Likeness of No One: (Re)presentingthe
First Emperor's Army
LadislavKesner
The subterranean army of several thousand life-size terracotta
soldiers, horses, and war chariots, under excavation
since 1974 from the site near the supposed burial chamber of
the first Chinese emperor, Qin Shi Huangdi (r. 246-210
B.C.), in Lintong, Shaanxi province, has become famous far
beyond the field of Chinese archaeology and early history as
one of the grandest archaeological discoveries of this century.'
Close to the tumulus which, according to historical
accounts, should contain the tomb of the First Emperor
itself, the terra-cotta army, situated in three subterranean
pits, is the most conspicuous part of the entire burial
compound, which also includes remains of a funerary precinct
with auxiliary burials, sacrificial pits, and many other
structures. According to the commonly accepted explanation,
the underground army was created as a replica of the
Qin army ("First Emperor's bodyguard sculpted in clay"):
clay soldiers and horses representQin Shi Huang's army and
stand in place of the real soldiers who could not have been
actually buried.2 There has not been much doubt about the
army's function either: it was buried to the east of the
tumulus to guard the tomb and to protect the emperor's
eternal sleep.
The striking realism (a characteristics that will require our
subsequent attention) of the army has lead some archaeologists
involved in the project and some authors to suggest that
the figures were modeled after living soldiers, that they were
actual portraits of individual warriors. The majority of
scholars, however, pointing out that the clay warriors' faces
conform to a certain number of stereotypes, is unwilling to
regard them as representations of individuals or to consider
the Qin terra-cotta figures as portraiture.3
Yet if the figures are not portraits (and it certainly cannot
be a priori assumed that they are), what then are they? If not
portraits, are they generic types? What exactly were they
supposed to represent? And what is the function of their
likenesses?
Clearly, the common (and often only implicitly assumed)
understanding of these figures, which considers them substitutes
for genuine models while not accepting them as
representations of individual persons standing as such models,
is inherently contradictory and raises more questions
than it purports to answer. The main reason seems to lie in
the fact that current opinions on the subject are based on
largely unexamined assumptions and uses of such concepts
as "portraiture," "substitution," "replica," and "realism"-in
short, on some taken-for-granted conceptions of representation.
Moreover, it seems that too much effort has been spent
on trying to explain the "meaning" of the army through
reference to contemporaneous religious practice, often enlisting
spurious evidence of texts, while too little attention has
been given to the forms themselves as the very medium in
which the work of art signifies.
My aim in this paper is to provide a more adequate
account of what the Qin terra-cotta figures represent, and to
determine more precisely the ontological and semiological
status of the figures and, by implication, of the whole
subterranean legion.4 To a large extent this will amount to
reckoning with the complexities of a conceptual and interpretive
framework, penetrating through the accumulated layers
of predetermined conceptions regarding the meaning of
tomb substitutes and seeing beyond the recurrent patterns of
assessing the style of the Qin figures as well as inferring their
status from such an interpretation of style. On the other
hand, it is hoped that the current elusive and unsettled issues
of portraiture, resemblance, construal of identity, and other
related problems of theoretical interest can be enriched
somewhat by attending to the ever more complex artistic
tradition, which has been literally surfacing from Chinese soil
in recent decades.
The Meaning of Mingqi and the Style of Qin Figures
Much as the sculptures of the Qin terra-cotta legion can be
perceived as unique in terms of their dimensions or style,
they are just one, albeit one of the earliest, instance of tomb
I would like to acknowledge the support of the J.
Paul Getty Foundation, whose postdoctoral grant
has enabled me to carry on research in China. I am
grateful to Richard Brilliant, Lothar von Falkenhausen,
David Keightley, Thomas Lawton, and two
anonymous reviewers for their criticial comments
and suggestions on previous drafts of this paper.
Unless otherwise indicated, translations are mine.
1. The best English-language scholarly introduction
is probably still to be found in Hearn. Of the
many catalogues accompanying exhibitions of the
Qin figures in Europe, North America, and Japan,
the most recent (Ledderose and Schlombs) contains
several essays on various aspects of the terracotta
army and a comprehensive bibliography
through 1989. The official report on pit no. 1
(Report) and Yuan both contain a comprehensive
corpus of data on the find. These can be complemented
by Legion, with a popular text and many
good-quality color reproductions. A selection of
articles published in Wenboduring the last decade
is reprinted in Yuan Zhongyi et al., eds., Qin Shi
Huang bing ma yong bowuguan lunwen xuan (An
anthology of papers by the Museum of the Figures
of soldiers and horses of the First Emperor of Qin),
Xian, 1989.
2. According to the most often accepted view, pit
no. 1 contains the Right Army of the Imperial
Guard in the capital Xianyang, pit no. 2 the Left
Army of the Guard, pit no. 3 the army headquarters
and pit no. 4 (unfinished) should have represented
the Central Army. See, e.g., Wang Renbo, "General
Comments on Chinese Funerary Sculpture," in
Quest, 41-42; and Yuan, 36. For another opinion,
see Wang Xueli, "Yifu Qindai de chenbing tu: Yilun
Qin yong kengde xingzhi qi biancheng" (A diagram
of Qin troops: On the character and deployment
of Qin figure pits), Wenbo, no. 5, 1990,
169-85 (pit no. 1 represents the deployment of the
Qin army formation, pit no. 2 a military camp, pit
no. 3 headquarters, pit no. 4 a battle array). See
also Zhang Zhongli, "Qing yong sanhao keng
xingzhi chulun" (My humble opinion on the nature
of pit no. 3 of Qin figures), Wenbo, no. 5, 1990,
107-11.
3. See n. 30.
4. See Freedberg, 77: "Much of the difficulty in
speaking about the ontological status of images in
our culture and, even more, in those of the past or
in other places arises from a failure to assess the
semiotics of the phenomena we have been consider-
ing."
116 ART BULLETIN MARCH 1995 VOLUME LXXVII NUMBER I
figures widely used in a mortuary context in early China. In
addressing the issue of the status and significance of the Qin
emperor's clay warriors, one is constrained by the existence
of the established, commonly held conceptions of what tomb
figures were meant to be and to do.
As in many other ancient cultures, furnishing graves with
an elaborate array of sacrificial goods was an essential
component of early Chinese mortuary practice. The late
Zhou ritual texts LiJi and Zhouli name a specific category of
objects destined to accompany the dead into the netherworld,
called mingqi (spirit vessels). The concept of mingqi,
both in classical Chinese texts and in contemporary scholarly
parlance, is rather elastic: it has been used to designate
artifacts serving as substitutes or surrogates for some "real"
object that is a living, functional, bigger, or more precious
entity. The concept thus covers a variety of objects, including
the imitations of utensils in cheap materials, but in the
ancient texts and since it has been used above all to refer to
the plastic images of humans and animals which began to
appear in the late Zhou tombs.5
Those sections of the Zhou ritual texts are mostly concerned
with the ethical dimension of the use of funerary
images. In one passage, however, Confucius himself is
quoted as objecting to the use of certain types of anthropomorphic
figures in funerals, fearing that they may prompt
the use of real human victims. Thus, in the sage's own words,
the efficacity of verisimilitude is underscored.6 On the basis
largely of a few passages scattered in early texts, Chinese and
Western scholars have explained the origins of mingqi by the
need to replace real human and animal sacrifices interred at
burials of nobility: hence they are commonly understood as
effigies or replicas.7
What status does such a concept imply for the mingqi
object? The figures were commissioned and created as
surrogate beings, substitutes in the sense of compensatingfor
the absence of real people, who, for various moral as well as
economic reasons, could not have been actually buried with
their master.8 This view is further augmented by the assumption
that these substitutes would act in the hereafter precisely
as their models did in this world and render the occupant of
the tomb the same service.9
A recentlyproposed opinion offers an alternativeconception
of the statusand meaning of the clayarmy.Questioning
the traditional "substitution account," it claims that the
figures should not be seen simply as substitutesfor human
sacrificial victims, or, in other words, as mere "equivalents"
of human beings put into the tomb only because for various
reasons it was impossible to bury living people themselves.
Accordingly, "the figures are substitutes for real soldiers as
little as Cezanne's still life is a substitute for a plate of fruit."'0
We can refrain for the moment from an obvious and (for
further discussion quite a pertinent) objection, namely, that
from the point of view of representation's required effects
and functions, Cezanne's still life could indeed be viewed as a
substitute for a plate of fruit. What this opinion implies is that
much as Czanne was probably not mainly interested in
painting a still life that would pass for a real plate of fruit, the
production of the terra-cotta army was not motivated by the
wish to create an illusion of the presence of real people, but
by the desire to acknowledge its status deliberately as an
artificial construct, to proclaim that it was of a different order
from real soldiers.
Such a view is based on the interpretation of certain
passages in late Zhou texts and some recent archaeological
finds, such as the tomb of Zeng Hou Yi, dated to 433 B.c.
These are taken as evidence for a major transition in the late
Zhou religious outlook and a new tendency in funerary
rituals, whereby the guiding idea became the separation of
the spheres of the dead and the living. At this time, the goal
of tomb builders was to represent all aspects of this world in
the netherworld, to produce within a tomb a comprehensive
replica of the human world. Therefore, the artificial substitutes
were regarded as even more appropriate, more fitting
for this task than the interment of real people, as they would
be able to depict all the required activities-embodied in
specific postures and gestures-far better than sacrificial
victims could."1 Thus, even though originally tomb figures
may have come into existence as surrogate people, by the
time of the construction of the First Emperor's tomb, they
were deemed artificial objects, intentionally commissioned
and made to proclaim their distinction from the living models.
On a closer view, one can discern a pattern of circular
inference operating behind both these accounts of the
significance of mingqi: the assumption about the meaning of
the tomb figures supports a certain reading of their formal
properties, while a description of the sculptural style itself is
generated from an a priori notion of status and meaning.
The traditional "substitution-for-absence" account of the
Qin army-and tomb figures generally-is tied to identifying
the style as "realistic." As the figures, according to this
5. For a general overview of history of mingqi, see,
e.g., Cai Yonghua, "Suizang mingqi guankui" (Observation
on funerary mingqi), Kaogu yu Wenwu, no.
2, 1986, 74-78.
6. "Confucius said of those who produced straw
souls [figurines] that they were human, but of those
who used wooden automata that they were inhuman,
because was there not a danger of the latter
leading to the use of human sacrifice?" LiJi, chap.
Tangongxia, repr. in LiJi Zhengyi, Shanghai, 1990,
I, 171b.
7. Zheng Dekun and Shen Weizhun, Zhongguo
mingqi (Chinese spirit objects), Harvard-Yenching
Institute Monograph, no. 1, Beijing 1931. See also
introductory essays in Quest.
8. A typical recent statement can be found in Wang
Renbo (as in n. 2), 39-41: "The burial of these
figures to replace human sacrifices was an important
sign of social transformation that affected
funerary rites.... The complete substitution of
funerary figures for human and animal sacrifices
was an innovative measure that liberated the productive
capacity of the society and raised the level
of the culture." Report, I, 162, puts forward essentially
the same explanation for the Qin emperor's
terra-cotta army. Lu Zhirong, "Qin Shi Huang ling
bing ma yong peizangzhi zhi yuanyuan tansuo"
(The origins of burial pits of soldier and horse
figures at the tomb of the First emperor of Qin),
Wenbo,no. 2, 1992, 50-54, links the Qin emperor's
legion to the armed sacrifical victims found in the
Shang royal tombs.
