Eric L. Hutton
Editor
Dao Companion to the
Philosophy of Xunzi
Editor
Eric L. Hutton
Department of Philosophy
University of Utah
Salt Lake City, UT, USA
Dao Companions to Chinese Philosophy
ISBN 978-94-017-7743-8
ISBN 978-94-017-7745-2
DOI 10.1007/978-94-017-7745-2
(eBook)
Library of Congress Control Number: 2016953488
© Springer Science+Business Media B.V. 2016
This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of
the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation,
broadcasting, reproduction on microfilms or in any other physical way, and transmission or information
storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology
now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this book
are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the
editors give a warranty, express or implied, with respect to the material contained herein or for any errors
or omissions that may have been made.
Printed on acid-free paper
This Springer imprint is published by Springer Nature
The registered company is Springer Science+Business Media B.V. Dordrecht
Abbreviations
In order to facilitate consistency of references across chapters and to aid readers
in locating passages from the Xunzi, the following abbreviations for references are
observed throughout this volume.
For the Chinese text of the Xunzi:
HKCS Lau, D.C. ࢹ⇯⡥, and F.C. Chen 䲣ᯩ↓, eds. 1996. A Concordance to
the Xunzi 㥰ᆀ䙀ᆇ㍒ᕅ. Hong Kong: The Commercial Press ୶उঠᴨ
佘. Cited according to the numbering system used in the concordance:
chapter number/page number/line number(s).
(Note: Not all authors in this volume follow the exact edition of the text
given in this concordance, so the listing of these numbers should not be
taken as an endorsement of that edition on their part but is rather primarily for reference purposes.)
For English translations of the Xunzi:
H
Hutton, Eric. 2014. Xunzi: The Complete Text. Princeton: Princeton
University Press. Cited as: page number, or page number.line number.
K
Knoblock, John. 1988–94. Xunzi: A Translation and Study of the Complete
Works, 3 volumes (vol. 1: 1988, vol. 2: 1990, vol. 3: 1994). Stanford:
Stanford University Press. Cited as: volume number in Roman numerals.
page number, sometimes followed by further reference given as chapter
number.paragraph number per Knoblock’s translation.
W
Watson, Burton. 2003. Xunzi: Basic Writings. New York: Columbia
University Press. Cited as: page number.1
1
Note: the pagination of this edition differs slightly from the earlier 1963 edition of Watson’s
translation.
xi
Chapter 1
Style and Poetic Diction in the Xunzi
Martin Kern
The Xunzi is widely recognized as a book of well-developed expository prose,1 even
though its literary style has been called, perhaps unfairly, “at best . . . indifferent” (Lau
1970: 8). Unlike other texts of early Chinese philosophy—Lunyu, Mengzi, Zhuangzi,
Mozi, and others more—it does not pervasively use anecdotes or dialogues to stage or
create its purported or real author as a particular persona (Lewis 1999: 58); its principal form is that of the discursive essay.2 Instead of appearing as a speaker in thirdperson anecdotes the way Kongzi, Mozi, or Mengzi do (and have their personas
created through these anecdotes), the Xunzi’s expository prose speaks from the perspective of “arguing for his ideas and against his opponents” (Denecke 2010: 180).3 In
the third century BCE, the Xunzi thus participated in, and contributed to, the rise of the
expository essay that can also be observed elsewhere, e.g., in parts of the Zhuangzi.4
One must be careful, however, not to overemphasize the text’s authorial voice as a
1
Paul Rakita Goldin expresses a common attitude toward the text: “Xunzi’s writing is succinct and
lucid, his philosophical positions original and reasoned” (Goldin 1999: xiii).
2
The only chapters that include mention of Xunzi (in the form of SUN Qing ᆛয, SUN Qingzi
ᆛযᆀ, or XUN Qingzi 㥰যᆀ) are 8 (“Ru xiao” ݂᭸), 15 (“Yi bing” 䆠)ޥ, and 16 (“Qiang guo”
ᕺ഻). In addition, the “Eulogy” (Knoblock) on Xunzi that may have come from a disciple or later
scholar and is appended to the final chapter 32 (“Yao wen” ) mentions him as SUN Qing.
Because they speak of XUN Kuang in the third person, these chapters are often taken as the works
of his disciples.
3
See also William G. Boltz: “[L]iterary or essay-like texts, authored by a single writer, in the way
we typically think of a text in the modern world, do not reflect the norm for early China but were,
at best, the exception” (Boltz 2005: 59).
4
Denecke might be overstating the case for the Xunzi when noting that its “new rhetorical format,
the expository essay, constituted a radical change, an innovation that was to fundamentally alter the
face of Masters Literature” (Denecke 2010: 180).
M. Kern (*)
Princeton University, Princeton, NJ, USA
e-mail: mkern@princeton.edu
© Springer Science+Business Media B.V. 2016
E.L. Hutton (ed.), Dao Companion to the Philosophy of Xunzi,
Dao Companions to Chinese Philosophy 7, DOI 10.1007/978-94-017-7745-2_1
1
2
M. Kern
personal one;5 in many instances, the seemingly first-person pronoun is not a firstperson pronoun at all but a general one that should be understood as “you” or “one”
(Harbsmeier 1997: 181–220). Either way, the expository chapters of the Xunzi reflect
a discursive and sometimes even combative style of argument that straightforwardly
addresses a series of topics and—exceptional in early Chinese rhetoric—does so with
“mundane pugnaciousness” (Harbsmeier 2001: 883). The individual chapters of the
received text—arranged first by LIU Xiang ࢹੁ (79–8 BCE) and then by YANG Liang
ὺَ (ninth century)—appear as separate monographs on a range of issues, even
though they are rarely, if ever, coherent from beginning to end.6
By examining in some detail a certain number of representative chapters, the
present essay argues for a more nuanced appreciation of the Xunzi’s style; specifically, by tracing the considerable stylistic differences between individual chapters,
it calls the idea of a single homogeneous “Xunzi style” into question. From the
perspective of style, the text emerges more as an anthology of varied writings of
Xunzian thought—if we admit to an overall philosophical coherence of the text—
than as a unified work. Thus, we may either allow that XUN Kuang 㥰⋱ employed
a considerable range of distinctive styles in his writings, or we may need to reconsider and broaden our ideas about the authorship of the Xunzi (or both). While the
present essay is not the place to address questions of authorship and authenticity, it
still offers observations that might be useful to any such discussion.
To begin with, the core of the Xunzi is considered to comprise chapters 1 through
24; by contrast, the final eight chapters seem considerably different in nature. As
scholars attribute the first twenty-four chapters to XUN Kuang and implicitly assume
their authorial unity and coherence, the later chapters have often engendered serious
doubt.7 The chapters in question include two separate sets of poetry (chapters 25
and 26, “Cheng xiang” ᡀ and “Fu” 䌖);8 one chapter of (in Knoblock’s count)
115 brief maxims (chapter 27, “Da lüe” བྷ⮕), four chapters of Kongzi lore in the
style of the Lunyu 䄆䃎 (chapters 28–31, “You zuo” ᇕ, “Zi dao” ᆀ䚃, “Fa xing”
⌅㹼, and “Ai gong” ૰)ޜ, yet with just a single brief parallel in the received Lunyu;9
5
Here, I disagree with Denecke’s analysis as well as with Knoblock’s translation.
In his introduction, Knoblock offers an extensive discussion of the textual history of the text
(K I.105–28).
7
In this respect, the Xunzi is not different from many, if not most, other texts of the early philosophical tradition.
8
For a detailed analysis of the fu poems see Knechtges (1989: 1–31); for a brief discussion of both
the “working songs” (cheng xiang ᡀ) and the fu, see Denecke (2010: 188–95); for the “Cheng
xiang” chapter alone, see Malmquist (1973b: 63–91) and Malmquist (1973a: 352–58). Scholars
disagree as to whether the content of these chapters is “philosophical” (Knoblock) or not
(Knechtges). For discussions of rhyme in the “Cheng xiang” chapter, see Li (2010: 89–93); Zhu
(1957: 42–47).
9
That parallel is the brief phrase zhi zhi yue zhi zhi ⸕ѻᴠ⸕ѻ (Knoblock 29.6) that appears as
zhi zhi wei zhi zhi ⸕ѻ⛪⸕ѻ in Lunyu 2.17 (“Wei zheng” ⛪᭯). By contrast, the four Xunzi
chapters of Kongzi lore have numerous parallels especially in Hanshi waizhuan 七∿ཆۣ and
Kongzi jiayu ᆄᆀᇦ䃎, and to some lesser degree in Da Dai Liji བྷᡤ䁈 and Shuiyuan 䃚㤁.
6
1
Style and Poetic Diction in the Xunzi
3
and a final chapter 32 (“Yao wen” ) that contains anecdotal lore regarding both
Kongzi and other early culture heroes. These eight chapters are extremely diverse,
with the two “poetry chapters” in both form and content showing clear affinities to
the southern fu style associated with the Chuci ᾊ䗝 (Lyrics from Chu) while also
being related to Warring States and early Han shui 䃚 (“attempts at persuasion”)
(Knechtges 1989: 21–31). It may well be for this reason—and especially for the
topos of the frustrated man bu yu н䙷 (“not meeting his time”) when the world is
morally corrupt and in a perverted state (Knechtges 1989: 21–31)—that in the bibliographic chapter of the Hanshu ╒ᴨ, QU Yuan ቸ (trad. 340–278 BCE) and
Xunzi are presented as the two originators of the fu genre. According to the account
preserved in the Hanshu, the genre at once originated with and culminated in the
works of these two authors, descending into a quick decline immediately thereafter
(Hanshu 1987: 30.1750, 1756).10
Despite its title, the present analysis will focus on the first twenty-four chapters
in the Xunzi, leaving the “Cheng xiang” and “Fu” chapters aside together with those
that follow them. There are several reasons for this decision. To begin with, the
heterogeneous nature and multiplicity of styles in these chapters has long been
acknowledged. One would be hard pressed to argue that the “Cheng xiang” and
“Fu” chapters belong to the core of the Xunzi. It was for sound reasons that LIU
Xiang had relegated them to the end of his Xunzi compilation (K I.106–10), and
even YANG Liang, who called them za 䴌 (“miscellaneous”), placed them in the
uneasy position between what he considered the authentic writings by XUN Kuang
and the additional body of material (chapters 27–32) that he attributed to later disciples (K I.112).11 Compared to the “discursive” Xunzi of chapters 1–24, the two
“poetry” chapters seem curiously out of place, and their designations as “Cheng
xiang” and “Fu” are dubious: while “Cheng xiang” is simply taken from the first
line of the chapter and obscure in its meaning (K III.169),12 the designation “Fu” did
not originate with XUN Kuang but was quite possibly chosen by LIU Xiang
(Knechtges 1989: 14–15).13 Moreover, it appears that the two “poetry” chapters
were not part of the Xunzi before LIU Xiang, or they were considered entirely marginal: unlike the majority of Xunzi chapters, they do not have any parallels in
Western Han literature save for a single snippet from the “Fu” chapter that is quoted
in the Zhanguoce ᡠ഻ㆆ (K I.105).14
10
The Hanshu bibliographic monograph “Yiwenzhi” 㰍᮷ᘇ, which in abbreviated form represents the catalogue of the imperial library at the end of the first century BCE, attributes twenty
pieces of fu to Xunzi.
11
The arrangement of presumed “inauthentic” material at the end of a Masters text is, of course,
standard procedure and hence expresses unambiguously what both LIU Xiang and YANG Liang
thought of the closing chapters.
12
For a more contextualized discussion of the term cheng xiang, see Kern (2003: 407–9).
13
Simply put, there was no literary “genre” called fu 䌖 in XUN Kuang’s time; see Kern (2003:
391–95).
