Bol/Wang Anshi and the Zhou li rev 0707 1
Wang Anshi and the Zhou li
Peter K. Bol
The importance of the Zhou li to the great Northern Song reform leader Wang Anshi
(1021-1086) is well known. He was not the first in his era to see in the Zhou li an all-
encompassing system for realizing the common good—Li Gou (1009-1059) had preceded
him with a fifty-one part series “On How the Rituals of Zhou Realized the Great Peace,”
although he had not written a commentary on the text.1 For some the point to be made is
that Wang used the Zhou li as a source for, or at least the justification for, the “New
Policies,” policies that greatly increased the state’s role in the economy, society, and
culture.2 I suspect that many of us have tended to see the Zhou li in the same light, but as
Jaeyoon Song demonstrates in his paper, a large number of Southern Song commentators
set out to disprove that the Zhou li in fact provided justifications for a strongly
centralized activist state.
Surprisingly, little has been written about Wang’s understanding of the Zhou li.3 It was
one of three classics for which special commentaries were prepared by the New Policies
regime (along with the Odes and Documents) and the only one of the three that Wang
claimed to have personally authored. Although his commentary, printed and widely
distributed, disappeared (possibly by mid Ming) it is quoted in many Southern Song texts
and the larger part of it was included in the Great Canon of the Yongle Reign Period
(Yongle dadian) in the early fifteenth century, from which it was retrieved and copied
into the Siku quanshu in the eighteenth century.4 Recently Cheng Yuanmin has
exhaustively culled Song and later texts for further quotations and comments on Wang’s
views, creating the best possible edition we can expect short of a Song imprint
reemerging.5 I am not sure how we can account for this lack of interest, but I suspect (to
generalize from my own case) that many, having looked at Wang’s commentary, found it
to be little more than that: a commentary on a rather dry text that consists mainly of lists,
devoid of argument and lacking a narrative. Moreover, the wealth of writings in Wang’s
literary collection and records of his activities as a chief councilor provide far richer
grounds for probing his ideas and intentions.
Contexts for Wang Anshi’s Commentary
I shall not pursue the effort to tie the Zhou li to specific policies. Instead, I propose to see
his commentary in two contexts. The first context is its intended use. The commentary,
which was finished in 1075, was part of the creation of curriculum for a newly instituted
national school system and a new examination system. Wang had abolished the various
memorization fields in the civil service examination and had replaced the regulated verse
Li Gou, 1981: ch. 6-14.
See, for example, Hou Jiaju,1987: 301-7.
3 An important exception is Azuma Jūji 1995.
Wang Anshi 2002.
Cheng Yuanmin 1986, vols. 3-4.
Bol/Wang Anshi and the Zhou li rev 0707 2
and regulated rhapsody in the first session of the prestigious Presented Scholar
examination with essays on the Classics. In support of the new education program the
court promulgated three new commentaries: on the Rituals of Zhou, Book of Documents,
and Book of Odes. Cheng Yuanmin has also collected remnants of the Documents and
Odes commentaries. In retirement Wang composed the Explanations of Characters (Zi
shuo), which explained the meaning of words used in the Classics. Wang submitted the
Explanations of Characters to the court in 1082 it was not officially printed and
distributed until 1094. The text is loss but a good number of examples of Wang’s
explication of characters appear in his Zhou li commentary.6 Thus the commentary was
part of a broader program of learning (xue 學) which would come to be known as the
“Wang Learning” 王學 or the “New Learning” 新學. Perhaps for strategic reasons
opponents of the New Policies were generally more critical of the curriculum for
excluding other points of view than of the commentaries themselves. 7
The second context is the larger shift that had been taking place in the writing of
commentaries on the Classics. It had first appeared in commentaries on the Spring and
Autumn Annals in the late eighth century and had grown in force Northern Song. This
shift is commonly described as a change from a focus on philological issues (xungu 訓詁)
to moral significance (yili 義理). Although this dichotomy unfortunately suggests that
Han-Tang commentators were uninterested in the significance of the ancient canon,
Wang’s advocates and critics clearly saw him as someone who was concerned with larger
moral issues: some objected that “he used yin-yang and nature and destiny (xing ming 性
命; i.e. that which is in things innately ) to explain [the Classics]” and some praised him
on much the same grounds: “He has thought through the principles of nature and destiny
and the way and virtue.”8
These two contexts involve related problems. First, what was it that the commentary was
supposed to accomplish? A textbook can be a vehicle for teaching content (what one
should know about a subject) and at the same time serve as a means of teaching a method
of analysis (how one should think about the subject). Wang made this distinction in 1066:
"Those today who are literati (shi) know they should learn, but some do not know how they
should learn."9 In the case of a Classic the text already exists; the question was always
what commentators wanted to do with it. Was it to help people read the original? To
show them how to interpret it? To advance an argument about what was truly significant
about it? Zheng Xuan’s commentary from the Han dynasty was widely available. Jia
Gongyan’s 賈公彥 (fl. 650) elaboration of that commentary (the shu 疏, also known as a
“subcommentary”), which drew on many other sources, was available as a separate text.
It had been printed in about 1000, although it was not published together with Zheng’s
Zhang Zongxiang 2005, reprints definitions of 618 characters found in the Zhou li and in
citations of Wang Anshi in other texts that Zhang collected in the 1940s. Is still not clear,
however, how the entries in the original text were composed. Zhang’s text unfortunately does not
specify the location of the citation in the Zhou li commentary or analyze its menaing in context.
Cheng Yuanmin 1978 . The Zi shuo is also discussed by Winston Lo 1976.
Cheng Yuanmin 1978: 252, 60.
Wang Anshi 1959: 82.863.