9. More specific explanations supplement this account.
Thus, it has been suggested that the Qin
terra-cotta soldiers reflected magical Daoist beliefs
that realistic, lifelike images could evoke real people
(H. Brinker and R. Goepper, Kunstschdtzeaus China:
Neuere Archaologische Funde aus dem Volksrepublik
China, Cologne/Hildesheim, 1980, 107).
10. L. von Falkenhausen, "Ahnenkult und Grabkult
im Staat Qin: Der religi6se Hintergrund der Terrakotta-Armee,"
in Ledderose and Schlombs, 46-47.
11. Ibid., 47. This view had been anticipated even
earlier; see D. Croissant, "Funktion und Wanddekor
der Opferschrein von Wu Liang Tz'u," Monumenta
Serica, xxIII, 1964, 109-10, n. 68.
12. Yuan Zhongyi, Qin Shi Huang ling bing mayong
(Terra-cotta warriors and horses at the tomb of the
First Emperor), 3rd ed., Beijing, 1986, 17; Report, I,
158-59 and Yuan, 305.
13. See Hearn, 368; and Brinker and Goepper (as
in n. 9), 107.
(RE)PRESENTING THE FIRST EMPEROR'S ARMY 117
view, were intended to replace real people or animals, their
creation was guided by the purposeful striving for a visual
verisimilitude as complete as possible. In the replicating
task, only the maximum visual accuracy, as literal a notation
as possible, would do. Such a thesis is presented in a most
straightforward fashion in Chinese scholarship, where the
"realistic style" of Qin warriors is interpreted as simply
determined by the images' purported status and identity as
funerary substitutes for real models.'2 But similarly, if less
assertively, molded views can be found in Western treatments,
too.13
David Freedberg has recently provided sophisticated support
for such an idea with his apt reminder of how magical
efficacity depends on resemblance, on the degree of visual
accuracy in the lifelike image.14 However, Freedberg is
explicitly referring to votive limbs and wax figures-that is,
images whose status as objects with magical properties is
securely determined. And that is certainly not our case. But
first of all, it is necessary to ask whether realism is indeed the
most salient and defining stylistic quality of Qin figures, as is
almost invariably suggested. More precisely, we have to ask
whether a purposeful striving for visual accuracy can be
detected as the most prominent stylistic feature of these
sculptures.
It should first be noted that the omnipresent predicates of
"realism" or visual "correctness" in the writings of Chinese
authors seem above all to bespeak their practice of using
such terms as value-laden concepts, whereby in the rigidly
Marxist realm of Chinese academic practice, "realism" per se
stands for a sign of artistic quality and evolutionary progress.'5
However, even for Western authors-who might be expected
to be devoid of such biases-the notion of realism seems to
be central to their perception of Qin figures, to the extent
that the style of the sculptures is most often simply labeled as
"realistic."'6 On the other hand, it also has been pointed out
that these clay warriors are better viewed as examples of
conceptual art and, even in some recent Chinese scholarship,
their style is characterized as incorporating both the "general"
and "particular" aspects of the form.'7
The label of realism may be useful and legitimate for an
overall description of the Qin emperor's army within the
evolutionary sequence of early Chinese art.18However, as a
normative description it is both insufficient and misleading
(as would be the opposite predicate "conceptual").19
The Qin warriors clearly exhibit a certain affinity with what
1 Armored Infantryman, detail (from Legion, fig. 77)
is conventionally, if not always correctly, termed conceptual,
or archaic, sculpture. The bodies are conceived largely in
linear and columnar planes, most of them being axially fixed
(Fig. 1). But even the figures of infantrymen, with their static
postures and frozen movements, do not display the degree of
rigidity and symmetricality characteristic of classical examples
of archaic Greek sculpture. The postures of certain
warriors have a fair degree of organic interplay of planes,
and escape the strict frontality and the awkwardness associated
with most of the standing figures.
14. See Freedberg, 157: "Their [magical images]
operational effectiveness is perceived as deriving
from the closest possible form of realism available
to maker and consumer"; see also ibid., 203, 245.
15. See, e.g. Fu Tianchou, The Underground TerracottaArmy
of Emperor Qin Shi Huang, Beijing, 1985,
11; Yuan, 366; and Wang Xueli, "Xionghunde qipo
xieshide yishu" (Vigorous spiritedness; realistic art),
in Zhongguo kaoguxue yanjiu lunji: Jinian Xia Nai
xiansheng kaogu wushi zhounian (Collected essays in
the study of Chinese archaeology: Commemorating
the 50th anniversary of archaeological work of
Mr. Xia Nai), Xian, 1987, 372-90, esp. 375-77.
16. For a representative statement of Western scholarship,
see, e.g., H. Brinker, "Chinso-Zen Portrait
Painting in China and Japan," in International
Symposium on the Conservation and Restoration of
Cultural Property: Interregional Influences in East
Asian Art History, Tokyo, 1982, 177: "The staggering
realism of the life-size terracotta soldiers . . . marks
a hitherto unknown level of achievement in naturalistic
depiction" (my emphasis). Another typical example
of such a perception can be found in Ledderose
and Schlombs, where the stylistic features of
figures of warriors are discussed in a section titled
"Der Realismus der terrakotta Figuren" (277-78).
17. G. Kuwayama, "The Sculptural Development of
Ceramic Funerary Figures in China," in Quest, 67;
Report, I, 142; and Yuan, 314.
18. On a similarly problematic use of the concept of
realism in the context of Roman sculptural portraiture,
see S. Nodelman, "How to Read a Roman
Portrait," Art in America, Jan.-Feb. 1975, 27-28.
John Baines notes that although the realism cannot
be rigorously or neatly defined, "it does not mean
that it is any less real or important as a phenomenon.
In part it may be a self-sustaining elaboration
on the forms inherited from earlier generations" (J.
Baines, "Theories and Universals of Representation:
Heinrich Schifer and Egyptian Art," Art History,vIII,no.
1, 1985, 16.)
19. Anthony Cutler has recently underscored the
role of the previous body of writing relating to
image in question ("Stalking the Beast: Art History
as Asymptotic Exercise," Wordand Image, vii, no. 3,
1991, 224). Qin figures are a marked case in point.
Ubiquitous interpretations of Qin terra-cottas as
"realistic" and "naturalistic" art seem to have been
generated more from the work of earlier commentators-faced
with the need to come to terms and
conceptualize the unprecedented appearance of
the figures and thus creating a self-perpetuating
tradition of description-than from critical assessment
and observation of their forms.
118 ART BULLETIN MARCH 1995 VOLUME LXXVII NUMBER 1
.............
.
10
12
iiiii~iiiiiiiiii6
2 Belt Hooks on figures from pit no. 1 (from Report,I, fig. 51)
The most striking and significant stylistic feature of the
Qin figures, however, lies in the way these conceptually
conceived forms are organically infused with details of the
utmost visual accuracy. There is a prominent interplay
(neither fully harmonious, nor entirely baffling and disturbing)
of stylized components of bodies with instances of a
literal notation of headgear, hairdos, costumes, armor, and
footwear. There are examples of striking visual realism in
Chinese art prior to the Qin terra-cotta army; but there is no
precedent for this massive deployment of verisimilitude,
evident especially in the rendering of garments and parts of
the armor and accoutrements. Belts and belt hooks, boots
and shoes, collar closures, buns and plaits are all uncompromisingly
transcribed into plastic form (Figs. 2, 3).20 As a close
look at the figure of the kneeling archer from pit no. 2
reveals, this perceptual realism includes such details as the
bending of the right foot on which the weight of the body
rests, as well as the texture of the sole, and the folds on the
inner side of the left sleeve (Figs. 4, 5).
Such consistent striving for visual verisimilitude does not
merely characterize one aspect of the sculptural style of these
soldiers. It can likewise be observed in figures of horses and
models of carriages found in nearby pits. It is further
enhanced by the presence of genuine bronze (and a few iron)
weapons held by the warriors, instead of painted or sculpted
substitutes. Again, the figures were originally elaborately
painted and although the polychromy in most cases has now
flaked off, enough remains to suggest that it was applied to
imitate the colors of the actual armor and garments worn by
'o
I
--------------------iiii!iil
--- ---------
-----------li~ii
3 Buns and Plaits on figures from pit no. 1 (from Report,I, fig.
66)
different ranks of soldiers.21 As the inclusion of real weapons
testifies, this iconic impulse extends to the point at which the
sign is effectively fused with the signified. This feature, as we
shall see, can be further observed in some other aspects of
the whole necropolis.
The effect of literally transcribed elements, however, is
balanced by stiffness, occasional distortion of proportions,
and in some cases by awkward exaggerations of body parts.
For all their realistic impact, most of the figures convey to the
contemporary viewer a sense of marionettelike artificiality,
rather than that of immanent, empirical reality, as mediated,
for example, by some fine examples of Greek sculpture. Such
characteristics and impact of style essentially inhere in the
subject matter itself-in the constraints of depicting fully
armed warriors, standing in military formation in certain
appropriate positions. Firstly, with the exception of heads
and hands, the bodies are entirely clad in tunics or armor,
their contours defined by garments. No underlying anatomical
structures, such as joints, and no muscular tensions are
visible whose depiction would imbue gestures and postures
with a more convincing approximation of reality. In short,
there is no sense of actual physis, no immediacy that can be
better expressed by the naked body. Secondly, as the figure
of a kneeling archer (Fig. 4) clearly demonstrates, the figures
could not have been depicted in just any random postures,
expressive of transient states of mind and thus of more
convincing individuality, but had to be represented with
specific gestures predicated by their function within the
entire formation.
(RE)PRESENTING THE FIRST EMPEROR'S ARMY 119
4 Kneeling Archer from pit no. 2 (from Legion,fig. 109)
The preceding analysis exposes the pitfalls of simply
summarizing the impact of the style of the Qin's emperor
army. The majority of scholars perceive these forms as
"realistic," assuming that they strive to be like real people,
horses, and chariots. Others choose to see the style as
mediating a sense of artificiality, referring to the sculptures
as self-labeled representations and proclaiming that they are
unlike real people and objects. Neither alternative seems
wholly satisfactory. There is no means of determining how
5 Kneeling Archer from pit no. 2, back view (from Ledderose
and Schlombs, fig. 229)
the style was perceived by the audience for which it was
destined-namely, the soul of the dead emperor.22 The few
passages in late Zhou and early Han texts that scholars have
traditionally squeezed dry in order to determine what kind of
reality funerary images represent are an unreliable guide.23
Not only are they notoriously ambiguous and open to
varying interpretation; their use to interpret funerary art is
also based on the uncritical assumption that they are prescriptive,
expressing normative ideas and attitudes, which repre-
20. For illustrations, see Report, II,figs. 92, 93, and
Legion, figs. 100-102 (boots); Report, II, figs. 94-97
(belt hooks); Report, I, 98, fig. 46 (collars); Report, II,
figs. 98-99, and Legion, figs. 104-6 (headgear);
Report, II, figs. 100-104, and Legion, figs. 196-214
(hairdos, buns, and plaits); Report, II, figs. 143,
144, and Legion, figs. 93-98 (hands).