14
The quote is in the chapter “Chu ce si” ᾊㆆഋ, “Ke shui Chunshen jun” ᇒ䃚᱕⭣ੋ.
4
M. Kern
The more interesting and more important reason to focus on chapters 1–24, however, is a different one: “poetry” in the Xunzi is not simply what can be found in
chapters 25 and 26. Just as in many texts of expository zhuzi 䄨ᆀ (“masters”) prose
from the Warring States, it is not a certain body of text bearing the stylistic distinctions of a particular genre; instead, it is a mode of speech, or language use, that
deeply pervades what is usually taken as “expository prose.”15 This mode of speech
is ubiquitous in the Xunzi. To some extent, though not nearly sufficiently, Knoblock’s
translation shows efforts to identify those “poetic” passages that are distinguished
by rhyme and meter. While still inadequate, this effort marks an important step forward, considering that scholarship on early Chinese philosophy and intellectual history has only recently begun to attend to linguistic form as important for thought
and argument.16 The occasional and sometimes even pervasive poetic style in the
Xunzi, and in so many other early Chinese texts of expository prose, is neither a
“genre” nor some sort of superficial, external embellishment of reasoned discourse
(let alone impediment to logical expression and interpretation, or, as Angus C.
Graham has noted for the Zhuangzi, a “collision of logic and poetry”) (Graham
1991a: 214). Instead, this style of diction is also an intellectual style. It is the very
medium through which large parts of the argument operate in the Xunzi, and as such
it fulfills—as style in any rhetorical tradition—functions of persuasion and even of
what philosophers are wont to call illocutionary force. Simply put, for the Xunzi and
other early Chinese texts, to speak in verse is to speak in the voice of traditional
authority and of an emphatic claim for truth.
Precisely because the Xunzi is considered a text driven by the desire for arguing,
and because that arguing is not simply logical or analytic, its literary style is central
to the quality not merely of its form of expression but of its argument itself.
Linguistic rhythm itself, as the youshui 䙺䃚 (“persuaders”) of the Warring States
period knew very well, carries a stylistic type of persuasive power by its mere formal structures of parallelism and repetition; this is especially true for a style of
15
Günther Debon has repeatedly pointed to the presence and significance of rhymed sayings
(“Spruchdichtung”) in early Chinese expository prose; see Debon (1996: 36–42); on rhymed sayings especially in the Xunzi, see Debon (2002 vol.1: 21–30).
16
In discussing the fallacies of the “rambling mode” in translations of the Zhuangzi, A. C. Graham
has offered the most cogent critique of negligence toward the poetic features of early “expository
prose,” summarized in the sentence “The effect of assimilating the verse to prose is almost always
catastrophic”; see Graham (1991b: 119–44, esp. 130–43). The groundbreaking work on rhyme in
early expository prose is Jiang (1993); see also Long (1962–63, repr. 2009: 182–283), and Tan
(1995: 12–19). The gradually increasing body of scholarship on the formal aspects of early Chinese
philosophical texts includes Rudolf G. Wagner’s analysis of “Interlocking Parallel Style,” see
Wagner (2000: 53–113), Dirk Meyer (2011), Raphals (1994: 103–16), Roth (1999), Queen (2008:
201–47), Baxter (1998: 231–53), Fischer (2009: 1–34), Boltz (2005: 50–78), Liu (1994), LaFargue
(1994); see also Kern (2014), Xu (1990: 58–64), and Morrison (1981: 391–420). Aside from
Knoblock’s translation of the Xunzi, Hutton’s new Xunzi translation also marks off the poetic parts
of the text, as does the translation of the Huainanzi by John S. Major et al. (Major et al. 2010).
1
Style and Poetic Diction in the Xunzi
5
arguing, ubiquitous in the Xunzi, that is built around analogical patterns and in this
conveys a strong sense of order.17 It is therefore that the present essay focuses on the
use of the poetic style in the discursive core chapters of the Xunzi.
Of course, concerns about the Xunzi’s style go much beyond “poetry” as a particular type of language use. What the text lacks in literary flourish (especially by
comparison with the Zhuangzi) or historical anecdotes (when compared to many
other early philosophical texts), it often gains in focus and stringency of argument,
as the individual chapters do tend to focus on their respective subject matter at least
for their larger parts. This relative stringency, combined with an explicit diction that
rarely uses esoteric anecdotes and elliptic sayings, makes the Xunzi into a text that
is relatively easy to follow; it grants few of the pleasures of reading the Zhuangzi but
also provokes none of the frustrations the Lunyu stirs in readers hoping to decipher
the meaning of some particular passages (not to mention their position within a
larger philosophical context). While scholars have considerable difficulty in situating the Xunzi’s philosophy between “Confucian” and “Legalist,” occasionally
resorting to phrases such as “realistic” or “authoritarian Confucian” (Rickett 1985:
3, 249, 412), they do not face the sort of wide-ranging diversity of thought that
forces them to assign any of the twenty-four core chapters to different authors or
“schools of thought.” In short, the first twenty-four chapters of the Xunzi are commonly taken as mutually supportive and non-contradictory, expressing different
aspects of a single coherent system of thought; their sometimes divergent viewpoints have been rationalized as coming from distinct periods (early, middle, late)
of XUN Kuang’s long life.18
On the whole, it also appears that while the text draws on a considerable amount
of traditional source material, including numerous pieces of proverbial wisdom and
rhetoric (K I.124–28), there is no direct evidence that it is pieced together from
materials common to a wider range of texts. In Knoblock’s view, the relatively large
number of sections that the Xunzi shares with the Hanshi waizhuan, the Da Dai Liji,
and the Liji is by and large the result of these texts borrowing from the pre-LIU
Xiang Xunzi material, and not the other way around (K I.105–6). By implication,
the Xunzi is then generally seen as (a) having existed in more or less its present form
before the early Han and (b) being not a composite or compiled text but a truly
authored and original one.19 These widely shared assumptions do not imply that the
17
Here, I allude to Ernst H. Gombrich for the power of formal structure in argument (Gombrich
1979); see also Bagley (1993: 34–55). For early Chinese rhetoric, see further Schaberg (2001:
21–56). As Schaberg observes, the “rhetoric of good order” applies to both speech and written
prose; I would add that expository prose with its implied authorial voice is indeed built upon the
techniques of persuasion that are first visible in discursive speech.
18
See Knoblock’s appendices “Composition of Each Book” in each of his three volumes.
19
On “composite texts” as a common phenonemon during the time of the Xunzi, see Boltz (2005:
50–78). In saying “truly authored and original,” I do not overlook that many scholars (e.g.,
Knoblock) have noticed what they call “borrowed” elements in the text. But to consciously borrow
existing language is an intense form of authorship as it implies thoughtful and intentional decisions
on the side of the author.
6
M. Kern
book was initially devised as a grand, unified whole as was the case with part of the
Lüshi chunqiu (dated in part to 239 BCE) and the entire Huainanzi (139 BCE); there
is evidence that at least some Xunzi chapters, or individual paragraphs, existed independently from their present context in the book. In fact, we have no reason to
assume that XUN Kuang thought of his writings as constituting a “book.” Any
attempt to see a particular order in the arrangement of the existing chapters is
defeated by the fact that the received Xunzi represents YANG Liang’s re-arrangement
of LIU Xiang’s earlier compilation, which in turn was not the “original Xunzi”
designed by XUN Kuang himself—a thing that most likely never existed in the first
place. By necessity, the object of our analysis is the received text, with at least some
of its chapters being internally in disarray. It may well be that some of the stylistic
incoherence is the result of textual deterioration at an early stage, perhaps comparable to what happened, say, to the “Ziyi” ㏷㺓 (“Black Robes”) text where the
received Liji 䁈 version is decidedly inferior to the two manuscript versions from
Guodian 䜝ᓇ and in the Shanghai Museum corpus, both dating from around 300
BCE.20
Be this as it may, it remains significant to observe distinct differences of style
between and within the individual chapters of the text. In some brief but illuminating comments on chapter 1, Knoblock has argued that the first seven sections of the
chapter (in his numbering, equaling roughly half of the chapter) are replete with
traditional material that “is widely attested in other works dating from Xunzi’s time
and later” while sections eight through fourteen are “mostly the original composition of Xunzi and as such [are] much more rarely ‘quoted’” (K I.124).21 Such a
conclusion should be phrased more carefully: whether or not the second half of the
chapter is indeed “the original composition of Xunzi” is, in fact, impossible to
decide. What cannot be disputed, on the other hand, is the fact that by and large, the
latter half of the chapter is not shared with other texts, while the first half overwhelmingly is. Why? To some extent, the answer to this question may be found in
the analysis of style—and such an analysis further reveals that the two halves of the
chapter have little in common and perhaps should not be conceived of as an integrated whole.
In the present essay, I examine chapter 1 in some detail. This chapter shows significant stylistic features one also finds elsewhere in the Xunzi. Following this analysis, I comment briefly on specific features in several other chapters that are
generally considered of central importance to the text as a whole. Whenever I quote
from the original, I arrange the text in a way that reveals its formal structures.
20
See Kern (2005: 293–332), and Kalinowski (2000–01: 141–48).
In the present essay, I do not always follow the divisions into sections as given in the CHANT
version (which is also largely coherent with Knoblock’s division). I indicate where I differ from
CHANT or Knoblock.
21
1
Style and Poetic Diction in the Xunzi
7
HKCS 1/1/3–5:
ੋᆀᴠ: ᆨнਟԕᐢDŽ
A junzi says: “In learning, one must not desist.”
䶂ǃਆѻᯬ㯽, 㘼䶂ᯬ㯽;
ߠǃ≤⛪ѻ, 㘼ሂᯬ≤DŽ
Blue is taken from the indigo plant, yet it is bluer than indigo.
Ice is made from water, yet it is colder than water.
ᵘⴤѝ㒙, 䕞ԕ⛪䕚, ަᴢѝ㾿, 䴆ᴹ᳤, нᗙᥪ㘵, 䕞֯ѻ❦ҏDŽ
A piece of wood as straight as a chalk line can be rounded [by steaming]
to become a wheel; its curvature [will then] conform to the compass.
Even when dried in the sun, it will not return to its [former] straightness.
The process of rounding by steam has caused it to be like that.
᭵
ᵘਇ㒙ࡷⴤ,
䠁ቡ⽚ࡷ࡙DŽ
Thus [it is said]:
If wood is aligned to the chalk line, it will be straight;
if metal is put to the whetstone, it will be sharp.
ੋᆀ
ঊᆨ㘼ᰕ৳ⴱѾᐡ,
ࡷ
⸕᰾㘼㹼❑䙾⸓DŽ
If the junzi
studies broadly and daily inspects himself on three counts,22
his understanding will be clear and his conduct without transgression.
22
My translation follows YANG Liang’s commentary and the parallel in Lunyu 1.4 (“Xue er” ᆨ㘼):
“Zengzi said: I inspect myself daily on three counts” (ᴮᆀᴠ:Nj੮ᰕйⴱ੮䓛nj); later commentators on the Xunzi have interpreted the word can ৳ (*N-sʕrum) not as san й (*srum; “on
three counts”) but as yan 傇 (*m-qʰr[a]m-s; “to examine”) and have further argued that the two
characters xing hu ⴱѾ (“inspect” plus directional preposition “at”) are an interpolation. Thus,
Knoblock translates as “the gentleman each day examines himself” (135). I see no need for this
emendation, nor can I think of a good explanation for the purported interchangeability of can and
yan.