Bol/Wang Anshi and the Zhou li rev 0707 3
commentary until the early Southern Song.10 Wang’s commentary was clearly informed
by Zheng Xuan and, in the examples we shall consider here either assumes Zheng’s
glosses or, when he does not, explains why his own reading is correct. In contrast to Jia
Gongyan, who made Zheng Xuan the focus of his own commentary, Wang was not
interested in arguing with Zheng or Jia. Wang spoke to the text of the Classic itself and
provided a way of seeing how it could be meaningful in the present. Intended or not it
was an example of how Wang thought about something.
But when we ask what it was that Wang was thinking about in reading the Classics, we
do not find Wang providing a single clear-cut answer. He tells a student that "If you wish
to illuminate the Way, then whatever departs from the Classics of the sages is not worth
being illuminated."11 This suggest that the Classics, and the Zhou li, are the only
framework for thinking about values. And yet he could also state, soon after leaving the
chief councillorship, that by being willing to take all knowledge into account he could truly
grasp the system in the Classics:
For long the world has not seen the complete Classics. If one were only to
read the Classics it would not be enough to know the Classics. I thus read
everything, from the hundred schools and various masters to [such medical
texts as] the Canon of Difficult Issues and the Basic Questions, the
pharmacopeia and various minor theories, and I inquire of everyone, down
to the farmer and the craftswoman. Only then am I able to know the larger
system (da ti 大體) of the Classics and be free of doubt. The later ages in
which we learn are different from the time of the Former Kings. We must
do this if we are fully to know the sages.12
This suggests that in fact ideas gained in the world outside of the text could be essential
to determining what was of value in the text, a rather different proposition than his advice
to the student.
This brings us to a second question and my primary concern. If we assume that a text is
meaningful, whether because it has coherent relationship to the world outside of the text
or because the text is adequate in itself, how can its meaningfulness be established? If
part of the job of learning is, as Wang said, to teach people “how they should learn,” then
how can a commentary propagate a method of establishing or discovering meaning that
others can share? Wang’s assumption had long been that there was a method to how the
sages had done things, they had a shared sense of what was fundamental to other parts
and of the order in which they should proceed. Early in his career he poses this problem
There were root and branch to the sages' ordering of the age. There was
what came first and last in their putting it into practice. The problems of the
world have been left uncorrected for a long time now; teaching and policy
have yet to be made according to the conceptions of the sages. We have lost
sight of the root, seeking it in the branch; we have taken what should come
last and put it first. And thus the world careens toward disorder. Now if it is
Cheng Yuanmin 1986: vol. 3, 30.
Wang Anshi 1959: 74.786.
Wang Anshi 1959: 73.779.
Bol/Wang Anshi and the Zhou li rev 0707 4
so that the world will not be ordered except through the means that the
sages used to achieve order, then to be considered a true literatus (shi 士 )
one must attend to how the sages achieved order. I want you gentleman to
relate in full the root and branch of how the sages achieved order and what
they thought first and last.13
And it is echoed in the Palace Examination of 1070, soon after Wang came to power:
When the sages exercised kingship over all under heaven all officials
fulfilled the duties of their offices and all affairs were correctly organized.
If there was something left undone then they did it, and whatever they did
succeeded. If there was someone left unreformed then they reformed him,
and whomever they reformed accepted it. The fields were opened to
farming; the irrigation channels were in good repair. Plants and trees
flourished. Fowl and beast, fish and snake, all realized their natures. Their
wealth was sufficient for making the rites complete, their knowledge was
sufficient for extending [the suasive power of] music. Their governance
was sufficient for making punishments accurately fit [the crime].
Gentlemen, what must be done to attain this?
But the question then made a crucial demand: candidate should approach the
question in terms of method:
The failings of the present moment may be said to be legion. The method
of repair must have a root and branch and there ought to be what comes
first and last in what should be done. This is something you, Gentlemen,
ought to know. 14 方今之獘，可謂眾矣。捄之之道，必有本末。所施
The idea that there is a method, a system, or a set of structural relationships that can be
known echoes the fundamental claim made in Wang’s famous “Myriad Word Memorial”
of 1058 in which he set out the method of the sage kings’ creation of perfect order in
antiquity and its application in the present. Wang’s argument was that ideas, understood
in terms of their internal coherence, transcended history. There had been change in
antiquity—a period that covered a millennium—and the particulars of situations facing
government and the responses of the sage kings differed, “but their conceptions in acting
Wang Anshi 1959: 70.747. This question includes language identical to that used in a letter to Zu
Zezhi 祖擇之 (1011-1085) datable to 1046, see 77.812.
Xu Song 1957: xuanju 7.19a-b.
Bol/Wang Anshi and the Zhou li rev 0707 5
on society and state were always the same in [their] root and branch and what came first
and last.”15 而其為天下國家之意，本末先後未常不同也
In this Wang was echoing Ouyang Xiu, who almost twenty years earlier had asserted
much the same thing in his call for programmatic reform, “On Fundamentals.”
There is root and branch to all human affairs and there is what comes first
and last in ordering them. The documents of [the sage kings] Yao and
Shun are sketchy, [yet] those in later ages who ordered the world always
took their models from the Three Dynasties [of antiquity] because they
had inferred the root and branch and knew what came first and last.16 天下
We might take the terminology of root-branch/first-last as pointing to a conceptual field
that would include contemporary terms such internal logic, deep structure, system, and
method. The question is what this means in practice, as part of a curriculum that aims to
teach students how to learn.
Wang did believe his commentary on the Zhou li had a role to play in this, as he
explains in his preface.
Literati have been blinded by commonplace learning for a long time. The
sage above has been concerned about this and has used the
learning/methods of the Classics (jing shu 經術) to form them. He has
further assembled Ru ministers to explicate their points for distribution to
the schools. Your minister has been responsible for the Offices of Zhou.
When the Way is present in the affairs of government, noble and lowly
have their proper place, last and first have their proper order, many and
few have their proper number, and slow and fast have their proper time.