21. See Report, I, 193-94.
22. By contrast, the living audience is crucial in the
study of some other early dynastic funerary monuments
and representations; e.g., the imagery of
Han funerary shrines was intended as much for the
live viewers as for the deceased himself. As some
recent studies have persuasively demonstrated, the
styles of these monuments and images are tied to
the concerns, aspirations, and ideologies of the living
audience (see Wu Hung, The Wu Liang Shrine:
The Ideology of Early Chinese Pictorial Art, Stanford,
1989, and M. Powers, Art and Political Expression in
Early China, New Haven/London, 1991.)
According to some passages in another ritual
text, Yi Li, the articles for the dead were also
displayed prior to entombment: "Mingqi are displayed
to the west of carriages . .. upon arriving at
the funeral grounds, they are arranged to the east
of the road, while the main ones are arrayed on the
northwestern side." In later times, especially during
the Tang dynasty, tomb figures and other
goods were probably destined as much for the eyes
of participants in elaborate public funeral processions
and rites as for the dead themselves. But there
is no evidence that the underground world of the
Lishan necropolis was intended for anyone other
than the First Emperor.
23. These are, in particular, passages from Li Ji
(compiled in the 2nd century B.C. from earlier
materials) on the use of funerary vessels and consciousness
of the dead. The key passage reads:
"Confucius said, 'In dealing with the dead, if we
treat them as if they were entirely dead, that would
show a want of affection and should not be done;
or, if we treat them as if they were entirely alive, that
would show a want of intelligence and should not
be done. On this account the vessels of bamboo
[used in burials of the dead] are not fit for actual
use; those of ceramic cannot be used to wash in;
those of wood are incapable of being carved; the
lutes are strung, but not in tune; the bells and
musical stones are there but they have no stands.
They are called 'spirit [funerary] vessels,' that is,[the
dead] are thus treated as if they were spiritual
intelligences," Liji, chap. "Tangongshang," in Liji
Zhengyi (as in n. 6), I, 143a; J. Legge, trans., The
Chinese Classics, 3rd ed., 5 vols., Hong Kong, 1960,
I, 148. For a discussion of the issue of the consciousness
of the dead from the philosophical perspective,
see A. Graham, Disputers of the Tao: Philosophical
Argument in Ancient China, La Salle, Ill., 1989,
15-18, 258-60.
120 ART BULLETIN MARCH 1995 VOLUME LXXVII NUMBER I
6 A Formation of Infantrymen in pit no. 1 (from Legion,fig.
56)
sentational practice simply followed. But there is little justification
for this approach in general and even less in the case
of the Qin emperor and his mausoleum.24
The Features
The stunning degree to which the features of the figures are
elaborated and the differences perceived-resulting in the
initial observation that "no two faces are alike" (Fig. 6)have
lead some to believe that the figures are portraits,
modeled after real people.25 But close scrutiny reveals that
physiognomic details and facial expressions are to a large
extent stereotyped. While certain figures exhibit a remarkable
degree of individuation, most faces can be best described
as aggregates of prefabricated expressive and representational
units (Figs. 7, 8).26
Sculptors worked with a certain repertoire of standardized
components, obtained through the use of a limited number
of molds. The individual appearances were assembled
through combination of these basic components, followed by
the hand finishing of details.27 This "editing" of faces was
naturally firmly intertwined with sculptural procedures, but
need not be looked at merely as a consequence of the
technical constraints of serial production.28 The most authoritative
Chinese treatments distinguish eight basic facial types,
according to the general contours of faces, each of which can
be realized in many different expressions. The heads are
thus said to embody both generality and individuality.29
Typically, most scholars now claim that despite all the
specific traits and seemingly endless diversity in the rendering
of faces, the figures are better viewed not as individualized
portraits of particular subjects, but as "types": "The goal
was to produce the appearance of individuality, without its
substance, realism without portraiture."30 Such a conclusion
seems to be supported by our viewing experience: for all
their differences, in most cases the figures do not create the
sense of specific beings. However, is such a conclusion any
better grounded than its opposite, the notion that the
warrior figures are portraits? In both cases, the determination
of the status of the image is derived merely from a purely
subjective assessment of visual properties. Yet is there any
less convincingly articulated visual appearance in the figure
of the Qin general (Fig. 9) than, for example, in the portrait
bust of Queen Nefertiti (Fig. 10) or in the Heian-period
sculpture of the high priest Roben (Fig. 11)? Do the Nefertiti
and Roben portraits mediate any more objective a transcrip-
24. Martin Powers in his study of Han funerary
shrines has recently argued that passages from LiJi
and other texts suggest a conventional and selfconsciously
representational funerary art (Powers
[as in n. 22], esp. 58-61). However, this account
seems to underestimate beliefs in the reality of the
afterworld and the potentials of visual realism
within the macrocosmos of the tomb (see also
Jonathan Hay's review of Powers, Art Bulletin, LXXV,
no. 1, 1993, esp. 173-74). There is little reason to
suppose that the commission of the First Emperor's
burial compound was predicated on ritual prescription
and Confucian ethical views, which may be
relevant for understanding the funerary art of the
Eastern Han literati. Such texts were in any case not
fully formulated at the time of the First Emperor's
death.
25. Nie Xinmin, "Yetan Qin bingmayong de zhuti"
(Discussion of a subject of Qin figures of soldiers
and horses), Wenbo, no. 5, 1986, 57; Li Dingxuan,
"Qin bingmayong de xieshi zhuyi yishu fengge"
(Realism of artistic style of Qin figures of soldiers
and horses), Wenbo, no. 1, 1985, 36; Wang Yuqing,
"Qin yong mianxing he biaoqing" (Facial features
and expressions of Qin figures), Wenbo,no. 1, 1984,
70-73; and Arthur Cotterell, The First Emperor's
Warriors.London, 1981, 28-29.
26. For clear details of faces and facial expressions,
see Legion, figs. 143-95.
27. For a fuller account of production methods, see
Report, I, esp. 163-92. E. von Erdberg, "Die Soldaten
Shih Huang Ti's-Portriats?" in D. Seckel et
al., eds., Das Bildnis in der Kunst des Orients, Stuttgart,
1990, 225-29, summarizes the procedures
involved in creating the heads of the figures.
28. I am struck by the persistence of this artistic
procedure in a Chinese context. The late imperial
manuals for artisan makers of ancestral portraits
contain a stock of standard formulas for various
types of noses, mouths, and eyes which could be
combined according to the specific conditions of
the patron's wishes to construct the likeness of a
person. See J. Bredon, Peking, Shanghai, 1931,
130. See also L. Kesner, "Memory, Likeness, and
Identity in Chinese Ancestor Portraits," Bulletin of
the National Gallery, III, 1993-94, in press. Lothar
Ledderosse ("Module and Mass Production," in
Proceedings of International Colloquium on ChineseArt
History, 1991: Painting and Calligraphy, Part 2,
Taipei, 1992, 826-28) mentions the procedures for
creating Qin figures and especially their features as
one example of module system, whereby the image
is constructed by manipulating a stock of building
blocks.
29. See Report, I, 142-50, figs. 74-76, and Yuan,
315-19. The English summary in Report says (498)
that "of the 1,087 clay figures unearthed, no two
faces are alike. Each face is a personal portrait"; the
term "portrait" is not used in the Chinese original
of this passage (142-43), which merely states that
each facial shape and expression is different. Most
recent Chinese sources show a certain inconsistency
in the matter; e.g., in Yuan, 317-18, it is said that
among thousands of figures "all of them have
different individuality ... it is difficult to find two
faces which would be identical," while earlier (316)
the claim is made that of some 700 figures so far
restored, there are approximately 100 kinds of
different facial types and about 300 different expres-
sions.
30. D. Keightley, "The Quest for Eternity in Ancient
China: The Dead, Their Gifts, Their Names,"
in G. Kuwayama, ed., Ancient Mortuary Traditions of
China: Papers on Chinese CeramicFunerary Sculptures,
Los Angeles, 1991, 23 n. 3. See also Hearn, 361.
For a strong assertion that Qin figures cannot be
considered as portraits, see A. Schlombs, "Die
Herstellung der Terrakotta Armee," in Ledderosse
and Schlombs, 97.
(RE)PRESENTING THE FIRST EMPEROR'S ARMY 121
7, 8 Heads of Officers; Charioteers and Armored Infantrymen (from Legion,figs. 154-66)
122 ART BULLETIN MARCH 1995 VOLUME LXXVII NUMBER 1
9 Head of a CavalryGeneral (from Legion,fig. 151)
tion of empirical reality, any more convincing a sense of an
individual than does the figure of the Qin general? Obviously,
all three images contain not a few conventionalized
features, dictated by the specific circumstances of their
commission and purported use, as well as aesthetic and
ideologic constraints.31
Limits and inadequacies inherent in the traditional Western
concept of portraiture, dominated by an emphasis upon
likeness and the reproduction of an individual's unique
physiognomy, have recently received much attention. As
Richard Brilliant has observed:
the quality and perspicacity of [the portraits'] reference
rather than the accuracy of depiction establishes the
normative standard of their being. Then, the degree of
likeness, comprehended historically as some requisite
quotient of resemblance, may vary almost without limit,
effected more by changing views about personal identity
and the function of artistic representation than by the
10 Portrait Bust of Queen Nefertiti (from TheEgyptianMuseum
in Berlin, Berlin, 1990, 97)
peculiar physiognomy or appearance of the Subject ...
likeness is never more than a represented approximation
that operates conceptually to fix transiency in an elusive
image.32
Such a perspective is particularly apt in the context of
Chinese figural representation.33 In tomb sculpture especially,
there are many instances of figures that have a much
greater degree of visual specificity than many images recognized
as "portraits." Yet unlike such paintings, where the
identity of the models can be inferred from the context, tomb
figures do not allow for any identification of the person
depicted.34
A realistic image, of course, often a priori implies the
existence of an actual being as a model. As interpreter, one
may be fully aware in theory that the degree to which the
image is elaborated is not indicative of its status, that "the
question of reference, in fact, is totally independent of the
degree of differentiation.35" Yet one's interpretative efforts
31. See Nodelman (as in n. 18), passim, and Davis,
41-46, on an interchange between conventional
factors and individual features in constructing "likeness"
in late Roman and Egyptian portraiture
respectively.
32. R. Brilliant, "Editor's Statement: Portraits: The
Limitations of Likeness," ArtJournal, XLVI,no. 3,
1987, 171-72. See also Brilliant, chap. 1, esp.
25-26. The classic formulation of the problem can
be found in E. H. Gombrich, Art and Illusion,
Princeton, N.J., 1961, 90.
33. See, e.g., Audrey Spiro's study of 4th- and
5th-century portrait tomb reliefs, Contemplating the
Ancients: Aesthetic and Social Issues in Early Chinese
Portraiture, Berkeley, 1990.
34. A typical example of a "contextual portrait" is a
representation of a woman on a central section of
the painted banner from the tomb of Lady Dai (ca.