8
M. Kern
The passage begins with the quotation of what David Schaberg has called a
“platitude persona” (Schaberg 2005: 177–96), namely, the figure of the anonymous
and unspecified junzi ੋᆀ that appears also in numerous other texts of the time and,
as Christoph Harbsmeier has noted, in general does not refer to any specific individual.23 The quotation of his saying “In learning, one must not desist” is a gesture
toward tradition: whoever the author of the chapter is, his opening words are not in
his own voice but draw on pre-existing authority that, furthermore, is not located in
a particular person but in a generalized junzi. His statement of received learned
opinion is then followed by two sets of analogies: the first, on blue/indigo and
ice/water is taken from the natural world; the second, and much more extensive one,
is from the realm of craftsmanship that also figures prominently in the rhetoric of
other early philosophical prose (De Reu 2010; Major 2014). Following the second
of these analogies, the text returns to a general, indeed apodictic, statement on the
matter of “learning”: “If the junzi studies broadly and daily inspects himself on
three counts, his understanding will be clear and his conduct without
transgression.”
In this sequence, the analogies in the middle part lead from the initial piece of
traditional wisdom toward a statement on learning as transformation of the self by
regular exercise of self-examination. The middle part is not built on explicit deductive logic but rather on the implicit inference from analogies, reinforced by repetition and parallelism, that by mere accumulation generate some rhetorical force. The
first analogies of blue/indigo and ice/water, for example, are ready-for-use, disposable items from the general store of rhetorical analogies; the second analogy—the
wood bent by steam and then remaining bent even when dried again—is a more
original comparison to a person’s permanent transformation by learning. It is followed by gu ᭵, an introductory sentence adverbial that often does not have a strong
logical force (as in “therefore”), as it does not function as the hinge between the
immediately preceding sentence or section and the subsequent one (Gassmann and
Behr 2005 vol.1: 96). Instead, it frequently serves as the introduction of another
piece of traditional wisdom: what follows gu (which I translate as “thus [it is said]”
to indicate that the following is again a quotation or otherwise marked speech)24 is
a general maxim, usually bound by rhyme or rhythm, that is supported by the preceding illustration. Here it is important to remember that we are not in the style of
deductive reasoning: while the maxim (in this case not rhymed, but a formulaic
couplet governed strictly by parallelismus membrorum) picks up the analogy of
wood, it actually takes it into the opposite direction. Now, wood is not bent but
straightened, because this is how it is parallel to the knife that is sharpened. Finally,
the text returns to the junzi but not necessarily to the one who was quoted in the
beginning, and his theme is not—as in the initial proverb-style saying—learning
that shall never end; instead, the focus is now on the regular practice of selfexamination by which the junzi will permanently transform himself. However, for
23
24
Personal communication, January 2012.
On different types of quotation (and pertinent references), see Kern and Hunter (forthcoming).
1
Style and Poetic Diction in the Xunzi
9
the analogy to wood to operate properly, one would have to assume that the junzi
acts on his inner self as on an object distinct from the examining mind: wood neither
examines itself nor bends or straightens itself; it is acted upon so as to become permanently changed.
In short, this sequence alone includes five sections that are all (a) mutually independent and (b) formed in different ways. Any of these sections could easily be
integrated into different contexts, and one might well want to ask how many of them
are original to the Xunzi.25 What holds the section together is the fact that it is
framed by two maxims associated with the junzi and his learning, even though these
maxims emphasize different aspects of his self-cultivation.
In the conventional division of the text,26 the first section ends right here. This
would be fine if it would not mean that the next section then had to start with another
gu—the sentence adverbial to introduce a concluding commonplace. Yet the reasoning for starting a new section here is not implausible: the following lines seem,
at best, only loosely related to the preceding text. Thus, the section introduced by gu
may indeed not belong to the first section—but in this case, it may be altogether
misplaced, or whatever may have preceded it originally is now lost. On the other
hand, one might argue that its weak connection to the preceding text in the current
version of the Xunzi is symptomatic of the entire first section which, as just shown,
is altogether loosely integrated and possibly assembled from various bits and pieces.
If the first section in its present form is indeed one author’s original composition, it
does not show him overly concerned with the cogency of his argument—or he relies
on an audience of insiders capable of generating from his style a stringent line of
thought.
Be this as it may, what follows the second gu is another piece of traditional
wisdom strictly organized by syntactical patterning:
HKCS 1/1/7–10:
᭵
нⲫ儈ኡ, н⸕ཙѻ儈ҏ;
н㠘␡䉯, н⸕ൠѻҏ;
н㚎⦻ݸѻ䚪䀰, н⸕ᆨѻབྷҏDŽ
Thus [it is said]:
If one does not climb a high mountain, one does not understand the height
of heaven.27
25
Here, I differ from Knoblock’s assessment which is purely based on the comparison of the Xunzi
with other transmitted sources. To say that these sources—especially the Da Dai Liji and the
Hanshi waizhuan—seem to be quoting from the Xunzi and not vice versa is not the same as saying
that whatever they quote did actually originate with the Xunzi.
26
As reflected in the CHANT edition as well as in Knoblock’s translation.
27
Clearly, the metaphor refers to the “height of the sky,” yet at the same time, the two sentences
here invoke the “heaven/earth” cosmology.
10
M. Kern
If one does not look down into a deep valley, one does not understand the depth
of the earth.
If one does not hear the words left by the former kings, one does not
understand the greatness of learning.
Ҿǃ䎺ǃཧǃ䊹ѻᆀ, ⭏㘼਼㚢, 䮧㘼⮠؇, ᮉ֯ѻ❦ҏDŽ
As for the children of Yu, Yue, Yi, and Mo: at birth they make the same sounds;
growing up, they differ in their customs. Education causes them to be like that.
䂙ᴠ:
ఏ⡮ੋᆀ,
❑ᙶᆹDŽ
䶆ޡ⡮ս,
ྭᱟ↓ⴤDŽ
⾎ѻ㚭ѻ,
ӻ⡮Ჟ⾿DŽ
(*-ə)
(*-ək)
(*-əp)
(*-ək)
(*-ə)
(*-ək)
An Ode says:
Ah, you noble men,
do not consider permanent your being at rest and at ease.
Be reverent and respectful of your positions,
be fond of those who are upright and straight.
Exert [your inner] spiritual capacity and adhere to it,
to increase your radiant blessings.
⾎㧛བྷᯬॆ䚃,
⾿㧛䮧ᯬ❑⾽DŽ
Among one’s spiritual capacities, there is none greater than the way
of transformation.
Among blessings, there is none more enduring than being without misfortune.
Once again, the passage is structured by rhythm and semantic parallelism;28 the only
rhymes are in the quotation of the final stanza from Ode 207, “Xiao ming” ሿ᰾
(“Lesser Brightness”). The initial passage following gu (“Thus [it is said]”) has no
discernable connection to anything before or after except for its praise of learning
from the words of the ancient sages. This, however, is then followed by an analogy
that resonates closely with the earlier metaphor of wood that is permanently bent.
28
The absence of rhyme, however, does not mean the absence of poetry; Baxter notes that “both
rhyme and semantic patterning,” especially including semantic parallelism, “are used as poetic
devices” in the Laozi (Baxter 1998: 237).
1
Style and Poetic Diction in the Xunzi
11
In fact, the lines that conclude the earlier analogy of bent wood and the later one of
acquired customs are strictly parallel:
䕞֯ѻ❦ҏDŽ
The process of rounding by steam has caused it to be like that.
ᮉ֯ѻ❦ҏDŽ
Education causes them to be like that.
If anything, this direct and likely not accidental piece of parallelism suggests that
the two parts of text do belong to a single section, even though there is additional
material in it. In other words, the strict parallelism signals both unity and, perhaps,
the addition of formerly unrelated material to that unity of expression. What follows
the second analogy—the quotation from the Odes as well as the final statement on
one’s spiritual capacity and blessings—is only partially related to the main theme of
cultivation through learning, namely, in its reference to transformation and possibly
also to being zhi ⴤ (“straight”), but not at all in its reference to blessings. In its strict
parallelism, the final statement appears once again as some sort of proverb and was
possibly independent from the Odes quotation with which it is paired here. In sum,
the first section, or sections, do not develop a cogent argument; instead, they embellish the principal thesis on the lasting influence of learning with various pieces of
traditional wisdom culled from different sources.
Consider now the second (or third) section of the first chapter:
HKCS 1/1/12–15:
੮ే㍲ᰕ㘼ᙍ⸓, нྲ丸㠮ѻᡰᆨҏDŽ
੮ే䏲㘼ᵋ⸓, нྲⲫ儈ѻঊ㾻ҏDŽ
I once spent the whole day thinking, but it was not as good as what I learned in
an instant.
I once stood on my toes gazing into the distance, but it was not as good as what
I broadly saw after ascending a place on high.
ⲫ儈㘼ᤋ, 㟲䶎࣐䮧ҏ, 㘼㾻㘵䚐;
丶付㘼બ, 㚢䶎࣐⯮ҏ, 㘼㚎㘵ᖠDŽ
By ascending a place on high and waving, the arm does not gain in length,
yet its visibility reaches into the distance;
By shouting with the wind, the voice does not gain in strength, yet its audibility
becomes more distinct.
ٷ䕯俜㘵, 䶎࡙䏣ҏ, 㘼㠤ॳ䟼;
ٷ㡏ᾛ㘵, 䶎㜭≤ҏ, 㘼㎅⊏⋣DŽ
12
M. Kern
To make use of carriage and horses is not to benefit one’s feet but
to go a thousand li;
To make use of boat and oars is not to gain ability with water but to cross rivers
and streams.
ੋᆀ⭏䶎⮠ҏ, ழٷᯬ⢙ҏDŽ
The junzi is not different by birth; he is good at availing himself
of external things.
This section consists of three statements, each composed of two parallel halves, and
a concluding pronouncement on the junzi. Each such prose “couplet” has its own
meter and rhythm, yet all three are unified in their extreme, mechanistic parallelism;
and one leads to the next through the continuous use of a key phrase (deng gao
ⲫ儈 in the transition from the first to the second statement, and the negative fei 䶎
from the second to the third). Each statement is an illustration of being “good at
availing oneself of external things”; all three are then summarized in the statement
on the junzi. What we see here, as before, is an accumulation of examples, a brief
catalogue of mutually independent units.
Looking back at the first two (or three) sections discussed so far, the recurring
element is the mention of the junzi:
1. A junzi says: “In learning, one must not desist.”
2. If the junzi studies broadly and daily inspects himself on three counts, his
understanding will be clear and his conduct without transgression.
3. The junzi is not different by birth; he is good at availing himself of external
things.
Indeed, if there is a discernable theme in the beginning of the Xunzi, together
with the emphasis on xue ᆨ (“learning”), it is the concern with the ideal of the
junzi:29 a person whose status is not inherent or inherited, but earned through effort
and the ability to act upon himself and to draw on external things. Importantly, this
ideal is an attainable choice—and hence can be argued for by way of persuasive
rhetoric. Strictly speaking, everything beyond the three statements on the junzi is
dispensable in the sense that any part of it could be dropped or replaced by something different. In these cumulative sections, the Xunzi does not develop an explicit
deductive argument; rather, the text pronounces itself three times on the junzi and
then, in seemingly random order, fills its columns with illustrations and formulaic
pieces of traditional wisdom. Strikingly, none of these pieces—and nothing in the
opening passages—involves the style of historical anecdote one is accustomed to
29
Goldin sees this as the theme of the entire Xunzi: “The overarching preoccupation that binds
together the diverse arguments and reflections in the text is the role of the noble man” (Goldin
1999: xi).