[I.e. ranks, priorities, amounts, and timing are all appropriate to the tasks
at hand]. Their institution and deployment depends on policy but putting
them into practice depends on the person. No time was better than that of
the Duke of Zhou for having people capable of filling the office and
offices capable of putting policy into effect. And of texts that can be found
in the written record no document is more complete than the Offices of
Zhou for policies that can be applied in later ages.
It must have been that ongoing practice exalted them and change and
continuation completed them, so that later ages had nothing to add. How
could that have been due to the efforts of [Kings] Wen and Wu and the
Duke of Zhou alone? It was like the cycle of the four seasons, day and
night lengthen and bring about cold and heat; it does not happen in a
Wang Anshi 1959: 39.410.
Ou-yang Hsiu 1961: waiji 4.411-13.
Bol/Wang Anshi and the Zhou li rev 0707 6
Since the decline of the Zhou to today, through a thousand and several
hundred years, the traces of that great peace have been swept almost
completely away. What scholars see is no longer the complete Classic. At
such a time to wish to explicate and make it clear is something of which,
although Your Minister dares undertake it, he knows the difficulty.
And,since he sees explicating and making it clear as difficult,he also
knows the difficulty of going back and restoring its [manner of]
establishing polices and accomplishing things…
Wang’s claim for the text is that the Zhou was a moment when the proper order of
things was better realized than at any other moment in history, and thus the text as
a more complete account than anything else available is worthy of attention. This
opens an interesting possibility. If Zhou government and the Zhou li represent a
successful effort to realize “the Way” as something that exists independently of
history then ideally we need to understand the Way on which the Classics are
based——and break free of the constraints of the text and period. As Wang said
in a passage quoted above, understanding how things work in the present is a
necessary aid to understanding the system of the Classics. Since the text is the
only surviving concrete expression of this achievement, the challenge is to get
from the text itself to its real foundation. This—a manner of interpreting the text
that reveals the conceptions on which it is based—is central to Wang’s practice of
Interpreting the Zhou li: Wang versus the Han-Tang Reading
What Wang is trying to accomplish with his commentary becomes more apparent
by contrasting his method of interpretation with the Zheng Xuan and Jia Gongyan
commentaries. We shall see that Zheng’s concern lies with determining what the
words in the texts refer to, he wants to find the correct connections between words
and things. Jia Gongyan, however, is concerned with Zheng Xuan’s text, he wants
Bol/Wang Anshi and the Zhou li rev 0707 7
to validate Zheng’s reading by appealing to a broader range of texts. It is clear
that Wang is conversant with Zheng Xuan’s commentary and, given that Wang
makes it clear when he departs from Zheng’s glosses, accepts Zheng as
authoritative. Wang is asking different questions, but answering them does not
require a rejection of Zheng Xuan.
Below I will take up two passages, a list of offices and ranks and a list of
occupations and economic activities. The passages represent apparently radical
difference in his method of analysis: the first is dominated by a graphic analysis
of characters and the second by a logical analysis of the connections between
sentences. As Wang reads them they also speak to two issue of great concern at
the time: the nature of the political elite and structure of the economy. Taken
together they substantiate the claim that there was such a thing as “Wang
learning” in the sense that he has an approach to establishing the normative
meaning of things that others can share.
1. Names and the political elite
The Offices of Heaven section, the first part of the Zhou li, begins by stating that it is the
king who establishes the realm and that he appoints a premier (zhong zai 冢宰) to direct
his subordinates and aid the king in governing the royal realm and principalities.17 乃立
天官冢宰。使帥其屬而掌邦治。以佐王均邦國. This is followed by a list of offices in
the Offices of Heaven division and the number of men of various ranks/offices who staff
them. The list begins with the chief minister himself. Wang treats this list as one
commentatorial unit, in contrast to the Zheng Xuan commentary which (more correctly
given the quota for appointments) divides the list into three sections (1-6, 7-8, 9-10).
Office Rank of incumbent Quota
1 大宰 Premier 卿 Qing: Minister 1
2 小宰 Vice-Premiers 中大夫 Middle Da-fu: 2
Ordinary Grand Master
3 宰夫 Chief Stewards 下大夫 Lower Da-fu: 4
Junior Grand Master
4 上士 Upper Shi 8
5 中士 Middle Shi 16
6 旅 multitude 32
7 府 warehouseman 6
8 史 manager 12
9 胥 staff 12
10 徒 serviceman 120
Citations of the Zhou li, the Zheng Xuan commentary, and the Jia Gongyan subcommentary are
from Zheng Xuan 1965.
Bol/Wang Anshi and the Zhou li rev 0707 8
Zheng Xuan’s commentary answers a series of questions:
a. Why does the text first say Zhong zai 冢宰 but now Da zai 大宰 for premier? Because
Zhong 冢 is used from the perspective of his control over other officials whereas Da 大 is
used from the perspective of his service to the king. Zhong is above Da. The top of a
mountain is called Zhong.
b. What is Lü 旅 doing in a list of the king’s ministers? Lü actually means multitude and
this would be the Lower Shi, who govern the multitudinous affairs. All of these figures
are ministers of the king, from the Da zai to the multitudinous Lower Shi.18
c. Zheng Xuan treats lines 7-8, 9-10 separately and makes clear that these are different
status groups, the Fu 府 are warehousemen and the Shi 史 scribes; they are appointed by
their superiors rather than the king.
府 治 藏 。史 掌 書 者 。凡 府 史 皆 其 官 長 所 自 辟 除。
The Xu 胥 and Tu 徒 are commoners providing labor service. The Xu, being more
talented and intelligent, lead gangs of ten Tu.
