publisher colophon

I approach the closely interrelated topics of classical Chinese poetics and theories of language from a somewhat unconventional perspective, namely the relationship between some early texts dealing with aural and optical illusions and the reading of poetic imagery in the Confucian scholar Mao Heng’s exegesis of the Book of Poems (Shijing, China’s oldest collection of poetry). Polemically, I suggest that the poetics of the second century b.c. and its interest in metaphoricity originated in a philosophizing on the phenomenon of illusion, that is, the deceptive resemblance (si 似) between disparate objects, the discrepancy between appearance and actuality. Illusion, thus defined, may appear as confusing or uncanny in everyday experience (the doppelgänger, the mirage, the philanderer posing as saint, or linguistic ambiguity); yet this clash between form and content is an essential aspect of Confucian ritualism (li 禮), observable in rules of mourning, or in the use of metaphorical poetry as ritualized discourse. [End Page 251]

1. Traditional Sinology and Early Chinese Poetics

1.1. Historical Background to Early Han Dynasty Poetics

The Book of Poems (Shijing) contains 305 of China’s oldest poems, dating from circa 1000–600 b.c. (the Zhou dynasty).1 I shall offer an analysis of how these were interpreted by Confucian scholars during the early Han dynasty (224–6 b.c.) and how Confucian poetics related to, and emerged from, certain philosophical discussions on ontology, epistemology, and language during the Warring States period (475–221 b.c.).

In the Book of Poems we find historical pieces, many of which narrate the foundation of the Zhou dynasty in 1099 b.c. We also find hymns once performed as parts of rituals in ancestral temples.2 Other poems are voiced by senior officials remonstrating with their lord, concerned that his behavior will ruin the state; still others describe, in a straightforward manner, the love between man and woman. As early as Confucius (551–479 b.c.), the Poems were considered a source of ancient wisdom, and a fifth-century-b.c. author calls the anthology a “storehouse of righteousness.”3

During the Han dynasty—a formative period for Chinese culture, both intellectually and with regard to social institutions—the early Zhou dynasty was regarded as a model of virtuous statecraft, a golden age ruled by sages. Along with other Zhou texts, the Poems were canonized (jing means “canon” or “sacred book”) and subjected to systematic commentaries by Confucian scholars.

In this context, the love poems posed an awkward problem: how could one find ancient wisdom in these descriptions of raw, lowly passion? To escape from this predicament, Confucian hermeneuts (like Judeo-Christian readers of the “Song of Songs”) allegorized and metaphorized the Poems: the boy-girl affair was interpreted as an allegory of the relation between a ruler and his subject (see section 3.2).

Most ingenious in the metaphorical exegesis of the Shijing was Mao Heng (early second century b.c.), allegedly a student of the great Confucian philosopher Xunzi (ca. 335–ca. 238 b.c.).4 Mao’s hermeneutic masterstroke [End Page 252] was to separate the natural scene that opens the typical Shijing poem from the human situation that the poem “is about” and to call the natural description a “xing” 興 (to raise; to begin).5 In Mao’s interpretation, the xing always relates to human action: the virtuous bird described in a poem resembles (ruo) and so makes us associate it with the virtuous young woman whom the poet praises.

In its final version, the Mao Edition of the Poems (Maoshi) included Mao’s Commentary and two texts by other authors: the “Great Preface” (Maoshi daxu), a short text on the origin, nature, and function of the Poems, and the “Minor Prefaces” (Xiaoxu)—short comments on the meaning and historical background of each poem. During the Han dynasty, the Mao Edition became the standard redaction of the Poems, and it is the only one to have survived intact.

In the chronicles and philosophical works of the Warring States period, quotations of the Poems appear in ways that obviously anticipate Han dynasty hermeneutics and its metaphorical mode of interpretation. For instance, Mr. Zuos Commentary on the Spring and Autumn Annals (Zuo zhuan) records many instances of diplomats discoursing by way of poetry: the custom was called fushi because these dignitaries would quote (fu), a stanza of a well-known poem (shi) that metaphorically and ideally corresponded to the present situation.6 Thus, by quoting a love poem whose female addresser says, “If you, boy, won’t come to me now / I’ll wade across the river to find myself another beau,” the diplomat would allegorically be saying, “If you, mighty King, will not protect our humble state / We’ll make a pact with your neighbor.” Here, as in later interpretations of Shijing love poetry, the male and the female roles of an amorous relationship stand for, respectively, the ruler and his subject.

Lastly, although I will emphatically question its alleged impact on Han dynasty poetics, the theory of “correlative cosmology”—to use the conventional sinological term—was undeniably an important part of much philosophy in the periods with which we are concerned.7 It sought to reveal the hidden system of “categorical correspondences” (lei) according to which the seemingly disparate phenomena of the world were organized.8 Mr. Lüs Annals (the heterodox compilation of philosophical essays from [End Page 253] 239 b.c. we shall use below) also contains several chapters on this subject.9 Here the notion of “cosmological response” (ying) is exemplified by musical harmony: just as a gong tone played on one lute “responds” to a gong played on another, so objects and phenomena of the same category (lei) naturally and spontaneously bond (Lüshi: 676, 1360).10 In other contemporary texts, the gong tone corresponds to, and forms one category with, earth (among the five elements), yellow (among the five colors), the ox (among the five domestic animals), and the impluvium (among the five parts of a mansion) (Granet 1999 [1934]: 309). A similar relationship exists between man and “Heaven” (or rather “nature,” tian), from which omens descend to foretell major changes (Lüshi: 675). Later in the Han dynasty, the concepts Yin (the soft, dark, passive, feminine), Yang (the hard, bright, active, masculine), and the Five Phases (the interactions of earth, fire, water, wood, and metal) were systematically used to explain the “logic” underlying the world’s changes.

1.2. Polemics with Traditional Sinology

As I aim at a new understanding of early Han dynasty poetics, let me first characterize the conventional sinological view of it. In this view, Chinese lyric (shi), emerging spontaneously from a bard in response to the immediate situation, polarizes with Western poetry, calculated and fabricated by a cunning craftsperson.11 The difference in origins supposedly entails a difference in rhetorical and epistemological status: Chinese lyric inherently lacks metaphors, those Western tropes that produce (new) meaning by drawing similarities between the seemingly disparate. To a reader steeped in the Western literary tradition, this theory feels strangely counterintuitive, for [End Page 254] early Chinese literature and hermeneutics at least seem to overflow with so-called metaphors, parables, allegory, allegoresis; it values indirect expression and is constantly concerned with the relation between appearance and actuality. Are these tropes mere hallucinations of an Occidental mind desensitized by the philosophy—and poetics—of Metaphysics?12

Sinology has traditionally provided three overlapping explanations for the supposed absence of metaphor. First, the spontaneity, and with it the absence of premeditation, in lyric inspiration simply preclude the construction of abstract tropes. Second, the origins of poetics in China have been sought in its early philosophy, and their linkage contrasts with that between Western poetics and its philosophical sources: “figurative meaning cannot be conceived of independently from a certain worldview” (Jullien 2000 [1995]: 166). With remarkable synergy, the notion of correlative cosmology has come to define the fundamental difference between the Chinese and the Western systems of thought. Against the Occidental, Platonic division of the world into phenomena and ideas, sinologists pose a cosmology whereby things of the same category (lei) constantly interact and exert influence on one another.13 According to this model, the Occidental notion of metaphor is inapplicable to the Chinese figures of speech that to the untrained eye may resemble it. In Western poetics, the ontological difference between the earthly chair and the ideal chair yields the division of the metaphor into sensuous vehicle and abstract tenor: “The metaphorical exists only within the boundaries of [Western] metaphysics” (Heidegger’s Der Satz vom Grund, [End Page 255] quoted approvingly by Yu 1987: 17).14 In creating a metaphor, the Western poet thus abstracts meaning from one domain and imports it to another. By contrast, since Chinese cosmology not only lacks the notion of a “higher reality” but also assumes that all things spontaneously slot into natural categories, Chinese lyric can only link (images of) objects metonymically, hence also the gulf in the appreciation of the “literary work.” Whereas the Westerner searches for a spiritual meaning abstracted from the letter, the Chinese tradition accepts the poem as a truthful, literal expression of a historical person’s actual experiences: “it is the writer” (Owen 1992: 27).15

Third, it is claimed that China’s budding linguistics, just like its poetics, was in thrall to the cosmological thought pattern so that it operated with only two terms, “name” (ming 名) and “object/referent” (shi 實). The bond between Chinese names—“immutable cosmological entities”—and their referents was considered the unavoidable result of a spontaneous natural process, unlike the Western sign, in which the concrete signans and the abstract signatum coincide arbitrarily (Miller 1993: 425).16 A philosophical system based on these assumptions tends not to problematize “truth” or to distinguish false appearances from the “reality” they would be misrepresenting (Ames and Hall 1998: 110). [End Page 256]

1.3. Cosmology and Poetry

Thus Chinese lyric is said to lack metaphoricity, abstraction, and the status of the Western poem as an object “made.” Moreover, the structure of the Chinese language itself is taken to guarantee the bond between word and referent. Let us briefly elaborate on these claims to facilitate the ensuing discussion of the “poetics of cosmology.”

The sinological vision of a bard responding to the world in spontaneous poesy has been founded—incorrectly in my opinion—on the notion of “cosmological response” discussed above. The Chinese bard is depicted as an empty, depersonalized medium for cosmological influences; he or she produces lyric that is of the same kind as the situation that stirs him or her into lyric action: sad times thus produce sad songs and happy times happy songs.17

To learn how such a cosmological poetics works in practice, let us consider an example drawn from its most sophisticated exponent, Pauline Yu. The first piece of the Book of Poems, in Mao Heng’s interpretation, describes a “virtuous” bird as an image of the equally “virtuous” young lady whom the poem celebrates. For Yu, however, this is not a case of metaphoricity. Why? Simply because the bond between bird and woman, she claims, must be understood as having existed prior to and independent of the poet’s description of it: “the connections between subject and object . . . are viewed in the Chinese tradition as already preestablished” and “analogies already exist, to be discovered by the poet, not manufactured” (Yu 1987: 33; 1981: 224). Consequently, the Chinese “poet” is not a clever maker of metaphors: he or she merely responds in literal language to a world already full of correlative correspondences, an already categorized world.18 We are thus faced with a different way of constituting, or handling, similitude.19 A nonidealistic tradition (such as pre-Buddhist China) can do no more than assemble objects “of the same kind” into categories (lei), metonymically, like a child putting apples in one basket and pears in another. Epistemologically speaking, this Chinese trope does not, and cannot, lie. When the English speaker says that he “fancies that bird,” he is literally lying, since he is not referring to a fowl but to a woman. By contrast, when the Chinese bard depicts a certain bird [End Page 257] known for its virtuous behavior in order to indicate the virtue of his or her female protagonist, he or she is merely presenting one-half of a cosmological unity.

The “Romantic” notion of the inspired bard involuntarily emitting lyric, producing natural objects rather than fabricated artifacts, is a cherished topic in twentieth-century sinology. One scholar (Chen Shih-hsiang 1974: 387) describes a primeval Chinese bard for whom “to speak was to be a poet,”20 while another deprives the poet even of the faculty of discovery (granted by Yu): the poem “compose[s] itself through the poet” (Mair 1983: 3).21 This theory has been reinforced by misreadings of central texts from the same period, most notably the “Great Preface” and the “Minor Prefaces.” I have elsewhere tried to demonstrate that these two texts in fact describe poetry (shi) as consciously fabricated (zuo) tokens of ritualized communication, designed to express the poet’s aim (zhi) in an oblique manner, so as to bring about a change in the world without offending the addressee (Svensson 1999). In Western terms, shi is rather classicist more than romantic in its cool, calculated use of metaphors. It is profoundly rhetorical in its concern with form, with the possibilities and limits of language, with the processes of coding and decoding.

1.4. The Paradox of the Chinese Illusion

In what follows, I shall question the impact of correlative cosmology on early Chinese poetics, arguing that a debate on illusion constitutes the intellectual background to Mao’s hermeneutics and that a rhetoric of illusion is already implicit in the Poems themselves. More generally, I shall counteract the blurring of philosophy and rhetoric and problematize the reduction of the Chinese and Western traditions to a simplistic dichotomy of cosmology versus metaphysics.