168 B.c.) at Mawangdui, which is identified as a
depicition of the tomb master and thus can be
regarded as the first portrait of a specific historically
known personality in Chinese art. See L.
Kesner, "Portrait Aspects and Social Functions of
Chinese Ceramic Tomb Sculpture," Orientations,
xxil, no. 8, 1991, 33-42, for a comparison of this
painting with contemporary tomb figures.
35. E. H. Gombrich, "Meditations on a Hobby
(RE)PRESENTING THE FIRST EMPEROR'S ARMY 123
11 High PriestRoben(from TheGreatEasternTemple:Treasures
ofJapaneseBuddhistArtfrom Tddai-Ji,Chicago, 1986, 107)
are not exempt from the psychological constraints of the
viewing experience. Thus, the small, crude effigy with an
undifferentiated face, a figure from an Eastern Zhou tomb,
which directly precedes the Qin emperor's terra-cotta figures
in a similar funerary context could equally have been
intended to represent some particular person (Fig. 12).36 It,
however, does not particularly prompt questions about the
identity of its model, the figural aspect taking precedence
over the portrait, the general over the unique. But in the
presence of an image with prominently articulated features,
which declares itself authoritatively, it is difficult to escape
the notion that it must have been created to imitate a definite
human model. The Qin soldiers declare themselves as
representing some specific preexisting prototype in a way
that a less articulated image does not.37 But the central
question remains: who (or what) exactly are the figures of the
underground legion substitutes for? The issue, in other
words, is to establish the nature of the reference, to discover
whether there were real individuals as models for the clay
warriors. Furthermore, if we are to decide whether these can
qualify as portraits or not, the task is to determine what
established the individual's ontological status in the specific
historical context of the production of the Lishan necropolis.
Likeness and Identity
If, as we have seen, no particulardegree of resemblance is
required for the visual construction of the subject'sidentity,
any figure of the terra-cotta army could be considered an
individual portrait. In fact, some pieces with distinctlypronounced
featureswould seem to lend themselves to such an
interpretation. It is only when we shift our focus from the
individualfigures to the underground armyas a whole, and
when the methods and procedures of its production are
taken into account, thatwe are led to the conclusion that the
army could not have been conceived and created as an
aggregate of individualsoldiers'likenesses.
It is true that not much is known of the circumstancesof
the making of the terra-cotta army. Yet one important
inference,with a directbearing on the issue of the statusand
identity of figures, can be made. According to a passage in
Sima Qian's Shi Ji (Records of the historian), compiled
around 100 B.C.,construction of the necropolis startedwith
the First Emperor's accession to the throne.38Calculations
based on an assessment of the production methods and
technique suggest that the creation of the army must have
proceeded over a period of severalyears (tenyearswould be
a minimal estimate), and even at the time of the First
Emperor's death it probably had not been completed.39
Whateverthe precise time span of the creation of the army,
and even if some of the figureswere informedbythe likeness
of an actual person, it would not have been possible to
transcribe the particular appearance of more than seven
thousand men, one after another, into plastic form. Indi;a~Bsa~~~
'-'j~-~~~aai~~s~~~~~ ::'::::::::::::::'::::::::-i
::i:::j::::::::::::::::::::i:::::::::::::i::::::j:::::j::::::j:~:~:Bu::::~:_::~ ~~:&i:::::--::~i;-:-~:_~ ~ :-~~~vs-:j""j::'":::::::::::::::: :a_:-:::::::-:a:_:-::::~~-
::::-:::'::':-::':?:::::::::::::::::::~_:~-:i~~~~--~~~'-~'-sai'--a:.:--:::::-"~~~::'::
:.: ::j:::,:i::i::::::jji:j:::jj:::::::j::j::::::-:-::j:_:,:,:::i:~~~::--:ix:_:::,~::::j::i::::::::::::::-:-::i::i:::j:::,::~~:~~~-:'~~~~~~5:_:-~~-r:~_::::::::j--~~~~---:~:~_:$~il: :::'::::::::':i:~a-~ai
::::::......_::::ia'_ R ,, :::::::::::::::::j::::j::::::j::::i::i::~D~::::~:_~~i ~~--~~:,j:-i~~:~a,:-~~:::::-:::j:i::i:i:~r:::,:,:i::::::::~~_--xa~,:,:,:
:::::j::::.::::::::::::i:::::::i:::::::i::i:::::::j:::::j::i::::~*
:::::::j::::j:::::j::i:j:::::::::i:::j:::::
~~:aaa~~ ~B~::::::::-:::j::::::~-:::~,--:-::--_-
a
a :::
::j::::i:::::::::::j::~
a
.:..::,
~-:~~~:--:--:-:--:--:n:~:::::::::::
i::::_:~:_-i~::_:_:::i::j::_:::::_:i,:j:~~-::::::-:_:j::_~::::~-':::-i::::-:_:-:l ~~::: :::-j:::'_:-::~6:-:-i~~:~~:_:_~~::::::::~~~:~~::-'a~:~a~:::~:~~~::i:a:~ ~:-~ ~~~~,,::_::Q~i:~-i:::~ ::::::~::--:--:i:--:--::-:~:-::-:_
.. ?:jj::j:i::::j:::jji:j::4:_:j:::::::~:_~i_::::i:-:~::::~~ :-~-~~~X~~ ~ ~-~~::-::j::j::::::-:-:j::i::::::-::-:j:
.::i-:-::i::::i:--:::j::i:j::a:_::j:~:::~~:_~~~;~::_:::_i:j:-~~ ~~- :-::::::::::::::::-::j:--:i::i:i~~~:::_::::~:::-:?: -:~:::::::::j::i:-:i::i:_:j:::::::::i:j::::::
"~ s~~~::-~j::::-:-:-:,:-:i:__-:-:::::j:~n-~-::--a:_:~-~~1-::~-j:~:j:':j--E~----d~:--:::::::::~::_:~--::--::-:-:-::~_:_:i::i::_:-::j:~
j:i::i:~-~~:~e--a~:_~~-~i~~-ij::::~:i:::-:?:-::i--:~:--:::-:::-:-::::::':::':::::::::::::':::::'::::::::-:::::::':-::i:::
-::i:::
a
~vM ~::::::::::::jj::::::j::::::~ :i
,:::,:~
aj,$-asa,
~ "
~i::i:::::::::-::::::::::::_
:~_:BB-~I-$&~J ~~Js-~~~W
j-a:~a-:-a:::-::::::_:I:::::j::i: ::::::::::::j~~::::::i::::-:-:::::::::i::
~:::::::::: :r-:-i-:j :~~-_:i:~::::::::::::::jjj
-:~~ ::::::-:~ ~::::~
:::::::::::::~:::::':::.:.... -:::-::::::--:'~~:-:::-:~~~~ ~~lk3~~,-- j,::::j,:i::::::::j:::j:--::~::::j:::::::i:_::::::
-~U~.~;:':~;"~ ~-~B~~~::::::::~:--i::~a:,_-i:_::,::_:_:_-~~~~8::-_:::::~~i~-:-::~::~:::i ~~~~P:-i::::i:::::::j:::;~~_:-i~i~~~:~~~~~~: ~y~g:--::;-:-:----:--:-:~~sz~~~i:j::j:::_:i::j-:i::i:_:-:::::::i~~-~:
:,:-,:,-,::::,,:-:-:::'::::::-::-:~x:;;-i::: i::-:-::i::::::-:;-::,~~~~?:-:;~~~k~BI~~I_~~~B::::::::::::::::::::::::::::
a~U~~s~~~~~~ ~:~a~~-:a:,i:~:i:_:,,:,::i:-i-:i:-:::i::i::::::j:~ :i::i:::ji::j:4i:::::i~~~_~:~B__j:~_:~~::j:_:~~~-~~~~~:~~:::~::--:~:-:- -::::--a:$~--;~~::--:::::::::::j: :~:_:::-i:_::_i:_
::::::::j:::i::j~~~~~~~~~~~:~~~_: a~~~;_:i:_:i::j~l~-,:_~:~:::i:-i:a~~~~~~:j::i:i::~:i:_:,~.: ~ :: ::i::::~-:~ ~~ki~~,,::~::~:--~~:-ai:':j::-:::i:_:j:?::::::;:::::'::'~'::~::--~~~~~a::i:~,i:-ii:,i:,-::::, :::':::::':::-:c~:~::~-::-~:::~-:
12 Pottery Figurines, 5th-4th century B.C.(from Yuan, fig.
173)
Horse or the Roots of Artistic Form," in Meditations
on a HobbyHorse and OtherEssayson the TheoryofArt,
London, 1963, 3.
36. Anthropomorphic figurines were unearthed
from at least six tombs of Qin state, predating Qin
Shi Huangdi's mausoleum. See Hu Lingui, "Zaoqi
Qinyongjianshu" (A brief review of the early phase
ofQin figures), Wenbo,no. 1, 1987, 23-25.
37. The psychological justifications for this are
excellently treated by Freedberg; see esp. chap. 9,
201.
38. The passage is quoted below, at n. 53.
39. The fourth, much shallower pit found in the
immediate vicinity of three pits is generally interpreted
as the unfinished site for the fourth part of
the underground army, thereby implying that the
whole project had not been completed at the time
of Qin Shi Huang's death. For an alternative
opinion, which views pit no. 4 as a source of soil
used for the construction of the army, see Dang
Shixue, "Si hao keng shi weijian chengzhi yong
keng shuo zhiyi" (Calling into the question the
theory that pit no. 4 is an unfinished pit of figures),
Wenbo,no. 5, 1989, 65-71.
124 ART BULLETIN MARCH 1995 VOLUME LXXVII NUMBER I
r4-
12104
13 14
16 17 18
13 Incised and Stamped Characters found on figures from pit
no. 1 (from Report,I, fig. 115)
vidual likenesses of all those who composed the army at a
given moment prior to starting of the project could not,
therefore, have been intended. Thus, the main reference of
each figure is not to a unique individual, as revealed in the
contingencies of his physiognomy.40
Identity, to be sure, can be conveyed by means other than
through the depiction of distinct physiognomic traits.41 One
common means by which an image can be established as the
representation of a specific individual is through the identifying
inscription.42 Such a practice was widely utilized in early
Chinese portraiture, where persons were often differentiated
not by a conveyance of their respective likenesses but by an
inscription that named them.43 Of more than one thousand
fully restored Qin figures, more than five hundred indeed
bear incised or stamped characters (Fig. 13). These inscriptions,
however, have plausibly been read as "signatures" of
master potters, or, in some cases, as numerical symbols
serving to assign figures into certain groups, probably in the
process of their assembling.44 Not even in the figures of
high-ranking officers or generals do inscribed characters
point to individual identities. This absence of any demonstrable
attempt to identify a particular person within the
entire army is striking and suggestive. Accepting that the
army represents aspects of reality relevant to the beholder
for whom it was destined, one can conclude that in these
terra-cotta figures, the depiction of persons in the sense of
actual human beings was not intended. To the emperor, for
whom the clay army was created, the vital component of his
generals' (let alone his soldiers') capacity was obviously not
their individual personalities.