此 民 給 徭 役 者。若今 衞士矣 。胥 讀 如 諝 。謂 其 有 才 知 為 什 長
Jia Gongyan elaborates on Zheng’s statements. For example, his comment on Zheng’s
explanation for the change from Da zai to Zhong zai elaborates on Zheng’s explanation
largely by adducing other passages from the Zhou li in support of the idea that when a
chief minister is spoken of in his role as managing other officials on behalf of the king he
is called Zhong zai rather than Da zai. The point is to show that in fact the Zhong zai and
Da zai refer to the same figure, thus removing any possible confusion.
In contrast, Wang Anshi ignores the question of why the text shifts from Zhong zai to Da
zai. He begins by taking up Zheng Xuan’s inquiry into the status of the figures listed.
Zheng had identified the Lü as Lower Shi; Wang adds to this a definition of the Qing: if
the Xiao zai (Vice-Premiers) are Middle Da-fu then the Qing ought to be an Upper Da-fu.
Noting that the “Regulations of the King” chapter of the Book of Rites states that “The
Upper Da-fu of the Feudal Lords are Qing,” Wang concludes that “This ought not only be
so for the Qing of the Feudal Lords.” 大宰卿。小宰中大夫。則卿上大夫也。王制
曰。諸侯之上大夫卿。盖非特諸侯之卿為然也。19 Wang does not say why this
matters, but the effect of this is to say that the political elite was composed of Shi and Da-
fu, terms that would immediately resonate with his own readership, who referred to
themselves as shi dafu (scholar officials) and that there was no third layer of officials
apart from them.
I do not treat the last line of Zheng’s commentary: 王之卿六命其大夫四命士以三命而下為差.
Citations of Wang Anshi’s commentary are from Wang Anshi 2002. Emendations, noted in
brackets, follow Cheng Yuanmin 1986.
Bol/Wang Anshi and the Zhou li rev 0707 9
Wang goals in this section lie less with the text itself or the Zhou dynasty political system
than with the question of what we can know about the nature and function of those who
led the government. An historical inquiry might address this question through an analysis
of the Zhou system as found in the Zhou li and others texts—as Jia Gongyan’s collection
of citations demonstrated, the term Zhong zai was used with reference to one had
authority over the official apparatus of a state. Wang, however, is interested in meanings
that transcend their historical context and to this end he sets out to demonstrate that an
analysis of the graphic structure of the written character tells us something of timeless
significance. At the most fundamental level this makes the Zhou li itself largely irrelevant
as a means of judging the validity of Wang’s claims. This is his first case:
The character Qing is from ; is to memorialize. And it is from 卩;
卩 is to stop. The left is from and the right is from 卩; this is the idea
of knowing when to advance and when to stop. And it is from 皀 [in the
center], the qi (energy/steam) of millet. Millet is a product of the earth, it
has the way of nurturing people. This energy/steam [as signified by the 皀
element in the middle] is able to reach upwards. A Qing has the way of
nurturing people and he reaches upwards [to the king] but he is of the
Earth category, therefore the character is like this.
卿之字從 奏也。從卩。卩止也。左從 。右從卩。知進止之
Wang analysis of the graphic form is the stated justification for his definition of the basic
principles that should guide a Qing as chief minister: he is someone who knows how to
provide for the popular welfare, but he must also be able to make a connection to the
ruler above. The claim that he is of the Earth category suggests that the Qing should be
seen as the counterpart and balance to the king as son of heaven. Note Wang’s conclusion:
the character has this form because it reflects the qualities that, in the proper order of
things, a Qing as a high Da-fu should have.
Having explained the Qing, Wang turns to the other ranks in the passage: the Da-fu and
The character fu 夫 [distinguished person, husband] and [the character] tian 天
[heaven] are from yi 一 and da 大. That is why the fu/husband is heaven to the
wife. Heaven 天 is big 大 and has nothing above/superior to it, therefore yi/one 一
is on top of the da/big 大. Now although the fu 夫 is yi/one 一 and da/big 大 he is
not without anything above like heaven, therefore the yi 一 cannot be on top of
da/big 大 [as in the case of tian/heaven 天, but must cut through the da/big as in fu
夫]. A fu leads others with intelligence; A Da-fu is the greater of those who leads
others with intelligence.
Bol/Wang Anshi and the Zhou li rev 0707 10
The character shi 士 [in ancient times a term for the lowest noble rank; later a
term for one who serves; in Song the term for the educated elite who seek to serve
in government] and [the character] craftsman 工 and talent 才 all are from 二 [two
horizontals] and from 丨[a vertical]. Talent 才 reaches everywhere, therefore it
reaches out above and below. A craftsman 工 merely prepares human utensils,
therefore he does not reach out above or below. A shi 士 is not a complete talent
and thus he ought not to reach out in either [direction], but he aspires to the Way,
therefore he reaches out above. A shi is one who serves others, therefore 士 is also
glossed as “to serve 事.” If he serves others then he is not yet able to use
intelligence to lead others. He is not one others serve. Therefore one who is not
yet married is called a shi 士.
Lower shi are called lü 旅, because they are a multitude. The character lü 旅 is
from 㫃[flag] and from 从[follow]. As for a multitude, it follows the flag signals.
Since they follow the flag signals they follow others and do not direct themselves.
If Lower shi are lü 旅 then they also follow others and do not direct themselves.
The character warehouse[man] fu 府 is from 广 [roof] and from 付 [entrust to].
The roof 广 is for storage, the entrusting 付 is for entrusting goods to him.
The character manager 㕜 is from 中 [center] and 又 [assist]. He [the king]
establishes offices and apportions duties in order to create a center for the people.
What the manager㕜is responsible for is below, he merely helps them [i.e. his
The character staff 胥 is from 疋 [foot] and from 肉 [flesh]. The foot 疋 [is used]
because it is the lower appendage of things. Flesh because it is also able to nurture
people. The [staff] merely assist them. Therefore staff is also glossed as “assist.”
Qing 卿 is from 皀 （millet energy） and staff is from 肉 (flesh). Both have the
meaning/principle of nurturing people. Thus all that the king establishes is simply
to nurture people.
