[In the essay below, Riegel analyzes three poems about Confucius's time of exile.]

Among the many stories and tales which constitute the legends of Confucius's life, the most well-known and dramatic involve the Master's suffering and hardships during his exile from Lu.… According to the account set forth in the Tso chuan … and Shih chi …, Confucius departed, or perhaps was banished, from his homeland in 497 B.C., after a failed attempt, supported by him, to dismantle the defensive walls around the cities controlled by the powerful Chi … , Meng … , and Shu … families.¹ Along with a small group of followers, he wandered for thirteen years, travelling through Wei … , Ch'en … , Ts'ai … and other states which lay to the west and southwest of Lu.²

Tales of his trials, usually introduced with such conventional phrases as, "When the Master was in Wei" or "When Confucius was in danger somewhere between the states of Ch'en and Ts'ai … ," are found in numerous ancient sources including his collected sayings, the Lun yü … , as well as the Meng tzu … , Mo tzu …, Chuang tzu … , and, of course, the Shih chi chapter which Ssu-ma Ch'ien … (ca. 145-90 B. C.) devoted to Confucian lore, the "K'ung tzu shih chia".…³ It is conventional among those who have made a careful study of Confucius's biography to accept the accounts of the Lun yü and Meng tzu as "true" while dismissing the hostile Mo tzu and Chuang tzu stories as fictive hyperboles. Since the "K'ung tzu shih chia" is eclectic and its contents confused by Ssu-ma Ch'ien's attempts to fashion his materials into a consistent chronological whole (and perhaps by subsequent textual tampering as well), its versions have been judged unreliable unless attested by the Lun yü and Meng tzu.⁴

There can be little doubt that, because works such as the Chuang tzu and Mo tzu are openly hostile to Confucian teachings, their authors took great liberties with the received tradition, creating a Confucius to serve as the butt of their criticisms. It can at least be allowed that, in contrast to these, the books of the Lun yü are, with certain notable exceptions, friendly to Confucius, as is the Meng tzu, and hence may be more faithful representations of the earliest versions of the Master's years of exile. Yet, while Confucius's exile may have been an actual historical occurrence, it is doubtful that a claim of strict historicity can be attached to any of the accounts which relate it. All are complex and highly crafted literary pieces involving what we would label, from our perspective, fact and fiction. All exhibit the ingredients of "historical romance."⁵ I propose to demonstrate this by showing that some of the literary remains of Confucius's life consist of bits and pieces of ancient poetry which in their origins had nothing to do with Confucius and even predated him.

Among the Pei feng … of the Shih ching … is a group of three songs—No. 34: "P'ao yu k'u yeh" … , No. 36: "Shih wei" …, and No. 37: "Mao ch'iu" …⁶—which share striking thematic similarities with some of the lore surrounding Confucius's famous thirteen-year exile from Lu. More specifically, the language and structure of these three early songs are similar to, and in some cases identical with, the language and structure of certain later anecdotes about the exile now found primarily in the Lun yü and the "K'ung tzu shih chia," and, to a lesser extent, in the other sources of exile lore mentioned above. It is proposed that the parallels are close enough to warrant our regarding the songs as the models, or thematic archetypes, for the Confucian tales. It will be argued that because of their provenance in a literary compendium which according to scholastic traditions was much admired by Confucius, though remote from him in their original purposes, such literary sources were nevertheless judged, by those who would define the pattern of the Master's life, to be suitable examples of his behavior and teachings. By examining these songs in their original settings and then comparing them with the stories they have influenced, it is possible to provide a relatively detailed example of how Confucian legends utilized such ancient materials.

Together with the Wei feng …, all of the Pei feng and Yung feng … have long been identified as songs about the ancient state of Wei.…⁷ This is in part due to the tradition that Pei … and Yung … were appanages of Wei. More importantly, their association with Wei is drawn from the observation that, along with the Wei feng, many of the Pei and Yung songs refer to place names in Wei while some allude to historical personages who can with some certainty be identified with Wei figures known from other accounts. (Although Marcel Granet, Bernhard Karlgren, Arthur Waley, and other Shih ching authorities of European training have tended to discount the significance of regionalism within the Kuo feng songs in general, other scholars, most notable among them Shirakawa Shizuka …, have emphasized the regional origins of the Kuo feng as an important feature that should not be ignored.⁸) It is in any case the concern of the various "Small Prefaces" which introduce the songs, of Mao Heng …, patriarch of the Mao School version of the Shih ching and supposed author of the Ku hsün chuan … commentary, and of the great Han scholiast, Cheng Hsüan … (A.D. 127-200), to explain the meaning of the three songs and reconstruct their original settings in terms of people, places, and events, in Wei. They are naturally unconcerned with the purposes to which the songs may have later been put and we should not expect from them any notice of the songs' influence on Confucian lore.

It should also be noted that the numerous studies of the legends of Confucius, including the monumental works of Ts'ui Shu … (1740-1816) and (somewhat more recently) Fujiwara Tadashi …, do not mention the contribution of the three songs to Confucian traditions.⁹ To the extent that I try to expose this unappreciated connection between later lore and its literary antecedents in the Shih ching as well as to formulate the reasons why the connection was made, the study to follow represents a new, and perforce tentative, method of considering and evaluating the nature of the stories of Confucius.

In the analysis of the three songs, each will be taken up in the order in which it appears in the Shih ching. Each song is translated, its meaning paraphrased, and its connection to stories about Confucius's exile examined.

Song 34: "The Gourd Has Bitter Leaves"

"'The gourd has bitter leaves,'
   the ford is deep to cross."
"Where deep step on stones;
  where shallow wade."