Some authors, who have previously argued against the
notion of the Qin terra-cottas as portraits, maintain that in
fact it was precisely the sculptors' goal to evoke and express
the ideal of a uniformed, disciplined army of the Qin state,
where any presentation of individuality would contradict and
undermine such a purpose.45 It remains to be fully examined
how the tenets of the Legalist doctrine on which the Qin state
was organized, and which suppressed individualism in favor
of strict social regimentation, standardization, and subordination
to prescribed patterns of behavior, might affect the
construction of the underground army. On the other hand,
in any army the fulfillment of the soldier's role significantly
deemphasizes the qualities and characteristics of its individual
members. This fact alone, therefore, would hardly
suffice to explain the specific features of the Qin emperor's
army.
The conclusion that these figures do not portray unique
individuals does not automatically imply a negative answer as
to whether they are portraits. In itself it does not answer the
question of what they were meant to represent, or what was
their semiological status. And other questions remain to be
addressed. Why, for example, did the creators of the army
strive for such a variety in the rendering of features if the
establishment of personal identity through individual physiognomies
had not been intended? To address these issues,
we have to shift the focus of attention to a fact briefly
mentioned earlier-that in the Lishan necropolis we are
dealing with several, ontologically separate objects. So far,
we have been mostly concerned with the figures of soldiers as
40. On the link between visual appearance and
personal identity, see, e.g., S. Shoemaker and R.
Swinburne, Personal Identity, London, 1989, esp.
49-51.
41. Classical evidence has been provided by Leo
Frobenius, when he published his famous double
portrait of Maori chief Tupa Kupa, one by a
Western painter, the other by the chief himself. In
the first image, a portrait was produced according
to the conventions of Western academic portrait
painting, showing the man's face and shoulders,
thus attempting to fix Tupa Kupa's likeness. In
Tupa Kupa's self-portrait, however, the design of
tattooed skin only is presented, the sign of a man's
identity, his "likeness" in the Maori context. See L.
Frobenius, The Childhood of Man, Philadelphia/
London, 1909, 35. Several recent studies have
focused on the various means through which visual
identity in a portrait can be established, complementing
or entirely substituting for the recording
of distinct physiognomic traits. See, e.g., T. Breen,
"The Meaning of "Likeness": American Portrait
Painting in an Eighteenth-Century Consumer Society,"
Wordand Image, vI, no. 4, 1990, 325-50.
42. The locus classicus for the individuation of
image by naming, in a context not entirely dissimilar
with that of early China, is to be found in
Egyptian art. See, e.g., Andrey O. Bolshakov, "The
Ideology of the Old Kingdom Portrait," Gotinger
Miszellanea, nos. 117-118, 1990, esp. 102ff.
43. The early development of portrait painting in
China is to a large extent a case of conceptual, stock
images being transformed into representations of
definite persons by inscription. The most notable
early examples may be the famous figures on a
lacquer basket from Lolang, dated to the 1st-2nd
century A.D., or the portrait of one Puqianjiu and
his wife on the wall painting from their tomb
("Luoyang Xi Han Bu Qianjiu mu bihuamu fachu
jianbao" [A brief report on the excavation of the
Western Han tomb with painted murals of Bu
Qianjiu in Luoyang], Wenwu, no. 6, 1977, figs.
33-34). Spiro ([as in n. 33], esp. 21-36, 103)
discusses some further examples, especially an
important group of 3rd-century tomb reliefs depicting
the so-called Seven Worthies of the Bamboo
Grove. Among tomb figures, inscribed examples
are rare. When they occur, however (e.g., a group
of figures from a Sui-Tang dynasty tomb at Zhengzhou,
each bearing a one- or two-character inscription
on the back), this most probably denotes the
individuality of the person depicted (see Kaogu
tongxin, no. 7, 1958, 56).
44. Report, I, 194-207, figs. 112-19.
(RE)PRESENTING THE FIRST EMPEROR'S ARMY 125
individual works of art, or as self-sustaining entities. Each of
them, however, is only one constitutive part of the whole
three-pit underground army, which, in turn, forms just one
component of the entire burial complex. Thus, much as the
significance of a larger unit is determined by its various
components, it is only through consideration of the entire
micrococm of the Lishan necropolis that the significance of
its constitutive parts can be grasped.
The Fictional World of the First Emperor's Necropolis
It has been shown above that each sculpture of the Qin army
can be regarded as an aggregate of stylistic and expressive
components with varying degrees of differentiation.46 While
the bodily aspect of the sculptures retains a considerable
degree of figural distance-something that could be conventionally
labeled as stylization-such elements as headgear,
garments, armor, and other attributes attain a close iconic
resemblance to their models. And some elements, as we have
seen-such as real weapons carried by the soldiers-display
no figurative distance at all; instead of being represented,
they are, properly speaking, presented.
The analogous situation can be discerned in the entire
burial compound. Whether or not one accepts the common
interpretation that the necropolis is to be understood as a
replica of the First Emperor's palace and a symbolic diagram
of the cosmos,47 it undoubtedly was designed as and purported
to be a complex whole, made up of many compo-
nents.
Let us briefly and selectively consider some of its major
constitutive parts:48
(1) In the underground pit to the west of the tumulus,
inside the inner precinct, two bronze carriages have been
discovered, each pulled by a team of four horses, complete
with the bronze figure of a driver. The chariots are rendered
with minute attention to the structural and painted details, as
are the teams of horses. They are, however, not real carriages,
but replicas half the size of actual ones. In the same
pits as the chariots, large quantities of organic remains of hay
have been found, suggesting that these underground structures
represent depots.49
(2) To the east of the outer wall of the compound, some
one hundred pits were unearthed, containing horse skeletons
and kneeling terra-cotta figures of grooms. The inscriptions
on the accompanying ceramic vessels identify these
structures as palace stables.50 Another large group of horse
skeletons was buried on the western side of the tumulus
between the inner and outer walls.
(3) In the same area, to the west of the tumulus, another
cluster of some thirty small pits contains clay models of deer,
birds, and other animals and/or a kneeling figure. They are
usually interpreted as symbolizing the emperor's parks and
forests.51
(4) On the eastern side of the necropolis, nineteen tombs
were found, containing human remains, which in all probability
were those of high officials and royal retainers destined to
follow the First Emperor in death and executed at his
funeral.52
Finally, there is the underground tomb itself, which,
according to a famous and much-quoted passage from the
Han-dynasty historian Sima Qian's Shi ji, represents the
entire world:
As soon as the First Emperor ascended the throne, the
work [on his mausoleum] at Mount Li began.. .. The
artisans dug through three subterranean streams, which
they sealed off with bronze in order to make the burial
chamber. Inside they carried [models of?] palaces, pavilions,
and the hundred officials, as well as precious utensils
and valuable rarities. ... With mercury myriad waterways
of the empire, the Yangzi and Yellow Rivers, and the great
ocean itself were created and made to flow and circulate
mechanically. The heavenly constellations were depicted
above, while the geography of the earth was laid out
below.53
The presence of high doses of mercury in the nearby soil,
verified through recent scientific examination, lends some
credibility to this historical narrative, suggesting that the
entire microcosm was indeed created underground.54 According
to the same source, the First Emperor's concubines and
the workers involved in the construction of the Lishan
necropolis were likewise buried inside the tumulus.
There is much to be gathered from considering these
features and the description of the mausoleum. Obviously,
for whoever designed the entire burial compound it made
sense to use different representational idioms, to employ
elements with varying degrees of figuration. The components
taken from real life, such as human and animal
sacrifices, actual weapons, or hay-that is, those aspects
employing no figural distance from their referents (fusing
45. See, e.g., von Erdberg (as in n. 27), 228.
46. For a very useful concept of figural difference,
see R. Shiff, "Phototropism (Figuring the Proper),"
in Retaining the Original: Multiple Originals, Copies,
and Reproductions, Studies in the History of Art xx,
Washington, D.C., 1989, 162: "no description or
picture can copy its model, its original, without
differing from that original; and the difference or
distancing is created as if through the deployment
of a mode of figuration."
47. Hearn, 169-71.
48. For a reconstruction of the Lishan site based on
the situation in the early 1980s, see R. Thorp, "An
Archaeological Reconstruction of the Lishan
Necropolis," in G. Kuwayama, ed., The GreatBronze
Age of China: A Symposium, Los Angeles, 1983,
72-83. See also Report, I, 1-63; Li Dingxuan, "Qin
Shihuang ling kaogu tanpin" (Discussion on the
archaeology of the tomb of the First Emperor of
Qin), Wenbo,no. 5, 1990, 261-65.
49. These chariots were interpreted as auxiliary
vehicles for the ghost of the emperor to be used on
his inspection tours. See Yuan, 28-29; and "Qin
Shi Huang ling yihao tongchema qinglijianbao" (A
brief report on the clearing of No. 1 bronze chariot
with horses at the tomb of the First Emperor of
Qin), Wenwu, no. 1, 1991, 1-13.
50. "Qin Shihuang ling dongce majiukeng zuantan
qingli jianbao" (Brief report on probing and clearing
horse pits on the eastern side of the tomb of the
First Emperor of Qin), Kaogu yu Wenwu, no. 4,
1980, 31-41; Zhao Kangmin, "Qin Shihuangling
dongce faxian wu zuo majiukeng" (Five horse pits
found on the eastern side of the tomb of the First
Emperor of Qin), Kaogu yu Wenwu, no. 5, 1983,
23-26; and Yuan, 29-34.
51. "Qin Shihuangling yuan peizang zuantan qingli
jianbao" (Brief report on probing and clearing
burial pits within the funerary park of the tomb of
the First Emperor of Qin), Kaogu yu Wenwu, no. 1,
1982, 25-29; and Yuan, 34-36.
52. "Lintong Shangjiaocun Qimu qingli jianbao"
(Brief report on clearing Qin tombs at Shangjiao
village, Lintong), Kaogu yu Wenwu, no. 2, 1980,
42-50.
53. Sima Qian, Shiji (Records of the historian [ca.
100 B.c.]), Beijing, 1959, vI, 265.
54. See Chang Yong and Li Tong, "Qin Shihuang
ling zhong maizang gongde chubu yanjiu" (Preliminary
research on the mercury buried in the tumulus
of the First Emperor of Qin), Kaogu, no. 7, 1983,
659-63, 671.
126 ART BULLETIN MARCH 1995 VOLUME LXXVII NUMBER 1
sign with signified)-were presented, while other elements
were representedin varying degrees of specificity. This appropriation
of the whole range of representational modes is
marked by a strong propensity for iconic resemblance. It
should be noted, at least briefly, that such a feature is by no
means exclusive to the Lishan necropolis. Precedents can be
found at some Eastern Zhou tombs, for instance, where the
bodies of human sacrificial victims were used alongside
figural images.55 Only in the burial of the First Emperor,
however, has this approach been fully developed, resulting
in an all-encompassing strategy that deployed varying modes
of representation to stage a reality appropriate for the
emperor's eternal sleep. Such constituting of the fictive
reality through various figurative modes, not mere "realism,"
seems to be the key formal feature for grasping the
significance of the Lishan necropolis.56
Discerning this feature and its significance should allow us
to overcome the confines of the previous accounts of the
meaning of the Qin terra-cotta army. Underlying these
accounts was basically the assumption that representation
must perforce be preceded by a coherent and consistent set
of beliefs and attitudes, that it is possible to be sure what kind
of reality the figures do represent. Thus, in what has been
dubbed here as a substitution-for-absence account, the goal
of the army's sculptors was to re-create reality by replicating
its aspects with as much verisimilitude as possible, given the
limits of medium, technique, and representational capabilites.