1
13
Style and Poetic Diction in the Xunzi
read in other early writings of expository prose, nor is there a single historical reference
to anyone. The text here is not organized by chains of deductive arguments and
conclusions but also not by the logic of narrative; nor is it in any way adjusted to any
sort of historical context. This characteristic is true for much of the Xunzi and shared
with a text like the Laozi 㘱ᆀ, but not with most other Warring States writings,
including most of the recently excavated manuscripts of expository prose.30
The fourth (or fifth) section of “Exhortation to Learning” offers yet a different
way of traditional discourse, namely, the use of rhyme that is found in many passages of the Xunzi (Debon 2002): without any introduction, this section is composed
of tetrasyllabic lines and almost entirely rhymed, invoking the formal patterns of
the Odes. The passage falls neatly into four sections of four lines each, which are
distinguished by particular syntactic structures, further emphasizing the divisions
already marked by rhyme change. These brief sections are four variations on a
common theme:31
HKCS 1/2/3–5:
⢙于ѻ䎧,
ᗵᴹᡰDŽ
῞䗡ѻֶ,
ᗵ䊑ަᗧDŽ
*-ə
*-ə
*-ək
*-ək
A
A
B
B
As the categories of things arise,
They always have something from which they begin.
As honor and disgrace arrive,
They always are manifestations of [the person’s] virtuous power.
㚹㞀ࠪ㸢,
冊ᷟ⭏㹩DŽ
ᙐធᘈ䓛,
⾽⚭ѳDŽ
*-uŋ
*-ak
*-iŋ
*-ak
x
C
x
C
Meat that is rotten brings forth worms,
Wood that is withered produces grubs.
When neglecting the self by being lazy and indolent,
Misfortune and disaster will arise.
30
“Still another characteristic, finally, which separates the Lao-tzu from much of early Chinese
philosophical discourse, is that it is entirely free of narration, in the sense that its statements are
general and not anchored to any particular persons, times, or places. There is no indication of who
is speaking, no direct reference to historical events. This contrasts strikingly with typical Confucian
discourse” (Baxter 1998: 240).
31
In my representation of the rhyme pattern, the small letter “x” represents a non-rhyming line. My
simplified representation of the rhymes is derived from William H. Baxter and Laurent Sagart,
Baxter-Sagart Old Chinese Reconstruction (Version 1.00), online at http://crlao.ehess.fr/document.php?id=1217. Accessed January 23, 2012.
M. Kern
14
ᕧ㠚ਆḡ,
Ḅ㠚ਆᶏDŽ
䛚を൘䓛,
ᙘѻᡰΏDŽ
*-o
*-ok
*-iŋ
*-o
D
D
x
D
What is strong gives itself as support,
What is soft gives itself for bundling.
When vileness and depravity reside in a person,
Resentment is what he brings upon himself.
ᯭ㯚㤕а,
⚛ቡ⠕ҏ,
ᒣൠ㤕а,
≤ቡ⓬ҏDŽ
*-it
*-aj
*-it
*-aj
E
F
E
F
Where firewood is spread out evenly,
Fire will seek out the driest.
Where the ground is leveled evenly,
Water will seek out the dampest.
㥹ᵘ⮷⭏,
⦨㗔✹,
⢙ᗎަ于ҏDŽ
*-eŋ
*-an
*-aj
x
x
x
As grasses and trees grow together with their kind,
As birds and beasts form flocks,
Each thing accords to its own category.
Taken together, these rhythmic and euphonic sections offer a series of analogies that
illustrate the principle of sympathetic resonance in the natural world: because things
respond to one another according to their lei 于 (“categories”) of natural disposition, actions have their specific and inevitable consequences. Rhetorically, the four
sections contain what seem to be snippets of conventional wisdom. They are persuasive for two reasons: as observations of the natural world and by the force of
sheer accumulation that amounts to a veritable catalogue of phenomena of natural
resonance. The altogether eighteen tetrasyllabic lines are capped with a concluding
statement of six characters that offers, by a process of induction, the abstraction of
the principle illustrated: “each thing accords to its own category.” From here, the
text moves closer to its conclusion, beginning with a summarizing shi gu ᱟ᭵
(“and for this reason”) that leads to yet another set of analogies on the same theme
before concluding with a three-line pronouncement introduced again by gu ᭵
(“thus”) that caps the entire fourth (or fifth) section of the chapter. At this point, the
1
Style and Poetic Diction in the Xunzi
15
text maps the social world onto the natural, claiming that we can choose our actions
but then cannot control their “natural” and therefore inevitable and predictable
consequences. The junzi must be cautious in speech and action because his behavior
may attract calamity according to the same principle of resonance that governs the
preceding analogies from the natural world:
HKCS 1/2/5–7:
ᱟ᭵
䌚Ⲵᕥ㘼ᕃ⸒㠣✹;
᷇ᵘ㤲㘼ᯗᯔ㠣✹;
ᡀ㭝㘼匕✹DŽ
䟟䞨㘼㲻㚊✹DŽ
And for this reason [it is said]:
Where the archery target is set out, bows and arrows will arrive;
Where the forest woods are flourishing, axes and halberds will arrive;
Where trees provide for shade, numerous birds will rest;
Where [things turn] sour and acid, gnats accumulate.
᭵
䀰ᴹᤋ⾽ҏ,
㹼ᴹᤋ䗡ҏ,
ੋᆀަᡰ・ѾDŽ
Thus [it is said]:
When speaking, one might invite disaster;
When acting, one might invite disgrace—
May the junzi be cautious about where to take his stand!
Once again, the statement on the junzi—which easily matches up with the three
earlier ones listed above—provides the closure of the entire section; it is as much a
reminder of the chapter’s topic proper as it serves as a device to structure the text—
indeed, a kind of punctuation mark.
It is remarkable how the individual sections discussed so far are not only selfcontained but also composed in different styles, ranging from what—on purely formal grounds of rhyme and meter—could be called “poetry” to the variety of prose
patterns. Thus they could be linked not only to different discourses (e.g., about the
natural world, the realm of craftsmanship, or moral behavior) but also to different
rhetorical figures and patterns of speech. It is unlikely that these passages were
original compositions by a single author; more plausibly, they were diverse expressions of traditional wisdom and as such readily available to the compiler of
“Exhortation to Learning.”
16
M. Kern
It is not surprising that these expressions found their way into a range of
different texts (K I.124–26). To give just one example, in Han times the statement
(or half-statement, as it is only part of a “couplet” here) “I once spent the whole day
thinking, but it was not as good as what I learned in an instant” is attributed to
Kongzi in the Da Dai Liji བྷᡤ䁈 and to Zisizi ᆀᙍᆀ in the Shuiyuan 䃚㤁.
Considering that already in the Xunzi, all these individual statements are not integrated with whatever follows and precedes them, it is not possible to identify their
origin; just as we see them used as bits of traditional lore re-appearing in Han texts,
they may well have preceded the Xunzi as well. The same should be held for similar
passages across many other chapters of the text; what finds itself as quotation or
parallel in Han texts may very well not be derived from the Xunzi but could have
been material that was “traditional” or “shared” already in the third century BCE
and entered the Xunzi as such.32 The traditional idea of XUN Kuang as the principal
origin of his text interprets ideological differences within the Xunzi as coming from
different periods of XUN Kuang’s life; and in a significant number of cases, it
requires the assumption that certain passages are misplaced from an “earlier” to a
“later” stratum or vice versa. Inescapably, this procedure may well be called a
classical case of “the biographical fallacy”: relying to a considerable extent on the
text itself, one reconstructs a coherent author whose intellectual biography then, in
beautifully circular reasoning, serves as the master tool to stratify his different ideas
chronologically.
The text-critical observations made so far can be extended to the entire first half
of “Exhortation to Learning.” In each paragraph, metaphors and analogies from
either the natural world or the realm of craftsmanship are lined up in series, no historical references are included, and the passages end with a brief statement on the
junzi; in two of three cases, this final statement can then be found, verbatim or
unmistakably related, in the Lunyu. In addition, the Odes are quoted twice, each
time explicitly (“An Ode says:”) and with six lines, and the entire text is punctuated
by gu ᭵ and shigu ᱟ᭵, each time gesturing toward established wisdom that
appears both conclusive and unquestionable. Aside from the Odes quotations and
the implicit gu and shigu gestures toward traditional authority, no other text is
explicitly invoked; thus, it is impossible to decide, for example, whether the Xunzi
is quoting from an early version of the Lunyu or whether the latter, at some subse-
32
For a convenient survey of such passages, see the appendices “Composition of Each Book” in
each of Knoblock’s three volumes. For materials shared between the Xunzi and various pre-Han or
early Han texts, Knoblock likewise notes that “there is no reason to consider the possibility of
direct quotation since we are probably dealing with traditional material ancestral to both the Xunzi
and these texts” (K I.125). However, Knoblock does seem to assume that where such material is
present in the Xunzi, it was consciously selected by XUN Kuang and hence was under his authorial
control.
1
Style and Poetic Diction in the Xunzi
17
quent stage, possibly during the Han,33 adopted the lines on the junzi from the Xunzi
or a third source.
It is only with section seven,34 a little more than half-way into of the first chapter,
that “Exhortation to Learning” adopts the diction of expository prose, starting with
a two-part rhetorical question: “Where does learning begin? Where does it end?”
(ᆨᜑѾ? ᜑѾ㍲?) The answer, introduced by a simple yue ᴠ (“it is said,” or “I
say”), begins once again with a pair of formalized statements:
HKCS 1/3/7–8:
ަᮨࡷ
Ѿ䃖㏃, ㍲Ѿ䆰;
ަ㗙ࡷ
Ѿ⛪༛, ㍲Ѿ⛪㚆ӪDŽ
“In its sequential order:
It begins with reciting the classics, it ends with reading out loud the ritual
[precepts].
In its meaning:
It begins with being a learned man of service,35 it ends with being a sage.”
It is possible that yue ᴠ, as understood by traditional commentators and translated
by Knoblock, means “I say.” However, yue may well mark the above pair of lines
as another “saying” of traditional origin,36 especially as the text that follows them
continues in free prose. What is emphasized by yue is only the paired statement, not
the entire section that follows. Such “marking” of speech is a common rhetorical
feature of early Chinese expository prose—in which case yue may indicate that the
brief maxim is precisely not in the author’s original voice but a piece of wisdom he
cites approvingly.
By contrast, this maxim is followed by a rare instance (in chapter 1) of several
sentences in unbound prose:
33
For a possible Western Han compilation date of the Lunyu, see Zhao (1961: 11–24), W. Zhu
(1986: 40–52), Makeham (1996: 1–24), Hunter (2012).
34
In CHANT; Knoblock’s section eight.
35
In social terms, shi ༛ refers to the lowest aristocratic rank; in the present context, it implies the
learned man of low aristocratic status (still above the unranked commoners) who is associated with
military or other service.
36
On the rhetorical use of such markers of direct speech, see Kern and Hunter (forthcoming).
18
M. Kern
HKCS 1/3/8–9:
ⵏぽ࣋ѵࡷޕDŽᆨ㠣Ѿ⋂㘼ᖼ→ҏDŽ᭵ᆨᮨᴹ㍲, 㤕ަ㗙ࡷнਟ丸㠮㠽
ҏDŽ⛪ѻӪҏ, 㠽ѻ⦨ҏDŽ
By truly building up effort for a long time, one enters [into the process of
becoming a junzi];37 learning is something which continues until death
and only then stops. Thus, while the sequential order of learning continues
to the end of one’s life, when it comes to its meaning, it is what must not be
abandoned for even an instant. Those who engage in it are humans;
those who abandon it are wild beasts.
Without transition, this is in turn followed by a rhythmic and rhymed account of
three of the classics:
HKCS 1/3/9–10
᭵
ᴨ㘵ǃ᭯һѻ㌰ҏ;
䂙㘵ǃѝ㚢ѻᡰ→ҏ;
㘵ǃ⌅ѻབྷ࠶, 于ѻ㏡㌰ҏDŽ
*-ə
*-ə
*-ə
Thus [it is said]:
The Documents are the essentials for government affairs;
The Odes are where fitting tones come to rest;
The Rites provide the great distinctions according to [social] rules, they
are the guiding principles of classification.
