, “Ssu-ma Ch™ien ˜ch™eng i chia chih
yen™ shuo,” Jen-wen tsa-chih 1984.4: 80. This sentence has also been interpreted
as “to complete the words of a single family,” with reference to the ¬lial attitude
shown by Ssu-ma Ch™ien to his father Ssu-ma T™an, who is credited with being
the one to begin writing the Shih chi. Others have pointed out that the meaning
of this statement resides in Ssu-ma Ch™ien™s effort to give the Shih chi an inde-
pendent status as a literary work by selecting different elements from several
schools of thought and by using sources that belonged to different traditions; see
Pai Shou-i, “Shuo ˜ch™eng i chia chih yen,™ ” in Li-shih yen-chiu 1984.1: 55“60.
To me it seems that the general thrust of these interpretations is to stress the uni-
tarian outlook and syncretic effort present in the Shih chi.
Édouard Chavannes, Les m©moires historiques de Se-ma Ts™ien (Paris: Ernest
Leroux, 1895“1905) 1985, 1: clvi“clvii.
Chavannes, M©m. hist., 1: cviii. 22 B. Watson, Ssu-ma Ch™ien, pp. 89“93.


came “systematic” history. Systematic history appeared after China had
become a centralized power, capable of “transmitting the experience of
statecraft.”23 This position attempts to reconcile the images of Ssu-ma
Ch™ien as both “continuator” and “innovator.”
The relationship between the appearance of the Shih chi and broad-
ranging social and political changes occuring in Ssu-ma Ch™ien™s time is
brought into focus by the arguments stressing the new social and political
conditions created in China after its political uni¬cation and the consequent
“need” for an improved organization of historical knowledge. According
to Hulsew©, the Chinese historian on the one hand reorganized the histor-
ical materials of the past in “better co-ordinated frames”; on the other hand,
new methods and forms were produced by the uni¬cation of the empire
under the Han, which were to remain standards for of¬cial history writing
for the following two thousand years.24 Dzo Ching-chuan™s theory, along
the same line of thought, can be summarized in the following three points:
(1) as interest in the destiny of the human community grew along with
the increase in the number of those who took part in politics, the ancient
formulas for the presentation of historical data appeared more and more
to be insuf¬cient to satisfy the greater demand; (2) there arose a need
for better organization of the old oral and written traditions; and (3)
enhanced organization of historical knowledge could be realized when a
new Weltanschauung was introduced in China with the birth of a uni¬ed
empire.25 In sum, Dzo considered history writing the fruit of the empire™s

Piet Van der Loon, “The Ancient Chinese Chronicles and the Growth of His-
torical Ideals,” in Historians of China and Japan, ed. W. G. Beasley and E. G.
Pulleyblank (London: Oxford University Press, 1961), p. 29.
A. F. P. Hulsew©, “Notes on the Historiography of the Han Period,” in Histori-
ans of China and Japan, p. 33. Very often we read that Ssu-ma Ch™ien set the
pattern for the historical literature of the next two thousand years. Although the
Shih chi did provide a model for later Standard Histories, this does not mean that
no changes occurred thereafter. The major developments can be summarized as
follows: (1) Fixing the boundaries of each work so that it focused on the history
of a single dynasty (the Han shu being the ¬rst properly dynastic history); (2)
appointing to of¬cial positions scholars whose task was to compile the history of
the previous dynastic period; and (3) from the T™ang onward, ensuring that nor-
mally the compilation of dynastic histories was done by a committee rather than
by a single person or by a family of historians; see Lien-sheng Yang, “The Orga-
nization of Chinese Of¬cial Historiography: Principles and Methods of the Stan-
dard Histories from the T™ang through the Ming Dynasty,” in Historians of China
and Japan, pp. 44“45.
Ching-chuan Dzo, Se-ma Ts™ien et l™historiographie chinoise (Paris: Publications
Orientalistes de France, 1978), p. 38. In this position we can recognize an echo
of what has been arguably the most in¬‚uential Western theory on early Chinese
historiography, expressed by Balazs in the famous formula that “history was


creation and the establishment of Confucian dogma, which called for more
“didactic, moralistic, bureaucratic compilations.”26 He saw Ssu-ma Ch™ien
as the initiator of history written for use by state of¬cials.27 This “socio-
logical” interpretation of early Chinese historiography has enjoyed consid-
erable favor among scholars writing in the West, whether they have seen it
as an offshoot of “orthodox political morality,”28 as a tool to ¬t the needs
of the bureaucratic class, or as a mirror of correct moral conduct.29
Yet the Shih chi was not intended as a bureaucratic compilation, was not
necessarily written for state of¬cials. The implied audience for Ssu-ma
Ch™ien™s labor is to this day a matter of conjecture. Some scholars have tried
to go beyond an interpretation that stressed either the political morality of
the Shih chi or its bureaucratic nature. Various arguments have been pro-
posed to explain what might have induced Ssu-ma T™an and his son to
embark on writing the Historian™s Records, but the motives and conditions
under which the work might have matured “ consciousness of history™s
meaningfulness for the present, potential access to recorded materials,
stimuli drawn from the lacunae and disorder of the previous historical tra-
dition, and personal ambition “ remain subjective interpretations, dif¬cult
to evaluate without there being some evidence provided by the author
Some sentences in the letter of self-justi¬cation sent by Ssu-ma Ch™ien to
his friend Jen An have been raised to the status of a manifesto, or public
declaration of intent, and naturally have been the objects of close scrutiny.
These are “to gather the old traditions scattered all over the empire”

written by of¬cials for of¬cials”; see Etienne Balazs, “L™histoire comme guide de
la pratique bureaucratique,” in Historians of China and Japan, pp. 78“94 (trans.
in English as “History as a Guide to Bureaucratic Practice,” in Etienne Balazs,
Chinese Civilization and Bureaucracy [New Haven: Yale University Press, 1964]).
David Johnson, “Epic and History in Early China: The Matter of Wu Tzu-hsü,”
Journal of Asian Studies 2 (1981): 271.
Dzo, Se-ma Ts™ien et l™historiographie chinoise, p. 136.
C. B. Sargent, “Subsidized History,” Far Eastern Quarterly 3.1 (1943): 134“38.
Homer Dubs, “The Reliability of Chinese Histories,” Far Eastern Quarterly 6.1
(1946): 31; Hans Bielenstein, “The Restoration of the Han Dynasty. With Prole-
gomena on the Historiography of the Hou Han Shu,” Bulletin of the Museum of
Far Eastern Antiquities 26 (1954): 81.
F. Kierman, Ssu-ma Ch™ien™s Historiographical Attitude as Re¬‚ected in Four Late
Warring States Biographies (Wiesbaden: O. Harrassowitz, 1962). For a critique
of Kierman™s theory, see also Antony Hulsew©, “Reviews of Gestalten aus der
Zeit der chInesischen Hegemoniek¤mpfe aus Szu-ma Ts™ien™s Historischen
Denkwürdigkeiten, by Erich Haenisch; Ssu-ma Ch™ien™s Historiographical Atti-
tude as Re¬‚ected in Four Late Warring States Biographies, by F. Kierman Jr.; ˜The
Restoration of the Han Dynasty, II,™ by Hans Bielenstein,” T™oung Pao 52
(1965“66): 182“99.