"How fully the ford swells,
         'Evil!' the pheasant cries."
"The swelling ford will not wet your axle;
 the pheasant cries out to seek her mate."

"How harmoniously honk the geese,
         when the genial sun first rises.
If a knight goes to take a wife,
   he acts before the ice breaks."

"Beckoning, beckoning, is the boatman,
         Others cross, not I!
         Others cross, not I!
         I await my friend."

Song 34 tells of a man who is hesitant to take a wife. The story is presented as an elliptical dialogue between the man and another person, perhaps his betrothed or a go-between, who tries, without success, to convince him to act.¹⁰ The song opens with the first person referring to himself metaphorically as "a gourd with bitter leaves." Because such gourds, when old, were inedible and merely ornamental, the speaker is casting himself as a useless old man.¹¹ The same person adds another figurative aphorism about a ford (chi…) in the river being too deep to cross. In the Shih ching, various tropes which have in common the image of crossing a river occur with some frequency in love songs where they express great passion or are euphemistic for seduction. For the first person of Song 34 to say figuratively that the river is too deep to cross means that he is, perhaps out of some unnamed fear, unwilling or reluctant to marry. A second person answers the first by refuting the figure of speech about crossing the river, with terms suitable to its imagery: "Where deep, step on the stones / where shallow, wade."¹² In this way the second person encourages the first to ignore the dangers and be brave and persistent.¹³

In the opening of the second stanza the first person continues to resist. Now to his earlier saying about the depth of the river he adds the expansiveness of the water, in order to emphasize figuratively his fear that he is not meet for marriage. To this he couples a line about the baleful cry of a pheasant—the pheasant's call is given as yao … which sounds like the words yao …"die young" and yao …"calamity"—which is meant to say that the times are not right. But again he is refuted in the terms of the tropes he has chosen. For, in response, the second person claims that the first has misunderstood the pheasant—yao … also sounds like yao … , the desirous call of the cricket in Song 14, "The Grass Bug"¹⁴—and that its call means that it is seeking its mate. He assures, moreover, that, as deep as the water may appear, it will in fact not even wet his axles.¹⁵

The second person continues to encourage the first to go (stanza 3), exhorting him, through two bright sayings about overcoming timidity and seizing the moment: if he waits, the friendly circumstances are bound to change, just as the geese honk harmoniously in the morning but not later in the day; if he delays, he will not be able to go, just as a knight seeking a bride will not be able to fetch her once the winter ice melts and the rivers flood. The first person remains unconvinced (stanza 4) and, in keeping with his earlier metaphors about fording the river, sadly portrays himself as one left behind on a shore awaiting a friend while the boatman ferries others across to happiness.

The language of the song, its structure as a dialogue, and its theme of indecision all closely parallel an episode in the account of Confucius's exile in Wei, as preserved both in the Lun yü—but in the disparate fashion typical of that book—and more completely and coherently in the "K'ung tzu shih chia."

Sometime toward the end of Confucius's first sojourn in Wei, roughly between the years 493 and 490 B.C. according to the traditional chronology, the Chin nobleman, Chao Chien Tzu … initiated an attack on his enemies, the Fan … and Chung-hang … families. Pi Hsi … , Steward of the Fan family stronghold of Chung Mou … located in Chin north of the Yellow River, rebelled and took the fortress, apparently as an expression of his alliance with Master Chien of the Chao.¹⁶ He then sent an envoy to Wei, inviting Confucius to join him. The passage describing Confucius's response is found at Lun yü 17.7 and in the "K'ung tzu shih chia."¹⁷

Confucius wished to go. His disciple Tzu Lu said,"I learned from you that, 'Into the company of any man who personally commits evil, the gentleman will not enter.' Now, Pi Hsi is personally holding Chung Mou in revolt. How can you consider going there?" Confucius replied. "There is indeed such a saying. But is it not also said of the truly hard that 'No grinding will ever wear it down'? Is it not also said of the truly white that 'No steeping will ever make it black'?"

Confucius then uttered words which closely mirror the opening metaphor of Song 34: "How is it that I am a gourd? How can I be merely hung as an ornament and not eaten." With these words Confucius expresses his great desire to join Pi Hsi. Yet he does not act. According to a passage immediately following the preceding in the "K'ung tzu shih chia" but occurring in a completely separate chapter in the Lun yü, Confucius retires to his home and plays the stone chimes to give expression to his feelings on the matter.¹⁸ A passerby carrying a basket—a man identified by all commentators as an unrecognized sage—declares upon hearing the music: "With such passion does he strike the chimes!" But when the music was over the listener did not care for what it expressed of Confucius's decision. "How stupid he is! How stubborn he is! Since no one recognizes him he just quits! Where deep, step on stones. Where shallow, wade it."

With these last words, identical with the last couplet of the first stanza of Song 34 and intended to encourage Confucius to be brave and persistent, the stranger succeeds in changing Confucius's mind: "The Master said, 'That is resolute indeed. Against such resoluteness there can be no argument.'"¹⁹ There occurs at this point in the "K'ung tzu shih chia" another, unrelated, passage on music occasioned by the story of Confucius playing the chimes. Immediately following it, the story resumes (the passage is not found at all in the Lun yü) with Confucius going to the banks of the Yellow River, intending to cross it and join with the Chao family of Chin.²⁰ In this way, the figurative line in Song 34 about crossing a ford has become in the Confucian legend a literal element of the story: we are meant to suppose that Confucius actually travelled to the river. At the very edge of the river, however, he "hears" the news that two virtuous ministers have been murdered by Chao Chien Tzu. Because of this, Confucius changes his mind once again and, in words which echo the sentiments of the first persona in the last stanza of Song 34, declares his decision not to cross the river: "How beautiful is the river! How very immense it is! That I, Ch'iu, do not ford (chi…), is fate." When asked why he did not go, Confucius explains that the murders are an example of yao… "perverse murders of the young and innocent," and that a virtuous man will not travel to places where such things occur. His calling the murders yao is significant for it reveals that the episode of having Confucius hear the baleful news is based on the Song 34 figurative saying about the ominous cry of the pheasant which is cited by the persona of the song to say that the times are not right for marriage.