This self-contained unit of three rhymes is then elaborated upon as follows, with the
learning of both the Music and the Springs and Autumns Annals added:
HKCS 1/3/10–12
ᆨ㠣Ѿ㘼→(*təʔ)⸓DŽཛᱟѻ䄲䚃ᗧѻᾥ(*N-kək)DŽ
ѻᮜ᮷ҏ,
′ѻѝ઼ҏ,
䂙ᴨѻঊҏ,
᱕⿻ѻᗞҏ,
൘ཙൠѻ䯃㘵⮒⸓DŽ
When learning reaches up to the ritual precepts, it stops. This is what is called
the pinnacle of the moral way and its virtuous power.
37
The various commentators cannot agree on the meaning of ru “( ޕenter”) here; see Wang (2005
vol.1: 26–27).
1
Style and Poetic Diction in the Xunzi
19
The reverence and refinement of the Rites,38
The moderation and harmony of the Music,
The breadth of the Odes and the Documents,
The subtlety of the Springs and Autumns:
Everything between Heaven and Earth culminates in [learning].
Nothing of this adds up to an actual argument; it is more like a parade of platitudes
sputtering forth from the “discourse machine”39 that reproduces itself in ever new
variations on a circumscribed theme, in this case, “learning.” But this, of course, is
the force of its argument: contrary to the celebration of XUN Kuang as an author
with an emphatic personal voice, the text, while certainly advancing an intellectual
position, here is emptied of any individuality or surprising thought: in a mantra-like
style both rhythmic and repetitive, it falls from one rhetorical pattern into the next,
generating and regenerating itself in a continuous and inescapable loop of statements that are asserted but not argued. Because the passage is not built as a linear
structure, its continuation is not predicated on what comes before and after any of
its parts. It can be rejected but never refuted.
HKCS 1/3/14–15
ੋᆀѻᆨҏ,
ޕѾ㙣,
㪇Ѿᗳ,
ᐳѾഋ億,
ᖒѾअ䶌DŽ
ㄟ㘼䀰,
㶑㘼अ,
аਟԕ⛪⌅ࡷDŽ
ሿӪѻᆨҏ,
ޕѾ㙣,
ࠪѾਓ;
ਓ㙣ѻ䯃,
ࡷഋረ㙣,
ᴧ䏣ԕ㖾гቪѻ䓰ૹ!
As for the learning of the junzi:
It enters the ear,
Manifests itself in the heart,
Extends across the four extremities,
38
Here and elsewhere, I translate the term li in two different ways: when standing alone, as
general “ritual precepts”; when being part of a list of what are clearly the liu yi ޝ㰍 (“six arts”),
or some of them, as the title of a text (i.e., the Rites). By the time of Xunzi, this canon of learning
was well-established, as is now proven by the manuscripts from Guodian 䜝ᓇ of around 300 BCE.
39
I borrow the term from Owen (2001: 175–91).
20
M. Kern
Takes form in activity and repose.
Gasping in speaking,
Slow in action,40
Altogether one can take him as model and rule.
As for the learning of the petty man:
It enters the ear,
Goes out through the mouth,
Yet between ear and mouth
There are just four inches of space—
How could it suffice to grace a seven-foot body!
The message of these lines seems clear: the junzi is thoroughly—indeed bodily—
transformed by learning, while for the petty man, learning has a mere utilitarian
purpose, exiting the mouth as fast as it enters the ear. Yet immediately thereafter
follows an interesting twist: the learning of the ancients is a thing of the past—yet it
is the learning of the junzi:
HKCS 1/3/17–18:
ਔѻᆨ㘵⛪ᐡ,
Ӻѻᆨ㘵⛪ӪDŽ
ੋᆀѻᆨҏ, ԕ㖾ަ䓛;
ሿӪѻᆨҏ, ԕ⛪⣒DŽ
᭵
н㘼䄲ѻۢ,
а㘼Ҽ䄲ѻDŽ
ۢǃ䶎ҏ,ǃ䶎ҏ;
ੋᆀྲೞ⸓DŽ
The learning of the ancients was for themselves;
The learning of today is for others.
The learning of the junzi is for gracing his person;
The learning of the petty man is for preparing sacrificial birds and calves.
Thus [it is said]:
To pronounce oneself without having inquired is called presumptuous;
to pronounce oneself twice when having inquired once is called garrulous.41
Being presumptuous is wrong; being garrulous is wrong.
The junzi is like an echo.
40
Once again, the commentators do not agree on the meaning of these words and lines; see Wang
(2005 vol.1: 30). The choices in translating duan ㄟ (interpreted as chuan ை) and ruan 㶑 come
down to the question of whether the two terms are similar (YANG Liang, whom I follow here) or
opposite (WANG Tianhai, who understands them as “urgent” versus “slow”).
41
Commentators disagree on the meanings of ao ۢ and za ; see Wang (2005 vol.1: 32–33).
1
Style and Poetic Diction in the Xunzi
21
What does it mean that the junzi is “like an echo”? YANG Liang explains xiang ೞ
(“echo”) as “responsive sound,” and later commentators have expanded this reading
to mean that the junzi responds in the precisely adequate way: if prompted (“asked”)
once, he pronounces himself once; if prompted lightly or strongly, he will respond
lightly or strongly, respectively. This is a fine reading, though I take the text
differently: the “asking” or “inquiring” is not directed toward the junzi but toward his
own action, and his subsequent pronouncement is properly limited to the extent of
his inquiry. In this sense, what he says is “like an echo” of what he has learned. But
there is more to the use of the simile “like an echo.” It invites the reader to act “like
an echo”—to respond with the perceptive mind of a junzi able to discern what is
conveyed through the Xunzi. Here, unlike with its unreconstructed platitudes before,
the text engages the reader by means of its “poetry”: in using a simile to describe the
junzi, the text demands an act of interpretation—a choice between different possible
meanings—to become understood. It is in this space of negotiated meaning that the
reader encounters the mind of the author. The following paragraph contains an
instance of (self?)-quotation and (self?)-commentary to exalt the role of the teacher:
HKCS 1/3/20–21
ᆨ㧛ׯѾ䘁ަӪDŽ
′⌅㘼н䃚,
䂙ᴨ᭵㘼н࠷,
᱕⿻㌴㘼н䙏DŽ
ᯩަӪѻ㘂ੋᆀѻ䃚, ࡷሺԕ䙽⸓, ઘᯬц⸓DŽ
᭵ᴠ:
ᆨ㧛ׯѾ䘁ަӪDŽ
“In learning, nothing is more effective than being close to a person
accomplished in it.”
The Rites and the Music provide models yet do not explain;
The Odes and the Documents provide precedents yet do not speak directly;42
The Springs and Autumns are terse yet not easy to grasp.
If one imitates how a person [of learning] practices the explanations of the
junzi one becomes widely revered and [one’s reputation] extends
across generations.
Thus it is said:
“In learning, nothing is more effective than being close to a person
accomplished in it.”
Here, the final gu yue is an argumentative conclusion: because the classics do not
lend themselves to an easy understanding, one needs to follow the instruction and
42
Here, I disagree with WANG Tianhai who glosses gu er bu qie ᭵㘼н࠷ as chen jiu er bu qiehe
xianshi 䲣㠺㘼н࠷ਸ⨮ሖ (“[they] array old precedents but do not conform to present reality”).
Instead, bu qie zhi н࠷ⴤ, of which I take bu qie to be the abbreviated form, is a way of indirect
(and ineffective) speech, as attested in Hanshu 51.2329.
22
M. Kern
model of a teacher (“a person accomplished in it”). The following passage in
unbound prose is the first—and indeed only—part of the entire chapter that presents
a sustained deductive argument. It also is the only longer segment of chapter 1
where linguistic patterning does not dominate the development and expression of
reasoning; instead, the passage is driven by the use of logical conjunctions, interrogative particles, and conclusive markers such as an ᆹ (“how”), er yi 㘼ᐢ (“and
this is all,” “merely”), ze ࡷ (“then,” “thus”), ruo 㤕 (“if”), and finally another gu ᭵
(“therefore”).
HKCS 1/3/23 – 1/4/4
ᆨѻ㏃㧛䙏ѾྭަӪ, 䲶⅑ѻDŽкн㜭ྭަӪ, лн㜭䲶, ᆹ⢩ሷᆨ䴌
䆈ᘇ, 丶䂙ᴨ㘼ᐢ㙣DŽࡷᵛцマᒤ, н⛪ݽ䱻݂㘼ᐢDŽሷ⦻ݸ, ᵜӱ㗙,ࡷ
↓ަ㏃㐟䑺ᗁҏDŽ㤕ᤸ㼈么, 䂈ӄᤷ㘼乃ѻ, 丶㘵нਟऍᮨҏDŽн䚃២,
ԕ䂙ᴨ⛪ѻ, 䆜ѻ⥦ԕᤷ⋣ҏ, ԕᠸ㠲哽ҏ, ԕ䥀佀༪ҏ, нਟԕᗇѻ⸓DŽ
᭵䲶, 䴆ᵚ᰾, ⌅༛ҏ; н䲶, 䴆ሏ䗟, ᮓ݂ҏDŽ
For the path of learning, nothing is more expeditious than devotion to a person
accomplished in it; to exalt the ritual precepts is second. If in the first place one
cannot devote oneself to a person engaged in it, and in the second place cannot
exalt the Rites, how would it be enough to only learn some miscellaneous
precepts or simply follow the Odes and the Documents! In this case, to the
end of one’s days, one could not avoid being nothing more than a parochial ru.
If one is bound to take the former kings as one’s source and benevolence and
righteousness as one’s basis, then the Rites will rectify the warp and woof, the
ways and byways. It is like when one lifts a fur coat by its collar, grasps it with
five fingers from within and then shakes it, the [hair on the entire coat] that
falls smoothly into its place cannot be counted. If one does not take guidance
from the statutes of the Rites and acts by merely relying on the Odes and
the Documents, it is as if taking a finger to plumb the depth of the river,
or taking a dagger to dehusk the millet, or taking an awl to eat a gourd—it
just cannot be accomplished! Thus: he who exalts the Rites, even if not yet
comprehending them, is an exemplary learned man of service; he who
does not exalt the Rites, even though being scrutinizing and discriminating,
is an undisciplined ru.
Here we do have the argumentative and authorial style in the Xunzi that modern
scholars hail for its incisive reasoning (or deplore for its pugnaciousness)—and it is
here, in particular in the invectives against lou ru 䱻݂ (“parochial ru”) and san ru ᮓ
݂ (“undisciplined ru”) where one might best discover XUN Kuang’s authorial voice.
To summarize the findings above, chapter 1 of the Xunzi, however, is not the
place where this voice speaks with full force. Instead, it is a chapter largely built
around individual and mutually unrelated sections that dance around a set of common themes: the importance of learning, the preeminence of the ritual precepts over
all other disciplines and modes of conduct, the distinctive character forms and practices that define the junzi. What the chapter lacks in systematic architecture, inner
1
Style and Poetic Diction in the Xunzi
23
coherence, and linear progression, it gains through its widely varied features of
rhythmic repetition, a certain range of metaphors and similes, and the appeal of an
overall declamatory style that is more the product of the “discourse machine” than
an actual discourse, and that gains its persuasive force from just that. It is also
exceedingly difficult to determine anything like XUN Kuang’s authorial voice
throughout chapter 1, if by this we understand agreeing or disagreeing with others,
using a coherent way of argumentation, or including emphatic utterances such as
exclamations or rhetorical questions. The general (though not complete) absence of
such features and the lack of a linear development from one section to the next
might suggest that we are dealing not with an authored chapter, but with a compiled
one that loosely connects elements from diverse sources.