(wang-lo t™ien-hsia fang-shih chiu wen); “to research the conduct of affairs”
(k™ao chih hsing shih); “to examine the patterns that lead to success and
failure, and to rise and fall” (chi ch™i ch™eng-pai hsing-huai chih li); “to
investigate the interaction “ or boundary “ between Heaven and Man” (chiu
t™ien jen chih chi); “to comprehend the changes of the past and present”
(t™ung ku chin chih pien), and the aforementioned “to form a single school
of thought” or “to complete the words of a single family” (ch™eng yi chia
chih yen).31 The ¬rst two sentences have been investigated more thoroughly,
in some cases leading scholars to conclude that the Shih chi “represents the
union between ˜interpretation™ and ˜criticism™ in the tradition of Chinese
historiography.”32 Whether there is such a convergence is doubtful, in the
light of the ambiguity inherent in the use of terms such as “interpretation”
and “criticism” in an early Han intellectual context. Noteworthy for our
purpose are the stress on the dual function of the historian as someone who
has to investigate both heavenly and human phenomena, the sense that
history changes according to “patterns,” and the holistic or synchretic
vision of an intellectual pursuit.
The philosophy of history of Ssu-ma Ch™ien is enclosed in a few key
terms, in particular, li (pattern, order, cause) and pien, “change,” whereby
some kind of disappearance of the old and appearance of the new seems to
be implied. The sentence t™ung ku chin chih pien suggests that there is a
sense of “law” that needs to be derived from the investigation of
“change,”33 which the historian must use to penetrate the general patterns

Han shu 62: 2735. Watson™s translation of the last three sentences is: “[I wish]
to examine into all that concerns heaven and man, to penetrate the changes of
the past and present, completing all as the work of one family” (B. Watson, Ssu-
ma Ch™ien, p. 66).
Yü Ying-shih, “The Study of Chinese History. Retrospect and Prospect,” trans.
Thomas H. C. Lee and Chun-chieh Huang, in The Translation of Things Past.
Chinese History and Historiography, ed. George Kao (Hong Kong: Chinese
University Press, 1982), p. 11. Chinese Marxist historians have interpreted Ssu-
ma Ch™ien™s claim for a separation between t™ien and jen as indirect evidence of
a materialistic orientation, consisting of the historian™s denial of Heaven™s
intervention in shaping the fate of individuals or states, and in his attention to
economic relations rathern than to “supernatural” explanations; see La
Chang-yang,” Lun Ssu-ma Ch™ien te li-shih che-hsüeh,” in Ssu-ma Ch™ien yen-
chiu hsin-lun, ed. Shih Ting and Ch™en K™e-ch™in (Cheng-chou: Ho-nan jen-
min,1982), p. 84; Jen Chi-yu, “Ssu-ma Ch™ien te che-hsüeh ssu-hsiang,” in Ssu-ma
Ch™ien yü ·Shih Chi’ lun-chi, ed. Li-shih yen-chiu pien-chi-pu (Hsi-an: Shaansi
jen-min ch™u-pan-she, 1982), p. 105.
“Change” was regarded as a “law” (fa) of history generating a type of “con¬g-
uration” characteristic of each particular phase (hsing); as such, “change” was
regarded as a key principle of the Ch™un-ch™iu period. See Ch™un-ch™u fan-lu, ch.
63 (“Wu-hsing pien-chiu”), 10a, 183a, 130a.

of historical development. Such “laws” of history preside over and deter-
mine the speci¬c character of each historical phase.34 The concept of li,
adopted by Ssu-ma Ch™ien to indicate the causes or patterns that determine
the unfolding of human (historical) events “ success and failure, rise
and fall “ was also a term that had particular signi¬cance in astronomy
because the calendar was supposed to embody the li of the natural world
and was key to cosmological conceptions associated with divinatory
These concepts pertained to both sides of the activity of the historian-
astronomer, that is, to his investigation of heavenly movements as well as
to his attention to human developments, and they point to the same essen-
tial pursuit of a knowledge based on empirical observation. This form of
knowledge, somewhat similar to what the Greek called autopsy, “seeing
things for oneself,” had as its preliminary goal the description and record-
ing of noteworthy occurrences. In other words, the central concern of Ssu-
ma Ch™ien, in the investigation of historical change, may have consisted of
the de¬nition of those characteristics of the “revolutions” from one epoch
to the next that determine the special quality of a historical period and that
can be regarded as the unique attributes of each “change.”36 The same
concept, at least during the Han period, could be associated also with astro-
nomical phenomena, as we ¬nd in the following passage of the Huai-nan
Tzu, “when there is a great danger for the state, there are changes (pien)

This is what Yang de¬nes as shih, namely, the particular circumstances (the
“trend” or “momentum”) of that historical period; see Yang, Yen-ch™i, “Ssu-ma
Ch™ien te li-shih ssu-hsiang” in Ssu-ma Ch™ien he Shih chi, ed. Liu Nai-he (Peking:
Pei-ching ch™u-pan-she, 1987), pp. 41“58.
Ho Peng Yoke, Li, Qi and Shu. An Introduction to Science and Civilization in
China (Hong Kong: Hong Kong University Press, 1985), pp. 153, 27. For the
concept of li as universal order, or pattern, in the Huai-nan Tzu, see Charles Le
Blanc, Huai-nan Tzu. Philosophical Synthesis in Early Han Thought, (Hong
Kong: Hong Kong University Press, 1985), p. 133. Note that in the same sen-
tence Chavannes translates li as “causes” (M©m. hist., 1: ccxxxvii). According to
Needham, li means: “natural pattern, the veins in jade, to cut jade according to
its natural markings; principle, order, organization”; see Joseph Needham, Science
and Civilization in China, vol. 2: History of Scienti¬c Thought (Cambridge: Cam-
bridge University Press, 1956), p. 228.
Although the premises are different, this notion of change is not far from that
found in the theory of “relative circularity” put forth by Ohama Akira, who main-
tained that, although Ssu-ma Ch™ien regarded history as a continuous circular
movement, there was room for deviation and crises. He concluded that “the most
important point in the conception of history of Ssu-ma Ch™ien is not the system
of laws in history, but rather its transformation”; see Ohama Akira, “Shiba Sen
no rekishi kan,” in his Ch»goku. rekishi. unmei: Shiki to Shitsu (Tokyo: Keiso
Shobo, 1975), pp. 185“89.

in the heavenly signs (t™ien-wen).”37 Pien was a universal analytical concept
applicable to the observation of natural and astronomical phenomena, as
well as to historical ones.
At the time of Ssu-ma Ch™ien, the ¬gure of the astronomer had not yet
been separated from that of the historian; Ssu-ma Ch™ien himself was both.
As Nakayama points out, “in China the term shih comprehended both pur-
suits. Ssu-ma Ch™ien [. . .] is best known as an of¬cial historian charged
with the compilation of court documents, but in his post as T™ai-shih-ling
he was also responsible for maintaining astrological records.”38 Indeed, in
his capacity as Grand Historian Ssu-ma Ch™ien took part in all activities
related to the observation of astronomical phenomena and the preparation
of the calendar, that is, the two basic pursuits of Chinese astronomy.39
Terms like hsing, li, and pien express the same level of cognitive experi-
ence, theoretical assumptions, and world vision whether they refer to
human or celestial phenomena. It seems reasonable to suggest that the
Grand Historian™s activity had not only a theoretical but also a method-
ological af¬nity with that of the astronomer (or astrologer). In other words,
we can presume that history, being epistemologically intertwined with
astronomy, partook of the same heuristic assumptions and methodological
tenets. The empirical recording and systematic organization of data was
essential to both. This may have been true of an older period as well, but
it is only with Ssu-ma Ch™ien that we ¬nd a systematic investigation and
organization of history, and it is therefore Ssu-ma Ch™ien (and perhaps Ssu-
ma T™an, to the extent that we can identify his contribution to the Shih chi)
who was responsible for the application of a method derived from the tra-
dition of the astronomer-astrologer to historical inquiry.
The “revolutionary” aspect of Ssu-ma Ch™ien™s activity as a historian
must rest, at least in part, on the application of methods of empirical inves-
tigation and observation of natural and human phenomena, which guided
astronomical calculations long before him, to the description and rational-
ization of historical phenomena, and in particular of those new phenom-
ena that emerged at the time of the uni¬cation of the empire.40 At least one
Huai-nan Tzu, 20: 2a.
Shigeru Nakayama, Academic and Scienti¬c Traditions in China, Japan and the
West (Tokyo: University of Tokyo Press, 1984), p. 4.
Xi Zezong, “Characteristics of China™s Ancient Astronomy,” History of Oriental
Astronomy, ed. G. Swarup et al. (Cambridge: Cambridge University Press, 1987),
p. 39. He also took part in the famous calendrical reform of 104 b.c. and was
at the head of the state astronomers in charge on the observation of the sky; see
Homer Dubs, “The Beginnings of Chinese Astronomy,” Journal of the American
Oriental Society 78.4 (1958): 298.
Some light on this subject has been shed by Tu Sheng-yün, who takes empirical
observation to be the basis of Ssu-ma Ch™ien™s understanding of a number of astro-
nomical questions, ranging from the simple identi¬cation of stars and planets to