What, then, defines illusion, and what relevance would this concept have for our understanding of early Chinese philosophy and poetics? There is in early philosophy a desire to draw clear-cut boundaries between the myriad things of the world. There is, conversely, a profound anxiety associated with the blurring of identities, and with objects that bear too strong a resemblance to each other (such as twins, or the doppelgänger and his or her original). [End Page 258] The deceptive resemblance between a thing and its debased copy is referred to by the character si 似, which frequently means “as if “ but whose usage as a philosophical concept comes close to the modern, pejorative notion of illusion: an unwanted discrepancy between appearance and actuality. Indeed, the all-important Mencius (mid–fourth century b.c.) quotes Confucius himself as saying: “I detest that which [only] seems to be, but is not [si er fei 似而非]. I detest the bristlegrass, fearing that it will disorder [luan] the young plants [which its leaves resemble]; I detest the lip server, fearing he will disorder Righteousness; I detest the glib tongue, fearing it will disorder Trust; I detest the [lascivious] tones of [the state of ] Zheng, fearing they will disorder [true] Music; I detest purple [as a mixture of red and blue], fearing it will disorder the [pure] Red; I detest the vulgarian with an aim to please, fearing he will disorder Virtue” (Ruan 1979 [1815]: 2780; all translations are mine, unless otherwise indicated).22

This classic passage prompts two comments. First, Confucius focuses here on individual objects and phenomena, trying to determine the status and identity of each. Clearly, the endeavor to distinguish the genuine article from the illusory imitation is beyond, beside, or simply different from the task of finding cosmological correlations that, according to contemporary sinology, determined Chinese poetics at its origins. Second, in his catalog of things that “seem to be but are not,” Confucius refers both to natural objects (plants and colors) and to humanmade fabrications (hypocrisy, second-order music, and lies). In this view, illusion may occur both spontaneously in nature and through human calculation: it is everywhere to be feared. Yet in spite of the negative phrasing, it is in these speculations on perverted language use, and on how nature sometimes imitates itself to a point of confusion, that we can begin truly to fathom the background to Mao Heng’s hermeneutics.

To indicate the linkage between early Chinese poetics and the philosophy of illusion, I shall develop my argument as follows. I will begin by discussing two tales of mistaken identity (one spectral, the other semiotic) from the essay “Mistrusting Resemblances” (“Yi si” 疑似)in Mr. Lüs Annals; then I will contrast the suspicion of illusion, duplication, and ambiguity expressed there with the idea that illusion is a prerequisite for ritual activity—an assumption implicit in Xunzi’s roughly contemporary “Discourse on Ritualism.” [End Page 259] In a remarkable analysis of Zhou dynasty funeral rites, Xunzi argues that ritualized illusion is not a corrupt copy to be exposed and discarded but a unique mode of expression, a rhetorical way of conveying something beyond the realm of (everyday) life. This theory, I argue, is fundamental for our understanding of early Chinese poetics and aesthetics. Indeed, with the conscious play on—the domestication of—that which “seems to be but is not,” Xunzi establishes a rhetoric and a metaphorics that cannot be accommodated within the prison of correlative cosmology.

I shall then return to the Annals and demonstrate how the positive notion of illusion also determines their theory of metaphor, tacitly developed in the essay “Treasures of Another Kind.” Having thus distinguished between the damnable everyday illusion and the illusion favored by rites and witty language, I will finally proceed to Mao Heng’s discussion of poetry qua ritualized discourse and try to unravel the poetics underlying his Commentary on the Book of Poems. Guided by the preceding discussion, I shall work my way toward the fourth stanza of “Zheng yue” (“Fourth Month,” poem 192), to the optical illusion described therein, and to Mao’s remark thereupon. By the time I reach the conclusion, I hope to have indicated a linkage between Mao Heng’s hermeneutics and his concept of the xing, on the one hand, and the philosophy of illusion as developed by the Annals and Mao’s teacher Xunzi, on the other.

2. Metaphor and Illusion in Third-Century China

2.1. Word and World

Like most literature surviving from the Warring States period (475–221 b.c.), Mr. Lüs Annals [Lüshi chunqiu] deals mostly with ethical and “political” problems: What are the duties of a ruler and the ruler’s ministers? Is rebellion against a dissolute royal house permissible? How does one, at the same time, maximize the happiness of the populace and keep them from growing too strong? Not surprisingly—since a peaceful society was thought to depend on, and establish, orderliness and precision—a prevalent topic is the necessity of unity and, correspondingly, the exigency of eliminating ambiguity, deviation, and excess.

As an example of the book’s political theory, we may consult an essay suitably titled “Sticking to One [Principle]” (“Zhi yi” 執一):

An army must be unified by a general, a state by a ruler, and the whole world by an emperor. The emperor must stick to one [principle] to keep the world together. Order follows from oneness; chaos follows from twoness. [Yi ze zhi, liang ze luan 一則治兩則亂].

(Lüshi chunqiu 1988: 1132) [End Page 260]

For reasons that will be apparent later, a variant of the final phrase appears in a similar chapter titled “No Twotiming” (“Bu er” 不二) as “Order follows from oneness; chaos follows from deviation [or ‘otherness’ yi 異]” (ibid.: 1134).

The work also discusses the impediments to unity (or “oneness”). The dry and pragmatic talk of “oneness” and “twoness” is directly related to a more disturbing theme, namely the paranoid suspicion that every single thing may be duplicated, that is, falsified with respect to its true self. Indeed, at one vertiginous point, the Annals come to question the very idea of cosmological categories (lei): “Things are often ‘so’ according to category [lei], but in fact are not ‘so’ at all 物多類然而不然” (ibid.: 1642).

The chapter named “Mistrusting Resemblances” (“Yi si”) is entirely devoted to the problems of divergence and duplication.23 “The similarity in appearance [si] between different objects is the cause of great confusion,” the opening line states (ibid.: 1497). Sometimes, the text continues, human perception cannot distinguish between appearance and the inner, true nature of an object. The deceptive play between surface and being may arise at all times and at all levels:

The [deceptive] similarity between ordinary stones and jade is what worries the trader in jade . . . .That people with shallow understanding but with sophistical discourse may seem to possess thorough knowledge, is what worries the virtuous ruler. The ruler of a declining state only seems wise and its ministers only seem loyal. The similarity between disparate things puzzles the ignoramus and spurs the sage to intense brain-racking. Thus, Mozi wept when he came upon a fork in the road [qi dao].


The image of the great thinker Mozi (fifth century b.c.) at the forked road sums up the entire discussion: Mozi weeps because the wrong road is, according to the appearance presented to his senses, identical to the correct one but leads to an unwanted, wrong destination. In a nutshell, the formula of Mozi’s conundrum, and of similitude in general, is that one form (shared [End Page 261] by the identical roads) corresponds to two different contents (the two destinations). And in the two folksy tales that follow this passage, the same formula is developed into an embryonic semiotic theory:

In Claremont . . . lived a ghost fond of assuming the shape of a passerby’s son, grandson or brother. One day, a fellow from the village travelled to the market and, upon his drunken return, the ghost assumed his son’s shape, led him by the arm and abused him. At home, and no longer drunk, the man approached his son: Im your father! How could you be so cruel! Why did you make fun of me when I was drunk! His son wept and kowtowed: No! I did nothing of the kind! Yesterday I went to collect debts in Eastville, just ask anyone! Persuaded, the father commented: Damn! It must be that ghost Ive been hearing about. He decided to return to the market the following day for another drinking bout, hoping to meet the ghost again and, if given the chance, stab him to death. And so, the next morning he went and got drunk again. The son, meanwhile, fearing that his father would be unable to return home, went out to meet him. But as the father saw his real son he drew out his sword and stabbed him. The old man killed his own son because his wits had been confused by the duplicating ghost.

(Ibid.: 1498)

Das ist des Pudels Kern!” We are beginning to recognize a pattern in this chain of events. The son’s appearance has been usurped, doubled, and when the father sees his son’s countenance on his first return from the market, he is the victim of an uncanny illusion. Once the father has realized that an alien entity has assumed his son’s shape, the stage is set for catastrophe. Now the story, as hitherto told, will repeat itself but with a stand-in for the ghost and so with a bitter twist. The second time the father returns to the village and sees his son’s shape, he will once more fall victim to an illusion, for two reasons. First, he has come to the bizarre conclusion that his son’s appearance can be spectrally duplicated and that the link between form and content is thus arbitrary. Second, the earlier series of events repeats itself: in the morning he goes to the market again, gets drunk again, returns home and—again—meets someone resembling his son. Our tragic antihero is so confident that he has learned the rules of the new game (son equals ghost) that he forgets that his son’s shape is now radically ambiguous and can contain the ghost as well as the real son. This second time, the ghost’s trick is not to assume the son’s shape.

This exemplary tale of illusion merits a translation of its formula into semiotic terms. What has been called “form,” “appearance,” “outer shape,” or “surface” corresponds to what pre-Saussurian semiotics termed signans, the concrete and material form of a sign(the actual sound or written mark).24 [End Page 262] Similarly, the “content,” the “thing,” the “essence” presented by way of that form or surface may be understood as the semiotic signatum, the “meaning” of the sign, attached to its signans. The ghost story, we may say, tells a tale about a signans (the son’s shape) that, suddenly and with frightful consequences, acquires a second signatum (the ghost) adjoined to its conventional one (the “being” of the real son).

In the present context (“Mistrusting Resemblances”), this story could be understood allegorically as a warning about a ghostly (or indeed, spiritual) meaning that may take over a word in the hands of “people with a shallow understanding but with sophistical discourse.” A warning, that is, issued by dogmatic Confucian scholars against metaphorical, ambiguous, illusory language and in favor of the “correct name” (zheng ming) celebrated by Confucius, Xunzi, and other authors in the Annals.25

2.2. Xunzi’s “Agreement”

My semiotic translation above links up with one aspect of Xunzi’s theory of language, namely, his claim that the relationship between word and referent is not “inherent” (or “fixed,” gu) but established by “agreement” (yue 約; Xunzi jijie: 414). Consider the second tale of identities in flux in the “Mistrusting Appearances” chapter. There the incompetent last king of the Zhou dynasty, You (781–771 b.c.), and his depraved queen Bao Si are seemingly the protagonists. The story implies that what ended the Zhou dynasty was the altered meaning of an alarm signal, a semantic perversion brought about by Bao Si’s sexual desire:

The two capitals of the Zhou Dynasty, Feng and Hao, were both close to the territory held by the barbarous Rong people. Therefore, [the royal house] and the various vassals agreed on [ yue] building a fort by the Royal Road, and putting on its roof powerful drums that could be heard for miles. In the event of an attack from the Rong, the drums would spread the news and the vassals’ troops would hasten to assist their emperor. When the Rong finally attacked, King You beat on the drums, and the vassal’s troops duly arrived. Bao Si was exuberant and took great pleasure in the spectacle. King You desired his queen’s laughter, so he pounded the drums a great many times and the troops kept coming to the capital, only to find that there was no attack. It went so far that, when the Rong finally did attack again, no vassal responded to the king’s drum. The king expired [End Page 263] at the foot of Black Horse Mountain, and the whole world laughed at his death. With the “No Attack” he lost the “Real Attack.” 以無寇失真寇

In fact, the story is not about King You and Bao Si but about the fate of a sign, the drum signal. The tale’s plot furthermore is so perfectly analyzable in semiotic terms that the sinologist, in constant fear of accusations of hermeneutic colonialism, need not hesitate to apply this “Western” method to an ancient Chinese text. If the sound of the drum is the signal’s signans, its two signata are the two antagonistic meanings alluded to in the last sentence: “Real Attack” and “No Attack.” The shift in the meaning attributed to the signal may be regarded from the perspective of the interplay between addressor and addressee. At first, the signal signifies “attack” for both king and vassals. When the king desires to amuse Bao Si with the sight of soldiers on the move, thus breaking the original agreement, the signal has the signatum “attack” for the vassals and the signatum “no attack” for the king and queen. Finally, the tables are turned and the signal signifies “attack” for the king and “no attack” for the vassals.

From this I draw three conclusions. In the semiotic system implicit in this text, a sign consists of two parts, identifiable with the signans/signatum dichotomy described above. Second, the relation between the two components—form and content, word and meaning—is arbitrary and explicitly described as the outcome of an “agreement” (Xunzi’s yue) between two or more parties, hence the possibility of a multitude of signata. Finally, this agreement formula stands in firm opposition to cosmological language theories claiming that the relationship between ming and shi, word and object, is not arbitrary but that the two are bound together by a cosmic link prohibiting any gliding of meaning.