To what extent are these features of structure and diction common to the Xunzi
as a whole? For the majority of chapters, Knoblock—basing himself on existing
Chinese scholarship—has proposed more or less severe instances of textual corruption, primarily in the form of misplaced passages that do not connect well with their
immediate environment and instead appear to belong together with material in other
chapters. Because of his conviction that certain positions reflect the thinking of
certain periods within XUN Kuang’s long life, Knoblock further argues that individual chapters contain material not only on different topics but also from different
periods.43 While this may be true, one wonders how such textual confusion within
the chapters may have come about. Is the proposed rearrangement the attempt to
reconstitute a unity of thought and coherence of argument that in its proposed form
may never have existed in the first place? Or does the lack of integration reflect a
case of deterioration from an earlier, more tightly constructed text? Either way, the
abrupt changes of topic that occur frequently even in the chapters considered most
representative of the Xunzi’s thought suggest the relative independence of smaller
textual units within the chapters, and an overall rather loose structure of argument;
as Knoblock has pointed out, the Xunzi contains a considerable number of section
titles within the individual chapters, suggesting that the sections such titled had
once circulated independently (K I.107, 121, 123–24); moreover, in Knoblock’s
words, “we know from the history of the text that LIU Xiang and not Xunzi is
responsible for the original order not only of the books but also of sections or paragraphs within the books” (K I.123). Generally speaking, however, a passage that
follows logically from the previous one and, in turn, is the necessary basis for the
subsequent one, is less easily transposed than a relatively isolated one that stands on
its own. The editor, or series of editors, who may have rearranged the original order
within a chapter presumably did not willfully vandalize the text. He or they must
have considered the current arrangement as the most plausible or helpful one. This
might seem improbable if the chapters had been tightly constructed to begin with;
on the other hand, as the above-mentioned case of “Black Robes” reveals,44 it is by
no means impossible.
43
44
See the summarizing comments in his “Composition of Each Book” appendices.
See p. 6 above.
24
M. Kern
An example for the lack of textual integration is chapter 22, “Zheng ming” ↓,
which together with a handful of other chapters seems to express the ideological
core of the Xunzi. Stylistically, the chapter shows numerous similarities with chapter 1 discussed above: it contains few paragraphs in unbound prose, while the large
majority of passages is composed in short, highly formulaic patterns of rhythmic
parallelism that delineate the given passage from its immediate environment; a
number of short passages are rhymed; the different paragraphs are not arranged in a
continuous argument; there are no historical anecdotes or other specific historical
references; a small number of connecting markers such as gu ᭵ or shi gu ᱟ᭵ are
used liberally throughout; in three cases, a paragraph is capped by a quotation from
the Odes; there are few instances of an individual voice that would make itself heard
through rhetorical questions or exclamations; the chapter—except for a few brief
paragraphs—does not engage in refuting competing theories or positions; and a
distinction is made between an ideal past and a chaotic present. Perhaps most
importantly, the chapter as a whole does not have a continuous theme: after about
two thirds of the text, the concern with zheng ming ↓ (“correct use of names”)
falls away almost entirely, with much of the remaining parts focused, once again, on
the junzi. Altogether, the chapter consists far less of explicit reasoning than of
apodictic pronouncements, often introduced by the formulaic fan ࠑ (“as a matter of
principle”). Thus, similar to the case of chapter 1, the persuasive force of “Zheng
ming” lies not in arguments derived from traditional authority or historical precedent, nor does it rest in the compelling rebuttal of others or in a tightly woven
sequential argument. Its rhetorical force gradually rises from the procedure of piling
up passage upon passage, dictum upon dictum, that in their accumulation overwhelm the reader by their diversity of linguistic patterns. This principle of copia et
varietas (“abundance and variety”), central to the European traditions of rhetoric,
literature, painting, and music, can be found in early Chinese shui 䃚 (“attempts at
persuasion”) of XUN Kuang’s time, and it defines, then in highly stylized poetic
form, the core of the early fu 䌖 (“poetic exposition”). The formal feature of
extended parallelism, ubiquitous in the Xunzi, where a topic is pursued through a
series of statements that are typically capped with a final pronouncement introduced
by gu, is closely related to the technique of the catalogue—a rhetorical figure that is
as central to the Xunzi as it is to the Western Han fu. An extreme example in “Zheng
ming” is the following, where the topic of “name” appears only in passing (HKCS
22/108/14 – 22/109/3, K III.129–30):
❦ࡷօ㐓㘼ԕ਼⮠?ᴠ: 㐓ཙᇈDŽࠑ਼于਼ᛵ㘵, ަཙᇈѻ⢙ҏ਼DŽ
᭵∄ᯩѻ⯁լ㘼䙊, ᱟᡰԕަޡ㌴ԕᵏҏDŽ
ᖒ億ǃ㢢⨶ԕⴞ⮠;
㚢丣◱ǃ䃯ㄭǃཷ㚢ԕ㙣⮠;
⭈㤖ǃ咩␑ǃ䗋䞨ǃཷણԕਓ⮠;
俉㠝ǃ㣜兡ǃ㞕㟺ǃ┿ᓞǃཷ㠝ԕ啫⮠;
⯮Ⲓǃࠄ⟡ǃ━䡩ǃ䕅䟽ԕᖒ億⮠;
䃚᭵ǃௌᙂǃ૰′ǃᝋᜑǃⅢԕᗳ⮠DŽ
1
Style and Poetic Diction in the Xunzi
25
ᗳᴹᗥ⸕DŽᗥ⸕,ࡷ
㐓㙣㘼⸕㚢ਟҏ,
㐓ⴞ㘼⸕ᖒਟҏDŽ
❦㘼ᗥ⸕ᗵሷᖵཙᇈѻ⮦㉯ަ于, ❦ᖼਟҏDŽ
ӄᇈ㉯ѻ㘼н⸕,
ᗳᗥ⸕㘼❑䃚,
ࡷӪ㧛н❦䄲ѻн⸕DŽ
↔ᡰ㐓㘼ԕ਼⮠ҏDŽ
This being so, for what cause does one take things as equal or different?
One might say: because of the inborn facilities of the organs. In general, when
things are of the same kind and the same disposition, the way the inborn
facilities of the organs perceive of them is also the same. Thus, when put side
by side, they resemble one another and are interchangeable; this is why they
are given an agreed name by which they correspond to one another.
The embodiments of form and the patterns of color are distinguished
by the eye;
The clear and muddy notes, the well-tuned reed pipes, and the unusual
sounds are distinguished by the ear;
Sweet and bitter, salty and bland, acid and sour, and the unusual tastes are
distinguished by the mouth;
Aromatic and foul, fragrant and stinking, fishy and fetid, rotten and
festering, and the unusual odors are distinguished by the nose;
Pain and itching, cold and heat, slippery and firm, light and heavy are
distinguished by the embodiment of form;
Explanation and precedent,45 pleasure and anger, sorrow and happiness,
love and hate, and desire are distinguished by the heart.
The heart has understanding by distinction. When there is understanding
by distinction, then
Because of the ear, it is possible to know sound,
Because of the eye, it is possible to know form.
Following from there, understanding by distinction is always contingent upon
the organ being properly impressed by what is of its category; only then
it is possible.
If the five organs are impressed but do not lead to understanding,
if the heart understands by distinction but without leading to explanation,
45
I do not follow WANG Tianhai, who reads shuo/shui gu 䃚᭵ as yue ku ᚵ㤖 (“happiness and
bitterness”), as proposed earlier by Qing commentators (Wang 2005 vol.2: 897). First, happiness
is already included in the complete catalogue of emotions here; second, by accepting the characters
(and words) shuo/shui gu 䃚᭵ I suggest that “explanation and precedent,” which are clearly outside the catalogue of emotions, are purposefully added to the latter to claim that these modes of
speech, too, are governed by the heart.
26
M. Kern
then there is nobody among men who would not be inclined to call this
“not understanding.”
This is the cause by which one takes things as equal or different.
For the point made here—the different organs are perceptive of different types of
things, which is the basis for the human capacity to perceive of these things as
belonging to different categories, and for being capable of fully perceiving of reality
altogether—one might find this passage extravagantly verbose. Perhaps the passage
is constructed as a forceful rebuttal of an implied philosophical adversary, but otherwise, no reader needed the extensive catalogue of different sensations and the
organs receptive to them, nor was it necessary to dwell on the question of “understanding” or “not understanding.” Rhetorically, however, the text exhausts its limited topic. Instead of offering some chosen examples or analogies, it says everything
there is to say, as is then finally suggested by the closing formula regarding yuan er
yi tongyi 㐓㘼ԕ਼⮠ (“the cause by which one takes things as equal or different”)
that verbatim responds to the question at the outset. Within this frame, the minidiscourse stands on its own, and it is complete—which makes it persuasive. It also
is eminently movable to fit different contexts, but within the “Zheng ming” chapter,
it is part of a larger discussion on “correct names” that imply correct distinctions.
Here, the keyword that frames the discussion (and runs throughout it), is yi ⮠ (“different”), which also connects the present passage to the immediately preceding one
(not cited here). To this larger discussion of correct distinctions, the present passage
contributes the argument that the principles of human understanding and action are
biologically determined, a point that the Xunzi also makes elsewhere. However,
what makes the passage compelling and unquestionable—its rhetorical architecture
and sense of completeness—is also what facilitates its potential displacement.
The feature of the catalogue, frequent in shorter units of text regardless of the
topic at hand,46 is also operative at a larger level. In chapter 5, “Fei xiang” 䶎, it
appears in a rare instance (within the core chapters) of sustained references to historical figures (HKCS 5/17/10–24, K I.203–4)47 where it is followed, in short order,
by lists of “three” patterns of misfortune and behavior, respectively. Chapter 21, “Jie
bi” 䀓㭭, contains catalogues of what bi 㭭 (“blinds”) the human mind, examples of
sage rulers from the past and of their sensual perceptions, scholars who were
“blinded” by their particular convictions, and others more. Chapter 19, “Li lun”
䄆, is largely driven by a series of catalogues, beginning with the sensory organs and
everything that can be yang 伺 (“nurtured”), and then detailing the sumptuary rules
concerning ancestral sacrifices, the order of sacrificial offerings, funerary arrangements including the mourning garb, human emotions and the ways to display them,
tomb furnishings and grave goods, and so on. Another large catalogue can be found
46
Compare, e.g., the passage in chapter 3, “Bu gou” н㤏, that enumerates the qualities of the
junzi; see HKCS 3/9/15–17, K I.175.
47
Chapter 5 is unusual in containing many references to historical precedent, always connected to
well-known figures from the past. The chapter also contains a larger than usual number of quotations from the classics and anonymous sayings.
1
Style and Poetic Diction in the Xunzi
27
in chapter 6, “Fei shi’er zi” 䶎ॱҼᆀ. Here, the text starts out by denouncing
through formulaic repetition twelve groups of zi ᆀ (“masters”), divided in six pairs
before turning to the positive examples of Kongzi ᆄᆀ and Zigong ᆀᕃ, Shun 㡌,
and Yu . For five of the six pairs of aberrant scholars, the text concludes with the
same mantra-like formula: ran’er qi chi zhi you gu, qi yan zhi cheng li, zuyi qihuo
yuzhong ❦㘼ަᤱѻᴹ᭵, ަ䀰ѻᡀ⨶, 䏣ԕⅪᜁᝊ : ᱟ A, B ҏ (“Thus,
when their positions have precedent and their explications become well-formed,
they suffice to deceive and confuse the ignorant masses; such are [the masters] A
and B”). As in the other catalogues, the list is presented as being complete, final,
cast into fixed form—and is therefore compelling.