In summary, in both Song 34 and the foregoing Confucian anecdote the main theme is indecision as expressed in someone's hesitation to cross a ford. At several points in the extended tale Confucius is made to say or do something which either identically repeats or closely paraphrases the imagery of Song 34—especially noteworthy are the mention of the gourd, the allusions to crossing a river, the identical couplet about fording it no matter the depth, the reference to baleful news heard at a river's edge, the vastness of the river, and the final decision not to cross. In a manner similar to Song 34, the anecdote consists of a series of dialogues. Moreover, the order of Confucius's actions and utterances, as they are recorded in the more complete "K'ung tzu shih chia" version, follows that of the Song 34 narrative.

Song 36: "One so reduced!"

Oh, One so reduced, so reduced!
 Why not return?
If it were not for the lord's misfortune,
   why would we be here in the open?

Oh, One so reduced, so reduced!
 Why not return?
If it were not for the lord's impoverishment,
  why would we be here in the mire?

Song 36 is addressed to an exile by his followers. Both the "Small Preface" and Cheng Hsüan agree that this piece and Song 37 are a pair which refer to the exile of a certain Lord Li … who was driven from his territory by the Ti barbarians and fled to Wei.… Aside from their commentary, there is in the Tso chuan a brief mention of how, in 594 B.C., a Lord Li was established by the Chin army in Ti lands.²¹ But nothing is said of his having been driven out by the Ti nor is there mention of his having gone to Wei. Still, it is significant that Songs 36 and 37 were recognized by Shih ching scholars as a pair, for it is as a pair that they influenced Confucian lore.

We may summarize generally what Song 36 reveals about the character of the exile and his relationship with his followers who address the song to him. Their calling him Shih Wei …, "One Most Reduced," identifies him as a once-prominent figure who has lost his prestige and is living in humble circumstances.²² The question "Why not return?" reveals that the man could give up his exile if he chose to do so, but that he is perhaps stubbornly resisting repatriation and thus causing his followers to exhort him. His companions blame his "misfortune" and "impoverishment"²³ for their being "exposed" and "in the mire," hyperbolic metaphors for abject circumstances.²⁴

The theme of the stubborn exile accompanied by unhappy followers is also found in the account of an incident that took place during Confucius's time of homeless wandering, when he was in the small state of Ch'en. In 489 B.C., Wu … attacked Ch'en and it may be that Confucius and his followers were in fact caught in the crossfire. According to the legend, as embroidered in numerous sources, the small band was starving and near death. There is no specific mention of these troubles in the "K'ung tzu shih chia," only an elliptical reference to how Ch'en was, at the time, plagued by bandits. A story in the Hsün tzu … refers to Confucius's impoverishment when in the area of Ch'en and Ts'ai and has Tzu Lu wondering why it is that Heaven has rewarded Confucius's merits with such catastrophe.²⁵ The use of the terms "impoverishment" (o … ) and "catastrophe" (huo … ) may reflect the influence of Song 36.²⁶ The equally fictive Lun yü 15.1 says of Confucius's disciples that, when their food ran out, they were so ill "none could rise to his feet," a phrase reminiscent of the Song 36 followers complaining metaphorically of being exposed and "in the mire." The passage relates that Tzu Lu then came to the master and asked indignantly, "Does the gentleman suffer impoverishment?"—a question which appears to take its inspiration from the Song 36 reference to the impoverishment of the exile portrayed in that song.²⁷

Song 37: "Long Hair Hill"

'The kudzu on Long Hair Hill,
 how long its joints extend.'
Oh, my brethren!
 how many have been the days?

Wherever I have rested,
 I have always had friends.
Wherever I have tarried,
 I have always had helpers.

The fox-furs, crazed and confused,
 complain that chariot comes not to the east.
Oh, my brethren!
 I am without friends or comrades.

Oh, my pretty little things!
 sons of the vagabond bird.
Oh, my brethren!
 billowing sleeves and ear-plugs.

Song 37 records the sad, embittered words of an exile—again the "Small Preface" and Cheng Hsüan identify him as Lord Li—whose followers have tired of his cause and are abandoning him for a life of convenience and prestige. He begins by comparing himself to Mao Ch'iu …"Long Hair Hill,"²⁸ an oddly shaped mound—Mao says of it that "it is high in the front with a depression at the rear"—on which there hangs a mass of dangling kudzu vines (the hill's "hair") whose length symbolizes how the persona's followers have distanced themselves from him. Observing aloud to his followers—whom he calls throughout his "brethren" (shu po … )—that his exile has been long, he claims for himself that wherever he has chosen to visit or remain he has been able to have with him men who share his values and are meritorious.²⁹

But his situation has changed and thus, in stanza three, the persona quotes a saying, "The fox-furs are crazed and confused."³⁰ "Fox-furs" is a metonymy for noblemen; to say that they are "crazed and confused" (meng jung …) refers to their acting wild and abandoned, as if their heads were hooded and they could not see.³¹ The line alludes metaphorically to the persona's adherents abandoning him. Following this is a problematic line in which the persona appears to relate how his followers, dissatisfied as they are with him and his cause, complain³² that a chariot has not come to the east, which presumably refers to the fact that there has been no invitation from his homeland to return.