In this tale, the movement between convention and deviation takes place, if not yet at the level of language proper then at least at the level of another representational system. The “(deceptive) resemblance” (si) of which the chapter’s title speaks refers here to the alarm signals outward resemblance to itself, while its meaning is ever volatile. Through human manipulation the percussional sign deviates from its customary meaning while it retains its habitual appearance (the drum sounds the same), very much like the ghost’s manipulative replication of the merchant’s son (the ghost looks like the original). In sexual terms, the signans copulates with two signata simultaneously, a promiscuity with disastrous results. Moreover, and not without interest for us, what the Annals describe—in panic and in warning—is the formula for metaphor. [End Page 264]

2.3. Xunzi’s Funeral Scene: Illusion Becomes Mimesis

So far, the Confucian voice has spoken anxiously about the sign whose customary usage is abruptly altered so that “war” suddenly means “peace” and a father suddenly kills his beloved son. Any everyday visual or aural impression, this voice tells us, may be an illusion deviating from the order of things as we have come to know them. Death, likewise, breaks the everyday pattern and puzzles us. And perhaps to some extent our confusion stems from the perplexity we feel vis-à-vis death’s own illusion, its cruel little joke, the corpse. The “remains” of a person are precisely that exterior form, that image, which remains after its content is gone. In semiotic parlance, the corpse is a signans without a signatum, or is with a new and more abstract signatum. The body on the slab resembles the person who walked among us but is, in fact, not.

Xunzi’s “Discourse on Ritualism” (“Li lun” 禮論) contains a fascinating description of the funeral objects with which Zhou dynasty graves were furnished. Unlike most contemporary treatises on the subject, this passage goes on to explain the strange nature of the interred goods and interprets their significance as ritual objects. Xunzi probably did so because the meaning of the by then ancient and ossified burial rituals was not wholly clear to the people of the third century b.c., who at the same time assumed that every step, gesture, and movement in these rites somehow embodied the wisdom of antiquity. It is well known that the concept of “ritualism” (li) occupied a preeminent position in Confucian thought and that ritualized behavior involved not only a precise adherence to age-old ceremonies but, more specifically, a partiality for, so to speak, an “indirect” mode of expression. I already have drawn attention to the practice of “poetry quotation” (fushi) in the Warring States period, when diplomats would quote out of context a well-known poem that would thus gain a new, metaphorical meaning. Now I will argue that a similar logic is at work in Xunzi’s hermeneutics of entombment.

The immediate occasion for Xunzi’s treatise seems to have been the philosopher Mozi’s radical demand to discontinue the absurdly expensive and excessive funeral rites advocated by the Confucian camp (Xunzi jijie: 371). Xunzi responds that even birds and beasts mourn their dead by returning to their old dwellings (gu xiang), howling and moaning at the place they used to share with the dead one, before they can leave the sorrow behind (qu zhi). And if this is so, Xunzi asks, how could humans, with their greater intelligence, refrain from treating a dead person as well as an alive person? On a casual first reading, this seems to be the message of Xunzi’s essay, yet its real argument is quite different and is much more refined.

The funeral rite, Xunzi begins, [End Page 265]

embellishes death with life. One should accompany the dead [to their final resting place] in much the same manner as one did when they were alive, following the same principles in death as in life, in their absence as in their presence: one principle for the beginning and the end.

(Ibid.: 372; translation following Dubs 1928: 235)

Were it not for the notion of “embellishing” the mortal realm with objects drawn from the realm of life, this opening passage might read as describing the denial of the death of the loved one—a grave misunderstanding.26

Immediately after death has occurred, bathe the head and body, tie the hands, and put food in the mouth, just as in life [xiang sheng zhi]. If you do not wash the hair, then wet it and comb it three times only. If you do not bathe the body, then wet a towel and wipe the body three times only.


But to our bewilderment, in the next passage the principle of “as in life, so in death” is suddenly challenged:

Fill the ears and put in the ear plugs. Provide raw rice for food, put withered bones in the mouth, going against the way of doing things in life [ fan sheng shu].

(Ibid.; emphasis added)

From now on Xunzi is concerned with the significance of the rites for the mourner, not the dead. As he later says (ibid.: 366), they serve to “emphasize [or ‘reduplicate’] the grief [zhong/chong ai 重哀],” referring perhaps to the painful repetition of death’s advent that will eventually turn into an acceptance of the new situation and a reorientation for the mourner—the process Freud called a work of mourning (Freud 1982 [1917]).

Alerted by the tales of illusion in the Annals, we recognize the form/content dichotomy as the central theme in the following passage, wherein Xunzi analyses—dissects—the spectacle of the Zhou grave chamber. Death, Xunzi claims, is ritually conceptualized and presented by a trompe l’oeil display in which everyday objects (clothes, jars, musical instruments, bamboo) have been rendered useless by the absence/destruction of an unobtrusive but vital component. Thus they have their usual appearance but not their usual being.

The clay water and wine jars are empty and not filled [xu er bu shi].There is a fine bamboo mat but no bed. The wooden articles have not been finished, the pottery articles have not been shaped into complete objects, and the articles made [End Page 266] of bamboo and reeds have no core. The wind-instruments are all present, but not attuned to each other; the zither’s strings are there but have not been tuned in concert. The chariots are interred but with the horses facing the wrong direction. This announces that [these objects] are not used [gao bu yong ye]. The “objects used in life” [sheng qi] go into the grave, symbolizing the shift from one road to another [xiang xi dao 象徙道]. [These objects] are full of omissions and never brought to their perfection [lüe er bu jin]. They have appearance, but no utility. The chariots are led to the tomb and buried, [but] the metal rein-ends, the reins and the horse collars do not follow, in order to show [ming] that [these objects] are not used. The announcement of the shift of road and the display of uselessness all serve to emphasize/double one’s grief. Hence, the “objects used in life” have a certain polish but no utility, and the “luminous objects” [ming qi]have outward appearance, but no usefulness . . . .Thus the interment is the respectful concealment of [the person’s] form [xing].

(Xunzi jijie: 368–71; translation following Dubs 1928)

This prosaic translation at least hints at the highly rhetorical style of the original: the repetitions, the puns, the subtle analogies. And, obviously, Xunzi’s analysis of how appearance relates (to) actuality is very different from, and more complex than, that found in the Annals.27 True, the funerary goods have the appearance of their mundane counterparts in the warm sun, but a human, premeditated act has deprived each of its original being (its function as a usable utensil) in order that it be turned into a sign, to be made to “stand for something else.”28 If the Annals abhor the discrepancy between form and (conventional) content, Xunzi’s description of a very similar process rather praises it as part of the rhetoric of Ritualism—as a truer way of speaking (of) death. The funeral chamber is thus not a parasitic duplication of an original, “real” room. Rather, it offers a rhetorical statement.

Xunzi’s analysis urges us both to focus on the being of the illusion itself and to ask ourselves what the movement from functional object to uselessness actually means. An answer emerges from the last sentence quoted above, which speaks of the corpse as the form (xing 形) of the dead person: the form of the once living individual is interred along with the forms (mao, synonymous with xing) of the once usable objects. The funeral goods with their eerie ambience are thus not only illusory images of functional utensils but, moreover, imitations of the corpse, declaring it now useless and without function. But why this doubling, this second (and literal) reproduction that the death objects perform, this time of the corpse? We find the answer in [End Page 267] Xunzi’s remark about the “doubling of grief “ (chong ai): the unperfected funeral goods serve as a guarantee that the mourner will not reverse his or her journey into the underworld. Otherwise, the mourner might mistake the objects in the crypt for usable objects and thereby be induced to redirect his or her attention from the work of mourning to everyday life.29 (The real reason why the objects were left half-finished may, of course, have been that they simply were not intended for any practical purpose but were buried as symbolic utensils to be used in the afterlife.)30

At a more general level, Xunzi conceptualizes the phenomenon of death, not by anthropomorphizing death into the Grim Reaper but, negatively, by signifying death as the lack of an unspecifiable entity, rendering the dead person useless, just like a chariot that cannot be properly used without horses. This is further illustrated by the funeral objects (fashioned after the corpse) that “have appearance, but no utility” and, of which, the jars—empty [End Page 268] of wine and water—vividly exemplify the deceased person’s absence from the pleasures and necessities of daily life.

But the third act of imitation, this time of illusion itself, remains to be accounted for. The flaw in the objects that illustrates the corpse’s uselessness can be detected only at close range. But when we—having understood the flaws in the utensils—return to our original, distant point of view, the elaborate display of faulty goods (i.e., the creation of an illusory scene) suddenly illustrates the illusion produced by the corpse itself. Just as, from a distance, the burial goods may look like everyday, functional objects, so a reclining corpse may be taken for a sleeping—but living—body. The funeral objects thus mimic both the corpse’s emptiness or uselessness and its uncanny, illusion-inducing resemblance to its former owner. Moreover, the flickering between illusion and illustration means that the two are coupled but never simultaneous. The rhetorical effect is, instead, now here/now gone; now living/now dead. If we push Xunzi’s interpretation of the burial scene to its extreme, then this eerie vacillation between appearance and reality generates a sublime mimesis of the mourner’s despair: it cant be true/it is true; she cant be dead/she is dead.

We learned from the Annals that an illusion is the traumatic disturbance of the conventional meaning routinely attached to an external appearance. Here, Xunzi rather celebrates the change of meaning and function in the funeral goods. Why? Two things make this change understandable. The first is the ritual setting, which allows, and quite possibly demands, deviations from the normal laws of literalness (as in the fushi practice). Second, a playful answer is provided by the text itself: the funeral objects, it says, go into the grave, “symbolizing the shift from one road to another.” In order to explain how the ritual perversion of everyday objects produces new meaning, Xunzi plays with the ambiguity of the word dao: “road,” “way,” “method,” “the Way” (Tao), “order,” “speech.” Thus, the phrase xiang xi dao comes to mean “representing the change of dao.” In other words, the funeral objects not only signify a shift from the material road with which we associate the horse and chariot or, metaphorically, from the old order made obsolete by the arrival of death; they also indicate a change in the conventional mode of speech. Within the death ritual, the straightforward language of the living is no longer applicable simply because such language cannot adequately describe death or the lack death produces in the human body. Instead, we must resort to the ritualized and metaphorical language that emerges from the illusion of the corpse (exactly the claim that Xunzi proves in the pun on dao).

The possibility of a crossing over from illusion to illustration is significant, for it can serve as a basis of a (revised) theory of early Chinese rhetoric. Here unconventional (“metaphoric” or “symbolic”) meaning is achieved [End Page 269] through an act of deprivation or depreciation that brings about a momentary illusion which, in its turn, produces the deviant meaning sought by the skilled poet. Accordingly, the Chinese rhetorician starts off with the full presence of a sign (a drum signal, a body, an object, a word), its form and its conventional meaning, only to eliminate the latter in a violent act that conserves the sign’s form but produces a new meaning. “Violence” relevantly connotes illicitness and intrusion: the shift of signification amidst continuity of form in the texts from Mr. Lüs Annals, and the body’s similar transformation into a corpse, are both described in negative terms. However, Xunzi’s analysis of the entombment rites and their production of meaning must be considered in a wholly different light, for in the realm of rites and death, the transformation of the form/content nexus is praised. Poetry, certainly, originates in a similar semiosis, but does that mean that poetic language too is a ghost language, a death language?

3. That Which “seems to be but is not”: Illusion as Trope in the Mao Commentary

3.1. Treasures of Another Kind

As a transition to poetry proper, within the Book of Poems, I will consider a paradoxical statement in the essay “Treasures of Another Kind” (“Yi bao” 異寶; Lüshi: 666). We have already encountered the “otherness” (yi), referred to in the essay’s title, in the motto “Order follows oneness; chaos follows otherness.” And in the Confucian dialectic of convention and deviation, yi (“to differ,” “deviate,” “other”) assumes the same role as xi (“to move”) and qi (“forking”) in that they all represent the divergence from a hitherto followed dao—a road, principle, or meaning. Remarkably, the following passage not only speaks in favor of “deviation” but also demonstrates, by its rhetoric, the need to deviate from conventional language.