Bol/Wang Anshi and the Zhou li rev 0707 11
The character serviceman 徒 is from 辵[run] and from 土 [ground]. The
servicemen have no vehicle with them. In running and walking they are in contact
with the ground. Thus to go without a vehicle is called tu xing 徒 行, going on
Mr. Zheng thinks the warehouseman, manager, staff, and serviceman are all
appointed by their superior officials. This is because from the lower shi on up all
are appointed by the king. But King Mu commanded the Grand Tutor: “Carefully
select your officials.” So that although they were appointed by the king the one
who was in charge of them got to select them. Although the warehousemen,
manager, staff, and serviceman were not Shi the Former Kings did not
discriminate between statuses in employing people. The humble were not
embarrassed to serve the Shi and Da-fu; the noble did not resent working for the
son of heaven.
With his analysis of the graphs Wang arrives at some conclusions about the nature of the
bureaucratic leadership: at its apex it must further the welfare of the general populace
below and influence the king above, at the next level it consists of men who leadership is
based less on moral commitment then on their talent and intelligence, and at the lower
level they are people who follow directions but aspire to ideal purposes. There is not,
however, an absolute distinction between the officials and those in technical and service
positions below them. The Former Kings employed people without respect to their social
status—more of a Song idea than a Zhou one—and everyone in the system accepted the
fact that he was part of a hierarchical order and serving someone else.
Wang acknowledged Xu Shen’s 許慎 (c. 58-c.147) Explanations of Single Elements and
Compound Characters (Shuo wen jie zi 說文解字) of the Han dynasty as a precedent,
but states that he worked from his own insights. He justified his work in the memorial of
submission for the Explanation of Characters. Writing had been a preeminent concern of
the Former Kings because maintaining the right names and laws was essential to ensure
that people shared the same values. 凡以同道徳之歸一名法之守而巳。 Written
characters began as a means of transmitting speech, which had developed from the
expression of feelings, but “Although characters were instituted by people, the
fundamental nature comes from what is so-of-itself” 字雖人之所制。本實出於自然。
People were inspired by patterns and diagrams that were not of human origin and thus the
forms characters took were “the forms of what was so-of-itself,” their shapes, sounds, and
meanings were all so-of-themselves.” The fact that they were fundamentally of the
Bol/Wang Anshi and the Zhou li rev 0707 12
natural order and systematic meant that “Although immortals and sages lived in different
places and spoke in different accents and had different dots and lines [in writing], they
translated and comprehended them. Their meanings were unitary.” 20 故仙聖所居。雖殊
It follows from this, Wang explains in his preface, that “Whether the form is vertical or
horizontal, rounded or linear, slanted or straight, inside or outside, and left or right, all
has meaning. This is not something a person’s individual intelligence is capable of
人私智所能為也。The Qin’s adoption of clerical script nearly destroyed this; for Wang
such a monumental attack on the basis of morality in society was evidence that "Heaven
was allowing Our Culture to be lost." Thus the fact that he had been able to figure out how
meaning was inherent in writing, Wang concluded, was a sign that “Heaven plans to
restore Our Culture and is using me to aid its beginning. Therefore teaching and learning
must begin from this. Those able to understand this will already have nine-tenths of moral
As Wang practices it graphic analysis is a way of discerning the normative conceptions for
the figures to which the characters refer. These ideas are “moral” in the sense that they
define roles in an all inclusive, unified system, an integrated social order in which each
person finds a proper place and function. What is important, I think, is that the Zhou li can
serve as means of recovering these “natural” ideas because it reflects them, not because the
Zhou made them up.
2. The Social Order
The Zhou li defines the duties of the Premier (Da zai) with a series of numbered lists,
beginning with “Managing the six norms for founding a state and assisting the King in
governing the royal realm and principalities.” 掌建邦之六典以佐王治邦國 My second
example is one section of the description of the duties of the Premier, the first of five
sections of nine items each that deal with the economy and society.
“With the nine occupations he employs the myriad people” 以九職任萬民
1 三農 生九穀 growing the nine grains
2 園圃 毓草木 cultivating shrubs and trees
Wang Anshi 1959: 56.608-09.
Wang Anshi 1959: 84.879-80.
Bol/Wang Anshi and the Zhou li rev 0707 13
3 虞衡 作山澤之材 creating products of mountains and
yu heng marshes
forest- and marsh-men
4 藪牧 養蕃鳥獸 raising birds and beasts
5 百工 飭化八材 working the eight natural materials
6 商賈 阜通貨賄 accumulating and circulating goods
7 嬪婦 化治絲枲 transforming silk and hemp [into cloth]
8 臣妾 聚斂疏材 gathering ground and root vegetables
male and female servants
9 閒民 無常職轉移執事 no specialized occupation,
xian min moving to take up work
The passage names nine occupations (or possibly nine offices supervising these
occupations) and associates them nine kinds of economic activity: grain cultivation,
orchard and garden production, mountain and marsh production, fowl and animal
husbandry; handicraft production, trade, cloth production, vegetable gathering, and the
provision of labor.
Zheng Xuan’s commentary speaks to a single question: what things do the words in the
passage refer to. The issue is set by his citation of Zheng Sinong 鄭司農 (Zheng Zhong
鄭眾 d. 83 A.D.) who equates the san nong 三農 with three kinds of lands on which grain
is cultivated (flat, mountain, and marsh lands), lists the nine grains (e.g. millet), and lists
the eight materials and the terms for how they are worked (e.g. wood is carved).
Zheng Xuan proposes a different definition of the three types of land and the nine grains.