After further lamenting the disloyalty of his "brethren," the exile, in stanza four, brands them "turncoats." Calling them "pretty little things" (so wei … ),³³ a name which figuratively labels them junior and may convey some measure of contempt as well, he condemns them as traitorous and duplicitous by metaphorically identifying them as "sons of the vagabond bird," birds which, as Mao points out, are, "pretty when young but ugly when old."³⁴ In the closing couplet of the song the "brethren" are described as having "billowing sleeves and ear-plugs," the elegant and elaborate insignia of a high official, and thus they are identified as having accepted office somewhere.

The theme of the exile who expresses his need for loyal friends while criticizing and lamenting the way-wardness and disloyalty of his followers is also to be found in an account of an episode during the exile of Confucius. Sometime after the near starvation suffered by Confucius and his small band, referred to above, his companions began to break ranks with him. According to the "K'ung tzu shih chia" account of Confucius's wanderings, when Confucius and his party were still in Ch'en, Chi K'ang Tzu … , the newly empowered dictator of Lu, contemplated inviting him to return to his native state.³⁵ Warned by one of his advisers of Confucius's difficult personality, Chi K'ang Tzu decided instead to invite into his service one of Confucius's followers, Jan Yu.… Jan Yu immediately accepted. Confucius worried aloud that the invitation meant that Chi K'ang Tzu intended to employ Jan Yu in some grand office. But when Confucius saw Jan Yu, the disciple's dress and pretentious manner elicited from Confucius an acid comment also found in Lun yü 5.22 but without the introduction and setting provided by the "K'ung tzu shih chia".³⁶

Shall we return home? Shall we return home? The young boys of my party are wild and brazen. They show off their replete insignia of office without knowing how properly to cut them.

Confucius's question, "Shall we return home?" is a purposeful echo of the exhortation to the exile of Song 36 by his dissatisfied companions. The pleonasm k'uang chien … "wild and brazen" by which Confucius refers to Jan Yu is but a prosaic gloss on the rarer binom meng jung, "crazed and confused," used in the Song 37 saying about disloyal noblemen. Having Confucius refer to Jan Yu as one of the "young boys," an epithet used elsewhere by Confucius to address his disciples, parallels the "pretty little things" of the song. The "billowing sleeves and ear-plugs" of Song 37 become, in the prose anecdote, the less figurative "replete insignia" of Jan Yu.

The contents of Lun yü 5.22 also occur in the Meng tzu where Confucius is made to say: "My young sons are wild and brazen in entering and taking office. Have they not forgotten their beginnings?"³⁷ For reasons which are not entirely clear, the "K'ung tzu shih chia" treats Lun yü 5.22 and the corresponding Meng tzu passage as two separate entries.³⁸ Numerous authorities have recognized that this must be a mistake and that Confucius should be considered to have uttered his condemnation of his "young sons" only once.³⁹ It has not been noted, however, that the "K'ung tzu shih chia" places the Meng tzu version immediately following its elliptical reference to banditry in Ch'en. This pairing duplicates exactly the traditional coupling of Songs 36 and 37. If the two songs did indeed serve as a source for the anecdotes, this sequence of events in the "K'ung tzu shih chia" is perhaps the one intended by those who formulated the legend. Thus the story of Jan Yu's disloyalty and ambitiousness which parallels Song 37 should be joined with and viewed as a sequel to the tales of Confucius's earlier impoverishment which derive from Song 36.

One initial point suggested by the foregoing analysis of the literary origins of Confucian legend is that, in terms of coherence and completeness as defined by their relative closeness to the language and structure of the songs, the versions of the stories preserved in the "K'ung tzu shih chia" are often preferable to those of the Lun yü. This is especially clear in the case of the story of Confucius's indecision about fording the river to go to Chin. What appears in the "K'ung tzu Shih Chia" as a single narrative that parallels Song 34 is divided in the Lun yü into two separate passages, with the climax found in the "K'ung tzu shih chia" version omitted entirely. I am not suggesting that the "K'ung tzu shih chia" is a more reliable source for Confucius than the Lun yü. Nor am I discounting the distortions of the original lore which apparently resulted from Ssu-ma Ch'ien's passion for a chronologically connected narrative. I am claiming for the "K'ung tzu shih chia" that, in the instances studied above and in others as well, its arrangement and presentation of the anecdotes it shares with the Lun yü may be closer to earlier forms of the legend than that of the Lun yü.

We need not, because of some misguided faith in their antiquity, be wed to the disposition and division of passages in the Lun yü. While that book certainly had something like its present form as early as the time of Cheng Hsüan, who wrote the classical commentary for it, we cannot know how long before him this was true. Quotations of the Lun yü by title in the "Fang chi" … opuscule of the Li chi … and in the Han shih wai chuan … are not sufficient evidence in this regard, not only because they do not testify to the overall contents and division of passages of the Lun yü, but also because the dates of these sources are themselves problematic. They certainly do not prove that the extant recension of the Lun yü predates the compilation of the "K'ung tzu shih chia."⁴⁰

Since there are many quotations of Confucian sayings in the Meng tzu which do not appear in the Lun yü, it may be, as Arthur Waley seems to suggest, that during the late Chou, the body of Confucian sayings and lore was much larger than and of a rather different nature from the Lun yü.⁴¹ The "K'ung tzu shih chia" may adumbrate an alternate arrangement of Confucian tales found in this earlier body of lore, now lost but perhaps similar to the chronological presentation of the first two books of the Meng tzu, in which indications of chronology and concern for the coherence and completeness of the narrative were more in evidence than they are in the present Lun yü.