As for the second part of the title, the character bao (“treasure”) is ideographic to the extent that it represents various valuables—jade, a cowrie-shell, and earthen goods—found together under a roof. I use the word ideographic (tabooed by mainstream sinology) not to suggest that the notion of valuable objects stacked under a roof could only mean “treasures” (it might just as well have signified “storehouse,” for example) but to indicate that the linkage between 寶 and “treasures” is not completely arbitrary.31 I would [End Page 270] also claim that, once the link between bao and its meaning has been established, the components of the character cannot but influence the definition of what a “treasure” is.

“Yi bao” starts by stating: “It was not the case that the [wise] men of yore lacked treasures [bao]. What they treasured was ‘something else’ 古之人非無寶也其所寶者異也.” Later, in the story of the righteous Wu Yun, who fled from the state of Chu because of its corrupt government, there appears an old fisherman who helps Wu although he would have been amply rewarded by the Chu ruler had he captured the fugitive. The implicit question is, what did this humble but wise fisherman consider a treasure? Obviously not the piece of material wealth conventionally associated with the word “treasure” and the character bao but rather the nobler and more abstract “treasure” of assisting an upright man: virtue, a treasure within quotation marks.

What they treasured wassomething else’ (qi suobaozhe, yi ye). The text literally says that what the ancient sages valued was yi, “otherness,” “deviation.” And this is critical for our inquiry into Confucian theories on language and rhetoric, for this phrase articulates the paradoxical duality of the Annals: they damn illusion’s intrusion into everyday matters while embracing its occurrence in (literary) language. The very concept of yi, under attack elsewhere in the Annals, is here paradoxically associated with the ancient sages; it is “treasured” for its power to enrich language by the linguistic illusion called metaphor. We find in “Yi bao” exactly the same mechanism as in, most obviously, King You’s manipulation of the drum signal away from its conventional signification to an “other” meaning but also in Xunzi’s analysis of the funeral rituals. The rhetoric of “otherness” begins by presenting us with an anomaly: objects for practical use in a tomb or material riches in the hands of an empty-handed fisherman. And so this leads to an instant of conceptual chaos—the fisherman’s (material) treasure is an illusion— which transforms into clear, logical meaning once we discover that the entombed goods have been deprived of their usefulness or the “treasure” of the precious objects gathered under a roof. Thus, the illusion of the fisherman’s material wealth turns into a perfect illustration of his spiritual wealth. [End Page 271] In Xunzian terms, the sign (like the funerary object, the corpse sign) must “die” in order to be able to signify at another level.

Two further comments are called for here. First, despite the suggestion that transgressing conventional language is traumatic, these texts do not reject the power of language per se or assert that we would be better off without language (as with the celebration of silence in the Analects of Confucius, the Taoist Laozi, or, say, in the eighth of Rilke’s Duineser Elegien).32 Second, the ambiguity to which yi refers is primarily dialectic in that the sign’s new meaning is the negation of its original one. The fisherman’s spiritual “treasure” is the antithesis of a material treasure; the drum signal can mean only “attack” or “no attack”; Xunzi’s burial goods, formerly associated with “life,” now mean “death.” The similar usage (called ironic by the Western rhetorician) of the signans “treasure” in the sense of “antitreasure” furthermore suggests that, in the impoverished age of the narrator, the virtue of the sages could be conceptualized only as negative greed: his language had no positive name for that X which characterized the upright people of yore. (The same logic is at work when we, negatively, express our regard for a unique piece of art by calling it priceless [cf. Burke 1966: 419–28].)

3.2. The “xing”

The object of attention in the crowning part of my argument is a “tropological movement”: a subtle, rhetorical play on, and with, images of nature that Mao Heng detects in a Shijing poem, and that is directly linked to the themes of illusion and lying, and to the pattern of norm and deviation. Furthermore, this “movement” is intimately related to the trope that Mao Heng calls xing 興 (see 1.1).

Rather than a “concept” in the modern Western sense, Mao’s xing is better understood as a powerful method—or several slightly different methods—of transforming passionate love poetry into Confucian dogma. What will concern us below is the xing as explained in Mao’s meticulous comment on “Osprey Calling”—where natural imagery and human situation faithfully mirror each other—but also the negative turn into which Mao is forced when confronted with poems in which descriptions of nature and human being stand in ironic opposition. As demonstrated below, if the positive version of the xing (as in “Osprey Calling”) pertains to the pattern of illusion/illustration in using natural imagery as signs of “something else,” then the negative version doubles this operation by exposing the emptiness of the rhetorical gesture, the sheer hollowness of the metaphoric signans. [End Page 272]

3.3. “Osprey Calling”

The very first item in the Book of Poetry, “Osprey Calling,” begins thus:

關關雎鳩 Guan guan [cries] the osprey
在河之洲 On the islet of the river
窈窕淑女 The chaste and good girl—
君子好逑 A fine mate for the lord.
(Wang Xianqian 1988 [1915]: 8–9)

Mao comments on the two first lines:

This is a xing. “Guan guan” are harmonious calls. Although the osprey . . . is a passionate bird, the males and females live in separation from each other. The queen-consort [i.e., the “chaste and good girl”] takes joy in the virtue of her lord. Nothing is not harmonious or not in concord, and their sexuality does not degenerate into wantonness. [It is] sincere and firm, profound and deep, like [ruo] the ospreys’ living in separation. Things being like this, they can be used to influence and change the world.


Here we find the concepts that determine the Commentarys reading of the Poems: sexual moderation, hierarchical rigor, and social harmony. The negativity of Mao’s idiom (“no wantonness,” “nothing is not harmonious”) presents disharmony, fornication, and chaos as the prima materia to be fought and negated. Thus Mao underscores the fragility of the “kingly way” and discloses the Confucian fear of sexuality, excess, licentiousness, disturbance of hierarchy. (Sex, we observe, is the very pivot of human society: if man and woman only regulate their sexual behavior, the primary condition for a paradisal society is fulfilled.)

“Osprey,” in Mao’s reading, refers not only to the bird described as passionate yet chaste and virtuous. It has a second and more important signatum: the queen-consort (i.e., the “good girl” of line three) celebrated for her chasteness. The bird and the woman are linked in the text by means of similitude, since they allegedly share the characteristic of having a well-ordered sex life. This linkage is manifested quite clearly by Mao’s use of the metaphorical copula ruo 若, “like” or “as.” Although “Osprey Calling” itself formally juxtaposes nature and humankind, from Mao’s “xingish” perspective the bird qua bird is nonessential; its only function is to bring out the “virtue” he attributes to the queen. If it did not have a feature pertinent to the context, the osprey’s presence in the poem would in Mao’s eyes simply be incomprehensible—its crying cacophonous.

The “xingish” poem thus banks on a rhetorical moment of suspense positioning the reader between illusion and illustration, appearance and being. Like the funerary goods in Xunzi’s essay, the xing sign (i.e., the imagery of [End Page 273] the first two lines) retains its form while deprived of its conventional content and so acquires new signification. On a first reading, the xing is construed as a literal description of an osprey. But the Confucian reader’s second glance grasps the purely rhetorical connection between nature and humankind, decodes the metaphorical meaning, and finds therein the poem’s message of virtue.33

3.4. Negation: The Ironic Xing

This model (natural descriptions reflecting human situations) accurately describes the majority of figural interpretations that Mao calls “xing.” But a Shijing poem may playfully negate this scheme.

Consider the first stanza of “On the Mountain Grows Nutgrass”34 (“Shan you fusu,” poem 84):

山有扶蘇 On the mount there’s nutgrass
隰有荷華 in the swamps there are lotus flowers
不見子都 I do not see Zi Du
乃見狂且 I only see a mad man!
(Ibid.: 354–55)

Mao comments that the first two lines are “a xing . . . ,saying that high and low, big and small all obtain what is appropriate [for them]” (ibid.). In Mao’s semiotic system, the high mountain signifies high position (and thus “ruler”), while the low-lying swamps refer to the lowly minister. What this xingtrope is metaphorically describing, then, is the fortunate situation wherein a Superior Man (junzi) rules and the petty man (xiao ren) occupies a lowly position, just as the nutgrass and the lotus flower are found in their respective natural and proper habitats (Mao Commentary, ibid.).

Accustomed to the rhetorical code of “Calling Osprey,” the reader would expect lines three and four to offer a positive reflection of the xing by describing an orderly human situation. Ironically, however, these lines describe a state of confusion where the high position due the righteous person Zi Du has been usurped by a “mad man.” Thus, when the stanza has been read in its entirety, we understand that the xing’s “promise” of human order [End Page 274] was but illusory. But why this rhetorical twist? Mao’s brief comment underscores the text’s ironic tension, asserting that Zi Du was “well-loved and praised by his time” (ibid.). Thus, the narrator expects an upright man on the throne and is shocked to find it occupied by the “mad man.” The inversion of the order of things, however, is manifested rhetorically, since the xing itself frustrates the reader’s belief that he or she had cracked the code. In other words, the perversion of the xing model here subtly illustrates the perverted ways of a chaotic world. Likewise, the focus on the xing’s illusory being echoes the concern with lies and false appearances found in “Mistrusting Resemblances.” Like the xing in “Nutgrass,” the “mad man” assuming the posture of a True King is only a hollow image, an illusion. We soon will see how very central the appearance/actuality, lie/(metaphoric)truth, and xing/ironic-xing dichotomies are to Mao Heng’s poetics.35

3.5. Lies and Metaphoric Truth

The Shijing poem “Fourth Month” (“Zheng yue,” poem 192) is a long lamentation about governmental chaos. According to the extant commentaries, during the early Han dynasty the poem was understood as a critique directed by high-ranking noblemen (dafu) against the same King You who brought about the end of the Zhou dynasty by beating the war drum one time too many. It is a reading not unwarranted by the poem itself, for the eighth stanza does mention the wicked Bao Si, by whom, in Mao’s words, “King You was led astray” (ibid.: 670). For us, however, the question is not the exact temporal or geographical situation referred to but the literary technique employed by the narrator. For, although the Commentary does not define it as a xing, “Fourth Month” persistently employs a figure of speech that at one point prompts Mao to provide us with a clue to his poetics.

The poem starts by describing a disturbing anomaly:

正月繁霜 It is the fourth month, ample is the frost
我心憂傷 my heart—grieved and hurt
民之訛言 The false words of the populace
亦孔之將 are heavily exaggerated.
(Ibid.: 665) [End Page 275]

The zheng month (literally the first month), Mao says, refers in fact to the fourth month, a time of year that normally does not see any frost.36 We should immediately distinguish between two possible readings. On the one hand, we could interpret these lines literally and decide that the speaker is shocked and distressed at nature’s untimeliness, a deviation that, according to the ancient Chinese mantic tradition, could be interpreted as a disastrous omen.37 But we could also understand the image of frost in the fourth month as a trope used by the skillful poet to set the tone for a poem describing a world out of joint and deviating from the path of true virtue. As readers of the poem, we do not have to make a choice but can let the letter and the spirit exist side by side. As for our understanding of Mao’s reading of “Fourth Month,” what speaks against the rhetorical, more sophisticated interpretation is, of course, the plain fact that Mao does not label the image in the first two lines a “xing,” as he would normally do to indicate a phrase that should not be taken literally. Yet considering Mao’s comments on the later, similarly structured stanzas, we need to take a closer look at the lines.

If a poet represents the ground as covered by hoarfrost, not as an actual description of late spring but purely as an image of deviation and unnaturalness, would that poet not be engaging in rhetoric? And what is rhetoric’s master trope, the metaphor, if not literally a lie? And what theme do the last two lines introduce if not lying? And can it be a mere coincidence that the character eh 訛—meaning both “lie” and “rhetorical persuasion”—consists of the two components “word” (yan 言) and change (hua 化), thus indicating both a “transformed,” “perverted” word and a transformation brought about by speech? At this stage, let us merely remark that it would be amusingly ironic if the first two lines should literally be a lie, while the ensuing two lines accuse other people of lying.

As suggested by my polemical account of the poetics of cosmology, the distinction drawn between rhetoric and literalness is not mere scribble in the margins of the grand tradition of Shijing interpretation but the actual bone of contention between two factions of sinology. While I am inclined to describe both the Poems and their early Han dynasty exegesis as refined, rhetorical, and metaphorical, the general tendency is to present them as literal, concrete, and the result of a spontaneous outburst from the poet.