aspect of his work as historian can be regarded as the observational, or
documentary, stage, which was aimed at describing the various phases or
facts inherent to a certain topic. Recording historical events, and providing
an accurately documented description of them, was the main duty of Ssu-
ma Ch™ien as a historian, just as observing and recording celestial move-
ments was the ¬rst duty of Ssu-ma Ch™ien as an astronomer.41 As Ssu-ma
Ch™ien himself stated in his “manifesto” letter to Jen An, the historian-
astronomer was primarily called on to register and explain those phenom-
ena that indicated transformation and change.
How does this prolonged excursus on methodological issues affect our
discussion of Ssu-ma Ch™ien™s treatment of Inner Asia? I believe it is not
simply the existence of the Hsiung-nu “phenomenon” but more precisely
the search for observable phenomena and the ampli¬cation of the describ-
able world that informed the historian™s quest that led to the incorporation,
within the Shih chi, of an unprecedented wealth of information on Inner
Asia. Although knowledge of “nomadic peoples” through military con-
frontation, trade, and diplomatic relations must have been current in China
for at least two centuries before Ssu-ma Ch™ien, the appearance of a united
steppe empire at the doorstep of China was still a fairly recent phenome-
non, one that not only had grave repercussions at the state™s economic and
political levels but was also unprecedented in Chinese history. The great
concern caused by the novel formation of a uni¬ed, powerful nomadic
empire, and the many wars fought by the Han against it, quali¬ed the
Hsiung-nu as a topic worthy of investigation. But it was the application of
the astronomer™s method to this topic that generated the “paradigm
shift” from the moralistic or chronachistic accounts of the past to the
historically and ethnographically rich report that Ssu-ma Ch™ien was able
to produce. If other Han works of the same period failed to produce any-
thing even remotely comparable, this is because Ssu-ma Ch™ien was con-
cerned with and trained in the empirical acts of observing and recording to

the study of their movements for the sake of perfecting the calendar and the
sundial. According to Tu, “his [Ssu-ma Ch™ien™s] research in the natural sciences
trained his thought and methods,” a statement that has deep implications
for understanding his historical method; see Tu Sheng-yün, “Ssu-ma Ch™ien
te t™ien-wen-hsüeh ch™eng-chiu he ssu-hsiang,” in Ssu-ma Ch™ien he Shih Chi,
pp. 222“48.
Descriptive does not mean “objective.” I do not mean to imply here that the
narrative provided by Ssu-ma Ch™ien is free from ideological overlay. To the
contrary, the system of thought that the Grand Historian seems to subscribe
to weighed heavily in his construction of historical narratives. The “rational-
ization” of the phenomenon within the accepted code of values provided by
correlative thought in its Han Confucian formulation will be brie¬‚y tackled in
chapter 8.


a degree that only an astronomer could have achieved. As we will see, the
information he included in the Shih chi covered a large spectrum, from the
ethnographic to the economic, and from the cultural to the political and

The Hsiung-nu Described

Within a few years after Meng T™ien™s titanic expedition to the Ordos (215
b.c.), the north ceased to be, for the Chinese intellectual, either the habitat
of wondrous beings or an abstract philosophical construct. Soldiers and
laborers were dispatched to build and protect the extensive northern forti-
¬cations, and colonists were sent to open up the new land. The north
became a new frontier not only on the political but also on the cognitive
level. As its geographic horizon expanded, China™s need for knowledge
increased. The military disaster at P™ing-ch™eng and the humiliating peace
terms that the Han had to submit to added yet another dimension to the
problem of the north, which also appeared, suddenly, as an unprecedented
threat to the very existence of the Chinese nation. China needed “experts”
to manage the north. The debate on foreign relations that developed in
China starting from the early Han was based on knowledge that, although
sometimes couched in a language replete with classical allusions, had been
acquired only recently. This knowledge was mastered by military experts
and politicians who specialized in foreign and, in particular, northern
affairs. A new breed of soldier emerged, specially trained to ¬ght against
the Hsiung-nu. Statesmen studied frontier management and submitted
memorials to the throne proposing incentives and relief measures for the
people sent to colonize the region.
By Ssu-ma Ch™ien™s time, knowledge about the Hsiung-nu and other
nomadic peoples in the north and west had been accumulating for sixty
years thanks to war, trade, and diplomacy. Chang Ch™ien™s expedition to
Ferghana had brought back invaluable information, directly relevant to the
opening of political and economic contacts between China and the states
of the Tarim Basin and beyond. More important, Ssu-ma Ch™ien™s early
treatment of foreign countries opened the way to incorporation in the Han
shu of even more detailed information, which became available after the
establishment of the Protector General™s Of¬ce (c. 60 b.c.). Would we now
have this wealth of knowledge without Ssu-ma Ch™ien™s earlier investiga-
tion? The question may be idle, but only to a point; it may be useful to
focus on the issue that, before the Shih chi, there were no intellectual “con-
tainers” in which factual historical and geographical data could be stored
and transmitted. Ssu-ma Ch™ien introduced empirical criteria for the col-
lection of this type of information and the recording of historical events.


His achievements, in the areas of ethnographic enquiry, geographical doc-
umentation, and historical accuracy, are all the more remarkable consider-
ing that during his lifetime older conceptions had by no means been
jettisoned by the intellectual community.
In this chapter, I will examine the type of sources from which Ssu-ma
Ch™ien may have drawn his information on the Hsiung-nu and then move
on to analyze the three core features of the “descriptive” phase of the Grand
Historian™s treatment of Inner Asia: ¬rst, Ssu-ma™s representation of Hsiung-
nu society, including the nomads™ way of life, customs, cults, and military
and social organization; second, Ssu-ma Ch™ien™s attitude toward the effects,
on both politics and the economy, that the wars against the Hsiung-nu had
on China; and third, Ssu-ma™s treatment of geographical knowledge and
trade, which re¬‚ects a more general interest of the historian in the possi-
bilities that its expanding frontiers offered China. On this last point, I will
not deal with speci¬c geographical questions, but with the manner in which
the regions beyond the limina of the Chinese community are represented in
the Shih chi, and how the changes undergone by the northern and western
borders affected Chinese society.


personal acquaintance. Ssu-ma Ch™ien, like Herodotus in Greece, was
regarded as one of the most widely traveled men of his time.42 Most prob-
ably, considering that he had very little written material to rely upon, Ssu-
ma Ch™ien collected a considerable amount of geographic and ethnographic
information either by interrogating travelers or by traveling himself. When
he was twenty, he journeyed through southern China, possibly for as long
as four or ¬ve years. Then, after entering the civil service as a petty of¬cial
(lang-chung), he took part in several expeditions to the west and to the
south.43 Moreover, in his capacity as lang-chung and, later on, as Grand
Historian, he continued to travel as a member of Emperor Wu™s retinue,
following the ruler on inspection tours or when he traveled to perform ritual
ceremonies. Especially relevant to this study is the travel that Ssu-ma Ch™ien
undertook in the year 110 b.c., when he accompanied Wu-ti on a journey
to inspect the northern border. There, as he says in the biography of Meng

On Ssu-ma Ch™ien™s travels, see B. Watson, Ssu-ma Ch™ien, p. 48; Shih Ting,
“Ssu-ma Ch™ien yu-li kao,” in Ssu-ma Ch™ien he Shih chi, pp. 126“44; Nieh
Shih-ch™iao, Ssu-ma Ch™ien lun-kao (Peking: Pei-ching shih-fan ta-hsüeh, 1987),
pp. 12“24.
On this title, see Rafe de Crespigny, Of¬cial Titles of the Former Han Dynasty
as Translated and Transcribed by H. H. Dubs. An index compiled by Rafe de
Crespigny (Canberra: Centre of Oriental Studies, 1967), p 15.