In addition he defines several titles. For example, 2) pu yuan 圃園 is actually two terms,
pu refers to the place where fruit and melons are planted and yuan refers to the fencing
around it; 6) shang gu 商賈 refers to two kinds of merchants, traveling and stationary
(shopkeepers),who deal in valuables 貨 and cloth 賄. He also goes outside the Zhou li
and cites texts which support his definition. For example, 7) pin fu, the term pin is merely
a laudatory term for wife/fu; citing the Cannon of Yao’s reference to Yao giving his two
daughters as wives to Shun; 8) chen qie are terms for humble men and women, citing an
anecdote about Duke Hui of Jin 晉惠公 (from the Zuo zhuan). For Zheng Xuan the
Bol/Wang Anshi and the Zhou li rev 0707 14
equation of name with things in the world is the goal, for if those equations are clear then
the passage simply means what it says, nothing more needs to be said.
But in fact much more can be said, as Jia Gongyan demonstrates in a commentary over
ten times as long. Jia works within the framework of Zheng Xuan’s commentary, but by
doing so he is in fact shifting the focus from the relationships between words and things
to the relationship between words and words. His explanation of Zheng Xuan’s
disagreement with Zheng Zhong’s definition of the three kinds of agricultural land
(mountain, marsh, and flat land) is based on definitions of words: “It is because stones
are piled up that it is called a mountain and because water has spread that it is called a
marsh. They do not produce the nine grains, therefore the later Zheng does not follow
For Jia the major question is why Zheng says what he says, a question he answers by
surrounding Zheng with other texts that can buttress his statements. Thus, in the case of
agricultural land for example, Jia cites a definition from the Er ya 爾雅 and in support of
Zheng Xuan’s alternative list of the nine grains he quotes from a dietary text, the
“Monthly Ordinances” 月令 from the Book of Rites, and the “Birth of the People” from
Jia’s discussion of the other sections adopts these same strategies of defining words and
finding textual support. He is concerned with “meaning” and at some level shares Zheng
Xuan’s concern with knowing to which things in the world the words in the texts refer.
However, to a far greater extent Jia sees the words of the Zhou li, and the words of Zheng
Xuan, as existing in a realm of texts, so that it is the history of the usage of a word that
determines its most appropriate reference. Neither commentator asks whether there is
some larger lesson here that readers should learn, of something significant that is at first
unapparent that he can reveal to readers. Zheng Xuan’s task is straightforward: if he can
identify the referents of the words say then the text can speak directly to readers as an
account of the structure of Zhou government as an historical event and model. Jia’s
project is more complicated: he embeds Zheng Xuan in a context of texts in order to
show that Zheng is in accord with a larger textual tradition.
For the most part Wang Anshi accepts Zheng Xuan’s commentary and its claims about
the things to which the words in the text refer. In this passage, however, he differs on two
points, both of which turn out to be important to the argument he will make. First, he
wants 3) yu heng, which Zheng takes to be names of officials overseeing the mountains
and march, to be understood as a reference to the populations of these areas.
Mountain and marsh both are [under] the yu/warden22 yet it says the yu
and heng make the products of mountain and marsh. The mountain
yu/warden is in charge of laws for the mountains and forests, thus his laws
are applied in the mountains and forests. The river heng/warden is in
charge of overseeing the restrictions on rivers and marshes thus his
As the term is defined in the Kong Anguo’s commentary on the “Canon of Shun” in the Book
Bol/Wang Anshi and the Zhou li rev 0707 15
restrictions are applied on the rivers and marshes. The yu and heng are
officials of the mountains and marshes, but the creation of products of
mountain and marsh is the occupation of the populace. It follows that what
are here called yu and heng refer to the people in these areas.
I think his second case, 4) pin fu, is meant show that the term refers is to women as a
dependent group, because this fits his sense of the descending sequence of status groups
in this part of the text. “A pin is one who has a husband. A fu is one who has a mother-in-
law. When the husband has died and the mother-in-law is old then they have no
occupation, therefore those who are employed [in cloth production] are only the pin fu.”
This sets up a series of arguments about what we should learn from the Zhou li. The first
has to do with the sequence from 1) to 3) in which the different kinds of activities
represent different degrees of dependency on human intervention. The Zhou li is
revealing something about how the world of food production works in universal terms.
For the nine grains it says sheng/grow; for shrubs and trees it says
yu/cultivate; for birds and beasts its says raise/yang fan. The nine grains
are not able to grow themselves, they depend on the three types of farmers
to be grown. Shrubs and trees and able to grow themselves but are not able
to cultivate each other, they depend on gardeners to be cultivated. Birds
and beasts are able to cultivate each other but are not able to raise
themselves, they depend on pastoralists to be raised.
The second argument is based on the use of two verbs to describe the work in lines 4
Yang fan is to nurture/yang and then breed/fan them [i.e. birds and beasts].
Chi hua is to work/chi and then transform/hua them [i.e. natural materials].
Fu tong is to accumulate/fu and then circulate/tong them [i.e. goods].
Hua zhi is transform/hua and then order/zhi them [i.e. silk and hemp].
Ju lian is to collect and then store them [vegetation]
Wang’s point, I take it, is that the verbs involved reveal that there is a necessary two-part
sequence in the work process. Pastoralists must first raise the livestock to the appropriate
age and then they can it breed them; craftsmen must first prepare their hides, bones, ivory,
Bol/Wang Anshi and the Zhou li rev 0707 16
etc. (see below) and then they can transform them into utensils; merchants must first
invest in accumulating goods and then they can circulate them; women must first
transform silk and hemp into thread and then they can put them in an orderly arrangement
by weaving; servants must gather first vegetables and roots and then they can store them.
Here too, the point is that there are necessary processes in these activities that hold true
for all times and places.
Wang’s third argument has to do with a distinction between vegetable products that can
be both eaten and used, in contrast to animal products of which the edible and the useful
come from different parts.
As for the nine grains, shrubs and trees, and the products of mountain and
marsh: they are what people eat and use. As for the birds and beasts: their
flesh are for people to eat, their feathers, fur, teeth, tusks, bones, antlers,
sinews, and hides are for people to use.