I would propose that this early body of Confucian sayings and lore was not only extensive but also enjoyed wide and perhaps even popular circulation. This is already suggested, at least for the Han dynasty, by the incorporation of numerous rather simplistic cautionary tales involving the Master and his disciples in the Shuo yüan … , Hsin hsü … , and other eclectic collections of moralistic anecdotes gathered by Liu Hsiang … (ca. 79-76 B.C.). The K'ung tzu chia yü… , which purports to be a collection of ancient Confucian tales, might be another useful reflection of the extent and popularity of Confucian lore during the Warring States and early Han, were it not made suspect by what appear to be numerous interpolations and distortions added to the text to serve the selfish scholarly interests of Wang Su …(195-256) in his attempts to denigrate the authority of Cheng Hsüan.⁴²

It is thus significant in this regard that in the Fu-yang … , Anhwei, tomb of Hsia-hou Tsao … (d. 165 B.C.), the second Lord of Ju-yin … , there was discovered, among other very fragmentary manuscript remains, a board on both sides of which was written the table of contents of a collection of forty-six stories about Confucius and his disciples.⁴³ Though the text itself was unfortunately not found—if placed in the tomb it disappeared through deterioration—the table of contents gives the titles of its stories and these suggest that the work had much in common with the present K'ung tzu chia yü and was the sort of collection upon which the latter may have been based. In any case, this shred of evidence, found as it was in an area which in Han times was a cultural outpost remote from the centers of learning, does provide some indication of the wide circulation of Confucian tales. We learn by the serendipity of Chinese reporting on the find that one of the anecdotes had the title, "Confucius Wails upon Approaching the River" (K'ung tzu lin ho erh t'an … ).⁴⁴ Even if we cannot know in detail this early version of the story of Confucius's decision not to join the Chao family upon learning, at the edge of the Yellow River, of Chao Chien Tzu's crimes, it is gratifying to have some confirmation of its popularity.

That there should have been discovered in the songs of the Shih ching models and patterns upon which to base accounts of Confucius's life is not in itself anomalous. Even in antiquity the ancient canon of songs was highly admired as a mirror of proper words and behavior. According to numerous Lun yü passages, Confucius himself regarded the songs as a rich catalogue of sentiments, deeds, and expressions which should be emulated. This attitude toward the text is also seen in the frequent quotation of the Shih ching to illustrate a point of morality or decorum, a practice much in evidence in the ancient philosophical, historical, and ritual literature.

In the examples we have studied, however, the songs are not merely quoted but reworked and transformed into the skeletons of prose anecdotes. They serve as frameworks of plot and terminology which are then considerably expanded with other materials to fit what the author wishes to say about Confucius and his times. Perhaps the most notable aspect of this exploitation of Shih ching imagery is the way in which aphorisms and maxims, figures of speech typical of Shih ching style and clearly critical to the rhetorical function of ancient poetry more generally, become literal narrative in the legend. Thus, for example, the fear of fording a river in Song 34, figurative for a man's hesitation to take a wife, serves as a crucial element in the plot of whether Confucius will ford the Yellow River and join the morally suspect government of Chin; and the Song 36 use of "exposed" and "in the mire" as metaphors for abject suffering becomes a literal description of the crawling of Confucius' sick disciples. In Song 37, moreover, we have examples of how the more flowery examples of poetical tropes, e.g., "The fox-furs are crazed and confused" or "billowing sleeves and ear-plugs," are translated into more commonplace expressions.

Though transmuted by literal interpretation and paraphrase, the songs are still recognizable in the language and structure of the legend. This provides a subtle but explicit connection between Confucius and his spiritual forebears, the personae of the Shih ching songs. The weavers of the Confucian legend, his "biographers," have Confucius relive the songs by having him say their lines, think their intentions, and act out the gestures portrayed in them. It would be to miscast and underestimate their achievements, however, to say of these "biographers" that they have merely borrowed prestigious materials, or to accuse them of fabricating and fictionalizing Confucius's biography. What they have done is to discover in the Shih ching the ancient patterns of proper behavior which Confucius admired and to show how in his life the Master, too, adhered to them. The life of Confucius is made the summation of the Shih ching lessons he so revered; or, in the words of Lun yü 2.11, he has "reanimated the past" (wen ku … ).

Moreover, according to one ancient literary theory, the Shih ching songs should be counted among the great literary works created by those wrongfully banished, punished, or otherwise disaffected, to protest their fate and the circumstances of their age. Ssu-ma Ch'ien says:^.45

In the past, when Hsi Po …(i.e., Wen Wang) was caught in Yu-li he lectured on the Chou i. When Confucius was impoverished in Ch'en and Ts'ai he made the Ch'un ch'iu. When Ch'ü Yüan was banished he wrote the Li sao. When Tso Ch'iu … lost his sight (?) he possessed the Kuo yü. When Sun Tzu … was defooted he discoursed on the Ping fa. … When [Lü] Pu-wei … was demoted to Shu he transmitted the Lü lan … (i.e., Lü shih ch'un ch'iu) to his age. When Han Fei was imprisoned in Ch'in [he wrote] "The difficulties of persuasion" and "Lonely anger." The three hundred songs of the Shih are fundamentally works composed by the worthies and sages to express anger.