T™ien, Ssu-ma Ch™ien visited the long walls built by the Ch™in general
allegedly to contain the Hsiung-nu.44

oral accounts by chinese people. Just as Iordanes based his account
of the Huns, in Getica, on the information provided by Priscus, long-time
resident at the court of Attila, so did Ssu-ma Ch™ien rely on the written and
oral accounts of people who lived among the Hsiung-nu or, by being
employed in diplomatic or military missions, had acquired intimate knowl-
edge of them. One of the ¬rst was Ch™in K™ai, who was a hostage among
the Hsiung-nu even before the uni¬cation.45 Hostage exchange was a
provision regularly applied to treaties and an integral part of diplomatic
relations between states during the Warring States and during the Ch™in-
Han period.46
The Hsiung-nu, particularly during the ¬rst phase of the Han dynasty,
were a safe haven for rebellious Chinese leaders and dissatis¬ed military
commanders. Most famous is Chung-hang Yüeh (also read Chung-hang
Shuo), who ¬‚ed to the Hsiung-nu and provided them with inside knowl-
edge on how to conduct political negotiations with the Chinese. A speech
reported in the Shih chi in which he replied to the Chinese envoys™ claims
of superior virtue contains much information on Hsiung-nu life and habits.
This speech may have been recorded by the same Chinese of¬cials who had
gone to the Hsiung-nu court and who later circulated at the Han court,
where Chung-hang Yüeh was regarded as a renegade and a traitor. It is
worth quoting from, for behind the Confucian rhetoric that Chung-hang
Yüeh used to turn the tables on the Han diplomats, there is a close descrip-
tion of the Hsiung-nu society:
The Hsiung-nu clearly make warfare their [main] occupation; since the old
and weak cannot ¬ght, the best food and drink are given to the strong and
healthy, who then become the defense and protection [of the nation]; in this
way both fathers and sons can live long in security. Can one really say that
the Hsiung-nu despise old people? [. . .] According to Hsiung-nu custom,
people eat the meat of their animals, drink their milk, and wear [clothes made
with] their hides; the animals eat grass and drink water, therefore they move
about in seasonal cycles. This is why in critical times they practice riding
and shooting, while in peaceful times they enjoy themselves without other
engagements. Their communal laws are not burdensome, and are easy to

Shih chi 88, 2570; Shih Ting, “Ssu-ma Ch™ien yu-li kao,” 138; Derk Bodde, States-
man, Patriot and General in Ancient China (New Haven: American Oriental
Society, 1940), pp. 62“63.
Shih chi 110, 2885“86.
Han shu 94A, 3772; Shih chi chu-yi 2331. On hostage keeping, see Lien-sheng
Yang, “Hostages in Chinese History,” Harvard Journal of Asiatic Studies 15
(1952): 507“21.


implement. The relationship between ruler and subject is relaxed and simple,
so that ruling the whole country is just like [ruling over] a single person. When
a father, son or brother dies, they take their [i.e., the dead persons™] wives as
their own, because they hate to see a lineage die out. Therefore, even when
the Hsiung-nu meet with turmoil, the ancestral clan is always preserved.47
Prisoners of war were another source of information. Every battle pro-
duced for the victor, on either side, large numbers of prisoners that were
taken back to the victorious countries and used in several capacities, as
slaves, servants, or soldiers. In particular if they were persons of rank,
returnees must have provided valuable information, and Ssu-ma Ch™ien was
personally acquainted with military men who had extensive ¬rst-hand
knowledge of the Hsiung-nu, such as General Li Kuang and Su Chien, who
was a subordinate commander under Wei Ch™ing in the war against the

contemporary written documents and reports. Because the Hsiung-
nu had already been for several years a topic of heated debate at the Han
court, documents existed that Ssu-ma Ch™ien could and did mine to gather
the information he needed. Among these, the most most important, and
also the best known, are the memorials by Ch™ao Ts™o, which we have dis-
cussed in previous chapters.

hsiung-nu people in chinese society. Surrendered or captured Hsiung-
nu who had entered China constituted an additional source of information
for Ssu-ma. Two Hsiung-nu can be regarded as representative of the
nomads™ presence in China. One was the person who traveled to Central
Asia with Chang Ch™ien, a Hsiung-nu slave called Kan Fu. Over the thir-
teen years of Chang Ch™ien™s travels and captivity among the Hsiung-nu,
Kan Fu proved a loyal and resourceful aide to the Chinese explorer, and he
must have gained considerable notoriety to have been mentioned in Chinese
sources together with his famous master. The other person was Chin
Mi-ti, heir apparent to the throne of the Hsiung-nu king of Hsiu-ch™u, who
was captured by another Hsiung-nu chief, turned over to the Chinese, and
employed as a slave inside the Yellow Gate Palace to tend horses under the
supervision of the eunuchs.49

Shih chi 110, 2899“2900; Shih chi chu-yi 110, 2323 (Burton Watson, trans.,
Records of the Grand Historian by Sima Qian [New York and Hong Kong:
Columbia University Press and The Chinese University of Hong Kong, 1993], 2:
Shih chi 111, 2946.
C. Martin Wilbur, Slavery in China under the Former Han Dynasty, 206 B.C.“
A.D. 25 (Chicago: Field Museum of Natural History, 1943), pp. 306“307.


Ethnography of the Hsiung-nu

It is possible that Ssu-ma Ch™ien might have been regarded as a “bar-
barophile” by his contemporaries. In reporting the well-known apology for
the Hsiung-nu allegedly made by Chung-hang Yüeh, Ssu-ma Ch™ien showed
a view distinct from the orthodox faith shared by many of his contempo-
raries in the superiority of the Han rituals and civilization.50 Although the
Hsiung-nu may have been cruel, greedy, and arrogant, they also had their
own ways and traditions, and Ssu-ma Ch™ien™s sympathies may have gone
to those who, like himself, tried to understand them. Between Li Kuang,
the general who had trained his soldiers to ¬ght the Hsiung-nu using their
own methods, and the more orthodox Ch™eng Pu-chih, the historian
undoubtedly preferred Li Kuang.51 Ssu-ma Ch™ien, however, feared censor-
ship, and in the colophon to chapter 110 he admitted that he was not free
to speak openly. Even more signi¬cantly, Pan Ku did not mention Ssu-ma
Ch™ien™s name at the close of the chapter on the Hsiung-nu (94) in the Han
shu, when he discussed policies for dealing with the Hsiung-nu, such as
those proposed by Tung Chung-shu. Pan Ku quotes the Shu ching, Shih
ching, and Spring and Autumn Annals but does not mention the Shih chi,
although the ¬rst part of Pan Ku™s chapter is almost a verbatim copy of
chapter 110 of the Shih chi.
Pan Ku was clearly at odds with his predecessor™s views, and the fact
that he explicitly supported a more militant, forward policy is further evi-
dence of Ssu-ma Ch™ien™s paci¬st leanings.52 In Pan Ku™s description of the
Hsiung-nu, contrary to that of Ssu-ma Ch™ien, we ¬nd strong derogatory
expressions, such as that the Hsiung-nu had human faces but hearts of
beasts. These two almost antithetical approaches arose from differences in
methods and aims. Whereas Ssu-ma Ch™ien lived through a period of sharp
confrontation between the nomads and China (one that required China™s
urgent understanding of its neighbors), in Pan Ku™s age the Hsiung-nu had
become less threatening, China was more con¬dent of its own power as a
uni¬ed empire, and a stricter Confucian orthodoxy had asserted itself. Pan
Ku™s aim was to explain, and perhaps intervene in, the internal political
debate on foreign policy, rather than to investigate the nature of the Hsiung-
nu. Ssu-ma Ch™ien™s approach, in contrast, developed out of the experience