The experience of the authentic world, which a tomb
master could "believe in," can only be sustained if style is
sufficiently transparent. The alternate view, posing figures
not as substitute people but as self-labeling artificial creations,
has the goal of creating a reality of a different order, a
self-conscious representation. The style then becomes a
purposefully selected code, a mode of representation appropriate
only for depicting the otherworldly reality.
It surely need not be supposed that beliefs and religious
sentiments played no significant role in the creation of this
artificial eternal world. However, need we assume that the
determination of the status of the representation precedes its
coming into existence, rather than being enacted through
the process of depiction? Beliefs and motives, at any rate, are
always compromised by the actual conditions and circumstances
of image making.
What has been preserved of the First Emperor's necropolis
enables us to discern the artistic and ideological goal of its
creators: to produce a self-sustaining version of the world,
plausible enough to support the notion of authenticity. The
practical circumstances and constraints of the construction of
such a world must have decisively determined its format and
style. Thus, if it was possible some two centuries earlier for a
14 Charioteer, Standing and Kneeling Archers, and
Infantryman, seen from the front, side, and back (from Yuan,
figs. 49, 58, 79, 82)
ruler of the local state, Marquis Yi of Zeng, to have twentyone
young women (presumably his concubines and musicians)
follow him in death,'57if it was perhaps possible for the
First Emperor himself to have his concubines and artisans (or
even court retainers) buried with him, the same would not
have been possible in respect of an army of some seven
thousand fully armed men, along with horses, unless, that is,
they were replicated as clay models. Obviously, it was quite
possible (and desirable) to represent some horses through
the burial of actual animals, as it was preferable to present
real weapons and real hay, instead of their surrogates. But
ocean, rivers, and skies, as well as the army itself, had to be
depicted to become part of this realm. For such a fictional
world to be successfully and plausibly deployed, it had to
consist of all the elements deemed appropriate and neces-
55. E.g., Eastern Zhou tomb no. 7 at Changzi,
Shanxi province, which contained skeletons of at
least three sacrificial victims who accompanied the
master of the tomb in death, as well as four wooden
figurines. See "Shanxi Changzi xian Dong Zhou
mu" (Eastern Zhou tombs at Changzi county, Shanxi
province) Kaogu Xuebao, no. 4, 1984, 503-30, esp.
figs. 2, 10.
56. In a somewhat different context and sense,
DavidKeightleyhas identifiedwhathe describesas
"a totalistic desire" that seeks "to appropriate all
availableforms of expression, to speak with a full
range of dialects and accents"as a significanttrait
of Neolithic and Bronze Age Chinese art (Keightley,
1990a, esp. 19-23, 39.)
57. Hubei Provincial Museum, Zeng Hou Yi mu
(The tombsof MarquisYiof Zeng), 2 vols., Beijing,
1989.
58. See Freedberg's notion of "constitutive verisimilitude"
(i.e., a resemblance that reproduces)
striving toward reproduction of total reality, which
he contrasts with accuracy as a reflective category
(i.e., the accuracy that mirrors); Freedberg, 206.
59. Some authors associate the figures with specific
Qin military ranks, as their nomenclature was
recorded in extant texts (see, e.g., Chen Mengdong
(RE)PRESENTING THE FIRST EMPEROR'S ARMY 127
sary to (and presumably by) the emperor, for whom this
staged reality was created. But much as iconic resemblance
was desired and pursued, the efficacity of this fictional reality
was not predicated on the exact visual accuracy of its
constitutive parts, which in any case could not have been
achieved.58
Reference in the Clay Warriors
One aspect of ultimate importance in the real world was the
emperor's army, and for the surrogate army to become a
meaningful part of the fictional world it had to replicate the
roles and efficacity of its model. Our earlier finding that the
terra-cotta figures do not represent actual soldiers points to
an essential characteristic of the real army, and thus by
extension of its surrogate: in the emperor's eyes, his soldiers
were not important for who they were, but for what they could
do or what they could be usedfor. Their individual personalities
were fully encased in their roles. What mattered within
the confines of the underground realm were those aspects
that made each soldier a constitutive component of the
whole-his function, as it was embodied in position, gesture,
and the requisite attributes. And this is exactly what is
depicted in each figure.
The conceptual component of style, which imbues the
terra-cotta soldiers with a rigid, almost marionettelike appearance,
freezing them in certain (sometimes exaggerated)
postures and gestures, dominates their physical aspect. It
captures those movements and stances that spatially embody
and define-and thus differentiate-the most typical, meaningful
configuration in which the specific function of a given
individual within the whole unit was enacted (Figs. 2, 5, 14).
Any more convincing or psychologically more persuasive
kind of realistic depiction would be out of place here,
precisely as the aim was not to depict any random, subjective
posture or transient state of mind, but a relatively stable,
distinctive situation. In contrast, the descriptive, iconic component
of style literally transcribes the appearance of each
soldier's attributes-his headgear, outfit, boots, armor, weapons.
Equally it serves to differentiate his specific function
within the larger body of men. The figure of a kneeling
archer or a charioteer perhaps best illustrates how both
components of style, organically intertwined, refer to twin
aspects of the figure's model, through which its status was
established as an operating, functional component of the
army: the requisite posture and gesture, and the iconicity of
the attributes (Figs. 5, 15). This reference constitutes the
identity of each clay soldier: the specific rank and role/
function is the "who" represented-precisely the signified
we have been striving to disclose.59
It has already been mentioned that the essentially imper-
15 Charioteer from pit no. 2, detail (from Legion,fig. 88)
sonal, functional character of the clay army has a certain
correspondence with the Legalist tenets of the Qin state,
which stressed standardization, efficacity of the individual,
and subordination to prescribed social roles.60 However, to
understand how the formats and styles of the First Emperor's
eternal world developed, it is also necessary to look beyond
the specific social situation and institutions of the Qin state
and ideology to persistent, much older patterns of thought
and ritual action, which conspired in the creation of the
army. The use of tomb figures evolved within the context of
ancestor worship, within the broad framework of burial
practices and concern with the afterlife. Scholars studying
late Neolithic and Bronze Age mortuary customs have
pointed to depersonalization as a significant feature in early
Chinese religion and metaphysics. David Keightley, in particular,
concentrating on the detailed study of Shang oracle
bone inscriptions, has persuasively traced the loss of individual
personality involved in becoming an ancestor. In
Keightley's words: "Although [these] dead have power, they
have no personality."'61 Further implications of such a feature
have not been lost either: this bureaucratization and
ritualization of ancestor worship have been proposed as one
and Lu Guilan, "Qin ling bingyong jueji kao"
(Investigation of the ranks of soldier figures of Qin
[Shihuangdi's] tomb), Wenbo, no. 1, 1985, 14-20.
Further research will have to establish whether the
sculptures can indeed be so classified.
60. Maxwell Hearn has noted that the "profusely
detailed appearance of the terracotta army reveals
the same ardor for minutiae that pervades Qin's
precisely defined legal statutes" (Hearn, 368). There
is certainly a need for further attempts to determine
howthe contemporaryintellectualand cognitiveendeavors-such
as standardizationin writing,
measures,andweights--carried out duringQinShi
Huang's rule might have influenced the sculptors
of the terra-cottaarmy.
61. Keightley, 1990b, esp. 41-46, 68-70. On the
basis of arguments assembled in several of his
publications, Keightley supports his view of the
bureaucratizationanddepersonalizationof the dead
and death, especiallyin the practicesof conferring
temple names (describedas typological andjural)
on dead kings, of assigning functions to the dead,
and of using the dead body of the king itself to
legitimize the living king's political power and
decisions.See, too, D. N. Keightley,"TheReligious
Commitment:Shang Theology and the Genesis of
Chinese PoliticalCulture,"HistoryofReligions,xvII,
no. 1, 1978, esp. 218-20.
128 ART BULLETIN MARCH 1995 VOLUME LXXVII NUMBER 1
16 Formation of Infantrymen in pit no. 1, detail (from Legion,
fig. 133)
basis for the generalizing nature of early Chinese art and
aesthetics.62
The depersonalization and bureaucratization associated
with death and the dead obviously characterized not only the
object of worship itself-the dead ancestors-but also the
sacrificial victims destined to accompany the dead. Their
primary function was both to continue rendering appropriate
service to their master and to express, perpetuate, and
validate his/her ritual and social status. The evidence from
archaeological remains suggests that in this function the
sacrificial victims' own roles and status must have taken
precedence over their aspects as individual human beings.
To cite again one instance of negative, albeit significant
evidence: among a number of sacrifical victims unearthed
from Neolithic and Bronze Age tombs-sometimes with
well-preserved coffins and burial accoutrements-not once
was there any indication of the personal identity of the victim
being recorded.63 Tomb figures, which assumed the place of
actual sacrificial victims (substituting in a broader sense of
the term), were thus required to perpetuate the same task, to
represent a social/ritual identity, rather than a metaphysical,
personal one. The outward signs of function that integrated
an individual as a signifying, relevant part of the tomb
master's social (and hence ritual) realm had to be depicted.
Qin terra-cottas are one of the first and arguably the most
comprehensive embodiment of this logic.
We have established what the figures were meant to
represent; that is, what the main reference in each figure is.
And this finally positions us to address the question of what
this implies for the status of the figures. In other words, it
leads back to our initial question: can these soldiers be
regarded as "portraits"? The logical solution for those
authors who opposed the notion of Qin terra-cotta soldiers
as portraits was to assign them the status of types: "they seem
to be on the whole idealizedtypesrather than realisticportraits"
(emphasis added).64 Despite the recent awareness of the
blurred and fluid boundaries between the representation of
individual and type, this polarity still informs most discussions
of portraiture and has been crucial in discussions of
early Chinese figural art.65 Most early Chinese portraits are
characterized as typological in the sense that a type is used
for the depiction of a certain individual. The dichotomy of
individual versus type has recently been emphasized in an
attempt to contrast patterns of representation and thought
in early China and early Greece.66
What has been concluded here about the nature of style
and reference in the Qin figures would seemingly require us
to accept that they indeed represent types. First, it has been
ascertained that they are not representations of actual men.
Furthermore, in his posture, gestures, and attributes, each
figure embodies the most stable and intrinsic-hence typical-parameters
of a particular rank or specialization within
the army, be it general, officer, archer, carriage driver,
bowman, foot soldier, and so on. Each figure can also belong
to one of several facial types. Finally, some have also tried to
read physiognomic features as signifying the different regional
origins of those who constituted the Qin army.67 In
light of these findings it is possible to conclude that the
figures embody what might be considered typical qualities.
Yet I believe that they still do not qualify as representations
of generalized types and that it would be incorrect to assign
them this label. I would like to explore the possibility that
sculptures of the Qin emperor's army in fact challenge and
62. Keightley, 1990b, 68-70.
63. Also pertinent to our argument here is the fact
that in the royal Shang dynasty burials, various
classes of sacrificial human victims--corresponding
to their differing social ranks and functions
under the king when he was alive-were present.