To some extent this is a self-serving judgment on the part of Ssu-ma Ch'ien, for it provides him with antecedents to his own decision to continue his work on the Shih chi in spite of his disgrace at the hands of Han Wu Ti. It is of course most doubtful that the Chou i, Ch'un ch'iu, and Kuo yü were compiled under anything like the circumstances he describes.⁴⁶ Moreover, there are other accounts (some of them provided elsewhere in the Shih chi) of when and where the Lü shih ch'un ch'iu, Ping fa, and the discourses of Han Fei were composed. While the historical details of these characterizations are questionable, yet Ssu-ma Ch'ien is not greatly distorting the motivations that led to most of the works he cites. Certainly prominent among the intentions of those who authored the philosophical treatises he mentions was a passionate and often frustrated desire to reform what were perceived as the ignorant evils and mistakes committed by ancient rulers. The persona of the Li sao has undoubtedly suffered banishment. Reading the Shih ching one cannot escape the feeling that many, if not most, of its songs are the resentful expressions of the outcast and unhappy who were thwarted by the unresponsiveness of indifferent lovers, rulers, and gods.

It may have seemed to those who have given us the legend of Confucius's exile that the songs, as the expressions of the worthies of antiquity who had undergone similar deprivations, contained the proper prescriptions for the portrayal of the Master whose sufferings had inspired him to literary efforts: Ssu-ma Ch'ien claims that Confucius composed the Ch'un ch'iu … at that critical moment when he and his disciples were stranded and starving. Seen in this light, the songs defined the proper ways of portraying those special individuals, like Confucius, who were believed to have inherited the predicaments and sensitivities of the Shih ching poets. In this way these literary antecedents predetermine the circumstances and characterization as presented in the account of Confucius's life. They are the essential fundamental upon which the miscellaneous details of time, place, and personality—what we would call the real facts of Confucius's life—must rest. The process and its effects are not unlike the use by later Chinese historians of conventional sayings or formulaic characterizations—topoi—to introduce and identify the subject of a biography in terms of the classically prescribed roles, as well as to organize and give meaning to the more individualistic details of his life.⁴⁷

The occurrence in Confucian lore of influence from the Shih ching goes beyond historiographical practices, however, and is an example of a more broad-based and important phenomenon in Chinese literary history. The harking back to the language and imagery of the Shih ching occurs so commonly in later literature, especially of course in poetry, that it deserves to be identified as a literary convention which distinguishes and typifies the Chinese tradition. The significance of these occurrences often escapes due recognition. This oversight is prompted, perhaps unintentionally, when borrowings from the Shih ching are labelled mere "allusions." This is an empty term which unfortunately obscures the active and determining influence of the old songs on later literary expression by casting the Shih ching as somehow inert or by viewing it as an archaic touchstone or storehouse and the later author who makes reference to it as merely traditional or pedantic. If the examples of the use to which the Shih ching songs are put in Confucian lore are typical, they suggest that in literary studies more efforts should be directed toward uncovering the details of how the Shih ching and other canonical sources not only influenced the lexicon of later literature but also shaped its themes and content.

Finally, it should be emphasized that the selection of songs to serve in the capacity of models for Confucian legends was not haphazard. Since Songs 34, 36, and 37, but no others, were selected, this choice must have been due to features inherent in the songs which were concordant with what Confucius's biographers already believed to be facts about his life and character. It was undoubtedly important, for example, that the three songs were about Wei, the place where tradition said Confucius spent much of his exile, and that two of them, Songs 36 and 37, were already identified as a pair of songs about someone exiled from his home. More specifically, the hesitation and scrupulousness communicated by Song 34 are traits for which Confucius was well known, and sometimes mocked, elsewhere in the ancient accounts of him. Similarly the "misfortune" and "impoverishment" of which Song 36 speaks are leitmotifs which occur throughout Confucian legend.

More than either Songs 34 and 36, Song 37 seems to exhibit features which may have marked it as an especially fitting archetype for Confucian lore. Most striking is the song's opening metaphor about kudzu growing upon what is called Mao Ch'iu, "Long Hair Hill," for Confucius's given name was Ch'iu …, "Hill," and tradition said of him, similar to what Mao said of "Long Hair Hill," that he had a depression atop his head.⁴⁸ The song's persona refers to his followers as shu po, "brethren," a familial term similar to ti tzu …, "younger brothers," the name by which Confucius's corps of disciples was known. In Song 37 there is a complaint that no chariot returns to take the persona home, a detail which uncannily anticipates the fact that Chi K'ang Tzu decided not to invite Confucius to end his exile and return east to Lu. Moreover, the claim of the song's persona that he had always been able to attract and associate with like-minded people closely resembles several Confucian sayings—for example, Lun yü 4.25: "Virtue never lives alone; it always has neighbors." A song so replete with characteristic "Confucian" images and teachings may have seemed to the composers and compilers of Confucian lore not a mere literary mode but a magical precursor which prefigured and portended Confucius's fate.

Notes

It is a privilege and an honor to dedicate this study to Edward H. Schafer, Agassiz Professor of Oriental Languages and Literature, Emeritus, of the University of California, Berkeley. All citations of the Thirteen Classics of the Confucian Canon are to the edition prepared under the supervision of Juan Yüan … (1764-1849) and printed in Nan-ch'ang, Kiangsi, in 1816. For the Shih ching I also give the "Mao Number" and for the Lun yü and Meng tzu I also provide the chapter and section numbers used in the Harvard-Yenching Concordances. References to the Shih chi … are to the edition included in the Erh shih wu shih compiled by the I-wen Press of Taiwan. Unless otherwise indicated, the editions of all other Chinese sources consulted are those of the Ssu pu pei yao. My thanks to Paul W. Kroll for his comments on an earlier draft of this article.