The chapter that is singularly distinguished by fierce reasoning and a forceful and
emphatic (and hence distinctly individual) voice is chapter 23, “Xing’e” ᙗᜑ—the
rhetorical climax of the core Xunzi—where the text not only argues that human nature
is bad but also that the Mengzi’s opposite view is wrong. As noted by Knoblock, the
chapter is seriously damaged, and roughly the last third of the text is merely “tangentially connected with the main theme of the book” (K III.280). This having been said,
the first two thirds of the chapter do stand out by their rigorous and combative style.
The chapter contains the same abundance of stylistic patterns as other parts of the
Xunzi, and it is not short on formulaic expression: yong ci guan zhi, [ran ze] ren zhi
xing’e ming yi, qi shan zhe wei ye ⭘↔㿰ѻ, [❦ࡷ]Ӫѻᙗᜑ᰾⸓, ަழ㘵ڭҏ
(“Contemplating it from this perspective, it is clear that human nature is bad, and
that what is good is artificially brought about”) appears verbatim no less than eight
times, and in addition with two variations. But then there is more: a purportedly
direct citation of Mengzi (or from the emerging Mengzi tradition, though not matching up verbatim with anything in the current Mengzi text), followed by a rebuttal; a
wealth of conjunctions and sentence adverbials to develop fully formed arguments
(as opposed to mere pronouncements) that include not just the ubiquitous gu ᭵ and
ze ࡷ but also ruoshi 㤕ᱟ (“if this is the case”), ruo 㤕 (“if”), bi ᗵ (“invariably”),
ran ze ❦ࡷ (“this being so, then”), and others more; structures pointing out what is
“true” and what is “false” (shi ᱟ . . . fei 䶎), what is “ancient” and what is “today”
(gu ਔ . . . jin Ӻ); patterns of causation (“if A then B, if B then C, if C then D, and
hence . . . ”); the use of analogies; rhetorical questions such as he ye օҏ (“how?”)
and emphatic exclamations (qi 䉸 . . . zai ૹ); and staged dialogues with an interlocutor. While the “Xing’e” chapter contains the same rhythmic patterns as the
other chapters, these rhetorical patterns do not dominate the chapter or drive its
ways of developing an argument. In this analysis, I am not referring to the philosophical unity of the entire chapter (or even of the first two-thirds), nor would I
claim chapter 23 to be the philosophical core of the entire Xunzi.48 Instead, I wish to
48
As noted above, Knoblock has pointed out the problems with the textual integrity of the chapter.
For challenges to the relevance of the claim “human nature is bad” in the Xunzi’s overall philosophical system, or for the argument that the claim does not even belong to the original Xunzi, see Robins
(2001–02: 99–100). Robins himself, while firmly holding on to Xunzi’s authorship of the “core” of
the chapter, argues that it contains not one continuous argument but ten separate essays, possibly
reflecting Xunzi’s different ideas over the stretch of his long life. Whatever one might think of that,
the point is not relevant to my argument here, which is strictly on the stylistic level of the text.
28
M. Kern
emphasize the striking stylistic features of chapter 23 especially when compared
with a text like chapter 1. These features constitute a different mode of argumentation together with an intensified presence of a distinct authorial voice that speaks
with fierce conviction. Remarkably, with the exception of a single line toward the
very end, chapter 23 is among the very few chapters of the Xunzi that do not seem
to share any material with other pre-Qin and Han texts (Ho et al. 2005: 209–15). At
the same time, in post-Han times “Xing’e” has been by far the single most-referenced
chapter from the Xunzi and has come to represent the Xunzi altogether—so much so
that “readers of later centuries . . . seemed not to peruse much” of the entire text
beyond this particular chapter (Goldin 2011: 72).
From this, one might venture that it was not merely the disagreement with the
Mengzi that contributed to the Xunzi’s more marginal status in the tradition and that
in Song times, finally, led to Xunzi’s expulsion from the genealogical pantheon of
the Confucian orthodoxy as it became physically enshrined in the kongmiao ᆄᔏ
(“Kongzi Temple”) of succeeding dynasties. The double phenomenon that the Xunzi
is at its most forceful in its attack on the Mengzi, and that it is also here where the
textual voice appears most recognizable as that of a true author—always understood
as XUN Kuang—will not have gone unnoticed. Just where the text comes closest to
offering a strong argument by a strong author, it also is most vulnerable to rejection,
and to the punishment of its presumed author for having taken his stance.49 In
“Xing’e,” Xunzi and the Xunzi, rightly or wrongly, have long come to stand in for
each other.
As “Xing’e” with its hard-charging logic and emphatic expressivity seems to
reveal an individual voice (or constructs one rhetorically), it also reminds us that
much of the Xunzi does not. In fact, it seems to me that chapters 1 and 23 are strikingly different in nature: while the first, at least in part, appears as a compiled text,
the second appears as an authored one. There are many ways to rationalize how the
author of chapter 23 can also be the one who wrote chapter 1; perhaps he was a
master of many styles; perhaps his way of writing changed over the course of his
long life; perhaps the different topics suggested different forms of argumentation;
perhaps the deployment of traditional wisdom in chapter 1 is meant to exemplify the
cause of traditional “learning,” while the aggressive argument in chapter 23 is carefully crafted to reflect the harshness of its message—with both thus mimetically
representing on the linguistic level the philosophical meaning that is to be to
advanced.50 But any of these explanations would have to be explicitly appropriated
and defended in order to argue for the authorial unity of the Xunzi in the face of its
49
On being responsible and punishable as a hallmark of authorship, see Foucault (1979: 141–60).
As it happens, this is one of the characteristics of the early fu 䌖 (“poetic exposition”) for which
the Hanshu, as noted above, names Xunzi an exemplary early proponent; see Kern (2003: 383–
437). I am grateful to Eric Hutton for raising the bar here and below, as I discuss the issue of
authorship.
50
1
Style and Poetic Diction in the Xunzi
29
striking internal diversity. But why? What is the evidence (other than traditional
belief) for positing a single author for both chapters (and all other chapters as well),
and what is gained by trying to make that case—a case as forced and arduous as it
is with every pre-imperial Chinese text? How many problems are solved by the
assumption of Xunzi as the single author of the Xunzi, and how many problems are
created by it? Had Foucault not written his essay “What is an Author?” (Foucault
1979), we would have to invent something close to it in order to answer these
questions.
Consider again the initial parts of chapter 1 that follow the statement “A junzi
says: ‘In learning, one must not desist.’” As noted above, this pronouncement leads
to different metaphors from the worlds of nature and craftsmanship before offering
a set piece of traditional wisdom introduced by gu ᭵. Soon thereafter, we see
another gu introducing another, seemingly unrelated set piece which traditional
readers have understood as opening a new section of the text. As I have argued
above, it is not possible to begin a new section with gu; perhaps something is missing before, or the text might be in disarray. But there is also another way to make
sense of the structure of “Quan xue” (and many other parts of the Xunzi). What if
the chapter is not at all in disarray or incomplete—but simply appears as such
because we are expecting a linear progression?51 What if the mutually independent
passages were never meant to constitute such linear progression but, instead, were
compiled as parallel illustrations of the core ideas? In this way, the passages that
follow the different instances of gu are to be read not as one following the other but
as different and equally valid responses to the initial statement of the “gentleman,”
compiled from a larger repertoire of such responses. Such a repertoire may have
accumulated from different scenes of instructions some of which might even go
back to XUN Kuang himself, teaching his disciples; or it may have built up from
various discussions, oral or written, of “learning” that were associated with XUN
Kuang and his intellectual circle. There are numerous ways in which a repertoire
concerning the traditional topic of “Quan xue”—which is not at all unique to the
Xunzi—could have grown over time, and it is not difficult to imagine an editor compiling parts of it into the text we now have. Needless to say, any such editor would
have been attracted the most to precisely the kinds of metaphors, analogies, and
pieces of accepted and therefore authoritative wisdom that we find in the received
text of the Xunzi.
In other words, by expecting a certain type of argumentative logic across the different chapters, we may be misreading the text altogether. As soon as we abandon
the idea of the individual author and of the text as this author’s individually crafted
work, a chapter like “Quan xue” easily makes sense as a compilation of mutually
independent illustrations of the principal ideas associated with Xunzi and his circle.
51
The same question must be raised about chapter 23 which both Donald J. Munro and Robins—to
my mind rather disingenuously—have labelled “a mess”; see Munro (1996: 198), Robins (2001–
02: 157).
30
M. Kern
Moreover, unlike with the traditional reading of the text, we no longer struggle with
questions of authorship, ideological diversity, or stylistic incoherence. We also no
longer need to tie different parts of the Xunzi to the highly tentative reconstruction
of different periods of XUN Kuang’s personal life. With the exception of parts of
“Xing’e” and perhaps some other passages of careful disquisition, we can let go of
XUN Kuang as author. In return, we obtain a much richer vision of XUN Kuang as the
teacher who inspired the many different ways to think and speak about important
social, moral, and philosophical questions. This proposal must not be misunderstood as an argument about the authenticity of the text. To the contrary, it is only
with the common misreading of compiled texts as authored ones, and with the false
conflation of text and author, that textual authenticity is confused with authorship.
A clearer understanding of this confusion, finally, may also help us to recognize a
truly authored text when we see it—which most likely will not be chapter 1 of the
Xunzi.
Acknowledgments I thank Paul Rakita Goldin, Wolfgang Behr, Willard Peterson, Michael
Hunter, and Christoph Harbsmeier for their comments and bibliographic help. Only after the present essay was already in proofs, I took note of an excellent study of the first chapter of the Xunzi
(and related early texts on the topic of learning): Oliver Weingarten’s MA thesis, submitted to the
University of Hamburg (Weingarten 2004). Its analysis of the composition of the chapter (p. 46–59)
reaches many conclusions similar to my own, and it includes a superb, carefully annotated German
translation of the entire chapter (p. 103–14). In addition, Weingarten recently presented a paper in
English where he further extended his findings (Weingarten 2014).
Bibliography
Bagley, Robert W. 1993. “Meaning and Explanation.” In The Problem of Meaning in Early Chinese
Ritual Bronzes, ed. Roderick Whitfield, 34–55. London: School of Oriental and African
Studies, University of London. (One of the major statements on the function of ornament in
early Chinese bronze art.)
Baxter, William H. 1998. “Situating the Language of the Lao-tzu: The Probable Date of the Tao-teching.” In Lao-tzu and the Tao-te-ching, ed. Livia Kohn and Michael LaFargue, 231–53.
Albany: State University of New York Press. (A widely-read article that attempts to date the
Laozi by comparing some of its linguistic features to the Shijing and the Chuci.)
Boltz, William G. 2005. “The Composite Nature of Early Chinese Texts.” In Text and Ritual in
Early China, ed. Martin Kern, 50–78. Seattle: University of Washington Press. (An important
widely-received conceptual statement on the nature and structure of early Chinese texts.)
Debon, Günther. 1996. “Der Reimspruch im philosophischen Schrifttum Chinas.” In Debon, So
der Westen wie der Osten. Dichtung, Kunst und Philosophie in Deutschland und China, 36–42.
Heidelberg: Brigitte Guderjahn. (One of the few essays in any Western language dedicated to
the phenomenon of rhyme in early Chinese philosophical prose.)
Debon, Günther. 2002. “Verborgene Spruchdichtung im Hsün-tzu.” In Und folge nun dem, was
mein Herz begehrt: Festschrift für Ulrich Unger zum 70. Geburtstag, ed. Reinhard Emmerich
and Hans Stumpfeldt, vol. 1, 21–30. Hamburg: Hamburger Sinologie Gesellschaft. (A short but
insightful comment on rhyme in the Xunzi.)
Denecke, Wiebke. 2010. The Dynamics of Masters Literature: Early Chinese Thought from
Confucius to Han Feizi. Cambridge: Harvard University Asia Center.
1
Style and Poetic Diction in the Xunzi
31
De Reu, Wim. 2010. “How to Throw a Pot: The Centrality of the Potter’s Wheel in the Zhuangzi.”