See also Lu (as in n. 8).
64. Brinker (as in n. 16), 178. See also n. 30.
65. Baines (as in n. 18), 14: "Portraiture is a case of
how axial fixity might change into freedom through
detailed concentration on individuals rather than
types." On the interchange between portraits and
conventional figures, individuals, and types, see
also E. H. Gombrich, "Grotesque Heads," in The
Heritage ofApelles: Studiesin theArt oftheRenaissance,
Ithaca, N.Y., 1976, esp. 69-70; and J. Cahill, The
Compelling Image: Nature and Style in SeventeenthCentury
Chinese Painting, Cambridge, Mass., 1982,
esp. 108, 114, 130, 141.
66. "Early Chinese artists valued repetition of patterns
and they valued the type over the individual"
(Keightley, 1990a, 49). This theme is pursued by
Keightley, 1990b.
67. E.g., Legion, 163; and Wang (as in n. 25), 73.
68. The terra-cotta soldiers of the First Emperor
can thus be considered an extreme case of a more
general situation, characteristic of group portraiture
as practiced in both Western and Chinese art.
See Brilliant, 95ff: "It is the participation in the
group and the devices used by the artist to visualize
that participation that have such a high degree of
significance in establishing any member's identity."
(RE)PRESENTING THE FIRST EMPEROR'S ARMY 129
perhaps elude the conceptual dichotomy of individual versus
type so crucial to Western thinking on figural art.
Much as in the real world, the personal identities of
individual soldiers were of no consequence to the army's
operation; each figure in the terra-cotta army acquires its
status and significance only within the framework of the
whole and insofar as it contributes to the function of that
whole.68 On the other hand, the larger context, that is, the
army, is sustained only through the participation of a
definite number of entities, each embodying its own specific
role.
A look into the reconstructed formation of pit no. 1 helps
to elucidate this point. Postures, gestures, and basic attributes
(such as the type of armor, garment, or weapon),
which refer to the specific function of its model, establish
each figure as a member of a generic type (armored infantryman,
archer, etc.). However, within the formation of figures
belonging to the same type and thus conforming to one basic
stance and type of garment, no two sculptures are completely
identical; each member of the group is particularized in
some way. The identity of each figure is thus predicated on
its being differentiated, established as a specific segment of
reality within the entire body of the army. The status is
sustained through the effect of difference (Figs. 8, 16).
This also suggests an explanation of the variety of facial
expressions of the clay soldiers. Why, after all, were all the
faces not rendered in a stereotypical, undifferentiated (or at
least less particularized manner), if it is the other figural
aspects that carry the reference? The features, as we have
seen, do not convey distinct, individual physiognomies.
These faces are not (to use Richard Brilliant's apt phrase)
"repositories of primary signals of identity."69 Indeed, they
are not "likenesses" at all, and it is precisely because the
features do not carry a burden of reference to actual people
that they could be relentlessly manipulated and assembled
from a stock of prefabricated expressive units. And yet the
role of faciality in the Qin warriors is paramount. Facial
diversity was essential for creating (or at least enhancing) the
effect of difference. Within the spatial contiguity of figures of
one type, the face, differentiated from surrounding faces,
installs the plausibility and authenticity of each figure as a
distinct entity.
The First Emperor's terra-cotta soldiers may not be accepted
as portraits in the strict sense of the word. Like
Fragonard's portraits de fantaisie or some astonishingly specific
images from Egyptian tombs, they lack any reference to
individuals.70 Unlike those representations, however, for
reasons just outlined, the Qin warriors are not just typological
representations. Through their interaction in one spatial
continuum, several thousands of these images constitute the
composite portrait of an army. In this capacity, each figure
thus refers to a specific, irreducible element of reality.
According to Brilliant, "Fundamental to portraiture in the
vast repertoire of artistic representation is the necessity of
expressing the intended relationship between the portrait
image and the human original" (emphasis added).71 What
had to be revealed and reconstituted of the human originals
of whom the emperor's army was composed specifically does
not include any element of individuality; the personal identity
revealed in the image is nonexistent or, rather, is
completely consistent with the social/ritual identity.72 Yet, if
we can accept that the charge of portraiture lies primarily in
the notation of the specific segment of reality to which the
figural image refers, in sustaining distinctness of entity
within a given context, without necessarily having to contain
any qualities of individual personality, we should be prepared
to consider the Qin warriors as one instance at the
outer limits of the vast genre we call portraiture.73
The Subject of the Emperor and His Necropolis
Qin Shi Huangdi's burial compound was created during a
period of very significant development in the history of
Chinese art. Commencing some three centuries earlier and
culminating during the Han dynasty, around the first century
A.D., this development brought about the gradual
evolution of representational concerns, of psychological
realism. It is usually, albeit somewhat simplistically, described
as an evolution from decoration to representation.74
The terra-cotta army, as indeed the conception of the
whole necropolis, seemingly embodies a major innovation,
an achievement radically different from Bronze Age art.
Referring to Sima Qian's description of the subterranean
world of the emperor's tomb quoted above, Wu Hung has
recently characterized this change:
It reveals an artistic goal that would have been entirely
alien to the Three Dynasties [i.e., the Bronze Age].
Shang-Zhou ritual art did not portray any worldly phenomena
but aimed to visualize an intermediate stage between
the human world and the world beyond it, thus linking
these two separate realms. In the Lishan tomb, however,
art imitated actual things: there was an artificial ocean and
flowing rivers, and all images were arranged to create an
artificial microcosm of the universe.75
Another aspect of the Qin emperor's burial compound can
be pointed out as sharply contrasting with the art of the
preceding epochs: the very existence of monumental figural
69. Brilliant, 43.
70. For the status of the portraits de fantaisie, see
M. D. Sheriff, "Invention, Resemblance, and Fragonard's
Portraitsde Fantaisie,"Art Bulletin, LXIX,no. 1,
1987, 77-87. For Egyptian sculptures which may
qualify as portraits, see Donald Spanel, Through
Ancient Eyes:Egyptian Portraiture, Birmingham, Ala.,
1988.
71. Brilliant, 7.
72. See Wendy Steiner's suggestion ("The Semiotics
of a Genre: Portraiture in Literature and Painting,"
Semiotica, xxI, 1977, 111-19) that the uniqueness
of portraits among artworks comes from their
specific reference to an existing element of reality,
that is, a real person.
73. While supporting Brilliant's insistence on referentiality
as a necessary condition of portraiture, the
Qin soldiers seem to counter his claim that identity
constituted merely as a social artifact would be
incomprehensible to an artist and impossible to
represent (Brilliant, 12-13).
74. See, e.g., M. Loehr, "Some Fundamental Issues
in the History of Chinese painting,"Journal ofAsian
Studies, xxIII, no. 2, 1964, 186. William Watson
positioned the break with the tradition of formalizing,
nonrealistic art in the middle of the 1st century
B.c., but this whole process can be traced over many
centuries. See W. Watson, The Genius of China, exh.
cat., Royal Academy, London, 1973, 117; and
idem, Style in theArts of China, London, 1974, 52.
75. Wu Hung, "From Temple to Tomb: Ancient
Chinese Art and Religion in Transition," Early
China, xIII, 1988, 94.
130 ART BULLETIN MARCH 1995 VOLUME LXXVII NUMBER 1
workon such a scale, human figures having alwaysoccupied
a subordinateplace in earlierperiods. Mostnotablythen, itis
the presence of afeaturethathasbeen referredto here asthe
"iconicimpulse"-a totallyunprecedented systematicattention
to accuracy and verisimilitude in rendering reality.
Thus, to another writer, "the vividlylifelike soldiers constitute
such a radical departure from anything previously
knownfromChina. .. . something startinglynewand indicativefor
the future."76
However, is this notion of radicalinnovation, of a fundamentalbreakwitha
previousstage as simple and straightforwardas
the abovepassageswould seem to suggest?Whenwe
consider those features that specifically define the Qin
emperor's necropolis and the terra-cottaarmy, the mental
and technological processes involved in their creation, a
somewhatdifferent picture is obtained. To enumerate only
the most important points, which have already been discussed:
firstly, the pervasive sense of depersonalization,
brought about by the presence of several thousand figural
images, none of which, however, denotes any personal
identity, as well as by the total absence of any image of the
emperor himself;secondly,the emphasison orderlyarrangement
and placement; thirdly, the "totalistic"appropriation
of disparate figurative modes and idioms involved in the
creation of the underground world. Above all else, one can
discern the operation of "editorialcompulsion" to manipulate
a preexisting stock of expressive units, most clearly
present in the rendering of faces. Significantly, all these
elements define early Bronze Age (and to some extent the
preceding Neolithic) artistic production, most notably the
artof bronzeritualvessels in itsearlystage, duringthe Shang
dynasty.77It is only logical to find these analogies extending
to the realm of technology, since both the casting of bronze
vessels and the making of terra-cottafigures involve serial,
prescriptiveprocesses.78Definite analogies can also be projected
for the status of the artisan-craftsmanin both cases,
though the lackof space does not permitus a closerview.
The above-quotedobservationon the distinctionbetween
Qin and Bronze Age artistic practice is correct in its own
termsbut it remainsmisleadinglyincomplete: the goal of the
clayarmy'screators-to depictvastsegments of reality-was
indeed different from that of Bronze Age artisansand their
patrons. Yet the Lishan necropolis and the clay army
retained some defining, strategic characteristicsof earlier
art.
Although this is not the place for a full discussion of the
analogies and comparisons between the Qin and Shang
states, it can nevertheless be argued that there is a certain
correspondence betweenthe unified,centralizedstateof Qin
and the theocratic state of the late Shang period. A brief
comparison of the representational practice of the early
BronzeAge in relation to the ideologicalconcernsof the elite
will therefore provide the opportunity for yet another approach
to grasping the significance of the artificialmicrocosm
at Lishanand the role of the Qin emperor himselfin its
establishment.
For present purposes at least, early Bronze Age Chinese
art can be divided into twobroad groups. First,a distinctset
of elaborate and codified symbolicforms, especially bronze
andjade artifacts,central to the ritual activitiesof the elite,
functioned as a physicalmeans of access to the supernatural
realm. In this capacity,together with other typesof artifacts,
they also symbolizedthe religious and political authorityof
their owners. Second, it was surface decoration that enhanced
and strengthened the ritualefficacyof the object on
which it appeared. The repeated use ("use"implicitly subsumes
display) of ritual objects and images sanctified and
legitimized the politicalauthorityof the elite, thusvalidating
the existing socialhierarchyand maintainingthe sociopolitical
order.79
In this art, reference to the observablephenomena of the
naturalworldwas relativelyscarceand in any case played a
subordinate role. More important, the primary goal of
representational activity was not to depict the object of
worship-ancestors and spirits-nor indeed the king. In
strikingcontrastto the artof other ancient societies, there is
in early Chinese art a near-total absence of images of both
supernaturaland this-worldauthorities, and of their activities.
In early China, the status of ruler and elite-and thus
the stability of sociopolitical order-was to a large extent
establishedand maintainedthroughtheirabilityto appropriate,
control, and deploy various means of representation.