All this allows Wang to support what I take to be his central thesis: the sequence of items
in the Zhou li text reflects a necessary and logical progression in terms of the way the
economy works. In other words, the reason the Zhou did what it did and the text says
what it says is because for an economy to function successfully it had to be organized in
exactly this fashion.
The first item is the three farmers grow the nine grains;
The second is the gardeners raise shrubs and trees;
The third is the mountain and marsh men create the products of mountain
The fourth is the pastoralists nurture birds and beasts.
Wang continues his account of the necessity of the sequence:
The many craftsmen turn the products of mountain and marsh and
materials from birds and beasts into utensils for the populace; therefore the
fifth is the many craftsmen work and transform the eight materials.
When what the many craftsmen do is adequate for a person’s life, then
there ought to be merchants to assist them [in making their goods
available]; therefore the sixth is the merchants accumulate and circulate
Bol/Wang Anshi and the Zhou li rev 0707 17
In employing the populace male work is primary and physical strength
comes first. Widows and wives are women and weak. Therefore the
seventh is the wives transform and order silk and hemp.
Male and female servants are more humble; therefore the eighth is the
male and female servants gather and store vegetables and roots.
The idle people are the ones the other eight occupations rely on to
complete their work; therefore the ninth is idle people without constant
occupation who move to take up work.
He ends his discussion with a further reflection on this ninth item, which turns out
to be a more general statement on the nature of employment in the economy.
Now the work of the people of the eight occupations at times employs a
multitude, so they move about to take up work. How can they be few! In
my view those who have constant [occupations] have the
advantage/profits and those without constant [occupations] are employed.
This is the way of heaven.
In short, there must be a sizeable body of free labor, without constant occupations,
available to be employed by those with constant occupations in agriculture, textile
production, handicrafts, and trade.
This is not an insignificant statement, considering the alternatives that were being
proposed in the latter half of the eleventh century. Zhang Zai 張載 (1020-1077) had
hoped, prior to his death in 1077, to start an experiment, justified in part of his reading of
the Zhou li, aimed at restoring the well-field system, equalizing land holdings, and
creating self-sufficient agricultural communities under the leadership of Confucian elites.
We know, from Wang Anshi’s criticism of powerful families that acquired large
landholdings (which he called “engrossing families” 兼併之家) when he was chief
councilor, that he shared Zhang Zai’s concern with the monopolization of land. However,
Wang’s picture is not based on the idea of self-sufficient communities but on a
productive economy with a flexible labor pool with freedom of movement that is not
bound to the land by its dependency on landed elites. In contrast to Zhang Zai,Sima
Guang 司馬光 (1019-1086) saw inequality between rich and poor as inevitable and
defended it on the grounds that the rich had become rich thanks to their superior talent
and effort. He supposed that it was precisely through their dependency on the wealthy
that the poor survived. Sima was upset by the growth of free labor, the decline in the
percentage of the population that was bound to the land, and the growing role of trade
and handicraft production. To be sure Wang expanded the role of state institutions in the
commercial economy, in contrast to Su Shi and others who objected to the “bureaucratic
Bol/Wang Anshi and the Zhou li rev 0707 18
entrepreneurship” of the New Policies at the expense of private merchants.23 But the
crucial point is that in Wang’s view it is the economy as a system that creates wealth and
provides employment, a system in which grain-growing farmers are but one sector.
We can read these two passages from Wang Anshi’s Zhou li commentary as strategically
crafted to justify the New Policies, or the New Policies view of the nature of political
leadership and the economy. We can also read them as being informed by a more general
view, shared by many at the time, that read the Classics in light of contemporary beliefs
in the value of political leadership by men of learning and in the value of an open
economy with a strong commercial and industrial sectors that produced wealth for both
the government and general populace.
But we gain something of great importance toward understanding the relationship
between learning and politics at the time by taking seriously the final line of Wang’s
commentary on the socioeconomic order: “This is the way of heaven” 天之道也. Wang
is one example of the eleventh century search for guides to ordering the world that were
grounded in things themselves, in the world as something that existed independently of
human will, and thus were not the historically contingent creations of human subjectivity.
Wang held that the most important area for gaining access to this knowledge was the
Classics. He recognized that access through the Classics was mediated by the minds of
exceptional men, the words they created, and the texts they composed. This might seem
to throw the whole enterprise of gaining knowledge that had real value back into the
realm of culture and subjectivity (as Su Shi thought it necessarily did).
Wang avoids this conclusion with a combination of assumption, assertion, and method.
The assumption is that the world (and thus anything that is created that is true to the
principles of the world) is essentially coherent and systematic. The assertion is that the
written words and sage texts, as cultural artifacts thatmediate knowledge, are in fact true
to the principles of the world. This is true for the structure of characters, as we have seen
in his introduction to his Explanation of Characters; it is true for the Zhou li, as his
preface to that text makes clear; and it is true for the Classics as a whole, thanks to the
ability of Confucius to discern a single perfect system in texts that covered 2500 years of
historical change.24 The method, as we have seen here, is the project of finding
systematic relationships. He is doing the same thing when discussing the parts of a
character as he is when explaining the sentences that make up a text. The guides he
seeks—whether called the Way, the conceptions or intentions/yi 意 of the sages or former
kings, principle or pattern/li 理, or right principle/yi 義—are to found by discerning the
For “bureaucratic entrepreneurship” and diverse objections to it from Su Shi 蘇軾 (1037-1101),
Lü Tao 呂陶 (1031-1107), and Sima Guang, see Smith 1991. Zhang Zai’s economic feudalism is
discussed in Bol Forthcoming and Ong 2004.
As Wang explains in his essay “Confucius was Wiser than Yao and Shun,” Wang Anshi 1959:
Bol/Wang Anshi and the Zhou li rev 0707 19
system, structure, orderliness, connectedness, integration, or coherence in word and
language, but they exist in some sense outside of the text in the things themselves.