1. For the events leading up to Confucius's exile, see Tso chuan, Ting 10, 56.5a-6b and Ting 12, 56.9b-10b. Kung-yang chuan…, Ting 12, 26.11a, claims Confucius originated the plan to dismantle the cities. The Tso chuan, however, credits his disciple, Tzu Lu.… A further discussion of this material may be found in Homer Dubs, "The Political Career of Confucius," JAOS 66 (1946), 273-82.
2. A convenient chronological account of Confucius's wanderings, based on Lun yü and Meng tzu, is provided in D.C. Lau, tr., The Analects (Harmondsworth, 1979), 170-77.
3. H. G. Creel, Confucius and the Chinese Way (New York, 1960), 7-11 and 291-94, discusses these sources.
4. Ibid., 10.
5. The classic study of the roman in ancient Chinese literature is Henri Maspero, "Le Roman de Sou Ts'in," in Etudes Asiatiques publiées par l'Ecole Française d'Extrême-Orient à l'occasion de son 25e anniversaire, Volume II (Paris, 1925), 127-42. See also Maspero's China in Antiquity, tr. F. A. Kierman (Amherst, 1978), 357-65. In these studies Maspero attempts to show that much of what is said in early sources of the Warring States political strategist, Su Ch'in …, is the product of "pure imagination." He believes that the fragmentary stories of Su Ch'in derive not from a work of history but from the now-lost Su tzu … which he calls a novel or "a work of political philosophy in the form of a novel." Maspero argues not from the origins of the Su Ch'in legends but from chronological inconsistencies and internal contradictions. His conclusion, that Su Ch'in was fictional, should be regarded with some tentativeness, as Maspero himself seems to have done. Moreover, Maspero seems not to have fully appreciated how the competition of cults which grow up about a figure lead to great diversity in the sources for them which have come to us. This was certainly the case with the traditions surrounding Confucius. These and related questions are discussed by David Johnson in connection with the cult of Wu Tzu-hsü which grew and influenced legends about the great hero of the kingdom of Wu during ancient as well as medieval times. See his "The Wu Tzu-hsü Pien-wen and Its Sources," HJAS 40.1 (1980), 93-156, and 40.2 (1980),465-505.
6. The three songs would form a single series were it not for Song No. 35, "Ku feng" …, which has nothing to do with the others. That song should instead be read together with three other Pei feng songs, Nos. 30, 32, and 41, which form a set having to do with the winds of the four cardinal directions.
7. On the very fragmentary remains of a manuscript of the Shih ching, dating to ca. 185-165 B.C., and quite different in significant details from the Mao School version, there appears the state name Pei kuo. … This certainly suggests that the division of the Wei songs among three different states was not exclusive to the Mao School version, as was argued by some Ch'ing dynasty authorities, but instead was an ancient feature inherited by the various Han dynasty Shih ching traditions. (The manuscript was one of several discovered in 1977 in one of a pair of Former Han tombs located near Fu-yang … in Anhwei province. Another text from the cache is discussed below. For details on the Shih ching manuscript see Wen wu, 1984.8, pp. 1-21.)
8. See Shirakawa Shizuka, Shikyō kenkyū: tsuronhen … (Kyoto, 1981), 51-177.
9. Ts'ui Shu does, however, treat the legendary accounts discussed below in terms of their relative historicity in his Chu Ssu k'ao hsin lu … (TSCC), 2.47-51, 3.63-70. Fujiwara's useful two-volume study is entitled Kōshi zenshū … (Tokyo, 1931).
10. Arthur Waley, The Book of Songs (New York, 1960), 54, already renders the song as a dialogue.
11. Mao says that the bitterness of the leaves of the p'ao … "gourd" means that they cannot be eaten. The leaves evidently became inedible only very late in autumn when the gourd was old because the leaves of the hu…"calabash", with which Mao equates the p'ao, were delicacies offered the ancestors, according to Song 231, Shih ching 15C.3b. The Ho Yen … commentary at Lun yü 17.4a says that because the gourd cannot be eaten it is regarded as useless and thus hung on a wall where it remains.
12. Shuo wen chieh tzu … 11 A(2).15b defines li … as "stepping on stones to ford a stream." Ch'i …"wading" is defined more specifically by Mao and Erh ya…, 7.21a-b as "lifting the skirts." Because of these and other lines in the song which encourage the reluctant first person to act, the song is often taken as an expression of resoluteness.
13. At Tso chuan, Hsiang 14, 32.1 1a-b, and Kuo yü…,"Lu yü," 5.2b-3a, there is an anecdote in which Song 34 is recited by a man to illustrate his resolve to ford a river and pursue the enemy.
14. Shih ching 1D. 1a.
15. For kuei …"axle." see B. Karlgren, Glosses on the Book of Odes (Stockholm 1970). p. 118, loss 91.
16. For proof that Chung Mou was at this time a Fan family possession and not, as some old interpretations have held, already one of the Chao fortresses, see the comments of Huang Shih-san …, quoted in Shikikaichū kōshō…, ed., Takigawa Kametarō … (Kyoto, 1958), 47.46-47.
17. The quote is found at Lun yü, 17.3b-4a and Shih chi,47.14b. For a full translation of the latter, see E. Chavannes, Les mémoires historiques de Se-Ma Ts'ien (Paris, 1905), V, 347-48.