Shih chi chu-yi 110, 4: 2323“24. 51 Shih chi 109, 2869.
On Pan Ku™s theory of foreign relations, see Wang Gungwu, “Early Ming Rela-
tions with Southeast Asia: A Background Essay,” in The Chinese World Order,
ed. John K. Fairbank (Cambridge: Harvard University Press, 1968), p. 41; Chusei
Suzuki, “China™s Relations with Inner Asia: The Hsiung-nu, Tibet,” in The
Chinese World Order, p. 183.


and knowledge he had gathered while traveling, which made the historian
recognize that the Hsiung-nu constituted, in his age, an unsolved historical
problem, an anomaly that China had to face in the process of asserting itself
as a politically and culturally uni¬ed entity.
Various aspects of the Hsiung-nu way of life are carefully reported in
the Shih chi, including their social organization, rituals, and religious
beliefs. Characteristic of the historian™s narrative in this respect is the objec-
tivity of his observations and its remarkable freedom from the prejudices
quite common among Chinese intellectuals and politicians. In Ssu-ma™s
description of Hsiung-nu customs we can therefore distinguish two types
of information. The ¬rst type, direct, is information that Ssu-ma Ch™ien
provided himself, whereas the second, indirect, is information reported by
him as other people™s opinion. In the latter case, it is possible that Ssu-ma
Ch™ien was expressing his own thoughts using other people™s names to avoid
blame or to add greater weight to the opinions expressed. Nevertheless,
both contribute to give us a fairly accurate picture of Hsiung-nu customs.
For this analysis, “ethnographic” material on the Hsiung-nu has been
divided into the following categories: pastoral nomadism, burial customs,
society and laws, military training and war, state sacri¬ces and ritual, and

pastoral nomadism. Ssu-ma Ch™ien™s account of the Hsiung-nu is the
¬rst ever “ethnographic” account of nomadic peoples in Chinese history
even though pastoral nomadism, as a specialized economic activity, had
established itself in the steppe region to the north of China several centuries
before the Han. Ssu-ma Ch™ien™s account is clearly based on a detailed
knowledge of pastoral economy:
Most of their domestic animals are horses, cows, sheep, and they also have
rare animals such as camels, donkeys, mules, hinnies and other equines
known as t™ao-t™u and tien-hsi.53 They move about according to the avail-
ability of water and pasture, have no walled towns or ¬xed residences, nor
any agricultural activities, but each of them has a portion of land.54
This description of nomadic life, which became clich© in following
Chinese histories, presents some interesting points. First, Ssu-ma Ch™ien is
very speci¬c in listing all the animals bred by the nomads. This conforms
to the reality of pastoral life in places such as traditional Mongolia, where
the ¬ve animals commonly bred were horses, cows, sheep, goats, and
camels. The donkey and crossbred equines that cannot be readily recog-

On these animals, see Namio Egami, “The Kuai ti, the Tao you, and the Dao xi:
The Strange Animals of the Xiongnu,” Memoirs of the Research Department of
Toyo Bunko 13 (1951): 87“123.
Shih chi 110, 2879 (cf. B. Watson, Records, 2: 129).

nized were also part of the zoological inventory of the nomads. The breed-
ing and herding techniques for these animals, each of which requires a
special type of knowledge and care, could not have been a matter of rapid
evolution, and they strongly indicate that a mature pastoral nomadic
economy had existed in northern China for centuries. Second, in pointing
out that “each of them has a portion of land,” Ssu-ma Ch™ien seems to be
referring to individual nomads, though this is not to exclude that he meant
families and clans, because land rights were customarily recognized, in pas-
toral societies, on the basis of lineage. Ssu-ma Ch™ien™s phrase, which may
imply the existence of “private property,” should be interpreted as meaning
that the Hsiung-nu recognized for individuals and families the right to make
use of speci¬c pastures for their herds, even though the land was not fenced
and could not be purchased or sold.
If landed property rights, in whatever guise, existed among the Hsiung-
nu, it is unlikely that their movement in search of water and grass could be
a random one; Ssu-ma Chien™s statement that they had a “portion of land”
indicates that migrations were con¬ned to a given territory and implies the
notion of cyclicality, which is the most essential aspect of nomadic economy.
To my knowledge, this is the ¬rst written indication that the Chinese had
understood the seasonal regularity of the movement of the herds. The search
for water and grass symbolizes the nomadic lifestyle. In describing how
General Li Kuang organized his campaigns against the Hsiung-nu, it was
said that “he would make camp wherever he found water and grass.” This
expression suggests that he had adopted nomadic tactics of warfare to ¬ght
against the Hsiung-nu.55

burial customs.
In burials they use inner and outer cof¬ns, gold and silver [ornaments], clothes
and fur coats; however, they do not erect earthen mounds or plant trees, nor
do they use mourning garments. When a ruler dies, his ministers and concu-
bines are sacri¬ced in numbers that can reach several tens or even hundreds
of people.56
Ssu-ma™s description of burial customs is particularly interesting in view
of recent archaeological discoveries. A number of Warring States sites
attibuted to the Hsiung-nu, in particular Hsi-kou-p™an and A-lu-chai-teng,
contain large quantities of gold and silver ornaments. This funerary inven-
tory differs from those of earlier nomadic sites in northern China in the
absence of weapons. The Shih chi statement seems to con¬rm the attribu-
tion to the Hsiung-nu of burials that feature large quantities of gold and
silver ornaments.

Shih chi 109, 2869 (cf. B. Watson, Records, 2: 119).
Shih chi 110, 2892 (cf. B. Watson, Records, 2: 137).


The issue arising from the piece of information related to the presence
of an inner and outer cof¬n is more complex. Whereas the Hsiung-nu
burials in Inner Mongolia, unlike those of the nomads of South Siberia or
those of the Saka and Wu-sun of Sinkiang, are not marked by an external
mound, the inner and outer cof¬n is present only in a few of the Hsiung-
nu burials identi¬ed so far, which are far outnumbered by the more typical
vertical earthen pits. In contrast, the double cof¬n is a standard Chinese
method of burial. Therefore, the reference to the double cof¬n could lead
us to believe that, according to Ssu-ma Ch™ien, the Hsiung-nu had borrowed
a Chinese custom and were being made to appear indebted to China even
though, in reality, that was not the case. It is true, however, that in sites
such as Mao-ch™ing-kou, we ¬nd a double cof¬n.57 Moreover, it is possible
that a number of wooden cof¬ns have disappeared simply because of decay.
In fact, though not frequent, the pit with a cof¬n placed in a timber frame
(a second cof¬n) with no overgrave setting has been identi¬ed by archae-
ologists as one of the Hsiung-nu burial structures. The hypothesis proposed
by archaeologists such as Minyaev, who believe that different burial struc-
tures re¬‚ected social strati¬cation, would indicate that Ssu-ma Ch™ien was
not referring to all Hsiung-nu burials, but only to those of the aristocracy,
which were more elaborate than those of commoners.58 His information
was partly inaccurate, though, since in Inner Mongolia there are a number
of rich Hsiung-nu burials, undoubtedly belonging to prominent people,
with no double cof¬n. Nevertheless, it cannot be taken as a statement that
lacked historical reality, whose only purpose was to show a cultural bor-
rowing from China.
The last statement, referring to the killing of many scores of people sac-
ri¬ced at the death of the ruler, also does not ¬nd direct con¬rmation in
Hsiung-nu archaeology. However, no “imperial” Hsiung-nu grave has so
far been found, and therefore the evidence available is simply insuf¬cient
to either prove or disprove this statement. We may add that among the
Mongols it was customary to kill all people who witnessed the funeral and
burial place of a khan, to protect the secrecy of the location of the grave.
If such a tradition existed also among the Hsiung-nu, this might explain in
part Ssu-ma Ch™ien™s mention of a sacri¬ce en masse, though the speci¬c
mention of ministers and concubines would not be correct.