When the theme of lying reappears in the fifth stanza, it is impossible to [End Page 276] avoid the suspicion that the author is, in fact, using a rhetorical lie in his denunciation of linguistic trickery.

謂山蓋卑 [They? I?] say “a mountain is low”
為岡為陵 [But? That?] there are ridges and cliffs
民之訛言 The false [“transformed”] words of the populace
寧莫之懲 Why has no one put a stop to them?
(Ibid.: 667)

This is effective stuff: only a regime led by a corrupt ruler such as King You can claim that “up” is “down” and that a mountain—by definition high, with ridges and cliffs—is low. As with “treasure” in the Annals, here is an instance of linguistic abuse. (In the next stanza, the “populace” is reported as saying that heaven is “high” and that the earth is “thick”—unmistakable truths—but at that stage even the most self-evident statement by these language perverters can no longer be trusted.)

But are we to take this alleged quotation at face value? Should we really believe that the “populace” (a derogative term for immoral nobles) is literally trying to convince the narrator that this blatantly false statement is true? And why talk about mountains in the first place? I submit the hypothesis that the poet purposely chose “mountain” as a token of something high, paired it with the adjective “low” and put the entire contradictory phrase in the mouth of his enemies to illustrate their perverse and absurd behavior.38 Moreover, the “mountain-is-low” concept is perfectly consistent with the ironic vacillation between norm and deviation, expectation and disappointment that “Fourth Month” takes as both its theme and its rhetorical technique. The only conceivable drawback, from the poet’s perspective, is that he himself must be untruthful in his description of the perverters of truth.

If, for the poet, the phrase “a mountain is low” was an obvious example of a lie (the eh mentioned in the third line), then Mao’s unobtusive but remarkable comment on the first line displays a radically different attitude toward “lying,” since it demands that the key statement be taken metaphorically: “On the throne is not a Superior Man, but a petty man” (ibid.: 667). In Mao’s reading, what the phrase says is that the King (metaphorically represented by the mountain) is not the majestic figure he is supposed to be (the “mountain” is low). He may have the position of a king but not the [End Page 277] essence, just like the “mad man” in “Nutgrass.” More important than this familiar split between appearance and being is the fact that what the poet considered a lie Mao regards as true on a metaphorical level. No matter who— the poet himself or the “populace”—is speaking here (and Mao is unclear on this point), what Mao accepts as a legitimate mode of poetic expression is precisely the speech act that the poet condemns.

This insight gives rise to a new set of questions. What kind of lie is tolerable and even lauded, and which lies are despised? Might the schizophrenic word eh (“transformed speech, lie, persuasion”) help us understand the rhetorical principles followed by the Shijing poet and the Confucian hermeneut?

The eleventh stanza—exceptional in its illustration of threat and terror—is also based on the pattern of convention and deviation:

魚在于沼 The fish is in the pond
亦匪克樂 yet cannot rejoice
潛雖伏矣 Although by diving down it can hide on the bottom
亦孔之炤 yet the [light] greatly illuminates it
憂心慘慘 The grieved heart is in pain,
念國之為虐 thinking of the cruelty carried out by the state.
(Ibid.: 672; translation following Karlgren 1950)

As expected by now, the first line is deceptive, describing the fish in its proper and natural environment. The second line ironically deviates from the joy and harmony promised by the first line, while the third and fourth lines lead us down to the depths of the lake and also to a submeaning hidden below the textual surface. When “Crane Calling” (“He ming,” poem 184) says that “fish are swimming in deep waters/or [in the shallow water] at the islets,” Mao interprets it as a metaphor of the Superior Man (junzi) who should be “hidden yet obvious.”39 Mao’s argument is that the actions of a sage should influence the world without an obvious manifestation of power. The Confucian junzi should be but not be seen, and Mao accordingly comments that “fine fish hide in the depths; small, inferior fish swim around the islets” (Wang Xianqian 1988 [1915]: 672). In this metaphor of a state in turmoil, the “fine fish” hides at the bottom but remains threatened by an all-penetrating sunlight—a curious harbinger of Paul Celan’s Lichtzwang— bringing to vulgar attention the creature driven by righteous instincts. The conclusion is that this stanza is not a literal account of the behavior of fish [End Page 278] but a metaphor of how a corrupt government disturbs the natural behavior of the virtuous person and invades his or her most intimate space.

3.6. Cosmological Poetics and Abstraction

For the proponent of a cosmological poetics, the first and eleventh stanzas of “Fourth Month” literally describe omens sent down by Heaven as indications of the impending fall of the grand Zhou dynasty. That is, the cosmological worldview made the poet assume that such anomalies were commonplace when a ruling dynasty was about to fall. Consequently, the disturbed fish does not resemble Mao’s Superior Man in his distress at the corrupt regime. Rather, both fish and human are simultaneous, but different, symptoms of the disturbance of nature that coincides with the declining dynasty. Nor would a cosmological poet or hermeneut make an ontological distinction between fish and human in this text. The fish would remain a fish without being abstracted into a mere image of a human. In other words, natural phenomena and human actions coexist in the text on equal terms. The poet (or Mao) would link fish and human metonymically—not metaphorically, since they are parts of the same string of occurrences: in Pauline Yu’s terminology, they are of the same categorical correspondence (lei). The sinologist, this approach implies, must not succumb to the seduction of the Occidental tradition, with its penchant for abstraction and symbolism, on pain of falsifying the wholly other Chinese tradition by finding therein the cunning metaphors of the Western poet (Yu 1981: 219; Yeh 1985).40

Yu’s thesis has already been problematized by our quotation from the Annals (“things often appear to be ‘so’ according to category [lei], but are in fact not ‘so’ at all” [Lüshi: 1642]). Moreover, Mao’s claim that the woman in “Calling Osprey” “has the virtue [de] of the osprey” indicates that the xing enabled the Confucian hermeneut to link nature and human by a tertium comparationis (“virtue”) that has been abstracted from the literal (or “concrete”) descriptions of nature.

That a process of abstraction is an indispensable part of the xing is conclusively demonstrated by Mao’s interpretation of a natural image in “Sturdy” (“You bi,” poem 298). A description of zealous officials in a palace is interrupted by the following lines: “Flock-wise, the egrets / the egrets go alight” (Wang Xianqian 1988 [1915]: 1068). Mao comments that “the egret is a white bird. This [image] is used as a xing for the pure, white official” [End Page 279] (ibid.). The process of abstraction—“white” being the actual color of the bird and, with reference to humans, a synonym of “purity,” “immaculateness”—could simply not take place if the natural description were not subordinated to human action in the interpretative act. Here, the fallacy of cosmological poetics is apparent. If we accept the idea of correlative cosmology as the thought pattern underlying Mao’s hermeneutics, the white bird must be construed as morally “pure” (which is what neither poem nor commentary says), or the official in question must literally be “white” (through fear or paint).

3.7. That Which Seems to Be but Is Not

In his comment on the fourth stanza of “Fourth Month,” Mao is driven to pinpoint, identify, and name this pattern of promise and disappointment, of naturalness and deviation. The first lines go:

瞻彼中林 I look at the middle of that forest,
侯薪候蒸 [there is] only firewood and brushwood
(Ibid.: 667)

Mao (followed by all major later commentators) links this passage to Confucius’s outburst against false appearances quoted above and comments that “‘Firewood and brushwood’ means that this looks like [a forest] but that it isn’t [si er fei 似而非]” (ibid.).41 Thus, the description of the illusory forest pertains to the pattern of things deviating from one’s expectations in opposing genuineness and solidity (trees, forest) to lightness and inferiority (firewood as an inferior “version” of trees). Referring to the degeneration of the state, the poem says that what may look imposing from the outside is, on a closer look, mere rubbish: an illusion that vanishes once scrutinized. But Mao’s (or, rather, Confucius’s) words about that which “seems to be but is not” (si er fei) also apply to the rhetorical apparatus itself.

The first line describes how the narrator’s eye registers what appears to be a forest. That image is then contradicted by the statement that the “forest” consists merely of firewood and brushwood. Like tales in the Annals about the ghost and the drum signal, Mao focuses on the dialectic between appearance and essence, between seeming to be and being. For the rhetorician, it is precisely by describing a situation (that which is) through an incorrect word that the poet can produce an illusion and thereafter a metaphorical meaning. In this context, then, the word “forest” (lin) is a misnomer. A forest [End Page 280] is, by definition, an assemblage of trees, and a “forest” lacking trees cannot, strictly speaking, be a forest. However, it is not nature (or “Heaven”) that lets us down by deviating from its own presumed naturalness (as might have been argued with reference to the untimely frost described in the first stanza). Here, the deviation is calculated and takes place at the level of language or, in other words: at the rhetorical level.

The shift from conventional to rhetorical meaning is much more striking in the original because of the ideographic nature of the two characters bao and lin. As bao depicts certain valuable items under a roof, so the character that I have translated as “forest”—lin 林—is a recognizable depiction of two trees. Just as the poet wished, we actually see the trees of the forest and can hardly believe our eyes when, in the next line, we encounter only fire-wood and brushwood. This is the rhetorical moment put fully into practice, when the reader faces a paradox: is it a forest or just firewood and brushwood? But, again, the narrator is himself “at fault” for misnaming the firewood a forest. Unless we believe that this stanza is actually about a man mistaking firewood for a forest and that he wrote down his experience second by second, the first two lines must be part of the well-known rhetorical pattern of convention and deviation. Now the poet may be accused of the very crimes that he spends thirteen carefully structured stanzas denouncing: lies and usurpation. In Mao’s reading, in order to say (yan) that things today are not what they seem to be, the poet foregrounds a word whose conventional denotation he then abuses: he purposely calls the firewood and brushwood a “forest.” The si er fei, the linguistic illusion, has become a humanmade trope.

The poet who indulges in metaphoricity is thus a “misnamer,” a disturber of the alliance between signans and signatum, and the character for misnaming is precisely eh, “transformed speech,” “lie,” or “persuasion by words.” When the poet, in the fifth stanza, alleges that the “populace” says “the mountain is low,” he exposes a despicable act of lying, nameable by the word eh, via a sophisticated trope of persuasion that could also be referred to as eh.

The illusion that rhetoric generates is thus actually appreciated and useful as metaphorical meaning. Rhetorical tropes are lies for a good purpose, in contradistinction to the destructive “mutated speech” of ghosts, evil kings, or the “populace,” the meaning of which cannot be elevated into a higher significance. Consider now the first and last stanzas of “High South Mount” (“Jie nan shan,” poem 191), wherein it is clear that eh, in its benevolent form as persuasion, includes the accepted form of lying that we call metaphoricity (ibid.: 657–64).

In the final stanza of “High South Mount,” the narrator himself suddenly takes the floor and presents his aim: “I, Jia Fu, made this recitation . . . to [End Page 281] change your heart by words / to make you protect and cultivate the ten thousand city-states.” “To change by words” is my translation of eh as it appears in this context.42 According to Mao’s Commentary, and to what this lengthy poem itself reveals about the situation in which it was voiced, the narrator here is a nobleman of the declining Zhou dynasty exhorting the Grand Master Yin of Zhou to lead a virtuous life that the populace may emulate.43 To see how this admonition is carried out, consider the two lines that open the poem: High and steep is South Mount / With its stones piled high. A remonstration in ten stanzas directed to one of the most powerful men of the ruling dynasty would not begin with a literal—and therefore, in this context, nonsensical—description of a pile of stones. Instead, and most probably as an accurate reconstruction of the author’s intention, Mao interprets the majestic mountain as a “xingish” metaphor for the Grand Master Yin (Wang Xianqian 1988 [1915]: 657). Thus, just like the fushi, the rhetorical eh involves metaphoricity and analogy.

Furthermore, the formula of “si er fei,” by which Mao has instinctively described the rhetorical mode of “Fourth Month,” extends to the two tales in the Annals of the signal that “seemed to be” what it was not (i.e., an alarm signal) and the external shape of a person that was not what it “seemed to be” (first the beholder’s son, then a ghost). The formula also accurately describes Xunzi’s mise-en-scène of objects seemingly intended for the deceased but possessing a barely discernible flaw that resolves the paradox and transforms them (having “appearance but not utility”) into an illustration of the corpse and thus of death itself. It also describes the illusory use of “treasure,” a word that seems to refer to material riches but, in fact, does not.