Perpetuation of those activities that secured the elite's
position withinthe stable,immutable,andmanageableworld
orderwastantamountto claiming and assertingcontrolover
forms and modes of visual mediation;yet it did not include
the elite (or the object of their religious worship) as the
subjectof the depictiveactivities.80
The gradualreificationof visualpracticetowarddepiction
of real-life phenomena that took place in the later Bronze
Age gained momentum with the reestablishment of the
centralized, unified political power of the Qin state. Among
all the aspectsof theworldthatbecame legitimate subjectsof
representation, however, the ruler himself and his accomplishments
were not included. The situation followed the
pattern of the early Bronze Age in that the status of the
76. Joseph Alsop, The Rare Art Traditions: The History
of Art Collecting and Its Linked Phenomena
WhereverTheseHave Appeared, London, 1982, 218-
20. Alsop writes that the terra-cotta army of the
First Emperor provides a departure point for the
new development of Chinese art, and describes
"the reign of Qin Shihuangdi as the moment when
the second era of art obscurely, rather strangely, yet
demonstrably originated" (ibid., 217).
77. See Keightley, 1990a; and 1990b, esp. 52ff.
78. For a formulation of prescriptive technological
processes, as compared to holistic types of artistic
production, see Ursula M. Franklin, "The Beginnings
of Metallurgy in China: A Comparative Approach,"
in Kuwayama, ed. (as in n. 48), 96-97.
79. L. Kesner, "Taotie Reconsidered: Meanings
and Functions of Shang Theriomorphic Imagery,"
ArtibusAsiae, LXI,nos. 1-2, 1991, 29-53.
80. A comparison with ancient Egypt is particularly
instructive. The substantive body of Egyptian representation
was inextricably intertwined with the
ideological legitimization of rule as in China; it
depended, however, to a large extent on the dissemination
of an image of the king. See Davis,
passim, and esp. 203-24, for the ideological dimension
of canonical imagery. For the ideological
function of state-sanctioned royal likenesses in
ancient Egypt, see A. P. Kozloff and B. M. Bryan
with L. M. Berman, Egypt'sDazzling Sun: Amenhotep
III and His World, exh. cat., Cleveland Museum of
Art, Cleveland, 1992, 125-92.
81. Qin Shi Huang's dread of death is mentioned in
Shiji. According to the same source, the emperor
employed a number of alchemists/exorcists (xian)
to obtain an elixir of immortality. See Ying-Shih
Yu, "Life and Immortality in Han China," Harvard
Journal ofAsiatic Studies, xxv, 1964-65, 92-96.
(RE)PRESENTING THE FIRST EMPEROR'S ARMY 131
supreme political and religious authority was enacted and
maintained through the emperor's comprehensive control of
various orders of representation, without in any way depending
on the dissemination of his own image. The Shang elite's
privileged access to image making as a coded language, as a
specialized form of knowledge, was replaced at this time by
an even more concentrated control of and access to means
(cognitive, personal, technological) of depicting empirical
reality. The status of the First Emperor was commensurate
with his ability to deploy various modes of representation
through which the real world could be perpetuated in the
afterlife-in other words, with his ability to deploy artifice.
Thus, in burials of the Shang dynasty people followed the
dead king into the netherworld to render him appropriate
service and to symbolize his status. This custom was perpetuated,
albeit on a far less significant scale, by Eastern Zhou
nobility, and, as we have observed above, it was probably
partly used by the Qin emperor at Lishan. In fact, one
specific reference in the terra-cotta army must have been to
this practice, which at that time would have been still at least
dimly remembered. But where the Shang kings were able to
take life as a means of expressing and construing their status,
the First Emperor was capable of having humans and things
depicted as an alternative mode of presentation to achieve
the same end.
It has been repeatedly and variously stated that the
underground army embodies the ideology of the Qin state
and its Legalist doctrines and expresses the Qin emperor's
religious and supernatural beliefs. One objective in this
paper has been to point to some of the pitfalls inherent in
such accounts of the terra-cotta army, which essentially
position art as a vehicle for some predetermined belief or
philosophical concept. This is not to deny the role of
religious beliefs in the construction of the Lishan necropolis
and its inhabitants. What historians have claimed about the
Qin emperor's superstititous nature, his dread of death, and
his prolonged quest for physical immortality suggests that
there certainly might have been elements of belief in the
magic efficacy of the clay warriors involved in the commission
and construction of the army and the whole artificial microcosm.81
The First Emperor's psychic dispositions, anxieties,
and ego are as much-and perhaps more-relevant to the
army's origins as contemporaneous religious doctrines. This,
however, still does not help much to explain this remarkable
project. Were, after all, the clay soldiers any more real,
efficacious entities to the emperor than paper replicas of
electric rice pots, refrigerators, and servants are real and
functioning objects to the contemporary Taiwanese villagers
who are interred with them?82 How do we go about establishing
what those who commission such funerary goods believe
in? Continuity in these two instances of furnishing the dead
with representations of the significant aspects of their human
existence can be partly justified at least by reference to
religious beliefs and recurrent psychological dispositions.
These are insufficient, however, to account for the difference
between the clay soldiers of the First Emperor and the paper
servants of Taiwanese villagers today.
One substantial implication of regarding representation as
merely the expression of some belief or idea, rather than as
complete on its own terms, is to keep the gulf between
representation and reality wide open. However (and here my
argument assumes the form of a proposition), the significance
of the Lishan necropolis can only be grasped by
adopting the perspective recently embraced by Freedberg
toward the end of his book ThePowerofImages: to reconsider
the radical disjunction between the reality of an art object
and reality itself.83 The concern with locating some belief as
the underlying rationale for the terra-cotta army, then, must
be supplemented with a proper focus on the force of
imagination: not belief but make-believe played a crucial
role.84
The whole microcosm of the Lishan necropolis did not,
therefore, primarily reflect or express Legalist or Daoist
doctrines, the Qin state ideology, or the emperor's beliefs.
The Lishan mausoleum simply in itself constitutes the most
complete, self-sustained form of an idea.85 It occurs at a
critical juncture in the history of Chinese art, at the moment
when the coercive power of the ruler was matched with his
ability to command vast economic resources and, above all,
to exert exclusive authority over forms of visual mediation.
Because of such an unparalleled capability to orchestrate
representational modes, combining presentation with the
deployment of various figurative modes of representation, a
microcosm could have been created that sustained claims to
authenticity. The real world extended into the artificial one
to create a single continuous presence. In this continuum,
because of specific historical circumstances, the beliefs and
personal anxieties of one man were completely tantamount
to the ideological concerns of the state. The Lishan project
was naturally aimed at the future, designed to represent and
perpetuate a certain order of the world, to project it into
eternity for the needs of the emperor after his death.
However, it was during and through the long process of the
project's coming into existence that the very order it strove to
represent was being established and embodied in the first
place. The Lishan necropolis is a metaphor for the person of
82. See Emily Ahern, The Cult of the Dead in a
Chinese Village, Stanford, Calif., 1973, 226: "The
most elaborate array of paper objects I saw was for a
woman in the Li lineage .... Inside the front gate,
in the open courtyard, were replicas of an electric
fan, two television sets, an electric rice pot, a
washing machine, a car with large tail fins, a radio,
lawn furniture, and a refrigerator, along with several
servants wielding mops and brooms. ... Relatives
sometimes carry on long discussions while
gazing at the paper house, enumerating all the
comforts provided for the dead. Those close to the
deceased are clearly anxious about the precise
conditions of the ancestor's existence in the underworld."
See also A. Cormack, Chinese Births, Weddings
and Deaths, Peking [Beijing], 1923, 29-30.
83. Freedberg, chap. 15, esp. 436-41.
84. Kendall Walton has recently provided a thoughtful
account of representation, whose theoretical
thrust is most pertinent to our theme. It emphasizes
the role of the imagination (as contrasted to
belief) in the creation of fictional worlds; see K.
Walton, Mimesis as Make-Believe: On the Foundations
of Representational Arts, Cambridge, Mass. 1990,
esp. 39-42, 273: "We do not just observe fictional
worlds from without. We live in them."
85. The case of the First Emperor's mausoleum
(but more generally, of the whole of early Chinese
art) provides support for Whitney Davis's strong
critique of idealist, reductive notions of representation:
"rather than asking what ideas are expressed
in a canonical image, we should ask how ideas are
expressed in an image-which is another way of
asking simply what kind of ideas the images are and
not what kinds of ideas magico-religious or other
doctrines might be .... Representation is no more
and no less than a specific doctrine or belief in
itself' (Davis, 202-3).
132 ART BULLETIN MARCH 1995 VOLUME LXXVII NUMBER I
the First Emperor himself. His very subjectivity-though
nowhere figuratively depicted-was being enacted here in its
twin aspect of an individual with personal beliefs and anxieties
and the omnipotent First Emperor. Only thus established,
could he finally enter the hereafter.
Frequently Cited Sources
Brilliant, R., Portraiture, London, 1991.
Davis, W., The Canonical Tradition in Ancient EgyptianArt, Cambridge, 1989.
Freedberg, D., The Power ofImages: Studiesin theHistoryand TheoryofResponse,
Chicago, 1989.
Hearn, M. K., "The Terracotta Army of the First Emperor of Qin," in Wen
Fong, ed., The Great Bronze Age of China: An Exhibitionfrom the People's
Republicof China, exh. cat., Metropolitan Museum of Art, New York, 1980,
353-78.
Keightley, D., 1990a, "In Clear and in Code: Pre-Classical Roots of the Great
Tradition in China," paper delivered at the session "The Formation of
Great Traditions," Annual Meeting of the College Art Association, New
York.
, 1990b, Where Have All the Heroes Gone?: Reflections on Protagonists,
Art, and Culture in Early China and Early Greece, unpublished
manuscript, Berkeley.
Ledderose, L., and A. Schlombs, eds., Jenseits der grossen Mauer: Der Erste
Kaiservon China und seine Terrakotta-Armee,exh. cat., Museum am Ostwall,
Dortmund, GUitersloh, 1990.
Legion The First Emperor'sTerracottaLegion, compiled by the Archaeological
Team of Qin Shi Huang Mausoleum, Beijing, 1988.
Quest The Quest for Eternity: Chinese Ceramic Sculptures from the People's
Republic of China, exh. cat., Los Angeles County Museum of Art, Los
Angeles, 1987.
Report Qin Shi Huang ling bing ma yong keng yihao keng fajue baogao
1974-1984 (Report on the excavation of No. 1 pit of figures of soldiers
and horses at the First Emperor's mausoleum in 1974-1984), compiled by
Archaeological Institute of Shaanxi Province and Archaeological Team of
Qin Pits of Pottery Figures, 2 vols., Beijing, 1988.
Yuan Zhongyi, ed., Qin Shi Huang ling bing ma yong yanjiu (The research on
pottery figures of soldiers and horses at the First Emperor's mausoleum),
Beijing, 1990.
Ladislav Kesner, who received a Ph.D. from Charles University,
Prague, is a curatorof Chineseart in theNational Gallery,Prague.
He has written catalogues and articles on Chinese art; the latter
have appeared in Artibus Asiae, Oriental Art, Orientations,
and elsewhere[DepartmentofAsian Arts, National Gallery,zamek
Zbraslav, 15600 Praha 5, CzechRepublic].