society and laws. Hsiung-nu society was, to the status-conscious
Chinese, a remarkably egalitarian society, with little differentiation between

Thomas O. Höllmann and Georg W. Kossack, eds., Maoqinggou: Ein eisen-
zeitliches Gr¤berfeld in der Ordos-Region (Innere Mongolei) (Mainz: Verlag
Philipp von Zabern, 1992): Table 15, A.
S. Minyaev, “On the Origin of the Hsiung-nu,” Information Bulletin. Interna-
tional Association for the Cultures of Central Asia 9 (1985): 69“78.


commoners and “aristocracy,” as we have already seen in the memorial by
Chung-hang Yüeh cited earlier. Ssu-ma Ch™ien™s description of Hsiung-nu
laws indicates the same type of existence unencumbered by burdensome
social constraints:
According to their laws, those who draw the sword one foot [out of the scab-
bard] are sentenced to death; those who steal lose their properties; those guilty
of minor offences are ¬‚ogged, those guilty of major ones are sentenced to
death. The longest period in jail does not exceed ten days; the imprisoned
men in the whole country are very few.59
Finally, “They have no written language, and customary laws are only
The relatively freer and simpler existence of the Hsiung-nu is not con-
trasted unfavorably with Chinese society. On the contrary, behind the plain
description of these simple social and legal rules, one might see a veiled
criticism of the cumbersome legal system put into place by Ch™in Shih
Huang-ti and his legalist advisers. Although the Han rejected laws and pun-
ishments as the only tool of social order and cohesion, the daily life of a
Chinese subject was by no means free. Ssu-ma Ch™ien™s own opposition
to the excessive recourse to laws and punishment was openly stated in his
praise of the early Han, when “the meshes of the law were spread so far
apart that a whale could have passed through.”60 In Wu-ti™s time, instead,
the laws had become increasingly stricter and more oppressive, and myriads
of people were executed, often on trumped-up charges. Even more worry-
ing was the realization that the proliferation of laws and the increased use
of punishments did not protect the subjects from abuse and unlawful action,
often brought about by the very people who were supposed to enforce
the law. As Ssu-ma Ch™ien states in his ¬nal remarks to chapter 122, “from
the time of Zhang Tang [Chang T™ang]™s death on, the net of laws was
drawn tighter and tighter, and harsh penalties became increasingly frequent,
so that that the work of government of¬cials was gradually hampered and
brought to a standstill.”61 Yet, harsher laws and punishments were regarded
by him as a necessary evil, to be preferred to the total absence of them,
which would only favor unbridled tyranny. This was certainly a gloomy
picture of his own times, in comparison with which the simpler life of the
nomads must have seemed to match in some way the ideal of an effortless
social machine, uncoercive and yet fully functioning by virtue of its own
In this game of contrasting images, the historian paints a society, the
Han, that has still to ¬nd the measure of its own values, morality, and

Shih chi 110, 2879; Shih chi chu-yi 110, 2313 (cf. B. Watson, Records, 2: 137).
Shih chi 122, 3131 (cf. B. Watson, Records, 2: 380).
B. Watson, Records, 2: 407.

norms of social interaction.62 The “holier-than-thou” stance of the Chinese
envoys to the Hsiung-nu is not only ridiculed by Chung-hang Yüeh but
also questioned by Ssu-ma Ch™ien. This type of introspective, self-critical
attitude toward his own society is not overt, but transmitted to the reader
through the laconic description of a different system. Ssu-ma Ch™ien™s
description of the Hsiung-nu is not value free. If on the one hand his effort
to document other social realities demonstrates a new openness to the
outer world, on the other it is also anchored to a well-de¬ned ethical
basis, of which the reader must be conscious. It is in the silent contrast
between the description of the “different” and the consciousness of one™s
own cultural dimension that the historical narrative ¬nds its true and
most powerful message. The egalitarian, simple, harsh but fair, and
above all free existence of the nomads acquires a special attractiveness
only by contrast with that of the Chinese subject, whose life is fettered by
many laws and endangered by cruel punishments and whimsical law

military training and warfare. The Hsiung-nu military superiority,
at least in the use of cavalry, was evident to all the Chinese, but whereas
Ch™ao Ts™o and other theorists were interested in ¬nding ways to beat the
Hsiung-nu on the battle¬eld, Ssu-ma Ch™ien was interested in ¬nding the
reasons for their strength. In the description of the progress of the young
nomads from children shooting small animals to physical maturity, the
secret of their equestrian pro¬ciency and excellent marksmanship was
plainly explained:
As children they are able to ride sheep, and can shoot birds and mice with
bow and arrow. As they grow a little older, they can shoot foxes and hares,
which they use for food. Thus as adults they are strong enough to bend a
bow, and all can serve as cavalry soldiers. It is their custom to make their
living in times of peace by herding the domestic animals and hunting the wild
ones, but in critical situations everyone practices military skills in order to set
off on invasions. This is their inborn nature.63
In the Huai-nan-tzu, a text compiled during the reign of Wu-ti, considerable
thought is devoted to the law, a topic that had been for a long time a central
pillar of Chinese political philosophy. Whereas laws were considered necessary
by the authors of the Huai-nan-tzu, an effort is devoted to smoothing their edges
and con¬ning their potentially oppressive use, by recourse to the principle of
“rightness,” that is, consistency with what is desirable to the people. The search
for a balance between the “legalist” framework that informed the government
structure, with its laws and punishments, and the Confucian and Taoist empha-
sis of moral principles must have been deeply felt in contemporary Han society.
On this, see Ames, The Art of Rulership, pp. 138“40.
Shih chi 110, 2879; Shih chi chu-yi 110, 2313 (cf. B. Watson, Records,
2: 129).


It was truly their way of life, intimately connected with animals, whether
they rode, herded, or hunted them, that produced exceptional mounted
warriors. The expression “this is their inborn nature” (ch™i t™ien-hsing yeh)
which could also be translated as “this is their natural behavior,” is, in a
way, meant to reassure those who might have thought that these enemies,
almost invincible until Wu-ti™s time, were endowed with special powers.
Ssu-ma Ch™ien brings a disturbing and even mysterious fact back onto a
plane of rational understanding by clarifying, step by step, the essence
of nomadic military training, and how this was the result of a different,
but nonetheless natural, process of growth due to the pursuit of speci¬c
In the description of Hsiung-nu armament and tactics, Ssu-ma Ch™ien
does not indulge in long-winded comparisons with the Chinese. His narra-
tive is remarkably objective, and “moral” considerations are kept to a
minimum. He describes their weapons (“they use bows and arrows as their
long-range weapons, and swords and spears as their short-range
weapons”)64 and their habits when it comes to going to war:
At the beginning of a [military] enterprise, they observe the stars and the
moon; if the moon is rising they attack, if it is waning they retreat. [. . .] They
are skilled in the use of troops that lure the enemy into an ambush. As they
see the enemy they look for booty, [behaving] like a ¬‚ock of birds. When they
meet with hardship and defeat, they disintegrate and scatter like clouds. Those
who bring back from battle the body of a dead [Hsiung-nu] gain complete
possession of the dead man™s household and properties.65
The analogy of foreign enemies with beasts and birds ultimately goes
back to the classics, but in Ssu-ma Ch™ien™s age it was also a common way
to refer speci¬cally to the Hsiung-nu. In the memorial presented by Chu-fu
Yen to Wu-ti there are references to quasi-contemporary documentary
sources on policy making regarding the Hsiung-nu, which, on account of
the similarities with the Shih chi™s narrative, are very likely among the
sources used by Ssu-ma Ch™ien. It is said, for instance, that Li Ssu repri-
manded Ch™in Shih Huang-ti for sending troops into the Ordos by saying,
“it is impossible. The Hsiung-nu have no ¬xed cities or forts and no stores
of provisions or grain. They move from place to place like ¬‚ocks of birds
and are just as dif¬cult to catch and control.”66 Chu-fu Yen™s memorial also
quoted a warning that Kao-tsu had received from an imperial secretary on
the eve of his defeat at P™ing-ch™engch™eng: “It is the nature of the Hsiung-
nu to swarm together like so many beasts, and to disperse again like a ¬‚ock
of birds. Trying to catch them is like grabbing a shadow. In spite of all Your

Shih chi chu-yi 110, 2313; Shih chi 110, 2879.
Shih chi 110, 2892 (cf. B. Watson, Records, 2: 137).
Shih chi 112, 2954 (cf. B. Watson, Records, 2: 194).