Returning to my quarrel with cosmological poetics, it is the case that, in both classical Chinese and the English version, the phrase “si er fei” expresses, superfluously, what the notion of si, or “resemblance,” must denote on its own, namely, an external similarity between two (or more) objects. But by adding . . . er fei (“ . . . but not being”), Mao emphasizes the ontological inferiority of the thing “that resembles”: it can only delude the hasty, first glance but never the scrutinizing second look, for it is not the object that it resembles. A true cosmologist would never, like Mao Heng, take such an obsessive interest in illusion (often humanmade, as with the drum signal, [End Page 282] the funeral objects, the “treasure,” the “mountain,” and the “forest”) or in the power of representational systems to produce illusions and thus metaphors.

4. Conclusion

As an alternative to the prevalent poetics of cosmology, I have attempted to demonstrate how warmly ritual, poetic, and hermeneutic literature during the third and second centuries b.c. embraced the rhetorical play with literal and metaphorical meaning.

In “Mistrusting Resemblances,” and in Xunzi’s analysis of Zhou dynasty funerary rites, we found descriptions of two seemingly different events that, in fact, were identically structured. Both involved the doubling of a familiar object’s exterior, which resulted in the beholder’s inability to distinguish between original and copy. What may blind us to the similarity between the two kinds of illusion is the difference in the authors’ value judgment of it. Xunzi’s brilliant and poetic analysis of the funeral rites indicated that illusion—so feared in the Annals—could be usefully manipulated to express the otherwise inexpressible, and I claimed that this manipulation of signs is a rhetorical act. Xunzi’s insistence on the difference between being (or function) and form confirmed this hypothesis, and his interpretation of the useless funeral objects as an illustration of the useless corpse emphasized the critical rhetorical moment, when the very hollowness of the illusion is revealed and turned into a metaphorical truth. In the Annals’ “Treasures of Another Kind” we were surprised to find a full-fledged theory of metaphor as an awkward but necessary linguistic deviation. This enabled us to demarcate a space within Confucian thought where rhetorical and linguistic illusions were not dreaded but actually welcomed, yet often indicated by words and concepts denoting deviation and even falsehood: yi, xi dao, qi dao, eh. What makes some illusions so loathsome is their interference with and disturbance of everyday life, whereas in language, ritual, and poetry the illusion becomes acceptable.

What characterizes the illusion and generates its rhetorical potency is the deviation from the beholder’s expectations. In our analysis of Mao’s reading of “Fourth Month” we made the important discovery that the principle of deviation links and unites the three concepts eh, si er fei, and xing. The radical ambiguity of eh, meaning both “lie” and “rhetorical persuasion,” illustrated with the utmost clarity the Confucian weakness for the linguistic illusion, the transformed and/or transforming word.

Finally, let us see how illusion appears both as a theme and a rhetorical strategy in another Shijing poem. “The Great East” (“Da dong,” poem 203) [End Page 283] constantly contrasts appearance with actuality to illustrate how the state of Zhou has deviated from its former Way (dao), which was “[Smooth] as whetstone/straight as an arrow/Trod upon by the Superior Man/[merely] looked upon by the petty man” (Wang Xianqian 1988 [1915]: 727).

This poem, narrated by an impoverished and bitter aristocrat, begins with an image—a xing, according to Mao—that suggests wealth and abundance:

有饛簋飧 Full of cooked grain is the vessel
有捄棘匕 long are the thornwood ladles

How does this image of food utensils relate to the ensuing descriptions of empty spinning machines, suffering nobles, and the employment in high offices of lowly sons of “boatmen” and “servants”? Why this detailed depiction of full vessels and ladles? It is only in the final two stanzas that the imagery of utensils is resumed, bearing a striking resemblance to the broken goods in Xunzi’s grave chamber:

維南有箕 In the south there is the Winnowing Basket
不可以簸揚 but you cannot winnow with it
維北有斗 In the north there is the Ladle
不可以挹酒漿 but you cannot ladle wine or congee with it.
(Ibid.: 734)

The description of the two celestial constellations, the Winnowing Basket and the Ladle, is imprinted with the rhetorical pattern of that which “seems to be but is not.” The first line introduces a winnowing basket, and the second line mockingly gives us one that cannot winnow; the third line presents a ladle, and the fourth line gives a nonladling ladle.44 The dialectics of promise and disappointment thus explains the function of the first lines: just as the seventh stanza displays a winnowing basket and a ladle only to reveal immediately their illusory character, the xing speaks of food and plentiful food containers so that the rest of the poem can negate this wealth. And, as with all examples in this article, the latter trope builds on a correction of an initial miscomprehension. On a first reading, the wealth described in the xing is oddly inconsistent with the rest of the poem. However, when the last stanza has exposed the hollowness of the heavenly food utensils (which “seem to be but are not”; which “have form but not substance”), we understand that the vessel full of grain is as illusory as the Willowing Basket in the heavens, thus illustrating the narrator’s wealth (now gone), and [End Page 284] the house of Zhou (“Zhou” being no longer the authentic, virtuous Zhou). Consequently, the reason that Mao defines the two opening lines as a xing is to emphasize their illusory character: they are just images, emptied signs, transformed words.

It is in this historical and intellectual context—the Confucian discourse on illusion, rhetoric, metaphors, and lies—that we should understand Mao’s xing and his hermeneutic project.

Martin Svensson Ekstroem
Chinese Studies, Stockholm University


Works in Western Languages

Ames, Roger T., and David Hall
1987 Thinking through Confucius (New York: SUNY Press).
1998 Thinking from the Han (New York: SUNY Press).
1995 The Poetics, translated by Stephen Halliwell (Cambridge: Harvard University Press).
Baudrillard, Jean
1993 [1976] Symbolic Exchange and Death, translated by Iain Hamilton Grant (London: Sage).
Birch, Cyril, ed.
1974 Studies in Chinese Literary Genres (Berkeley: University of California Press).
Blanchot, Maurice
1996 [1955] Lespace littéraire (Paris: Gallimard).
Boltz, William G.
1994 The Origin and Early Development of the Chinese Writing System (New Haven, CT: American Oriental Society).
Burke, Kenneth
1966 Language as Symbolic Action (Berkeley: University of California Press).
Celan, Paul
1970 Lichtzwang (Frankfurt am Main: Suhrkamp).
Chen Shih-hsiang
1974 “The Shih Ching: Its Generic Significance in Chinese Literary History and Poetics,” in Birch 1974: 8–41.
Connery, Brian
1998 The Empire of the Text: Writing and Authority in Early Imperial China (New York: Rowman & Littlefield).
De Groot, J. J. M.
1989 [1892] The Religious System of China, 6 vols. (Taipei: Southern Materials Center, Inc.).
Deleuze, Gilles
1990 [1969] The Logic of Sense, translated by Mark Lester (London: Athlone).
Derrida, Jacques
1974 [1971] “White Mythology: Metaphor in the Text of Philosophy,” translated by F. C. T. Moore, New Literary History 6(1): 5–74.
Djamouri, Redouane
1993 “Théorie de la ‘rectification des dénominations’ et réflexion linguistique chez Xunzi,” Extrême-Orient Extrême-Occident 15: 55–74.
Dubs, Homer H., trans.
1928 The Works of Hsuntze [Xunzi] (London: Probsthain).
Fenollosa, Ernest
1983 [1936] The Chinese Written Character as a Medium for Poetry, edited with a foreword by Ezra Pound (San Francisco: City Lights). [End Page 285]
Freud, Sigmund
1963 [1919] “Das Unheimliche,” in Das Unheimliche: Aufsätze zur Literatur, 45–84 (Hamburg: Fischer).
1982 [1917] “Trauer und Melancholie,” in Studienausgabe, vol. 3, 197–212 (Frankfurt am Main: Fischer).
Genette, Gérard
1995 [1976] Mimologics, translated by Thaïs E. Morgan (Lincoln: University of Nebraska Press).
Graham, Agnus
1989 Disputers of the Dao (La Salle, IL: Open Court).
1991 “Reply to Roger T. Ames,” in Philosophical Texts and Contexts: Essays Dedicated to Agnus Graham, edited by Henry Rosemont, 286–91 (La Salle, IL: Open Court Press).
Granet, Marcel
1999 [1934] La pensée chinoie, with a preface by Léon Vandermeersch (Paris: Albin Michel).
Hansen, Chad
1992 A Daoist Theory of Chinese Thought: A Philosophical Interpretation (Oxford: Oxford University Press).
Henderson, John
1984 The Development and Decline of Chinese Cosmology (New York: Columbia University Press).
Huang Chichung
1997 The Analects of Confucius [Lun yu] (Oxford: Oxford University Press).
Jullien, François
1985 La valeur allusive: Des catégories originales de linterpretation poétique dans la tradition chinoise (Paris: Ecole Française d’Extrême-Orient).
1989 Procès ou Création: Une introduction a la pensée des lettrés chinois (Paris: Seuil).
2000 [1995] Detour and Access: Strategies of Meaning in China and Greece, translated by Sophie Hawkes (New York: Zone Books).
Karlgren, Bernhard
1931“The Early History of the Chou li and Tso Chuan texts,” Bulletin of the Museum of Far Eastern Antiquities 3: 1–58.
1964 Glosses on the Book of Odes (Stockholm: Museum of Far Eastern Antiquities).
1972 [1957] Grammata Serica Recensa (Stockholm: Museum of Far Eastern Antiquities).
Karlgren, Bernhard, trans.
1950 The Book of Odes [Shijing] (Stockholm: Museum of Far Eastern Antiquities).
Kern, Martin
Forthcoming “Shijing Songs as Performance Texts: A Case Study of ‘Chu ci’ [Thorny Caltrup],” Early China (25).
Knoblock, John, trans.
1988–1994 Xunzi: A Study and Translation of the Complete Works (Stanford, CA: Stanford University Press).
Legge, James, trans.
1960 [1871] The Chinese Classics. Vol. 4, The Shi King [Shijing] (Hong Kong: Hong Kong University Press).
Levi, Jean
1993 “Quelques aspects de la réctification des noms dans la pensée et la pratique politiques de la Chine ancienne,” Extrême-Orient Extrême-Occident 15: 23–54.
Levin, Samuel
1977 The Semantics of Metaphor (Baltimore, MD: Johns Hopkins University Press).
Lewis, Mark Edward
1999 Writing and Authority in Early China (New York: SUNY Press).
Liu, James J. Y.
1975 Chinese Theories of Literature (Chicago: University of Chicago Press). [End Page 286]
Loewe, Michael, ed.
1993 Early Chinese Texts: A Bibliographical Guide (Berkeley, CA: Society for the Study of Early China and the Institute of East Asian Studies).
Mair, Victor
1983 “The Narrative Revolution in Chinese Literature: Ontological Presuppositions,” Chinese Literature, Essays, Articles, Reviews (CLEAR) 5(1–2): 1–28.
Marney, John
1993 Chinese Anagrams (San Francisco: Chinese Materials Center).
Miller, Roy Andrew
1975 “The Far East,” in Current Trends in Linguistics, edited by Thomas A. Sebeok, 1213–64 (The Hague: Mouton).
1993 “Shih ming,” in Loewe 1993: 424–28.
Needham, Joseph
1956 Science and Civilisation in China, vol. 2 (Cambridge: Cambridge University Press).
Owen, Stephen
1992 Readings in Chinese Literary Thought (Cambridge: Harvard University Press).
1953 The Cratylus, translated by H. N. Fowler (Cambridge: Harvard University Press).
1952 Guide to Kulchur (London: Owen).
1970 [1916] A Memoir of Gaudier-Brzeska (New York: New Directions).
Pound, Ezra, trans.
1954 Shih Ching: The Classic Anthology Defined by Confucius (Cambridge: Harvard University Press.
Reding, Jean-Paul
1985 Les formations philosophiques de la rhétorique chex les sophistes grecs et chez les sophistes chinois (Bern: Peter Lang Verlag).
Riegel, Jeffery
1997 “Eros, Introversion, and the Beginnings of Shijing Commentary,” Harvard Journal of Asiatic Studies 57(1): 143–77.
Rilke, Rainer Maria
1962 [1922] Duineser Elegien (Frankfurt am Main: Insel-Verlag).
Röllicke, Hermann-Josef
1992 Die Fährte des Herzens: Die Lehre von Herzensbestreben (zhi) im Grossen Vorwort zum Shijing (Berlin: Dieter Reimer Verlag).
Rosemont, Henry, ed.
1991 Philosophical Texts and Contexts: Essays Dedicated to Agnus Graham (La Salle, IL: Open Court).
Saussy, Haun
1993 The Problem of a Chinese Aesthetic (Stanford, CA: Stanford University Press).
Schmidt-Glintzer, Helwig
1990 Geschichte der chinesischen literatur (Bern: Scherz).
Schwartz, Benjamin I.
1985 The World of Thought in Ancient China (Cambridge: Harvard University Press).
Sebeok, Thomas A., ed.
1975 Current Trends in Linguistics, vol. 13 (The Hague: Mouton).
Steiner, George
1975 After Babel: Aspects of Language and Translation (London: Oxford University Press).
Svensson, Martin
1999 “A Second Look at the Great Preface on the Way to a New Understanding of Han Dynasty Poetics,” Chinese Literature, Essays, Articles, Reviews (CLEAR) 21: 1–33.
Thompson, Stuart E.
1988 “Death, Food, and Fertility,” in Watson & Rawski 1988: 71–108. [End Page 287]
Tiffany, Daniel
1995 Radio Corpse (Cambridge: Harvard University Press).
Van Zoeren, Steven
1991 Poetry and Personality: Reading, Exegesis, and Hermeneutics in Traditional China (Stanford, CA: Stanford University Press).
Vandermeersch, Léon
1993 “Rectification des noms et langue graphique chinoises,” Extrême-Orient Extrême-Occident 15: 11–21.
1994 Etudes sinologiques (Paris: Presses Universitaires de France).
Vico, Giambattista
1968 [1744] The New Science of Giambattista Vico, translated by T. G. Bergin and M. H. Fisch (Ithaca, NY: Cornell University Press).
Waley, Arthur, trans.
1987 [1937] The Book of Songs [Shijing] (New York: Grove Weidenfeld).
Watson, James L.
1988 “The Structure of Chinese Funerary Rites,” in Death Ritual in Late Imperial and Modern China, edited by James L. Watson and Evelyn S. Rawski, 3–19 (Berkeley: University of California Press).
Watson, James L., and Evelyn S. Rawski, eds.
1988 Death Ritual in Late Imperial and Modern China (Berkeley: University of California Press).
Yeh, Michelle
1985 “Metaphor and Bi: Western and Chinese Poetics,” Comparative Literature 39: 237–54.
Yu, Pauline R.
1981 “Metaphor and Chinese Poetry,” Chinese Literature, Essays, Articles, Reviews (CLEAR) 3(4): 204–24.
1987 The Reading of Imagery in the Chinese Poetic Tradition (Princeton, NJ: Princeton University Press).
Zhang Longxi
1985 “The Tao and the Logos: Notes on Derrida’s Critique of Logocentrism,” Critical Inquiry 11: 108–31.