Normative ideas are accessible through the form (wen) but they exist of themselves.
From Wang’s perspective (a view that was not shared by everyone) this method was
shareable and it could be applied to the understanding of things in the world as well.
Although I think we can argue that some literati associated with the New Policies sought
systematic understandings of things in the world—Shen Gua 沈括 (1029-1093) comes to
mind—Wang himself did not suggest that the pursuit of what we might call “scientific
knowledge” should be divorced from the methodical analysis of the surviving ancient
The method, in Wang’s terms “learning,” is the means by which the individual cultivates
himself and extends what he has learned to others through writing and governing. Wang’s
early conviction that the sages’ mental apprehension of the Way, their teachings and their
governance, and their textual record were ultimately unitary in their “root and branch, in
what came first and last” and, at this level, applicable to past and present, seems never to
have left him. The problem, he explained at the time, was that too many of his
contemporaries did not grasp the proper order of things in their minds but confused first
and last, root and branch.25 His control over education and his new curriculum provided
an institutional means of rectifying the problem.
In considering the New Policies—and indeed the various attempts to transform society
through government in the Northern Song—we might pay more attention to the Zhou li
not as a source of policies but as a model for programmatic policy making, in contrast to
those who saw governance as responding to circumstances. There is, I think, a great deal
of evidence to support this. The various economic, social, bureaucratic, and cultural
policies of the New Policies era do fit together as a coherent whole and there was an
extraordinary attempt, at higher level, to govern through ritual and administrative codes
that spelled out how things were supposed to work in detail and to create sophisticated
models, even at the material level of calligraphy and architecture, that others could
Wang Anshi 1959: 77.812.
Discussed in Bol 2001.
Bol/Wang Anshi and the Zhou li rev 0707 20
Azuma Jūji 吾妻重二 (1995). Ō Anseki shūkan shingi no kōsatsu 王安石《周官新義》
の考察. Chūgoku kodai reisei kenkyū 中國古代禮制研究. Kominami Ichirō 小
南一郎. Kyōto-shi 京都市, Kyōto Daigaku Jinbun Kagaku Kenkyūjo 京都大學
Bol, P. K. (2001). "Whither the Emperor? Emperor Huizong, the New Policies, and the
Tang-Song Transition." Journal of Song and Yuan Studies(31): 103-34.
Bol, P. K. (forthcoming). Reconceptualizing the Order of Things in Northern and
Southern Sung. Cambridge History of China. Cambridge, Cambridge University
Cheng Yuanmin 程元敏 (1978). San jing xinyi yu Zishuo kechang xianwei lu 三經新義
與字說科場顯微錄. Qu Wanli xiansheng qizhi rongqing lunwenji 屈萬里先生七
秩榮慶論文集. Taibei, Lianjing chuban shiye gongsi 聯經出版事業公司: 249-
Cheng Yuanmin 程元敏 (1986). San jing xin yi ji kao hui ping 三經新義輯考彙評.
Taibei Shi 臺北市, Guo li bian yi guan 國立編譯館.
Hou Jiaju 侯家駒 (1987). Zhou li yan jiu 周禮研究. Taibei Shi 臺北市, Lian jing chu
ban shi ye gong si 聯經出版事業公司.
Li Gou 李 覯 (1981). Li Gou ji 李 覯 集. Beijing, Zhonghua shuju.
Hou Jiaju 侯家駒 (1987). Zhou li yan jiu 周禮研究. Taibei Shi 臺北市, Lian jing chu
ban shi ye gong si 聯經出版事業公司.
Lo, W. (1976). "Philology, An Aspect of Sung Rationalism." Chinese Culture 17: 1-26.
Ong, C. W. (2004). Men of Letters within the Passes: Guanzhong Literati from the Tenth
to Eighteenth Centuries. East Asian Languages and Civilizations. Cambridge,
Ou-yang Hsiu 歐陽修 (1961). Ouyang Xiu quan ji 歐陽修全集. Taibei, Shijie shuju.
Smith, P. (1991). Taxing Heaven's Storehouse: Horses, Bureaucrats, and the Destruction
of the Sichuan Tea Industry, 1074-1224. Cambridge, Council on East Asian
Studies, Harvard University.
Wang Anshi 王安石 (1959). Linchuan xian sheng wen ji 臨川先生文集. Beijing 北京,
Zhonghua shu ju 中華書局.
Bol/Wang Anshi and the Zhou li rev 0707 21
Wang Anshi 王安石 (2002). Zhou guan xin yi 周官新義. Hong Kong, Di zhi wen hua
chu ban you xian gong si: Zhong wen da xue chu ban she 迪志文化出版有限公
Wang Anshi 王安石 (1959). Linchuan xian sheng wen ji 臨川先生文集. Beijing 北京,
Zhonghua shuju 中華書局.
Wang Anshi 王安石 (2002). Zhou guan xin yi 周官新義. Hong Kong, Di zhi wen hua
chu ban you xian gong si: Zhong wen da xue chu ban she 迪志文化出版有限公
Xu Song 徐松, Ed. (1957). Song hui yao ji gao 宋會要輯稿. Beijing, Zhonghua shu ju.
Zhang Zongxiang 張宗祥 and Cao Jinyan 曹錦炎 (2005). Wang Anshi "Zi shuo" ji 王安
石《字說》輯. Fuzhou Shi 福州市, Fujian ren min chu ban she 福建人民出版社.
Zheng Xuan 鄭玄, Lu Deming 陸德, Jia Gongyan 賈公彥, Ruan Yuan 阮元 (1965).
Chongkan Songben Zhou li zhu shu fu jiaokan ji 重刊宋本周禮注疏附校勘記.
Taibei shi台北市, Yiwen inshu guan 藝文印書館. Scripta Sinica electronic