Majesty™s noble virtue, I fear any attempt to attack the Hsiung-nu will only
lead to danger.”67
Ssu-ma Ch™ien™s description of the nomads™ behavior in battle is accom-
panied by expressions of moral disapproval:
During a battle, if this is going well for them, they will advance, otherwise
they will retreat. They do not regard running away as something shameful;
they only care about li (pro¬t) and do not know of li (propriety) and yi
When they ¬ght in battle, those who have cut [enemy] heads or captured pris-
oners are presented with a cup of wine, and all the booty they have taken is
also given to them; the people they capture are made into slaves. Therefore,
in battle each man pursues his own gain.69
The statement about the Hsiung-nu™s shameless pursuit of loot parallels
the Tso chuan statement that the Ti are not ashamed of running away.70
The “barbarians™ ” greediness, while it can be related to a pre-existing
clich©, is set here in a much less ideological context, as if Ssu-ma Ch™ien
was relating common knowledge among the military people of his age. It
is, in other words, an opinion that is not necessarily derivative of the classic
tradition but is a prejudice of his own age, when the Chinese military people
prided themselves on ¬ghting “by the rules” and abiding by notions of
honor and sel¬‚essness. The Hsiung-nu had a different set of rules, and Ssu-
ma Ch™ien™s disconcerted judgment is similar to the cultural stance adopted
by the Greeks and Romans when they described the ¬ghting methods of
the steppe nomads.71
The ideal of the simplicity of “barbarian” life, opposed to the rule-laden
Chinese life, is proposed again, with reference to the military, in the biog-
raphy of General Li Kuang. His troops were loosely organized, never kept
in formation; when they camped there were few rules, and record keeping
was kept to a minimum; the only precaution he took was to send out scouts
on patrol. In contrast, the more traditional Ch™en Pu-shih kept his men con-
stantly busy, enforced a harsh discipline, and had his of¬cers constantly
writing reports. Ssu-ma Ch™ien™s comment is that both generals were suc-
cessful, and were not going to be caught unprepared, but the enemy was
more afraid of Li Kuang™s tactics, and the soldiers were happier to serve
under him.

Shih chi 112, 2955 (cf. B. Watson, Records, 2: 195).
Shih chi 110, 2879; Shih chi chu-yi 110, 2313 (cf. B. Watson, Records, 2: 129).
Shih chi 110, 2892 (cf. B. Watson, Records, 2: 137).
Ch™un Ch™iu Tso chuan chu, 1: 322.
Denis Sinor, “The Greed of the Northern Barbarians,” in Aspects of Altaic
Civilizations II, ed. Larry V. Clark and Paul A. Draghi (Bloomington: Indiana
University, 1978), pp. 171“82.


state sacri¬ces and rituals. Besides the structure of the Hsiung-nu
government, the Shih chi also illustrates other aspects of Hsiung-nu politi-
cal life, such as their state sacri¬ces and ritual practices.72 In many ways,
these rituals are reminiscent of Chinese ceremonies.
At dawn the Shan-yü leaves his camp and makes obeisance to the sun as it
rises, and in the evening he makes a similar obeisance to the moon. [. . .]
When they sit the place of honor is on the left side, toward the north. The
wu and chi days [i.e., the ¬fth and sixth of the ten-day week] are their favorite
ones. [. . .] Every year in the ¬rst month the important people hold a restricted
meeting at the Shan-yü™s court, and perform sacri¬ces. In the ¬fth month they
have a large gathering at Lung-ch™eng, where they sacri¬ce to the ancestors,
Heaven and Earth, and to their divinities. In autumn, when the horses are
fat, they hold a large meeting in which they encircle a forest (tai lin)73 and
calculate the number of people and livestock.74
From this passage we infer that the Hsiung-nu were using a calendar
based on the ten heavenly stems. They worshipped Heaven, the ancestors,
and their deities on the wu (¬fth stem) and chi (sixth stem) days. These stems
corresponded to the element earth and represented the middle, fortune, and
blessing. Politically, they represented the power to govern the tribes on the
four sides. In the Lü-shih ch™un-ch™iu and Li-chi it is explicitly stated that
the element earth corresponded to the center, and its days were wu and chi.
In the chapter “T™ien-wen” of the Huai-nan Tzu it is also said that “the
centre was the Earth; it was ruled by the Yellow Emperor who controlled
and restrained the four quarters; its planetary god was Chen-hsing, its
animal symbol was the yellow dragon; its note on the musical scale was
kung; its days were wu and chi.”75 The similarities raise the question of
whether there was any Hsiung-nu borrowing from the Chinese political

On the religion of the Hsiung-nu, see Hsieh Chien (Jiann), “Hsiung-nu tsung-
chiao hsin-yang chi ch™i liu-pien” [The Religious Beliefs of the Hsiung-nu and
Their Later Development] in Li-shih yü-yen yen-chiu so chi-k™an 12.4 (1971):
571“614. Important information on the rituals and sacri¬ces mentioned in the
foregoing passages is summarized in de Crespigny, The Northern Frontier, pp.
507“508, n. 15.
Shih chi 110, 2893. There are different interpretations of the meaning of tai lin.
Some say it means “to encircle a forest” (in the process of performing a sacri-
¬ce), others that it just means “forest,” and others claim that it is a geographical
name (see Shih chi chu-yi 110, 2319). De Groot locates it in the region of the
Han city of Ma-yi, in the Yen-men commandery; see J. J. M. De Groot, Chine-
sische Urkunden zur Geschichte Asiens I: die Hunnen der vorchristlichen Zeit
(Berlin and Leipzig: W. de Gruyter, 1921), pp. 59“60.
Shih chi 110, 2892 (cf. B. Watson, Records, 2: 137).
Quoted in Chen Ching-lung, “Chinese Symbolism among the Huns,” in
Religious and Lay Symbolism in the Altaic World, ed. Klaus Sagaster (Wiesbaden:
Harrassowitz, 1989), p. 67.


tradition, or whether Ssu-ma Ch™ien was deliberately referring to an ac-
cepted symbology of sovereignty to stress the concept that the Hsiung-nu
were an independent nation.76 This issue is somewhat involved, because we
do not have any corroborating source that may clarify the situation. The
Turco-Mongol people adopted in part the Chinese calendar, but that may
also have been a later development. But the expression of “bringing the cal-
endar” to foreign countries in Chinese sources is synonymous with politi-
cal and cultural expansionism. In this passage, however, no relationship of
subordination can be detected, and we can only speculate on the possible
implications. As this passage, taken in its entirety, seems to be consistent
with the “descriptive” narrative mode, we can formulate three hypotheses.
First, the Hsiung-nu, under the in¬‚uence of Chinese advisers such as Chung-
hang Yüeh, had started to make use of Chinese symbols of royalty such as
the calendar. Second, the source that Ssu-ma Ch™ien had relied on for this
information, possibly some Hsiung-nu captive or envoy, being familiar with
the Chinese rituals, had added that information to enhance the prestige of
the Hsiung-nu court. Third, an autochthonous calendrical tradition similar
to the Chinese actually existed among the Hsiung-nu.

language. The Shih chi reports an unprecedented number of Hsiung-nu
words. To be sure, these are still rather few, and they are insuf¬cient to
provide conclusive evidence of the type of language actually spoken by the
nomads.77 Nonetheless, the words™ inclusion represents a new level of
sophistication in the information that Chinese historical sources provide on
different cultures.
Very few “Hsiung-nu” words appear in works anterior to the Shih chi.
In the Yi Chou shu, chapter 36, “K™o Yin,” we ¬nd words such as ching-
lu, the Scythian dagger known in Greek sources by the name of akinakes,
and, in chapter 59, “Wang Hui,” there are two words that indicate some
type of horse, t™ao-t™u and chüeh-t™i.78 The portions of the Yi Chou shu