Modern Chinese Texts

Fu Wuguan
1993 Lüshi chunqiu yu zhuzi zhi guanxi [The Relationship between Mr. Lü’s Annals and the other philosophers (of the Warring States period)] (Taipei: Tung-Wu Ta-hsüe).
Li Zehou & Liu Gangji
1987 Zhongguo meixue shi [History of Chinese aesthetics] (Peking: Zhonghua).
Qian Zhongshu
1979 Guan zhui bian [The pipe-awler’s chapters] (Peking: Zhonghua).
Wang Fanzhi
1993 Lüshi chunqiu yanjiu [Studies in Mr. Lü’s Annals] (Huhhot: Neimeng gudaxue).

Premodern Chinese Texts

Chen Huan
1967 Shi Maoshizhuan shu [Annotations on Mr. Mao’s Commentary on the Poems] (Taipei: Hsüe-sheng).
Kong Yingda
1979 Maoshi zhengyi [The correct meaning of the “Mao Poems”], in Ruan 1979 [1815]: 259–630.
Lüshi chunqiu
1988 Lüshi chunqiu [Mr. Lü’s Annals], modern annotation and commentary by Chen Qiyou (Taipei: Hua-cheng). [End Page 288]
Mao Heng
1979 Mao zhuan (The Mao Commentary), in Ruan 1979 [1815]: 273–628.
“Maoshi daxu”
1979 “Maoshi Daxu” [“Great Preface”], in Ruan 1979 [1815]: 269–273.
“Maoshi xiaoxu”
1979 “Maoshi Xiaoxu” [“Minor Prefaces”], in Ruan 1979 [1815]: 269–627.
Ruan Yuan
1979 [1815] Shisan jing zhu shu [The thirteen classics, annotated and commentated] (Peking: Zhonghua).
1979; 1988 Shijing (The Book of Poems), in Ruan 1979 [1815]: 273–628 and Wang Xianqian 1988 [1915].
Shui yuan
1987 Shuo yuan [Garden of persuasion] (Peking: Zhonghua).
Wang Xianqian (1842–1918)
1988 [1915] Shi san jia yi jishu [The Three Schools’ interpretations of the Poems, collected and annotated] (Taipei: Mingwen).
Xunzi jijie
1988 [1891] Xunzi jijie [The works of Master Xun, collected and analyzed], edited and annotated by Wang Xianqian (Peking: Xinhua).
Zheng Xuan
1979 Mao zhuan Zhen jian [Annotations on the “Mao Commentary”], in Ruan 1979 [1815]: 273–628.
Zhuangzi jishi
1961 Zhuangzi jishi [The Works of Master Zhuang, collected and explained] (Peking: Zhonghua).
Zuo zhuan
1979 Zuo zhuan [Mr. Zuo’s Commentary], in Ruan 1979 [1815]: 1712–2188. [End Page 289]

To Kam Louie and the Asian Languages and Studies Department at the University of Queensland, my deep gratitude for two delightful years as a research fellow, during which this article was written. Parts of it have been presented in Brisbane, Stockholm, Sydney, Hong Kong, Taipei, Columbus (Ohio), and Los Angeles. Thanks to participants there for reactions of all kinds. Special thanks to Hedvig Atmer, Rod Bucknell, Michael Fuller, Bo Holmberg, Perry Johansson, David Keightley, Björn Kellgren, Torbjörn Lodén, Simon Patton, Bengt Pettersson, Haun Saussy, David Schaberg, James Unger, and Zhang Longxi for valuable comments. I am grateful for the detailed critique coming from the four readers for Poetics Today and Meir Sternberg. All remaining faults and idiosyncrasies are my own.


1. The Shijing is familiar to students of Western modernism from Ezra Pound’s translation, Shih-Ching: The Classic Anthology Defined by Confucius (1954). Other, more scholarly translations include Karlgren 1950 and Waley 1987 (1937). James Legge (1815–1897), in Legge 1960 (1871), includes brief summaries of Confucian commentaries on the Poems as well as translations of the “Great Preface” and the “Minor Prefaces.” Karlgren 1964 provides detailed, comprehensive, and reliable linguistic glosses on the Poems. Below, I give the number of each Shijing poem according to the sequential order used in Karlgren 1950, Pound 1954, and Waley 1987.

2. On the Poems as (used in) ancestral rituals, see Kern 2000.

3. Zuozhuan (Mr. Zuos Commentary), in Ruan 1979 [1815]: 1822.

4. For the traditional genealogy of the Mao school of Shijing interpretation, see Karlgren 1931 and the entry on The Correct Meaning of theMao Poems” in the Siku quanshu zongshu, reprinted in Ruan 1979 [1815]: 259–60.

5. Compare Riegel 1997 for an interesting discussion on the possible origins of Mao’s xingish readings.

6. See Helwig Schmidt-Glintzer’s (1990: 27–35) chapter on the Poems and the allegorical tradition. Chapters 4 and 5 of Jullien 2000 [1995] contain a useful discussion of fushi, although wholly within the framework of correlative cosmology.

7. Henderson 1984 and Needham 1956: 279–88 outline the history of correlative cosmology in early China.

8. I follow Pauline Yu (1987: 41–43) in translating lei as “categorical correspondences.”

9. Apart from speculations on statecraft, language, music, ritualism, epistemology, and cosmology, the chapters of the Annals are concerned with agricultural, legal, and military matters. The Annals, it would thus seem, give a fairly comprehensive picture of the contemporary philosophical debate. See Lewis 1999: 302–8; Fu 1993; and Wang Fanzhi 1993. The book is named after Lü Buwei (d. 235 b.c.), a prime minister of the state of Qin, who had sponsored scholars and rhetoricians to come to the Qin court. Mark Edward Lewis (1999) suggests that the Annals were compiled as a compendium to be used in the education of the young prince of Qin. Less than two decades after Lü Buwei’s death, the first emperor of Qin would conquer the other Warring States and unify them into “China” (the English word probably derives from “Qin”). The state of Qin perished in 206 b.c.

10. The two chapters in Mr. Lüs Annals are called “Responding to the Same” (“Ying tong”) and “Attracting the Same Kind” (“Zhao lei”).

11. According to Steven Van Zoeren’s (1991: 110) reading of the “Great Preface,” classical Confucian poetics holds that shi is “spontaneous and unmediated by artistry or calculation.” For a critical discussion of this theory, see Svensson 1999. François Jullien (2000 [1995]: 164) says that “Chinese poetry is . . . perceived as a phenomenon of incitement and has not embraced representation.” Similar views are expressed in Jullien 1989: 249ff.; Li and Liu 1987: 577; Liu 1975: 63–64, 69–70, 119; Owen 1992: 27; and Yu 1987. For alternative positions see Connery 1998; Röllicke 1992; Saussy 1993; Wang Xianqian 1988 [1915]; Zhang 1985.

12. Many other “comparative” studies attempt to sum up the Western tradition of literary theory and philosophy, not seldom getting trapped in a false East/West dichotomy by finding in the Chinese tradition merely inverse versions of the theories, terms, and concepts attributed to its Western counterpart. Although I have refrained from a general discussion of Western theories of illusion and metaphoricity, I should like to acknowledge here my debt to the “Occidental” concern (or fascination) with illusion and duplication that starts with Socrates’ ironic yet puzzled reflection on names that are all too similar to their referents (Cratylus, Plato 1953: 432b–c), appears again in Freud’s (1963 [1919]) theory of the unheimlich double, and reappears with latter-day philosophers of “simulation and simulacra.” Yet rather than Baudrillard’s conservative skepticism toward “simulation,” it is in Deleuze’s embrace of the simulacrum—breaking away from Plato’s second-rate, parasitic mimesis to establish an autonomous mode of expression—a certain affinity with Xunzi’s analysis of the rhetoric of Confucian burial rites (Baudrillard 1993 [1976]; Deleuze 1990 [1969]: 69–71 [for Deleuze’s theory of “sense”], 256–55 [for a deconstruction of the Platonic dichotomy of copy versus simulacrum], and passim). In addition, Blanchot 1996 [1955]; Derrida 1974 [1971]; Genette 1995 [1976]; Steiner 1975: esp. 214–35; and Tiffany 1995—to name only the most obvious works—have constantly forced me to regard Chinese theories of representation from new angles.

13. Referring to Ames and Hall 1987, Agnus Graham (1991: 287) concluded that “In the Chinese cosmos all things are interdependent, without transcendent principles by which to explain them or a transcendent origin from which they derive. . . . It is also without transcendent ends.”

14. Yu is here quoting Jacques Derrida’s rather more reserved quotation of Heidegger (Derrida 1974: 26 n.22). Now, if we were to switch from an ontological to an epistemological perspective, some early Chinese comments on figurative language sound remarkably, and perhaps disappointingly, like Aristotle’s comment about observing the similar in the dissimilar (Aristotle 1995: 1459a). Speaking of comparison (to use the most neutral translation of the character bi, which could also be rendered as “simile,” “analogy,” or “metaphor”), the fifth-century philosopher Hui Shi is quoted as saying that the “skillful speaker [literally “persuader,” shuizhe] uses what is known to explain [ yu] the unknown, and so makes people know it” (quoted from the first-century-b.c. collection Garden of Persuasion [Shui yuan] 1987: 272). Notice the key role, not least stylistically, played by the concept of knowledge (zhi): one gains knowledge of the hitherto unknown by comparing it to what is already known. The “cosmologist” would, of course, maintain that Hui Shi sees no ontological difference between the known (corresponding to the metaphorical vehicle) and the unknown (the tenor). Yet Hui’s definition of comparison (bi) suggests that we could use aspects of Western rhetoric to understand, not merely misconstrue, Chinese poetry and poetics. As Derrida (1974 [1971]: 25 n. 22) points out, the notion of metaphor as a means to knowledge is crucial for traditional Western theory, and Heidegger’s categorical claim (in Der Satz . . .) that metaphor—allegedly the archtool of Western metaphysics—banks on a false dichotomization of the sensible and the nonsensible touches upon a feature of metaphor that is “not the only, nor the first, nor the most decisive.”