Asian Philosophy 20(1): 43–66.
Fischer, Paul. 2009. “Intertextuality in Early Chinese Masters-Texts: Shared Narratives in Shi Zi.”
Asia Major third ser. 22(2): 1–34. (A contribution to the ongoing debate on the structure and
intertextuality of early Chinese texts.)
Foucault, Michel. 1979. “What is an Author?.” In Textual Strategies: Perspectives in PostStructuralist Criticism, ed. Josué V. Harari, 141–60. Ithaka: Cornell University Press. (A seminal and extremely influential essay on the problem of authorship, Foucault’s essay has triggered
hundreds of responses.)
Gassmann, Robert H., and Wolfgang Behr. 2005. Antikchinesisch—Ein Lehrbuch in drei Teilen.
Bern: Peter Lang.
Goldin, Paul Rakita. 1999. Rituals of the Way: The Philosophy of Xunzi. Chicago: Open Court.
Goldin, Paul Rakita. 2011. Confucianism. Berkeley: University of California Press.
Gombrich, Ernst H. 1979. The Sense of Order: A Study in the Psychology of Decorative Art. Ithaca:
Cornell University Press.
Graham, A.C. 1991a. Disputers of the Tao: Philosophical Argument in Ancient China. LaSalle:
Open Court.
Graham, A.C. 1991b. “Two Notes on the Translation of Taoist Classics.” In Interpreting Culture
Through Translation: A Festschrift for D.C. Lau, ed. Roger T. Ames, CHAN Sin-wai, and Mausang NG, 119–44. Hong Kong: The Chinese University Press.
Hanshu ╒ᴨ 1987. Beijing: Zhonghua shuju.
Harbsmeier, Christoph. 1997. “Xunzi and the Problem of Impersonal First Person Pronouns.”
Early China 22: 181–220. (A critical Review Article of Knoblock’s Xunzi translation, with
important linguistic insights.)
Harbsmeier, Christoph. 2001. “The Rhetoric of Premodern Prose Style.” In The Columbia History
of Chinese Literature, ed. Victor H. Mair, 881–908. New York: Columbia University Press.
Ho, Che Wah օᘇ㨟, CHU Kwok Fan ᵡ഻㰙, and FAN Sin Piu ›ழ⁉. 2005. The Xunzi with
Parallel Passages from Other Pre-Han and Han Texts lj㥰ᆀNJ㠷㉽ި╒ޙ〖ݸ䟽㾻䋷ᯉᖉ
㐘. Hong Kong: The Chinese University of Hong Kong.
Hunter, Michael. 2012. “Sayings of Confucius, Deselected,” PhD diss., Princeton University. (A
new, database-driven approach to the dating of the Lunyu.)
Jiang, Yougao ⊏ᴹ䃕. 1993. Ten Essays on Phonology 丣ᆨॱᴨ. Beijing: Zhonghua shuju.
Kalinowski, Marc. 2000–01. “Systèmes de croyances et de pensée du monde sinisé.” Annuaire
EPHE, Section des sciences religieuses 109: 141–48.
Kern, Martin. 2003. “Western Han Aesthetics and the Genesis of the Fu.” Harvard Journal of
Asiatic Studies 63(2): 383–437.
Kern, Martin. 2005. “Quotation and the Confucian Canon in Early Chinese Manuscripts: The Case
of ‘Zi Yi’ (Black Robes).” Asiatische Studien/Études Asiatiques 59(1): 293–332.
Kern, Martin. 2014. “Creating a Book and Performing It: The ‘Yaolüe’ Chapter of the Huainanzi
as a Western Han Fu.” In The Huainanzi and Textual Production in Early China, ed. Sarah A.
Queen and Michael Puett, 124–50. Leiden: Brill. Published in Chinese as Ke, Mading ḟ俜б.
2010. lj␞ইᆀNJⲴᡀᴨ㠷ཿᴨ:䄆lj㾱⮕NJㇷѻ⛪䌖. Collected Papers of the Center for
Ancient Chinese Classics and Archives of Peking University ेӜབྷᆨѝ഻ਔ᮷⦫⹄ウѝᗳ
䳶࠺ 9: 436–451.
Kern, Martin, and Michael Hunter. Forthcoming. “Quotation and Marked Speech in Early
Manuscripts,” in Early Chinese Manuscripts: Texts, Contexts, Methods, ed. Wolfgang Behr,
Martin Kern, and Dirk Meyer. Leiden: Brill.
Knechtges, David R. 1989. “Riddles as Poetry: The ‘Fu Chapter’ of the Hsün-tzu,” in Wenlin:
Studies in the Chinese Humanities, vol. 2, ed. CHOW Tse-tsung, 1–31. Madison: University of
Wisconsin Press.
Knoblock, John. 1988–94. Xunzi: A Translation and Study of the Complete Works, 3 vols. Stanford:
Stanford University Press. (Cited here as “K,” followed by volume.page number.)
32
M. Kern
LaFargue, Michael. 1994. Tao and Method: A Reasoned Approach to the Tao Te Ching. Albany:
State University of New York Press.
Lau, D.C. 1970. Mencius. Harmondsworth: Penguin Books.
Lewis, Mark Edward. 1999. Writing and Authority in Early China. Albany: State University of
New York Press.
Li, Binghai ᵾ⛣⎧. 2010. “Examination of the Title, Structure, and Concepts of the ‘Chengxiang’
Chapter of Xunzi”lj㥰ᆀᡀNJⲴㇷ乼ǃ㎀Ώ৺ަ⨶ᘥ㘳䗘. Jianghan Forum ⊏╒䄆
9: 89–93.
Liu, Xiaogan. 1994. Classifying the Zhuangzi Chapters, trans. William E. Savage. Ann Arbor:
Center for Chinese Studies, The University of Michigan.
Long, Yuchun 喽ᆷ㍄. 1962–63. “Rhymed Verse in Pre-Qin Prose” 〖ݸᮓ᮷ѝⲴ丫᮷. Chung
Chi Journal ጷสᆨ 2(2) and 3(1), repr. in LONG Yuchun 2009. Anthology of Philological
Studies in the Sizhuxuan ㎢ㄩ䔂ሿᆨ䄆䳶, 182–83. Beijing: Zhonghua shuju. (The most comprehensive survey of rhyme in pre-Qin prose, albeit far from complete.)
Major, John S. et al. 2010. The Huainanzi: A Guide to the Theory and Practice of Government in
Early Han China. New York: Columbia University Press.
Major, John S. 2014. “Tool Metaphors in the Huainanzi and Other Early Texts.” In The Huainanzi
and Textual Production in Early China, ed. Sarah A. Queen and Michael Puett, 153–98. Leiden:
Brill.
Makeham, John. 1996. “The Formation of Lunyu as a Book.” Monumenta Serica 44: 1–24.
(A seminal contribution on the problem of dating the Analects.)
Malmquist, Göran. 1973a. “A Note on the Cherng shianq Ballad in the Shyun Tzyy.” Bulletin of the
School of Oriental and African Studies 36(2): 352–58.
Malmquist, Göran. 1973b. “The Cherng shianq Ballad of the Shyun Tzyy.” Bulletin of the Museum
of Far Eastern Antiquities 45: 63–89.
Meyer, Dirk. 2011. Philosophy on Bamboo: Text and the Production of Meaning in Early China.
Leiden: Brill.
Morrison, Madison. 1981. “The Poetic Element in Lao Tzu.” Tamkang Review 11: 391–420.
Munro, Donald. 1996. “A Villain in the Xunzi.” In Chinese Language, Thought, and Culture:
Nivison and His Critics, ed. Philip J. Ivanhoe, 193–201. Chicago: Open Court.
Owen, Stephen. 2001. “LIU Xie and the Discourse Machine.” In A Chinese Literary Mind: Culture,
Creativity, and Rhetoric in Wenxin Diaolong, ed. Zong-qi CAI, 175–91. Stanford: Stanford
University Press.
Queen, Sarah A. 2008. “The Creation and Domestication of the Techniques of Lao-Zhuang:
Anecdotal Narrative and Philosophical Argumentation in Huainanzi 12.” Asia Major, third
series, 21: 201–47.
Raphals, Lisa A. 1994. “Poetry and Argument in the Zhuangzi.” Journal of Chinese Religions 22:
103–16.
Rickett, W. Allyn. 1985. Guanzi: Political, Economic, and Philosophical Essays from Early China,
Vol. 1. Princeton: Princeton University Press.
Robins, Dan. 2001–2002. “The Development of Xunzi’s Theory of Xing, Reconstructed on the
Basis of a Textual Analyis of Xunzi 23, ‘Xing E’ ᙗᜑ (Xing is Bad).” Early China 26–27:
99–158.
Roth, Harold D. 1999. Original Tao: Inward Training and the Foundations of Mysticism. New
York: Columbia University Press.
Schaberg, David. 2001. A Patterned Past: Form and Thought in Early Chinese Historiography.
Cambridge: Harvard University Asia Center.
Schaberg, David. 2005. “Platitude and Persona: Junzi Comments in the Zuozhuan and Beyond.” In
Historical Truth, Historical Criticism, And Ideology: Chinese Historiography And Historical
Culture From A New Comparative Perspective, ed. Helwig Schmidt-Glintzer et al., 177–96.
Leiden: Brill.
1
Style and Poetic Diction in the Xunzi
33
Tan, Jiajian 䆊ᇦڕ. 1995. “A First Examination of Pre-Qin Rhymed Verse” 〖ݸ丫᮷ࡍ᧒.
Literary Heritage ᮷ᆨ䚪⭒ 1: 12–19.
Wagner, Rudolf G. 2000. The Craft of a Chinese Commentator: WANG Bi on the Laozi. Albany:
State University of New York Press.
Wang, Tianhai ⦻ཙ⎧. 2005. Xunzi, Collated and Annotated 㥰ᆀṑ䟻. Shanghai: Shanghai guji
chubanshe.
Weingarten, Oliver. 2004. Ermunterungen zum Lernen—zu einem Motiv in der altchinesischen
Literatur. MA thesis, University of Hamburg. (A study of the first chapter of Xunzi together
with related and parallel texts in Shizi, Heguanzi, and Da Dai Liji.)
Weingarten, Oliver. 2014. “Xunzi 1, ‘Exhortation to Learning,’ as a Textual Collage.” Paper presented at the Conference “Reading the ‘Masters’: Contexts, Textual Structures, and Hermeneutic
Strategies.” Masaryk University (Brno, Czech Republic), September 2014.
Xu, Jie 䁡㎀. 1990. “Laozi and Philosophical Poems in Pre-modern China” 㘱ᆀ㠷ѝ഻ਔԓଢ⨶
䂙. Academics Monthly ᆨ㺃ᴸ࠺ 2: 58–64.
Zhao, Zhenxin 䏉䋎ؑ. 1961. “Who Indeed Compiled the Lunyu?” 䄆䃎ウㄏᱟ䃠㐘㒲Ⲵ. Journal
of Beijing Normal University (Social Science Edition) ेӜᑛㇴབྷᆨᆨ (⽮ᴳ、ᆨ⡸) 4:
11–24.
Zhu, Shizhe ᵡᑛ䕽. 1957. “Supplementary Explanations on Rhyme in the ‘Chengxiang’ Chapter
in the Xunzi” 㥰ᆀᡀㇷ丫䆰㼌䟻. Journal of Sun Yatsen University ѝኡབྷᆨᆨ 3:
42–47.
Zhu, Weizheng ᵡ㏝䥊. 1986. “A Detailed Discussion on the Compilation of Lunyu” 䄆䃎㎀䳶㝎
䃚. Studies on Kongzi ᆄᆀ⹄ウ 1: 40–52. (A seminal and widely influential essay on the dating of the Lunyu.)