18. See Lun yü, 14.39 (=14.15b) and Shih chi, 47.14b-15a, Chavannes, op. cit., 348-49.
19. The quote from Song 34 and Confucius's reply to it appear in Lun yü, 14.39 but have dropped from the "K'ung tzu shih chia."
20. Shih chi, 47.15a-b; Chavannes, 351-53.
21. Tso chuan, Hsüan 15, 24.11b-12a.
22. Wei… occurs in two senses in Song 36. It is in the first line of each stanza the noun, "obscure one, humble one." In the second line it is the negative, "if it were not for.…" The word shih … in general has two meanings in the Shih ching. Before verbs it is a modal. This aspect of the word has been studied in Ting Sheng-shu …, "Shih ching 'shih' tzu shou" …, Bulletin of the Institute of History and Philology 6.4 (1936), 487-95. When it occurs before nouns I take it to be an emphatic, related to other ancient words whose old pronunciations are similar to it. This seems to be what is suggested by Kuo P'u …, at Erh ya, 4.12a, who explains shih wei … as chih wei …, a gloss I have followed in my translation.
23. Ma Jui-ch'en …, Mao shih chuan chien t'ung shih …, 4.26b-27b demonstrates that ku … should be understood as "calamity" or "misfortune" and that kung … is a short form for ch'iung …, "impoverishment."
24. When Mao identifies chung lu …"in the open" and ni chung …"in the mire" as fortresses in Wei he is not defining the terms as toponyms but merely associating the scene of distress with strongholds in Wei.
25. Hsün tzu, "Yu tso, " 20.4b. At Mo tzu, "Fei Ju," 9.15b-16a, a passage with an almost identical formulaic introduction but intended to mock Confucius's hypocrisy claims that Confucius, who was usually scrupulous about his meals, ate meat given him by Tzu Lu even though he had reason to believe Tzu Lu had stolen it.
26. Cf. note 23 above.
27. Lun yü, 15.1b.
28. Mao … should be read mao.…
29. Following Mao's paraphrase.
30. For another instance of the saying, see Shih chi, "Chin shih chia," 39.8b, where it refers to the confusion of the Chin nobility when the three families usurped power and there was no single leader.
31. Meng jung …, or its variant meng jung … given in the Fu-yang Shih ching (see note 7 above), describes something completely covered over and concealed by a dense growth of vegetation and, by extension, the wild movements of those thus blinded. Because earlier scholars did not appreciate that "fox-furs" (hu ch'iu …) is a metonymy for the noblemen who wear them, meng jung has been mistakenly understood as descriptive of messy and unkempt fur.
32. I follow the Fu-yang Shih ching (see note 7 above) and read fei …"complain" rather than the negative fei… given in the Mao School version. The Mao reading has long been recognized as problematic. Ma Jui-ch'en, op. cit. 4.29b, for example, argued that the graph … did not here stand for the negative but for a homophonous demonstrative pronoun.
33. So wei… is but a variant of shao wei.…
34. The bird gets the name liu li …, "vagabond," from the way it rides about on the wind and does not perch. The identity of the tiny bird is uncertain, though it is probably related to the siskin and, as Ma Jui-ch'en, 4.29b-30a points out, not the owl.
35. Shih chi, 47.16a/b; Chavannes, 357-59.
36. Lun yū 5.10a/b. The passage has been taken to be a comment about some anonymous disciples who remained home during Confucius's exile, but the Shih chi makes it clear that it is Jan Yu whom Confucius is criticizing.
37. Meng tzu, 7B.37 (=14B.8a). The last sentence, which begins with the negative pu …, should be read as rhetorical question, a usage attested at Meng tzu 2A.2 and elsewhere.
38. The Meng tzu passage is repeated at Shih chi, 47.13b.
39. Liang Yü-sheng …, Shih chi chih i … (Kuang ya shu chü, 1887), 25.18b-19a.
40. There is an ancient tradition which ascribes the "Fang chi" to Confucius's grandson, Tzu Ssu.… This led numerous Ch'ing authorities to identify the "Fang chi" as a fragment of the latter's lost writings. In my doctoral dissertation, "The Four 'Tzu Ssu' Chapters of the Li Chi" (Stanford, 1978), I show that the ancient traditions are mistaken and propose that the "Fang chi" cannot be dated to much before the first century B.C.
41. Arthur Waley, The Analects of Confucius (New York, 1952), 22-23.
42. For the authoritative discussion of the problems surrounding the K'ung tzu chia yü as well as the differences between the Wang Su and Cheng Hsüan camps, see Gustav Haloun, "Fragmente des Fu-tsi und des Tsin-tsi," Asia Major 8 (1932), 456-61.
43. The Fu-yang cache, first discovered in 1977, is reported in some detail in Wen wu, 1983.2, pp. 21-23. (It included the Shih ching MS discussed above in note 7.)
44. Wen wu, 1983.2, p. 23, lists the titles of three of the stories and happens to include this one. One imagines that in some future publication we will learn the other forty-three titles included on the board.
45. Shih chi, "T'ai Shih Kung tzu hsü," 130.12a.
46. The line about the Kuo yü is garbled. Historical inconsistencies in the passage's other characterizations of the authorship of the texts it mentions are discussed in the commentary at Shikikaichü kōshō, 13.28-29.
47. See Herbert Franke, "Some Remarks on the Interpretation of Chinese Dynastic Histories," Oriens 3 (1950), 113-22, esp. 120-21.
48. Shih chi, "K'ung tzu shih chia," 47.2a.