15. Jullien (2000 [1995]: 142) claims that the Chinese reader is affected by the poem in the same manner as the poet was affected by the world in the first place: “incited by the world outside, he [the Chinese bard] in turn stirs up the reader’s emotions.”

16. Compare Miller 1975: 1217: “No name of anything, no word in the Chinese language, was thought to be of and in itself arbitrary, or in anyway the result of an arbitrary agreement on the part of the society employing it. Everything in the cosmos and on earth was the way it was, and every word, or name, was the word or name it was, for a reason: and that reason was a reflection of cosmic order.”

17. Jullien (2000 [1995]: 142) describes the Chinese bard as “incited by the world outside . . . [I]n China, poetry arises from a relationship of incitement rather than from a method of representation; the world is not an object for consciousness but a partner with consciousness in a process of interaction.”

18. For the most stimulating and convincing critique of the poetics of cosmology to date, see Saussy 1993: chap. 1.

19. The traditional sinologist will no doubt even consider “similitude” too Occidental a concept: “[to] say ‘X is like/unlike Y’ . . . is to move away from correlative towards analytic thinking” (Graham 1989: 338).

20. Chen is here (unconsciously) echoing Giambattista Vico’s (1968 [1744]) theories on the origins of human language and how primeval humans could not but mix categories and speak in figures that resemble latter-day “metaphors.” See also Levin 1977 for the related theory that poetic metaphors are not fanciful descriptions of the everyday world but literal descriptions of an other and autonomous poetic world.

21. The claim that early Chinese thought lacked abstraction is a commonplace in sinology. Compare Graham’s assertion (1989: 218) that Laozi’s “metaphors” are not “illustrations” but rather “thinking itself.”

22. I follow Zhao Qi’s and Sun Shi’s commentaries as given in Ruan 1979 [1815]. Note that in the Analects of Confucius (Lun yu 17.17), in a chapter supposedly earlier than the Mencius passage discussed above, Confucius is quoted as saying: “I detest Purple usurping the place of Red; I detest the tones of Zheng disordering classical Music; I detest the glib-tongued toppling states and noble houses.” (Huang 1997: 170, translation modified.) The earlier passage thus not only differs slightly in wording but also lacks the important statement about that “which seems to be but is not.”

23. In this period, the character 疑 (“to mistrust,” modern pronunciation: yi) often was used for its near-homophone 擬 (modern pronunciation: ni), meaning “to estimate” (Karlgren 1972: 252) or “to resemble,” “to imitate.” Thus, the title of this essay (“Yi si”) could be rendered also as “Imitation and Resemblance,” as demonstrated by Chen Qiyou’s thematic and philological comparisons with texts from the so-called Legalist school (Lüshi: 1498–99). Chen’s claim is further supported by the use, in the same essay, of the phrase yi si in this sense: “What follows in the wake of imitation and resemblance [ yi si] must be carefully studied” (ibid.) and by a similar use in the Book of Xunzi (“Zheng ming,” Knoblock: 22.5). However, since yi unambiguously has the sense of “to mistrust” in another of Mr. Lü’s essays, “[On the Necessity of Being a] Cautious Listener” (“Jin ting”; Lüshi: 704), its use in the present title (“Yi si”) is polysemic and possibly intended as a pun.

24. Ferdinand Saussure’s most original contribution to semiotics is the claim that the sign signifies through its difference from all other signs in the langue.

25. Compare Hansen 1992: 4–5 and chap. 9, which describe Xunzi’s “authoritarian response” to Mozi’s “sophistic” language philosophy. As hinted above, just like the Book of Xunzi, Mr. Lüs Annals contains a chapter called “Correction of Names.” For concise, reliable introductions to the Confucian correction (or “rectification”) of names, see Schwartz 1985: 91–94; Djamouri 1993; Vandermeersch 1993; and Levi 1993. Ezra Pound’s delight in the Confucian version of the concept of le mot juste is expressed in chapter 1 of Pound 1952 (zheng ming being transcribed as ching ming).

26. I will not here discuss the difference between the articles intended “for the living” (sheng qi) and those “for the dead” (ming qi). It is clear from the rhymed, penultimate sentence (“the ‘objects used in life’ have a certain polish . . . ”)that, for Xunzi, the two categories of things are alike in their illusory imitation of usable objects. Ming (shem ming zhi; ming bu yong)may mean “illuminate, indicate.” Sheng may, of course, mean “raw, unrefined.”

27. In the essay “Against Physiognomy,” Xunzi rejects the idea that a person’s appearance corresponds to that person’s personality (using the ungainly but sage Confucius as a prime example). Xunzi jijie:72–88.

28. Compare Maurice Blanchot’s (1996 [1955]: 346–49) discussion of the image, corpse, and broken utensil in “Les deux versions de l’imaginaire.”

29. Contentwise, there is a slight but substantial difference between Xunzi’s speculations on the significance and semiotics of the funeral rites and those expressed in the three canonical books on rites and ritualism that originate from roughly his time, the Zhou li, the Yi li, and the Li ji (all three included in Ruan 1979 [1815]). On two occasions in the Li ji, Confucius makes passing comments on the use of the ming qi. In the first passage (“Dan gong,” Ruan 1979 [1815]: 1289), he alleges that these strange, unperfected objects are the results of a necessary compromise between treating the deceased as forever extinct by burying them without any funeral goods (“inhumane, not permissible”) and treating them as though they were still alive through the interment of fully usable objects (“dim-witted, not permissible”). Although terse and somewhat obscure, this paragraph suggests that the ultimate motive behind the burial of the quasi-perfected goods was not primarily to please the manes but to allow the living to keep within the limits of “humanness” (ren) and, thereby, the ritual system. Likewise, in the second passage (“Dan gong,” Ruan 1979 [1815]: 1303), Confucius says that makers of the ming qi show a true insight into the principles of mourning (sang zhi dao) when they do not render the funeral goods into usable objects and when they make effigies of chariots and horses out of clay and hay, respectively, instead of putting real horses and chariots in the grave. In a similar fashion, he condemns the use of wooden replicas (yong) as “inhumane” (bu ren), since they bear too strong a resemblance to people of flesh and blood and, therefore, to the archaic, barbaric, and “inhumane” custom of human sacrifice. It is apparent that what concerns Confucius in this context, and what stirs his sympathies and antipathies, is human-kind’s adherence to ritual conduct and to “humanness” and not a concern for the spirit. Now, if both Xunzi and Confucius are concerned with the importance of the burial rites for the living, they differ markedly in their interpretations of the ming qi and, in particular, of the reasons why they must vary from objects intended for use by the living. For Confucius, as we have seen, the unhewn objects have the double purpose of enabling a person to treat the deceased “humanely” without appearing foolish while, at the same time, signifying the distance from life and, therefore and more importantly, from human sacrifices.

30. This is the standard interpretation of the burial or burning of “mock money and paper models of items to be used in the afterlife” (Watson 1988: 13). Compare J. J. M. De Groot (1989 [1892], 1: 26), who speculates that the destruction by fire of the funerary goods may be a vestige of a “more ancient destruction of real property.” More in line with Xunzi’s analysis, Stuart Thompson (1988: 75) points out that “the breaking of rice bowls” may “symbolize the separation of the deceased from the living.”

31. Twentieth-century sinology has never recovered from Ezra Pound’s flamboyant claim that the Chinese ideographic character is the ideal medium for poetry (Fenollosa 1983 [1936]), nor from his claim (Pound 1970: 46) that his friend, the artist Henri Gaudier-Brzeska, could immediately understand “the primitive Chinese ideographs (not the later more sophisticated forms)” merely by looking at them. If the backlash against Pound in some cases threw the baby out with the bathwater, Léon Vandermeersch (1994) and John Marney (1993) have made important contributions to our understanding of Chinese epigraphy and literature by studying the origins of Chinese writing in divination (Vandermeersch) and various forms of pun (Marney). Hansen 1992: 35–39 includes important insights on the ideographic nature of Chinese writing. According to William Boltz (1994: 176), Chinese scholars of the third century b.c. reacted against “trends . . . toward pure phoneticazation” which threatened the “natural order that was expected [in accordance with a correlative cosmology] to obtain between word and characters” by adding “semantic classifiers” to the characters in use.

32. For Confucius’s praise of silence in the Analects ([Lun yu] 17:18), see Huang 1997: 170.

33. To discover the imagistic structure of a poem and to interpret it are, of course, two different acts. The irony of the xing is that, although invented as a means of transforming a non-ideological text into Confucian dogma, it can easily end up as a weapon against Mao himself. Once Mao has indicated the parallelistic and metaphorical structure of “Ospreys Calling,” we cannot but identify the islet-confined and screaming osprey with the “girl,” thereby opening up the text for an erotic, “forbidden” reading that understands the girl as the unreachable object of the young man’s feverish dreams.

34. This follows Arthur Waley (1987 [1937]: 222) in translating fusu as “nutgrass.” Mao describes the fusu as a “tree”(mu).

35. In their comments on “Nutgrass” the “Minor Prefaces” give a different, but familiar, twist to the poem. They explain that, in the situation described by the poem, “What is considered beautiful [and consequently given high rank] is not really beautiful” (Wang Xianqian 1988 [1915]: 354). A comparison with Mr. Lü’s metaphorical usage of “treasure” is felicitous here since the Prefaces’ reading banks on a similar abuse of language, whereby the conventional meaning of a word (mei, “beautiful”) is used to refer to an object to which it is not applicable (the “non-beautiful” madman).

36. Several different calendars were simultaneously in use in ancient China. Here, Mao claims that the “first month” (the “zheng yue” of the poem’s title) in fact refers to the fourth month of the Xia calendar.

37. Zheng Xuan (a.d. 127–200), expounding Mao’s Commentary, argues that the untimely frost is Heaven’s response to “rash deeds” committed by depraved rulers.

38. A few centuries later, Hui Shi (see note 14 above) would indeed make a word for himself by claiming that “Mountains are as low as the earth, and heaven as flat as a marsh” (Zhuangzi jishi: 1102; Reding 1985: 357–60). For Xunzi’s rebuttal of this “sophism,” see Xunzi jijie:421 (“Zheng ming”).

39. Mao’s reading is probably derived from a citation of “He ming” by Xunzi, see Xunzi jijie: 128 (“Ru xiao”). The “fine” fish is clearly another variation of the same “hidden but obvious” trope. For the connection between Xunzi and Mao, see Chen Huan 1967.

40. See Svensson 1999. Again, the Chinese tradition is not monolithic. We should recognize the difference between, say, the poetics of Kong Yingda’s (a.d. 574–648) Tang dynasty “Preface to the Correct Meaning of the “Mao Poems” and that expressed in the Han dynasty “Great Preface.”

41. This follows Zheng Xuan’s (Zheng 1979: 441–44) and Kong Yingda’s (ibid.) comments on Mao’s Commentary. Note that Zheng Xuan agrees with Mao that this is a rhetorical trope (and not a literal description) “indicating [ yu] that the Royal Court is now a gathering place for petty men, although it should really be full of sages” (ibid.).

42. “Change” (hua) is Zheng Xuan’s interpretation. See Wang Xianqian 1988 [1915]: 664.

43. Min (“populace”) often denotes not only the demographic group of the lower classes but also the plebeian instinct to copy, mindlessly, the external world or, in other words, to be overwhelmed by the tendencies of the contemporary world. The Confucian contempt for the inherently shapeless and perfectly moldable populace is expressed in the Qi school comment on “High South Mount”: “If you love righteousness, the min will cherish [turn toward] humanness [ren] and their customs will be fine; if you love profit, the min will love what is perverted and their customs will degenerate.” See Wang Xianqian 1988 [1915]: 657.

44. For a listing of similar tropes in the Chinese tradition, see Qian 1979: 153–56.

Additional Information

Print ISSN
Launched on MUSE
Open Access
Archive Status
Archived 2005
Back To Top

This website uses cookies to ensure you get the best experience on our website. Without cookies your experience may not be seamless.