On the development and use of the calendar among ancient Turco-Mongol
peoples, see Louis Bazin, Les Systemes Chronologiques dans le Monde Turc
Ancien (Budapest: Akad©miai Kiadó, 1991).
The possible etymology of Hsiung-nu words is discussed in the following works:
J. Benzing, “Das ˜Hunnisches,™ ” in Philologia Turcica Fundamenta (Wiesbaden:
Franz Steiner, 1959), 1: 685“87; E. G. Pulleyblank, “The Hsiung-nu Language,”
Asia Major, n.s. 9 (1962): 239“65; G. Doerfer, “Zur Sprache der Hunnen,”
Central Asiatic Journal 17.1 (1973): 1“50; O. J. Maenchen-Helfen, The World
of the Huns (Berkeley: University of California Press, 1973), pp. 372“73;
O. Pritsak, “The Hsiung-nu Word for ˜Stone,™ ” in Tractata Altaica (Wiesbaden:
Harrassowitz, 1976), pp. 479“85; H. Bailey, Indo-Scythian Studies. Khotanese
Texts VII (Cambridge: Cambridge University Press, 1985), pp. 25“41.
On ching-lu, see Jaroslav Prusek, Chinese Statelets and the Northern Barbarians

in the Period 1400“300 B.C. (Dordrecht: Reidel, 1971), p. 133.

where these words appear, however, are, like the bulk of Yi Chou shu itself,
not ancient, dating to around 300 b.c.79 This date may possibly mark the
beginning of the acquisition of knowledge about northern languages in
China, a knowledge that appears to have increased considerably by the time
of Ssu-ma Ch™ien.
Inner Asian words appear in the Shih chi and Han shu in various con-
texts. Titles are the most common, such as “queen” (yen-chih) or the
various “kings” at the court of the ch™an-yü.80 These could not be trans-
lated into Chinese without incurring some ideological or terminological dif-
¬culty. Another type of words is those usually de¬ned as “cultural” words,
that is, words speci¬c to a given culture and lifestyle. Examples of these
words are “wagon” (fen-wen), “bag” or “basket” (chia-tou), the already
mentioned “dagger” (ching-lu), “tent,” possibly a yurt, (ch™iung-lu), and
“kumiss” (lo), another type of fermented mare™s milk was called t™i-hu,
“dried curd” (mi-li), and “fat” or “butter” (su).

Geographic Expansion and Trade

In the Shih chi place names, distances, and topographical information all
appear to conform to a high standard of accuracy, and for the ¬rst time
geographic information beyond the boundaries of the Central Plain became
a necessary ingredient of the historical narrative.81 Ssu-ma Ch™ien™s

Edward Shaughnessy, “The I Chou shu,” in Early Chinese Texts, pp. 229“33.
The word wang, usually translated as “king,” was a common one in Han society
and indicated a variety of peoples, including the chu-hou-wang, the sons of the
emperor, probably better translated as “princes,” and other nobles.
There is extensive scholarship on the correct identi¬cation of place names. The
toponomastics of foreign lands is naturally important in assessing the level of
information possessed by the Chinese during the Han dynasty and also the lin-
guistic identity of the people living in those distant regions. However, in the
context of Ssu-ma Ch™ien™s approach to foreign peoples, my objective is to look
at the organization of geographical information rather than attempt to identify
names of localities and regions. Most of the geographical research on Inner Asia
has concentrated on chapter 123, on the Western Regions. Important studies
include A. F. P. Hulsew©, “The Problem of the Authenticity of Shih-chi,” T™oung
Pao, 66 (1975): 83“147; A. F. P. Hulsew© and Michael Loewe, China in Central
Asia. The Early Stage, 125 B.C.“A.D. 23. An Annotated Translation of the Chap-
ters 61 and 96 of the History of the Former Han Dynasty (Leiden: Brill, 1979);
J. R. Gardiner-Gardner, “Chang Ch™ien and Central Asian Ethnography,” Papers
of Far Eastern History 33 (1986): 23“79; D. D. Leslie and K. H. J. Gardiner,
“Chinese Knowledge of Central Asia,” T™oung Pao 68.4“5 (1982): 254“308;
Edwin G. Pulleyblank, “Chinese and Indo-Europeans,” Journal of the Royal
Asiatic Society (1966): 9“39; and id., “Han China in Central Asia,” International

rationalistic attitude was no longer satis¬ed with a vision of the geographic
space as an accessory to mythology and a repository of legendary and fan-
tastic accounts, whose images no longer responded to the increased data
and precision demanded by the new political situation of a uni¬ed and
expanding empire. The motives for the transition from a mythological geog-
raphy to an “exploratory geography” should be sought in a direction dif-
ferent from the literary tradition.82 The main factors that contributed to the
“discovery” of foreign regions and peoples (as consciously investigated and
realistically described agents in historical events) can be summarized as
First, after 221 b.c. Chinese intellectuals and statesmen could ¬nally rec-
oncile the notion of a common culture with that of political and adminis-
trative unity. The passage from a state of fragmentation to one of uni¬cation
and centralization embodied in the concept of t™ien-hsia was the essential
precondition for the development of geographical knowledge. To be able
to look beyond its “national” boundaries, China had to become a single
political body and thus abandon the inward-looking attitude characteristic
of the segmented community of the Warring States period.83 The vastness
of the territory under a central administration, the increased control
imposed by the Han emperors over regional centers of power, and the
ampli¬cation of the bureaucracy are all elements that called for a more
precise knowledge of the land.
Second, China™s territory had been increasingly expanding into foreign
lands ever since the last phase of the Warring States period.84 The military
push that took place in that phase on the one hand brought the Chinese
people into closer contact with foreigners, and was responsible, especially
along the frontiers of the northern states, for the creation of amalgams of
different ethnic groups. On the other hand, the military reorganization of
the frontier, especially in terms of forti¬cations, road building, and estab-
lishing garrisons for guarding strategic places, had called for a better and

History Review 3 (1981): 278“96; Yves Hervouet, “Le valeur relative de textes
du Che-Ki et du Han-chou,” in Melanges de Sinologie off©rts a Monsieur Paul
Demieville, part II (Paris: Bibliothèque de l™Institute des Hautes Etudes Chinoises,
1974), pp. 55“76; Paolo Daf¬nà, “The Han Shu Hsi Yu Chuan Re-Translated. A
Review Article,” T™oung-pao 68.4“5 (1982): 309“39; and Kazuo Enoki, “On the
Relationship between the Shih-chi, Bk. 123 and the Han-shu, Bks. 61 and 96,”
Memoirs of the Research Department of Toyo Bunko 41 (1983): 1“31.
R. Mathieu, “Fonctions et moyens de la g©ographie dans la Chine ancienne,”
Asiatische Studien 36.2 (1982): 139.
Mathieu, “Fonctions et moyens de la g©ographie dans la Chine ancienne,” p. 125.
Particularly signi¬cant is the expansion of Ch™in in the southwest. On the con-
quest of the states of Shu and Pa and their signi¬cance in the uni¬cation of the
Chinese empire, see Steven Sage, Ancient Sichuan and the Uni¬cation of China
(Albany: State University of New York Press, 1992), pp. 107“17.

more precise knowledge of the terrain. From what we know from Ssu-
ma Ch™ien™s accounts, maps were already used by the Ch™in in the year
227 b.c.85 Although this does not exclude the development of cartography
long before then, it seems that the most remarkable advancements in map
making “ or at least our knowledge of it “ do indeed go back to the early
imperial period.86 Two Han military maps found at Ma-wang-tui™s tomb
no. 3 testify to the high level of technical specialization reached in this ¬eld.
The maps include not only indications of encampments and forti¬cations


. 11
( 15)