Sacred Texts  Confucianism  Index  Previous  Next 
Buy this Book at

The Book of Poetry, tr. by James Legge, [1876], at

I have included the first two parts of the Introduction from the 1876 edition.—JBH






Statements of Chinese scholars.1. Sze-Ma Ts‘ën, in his memoir of Confucius, says:—The old poems amounted to more than 3000. Confucius removed those which were only repetitions of others, and selected those which would be serviceable for the inculcation of propriety and righteousness. Ascending as high as Sëeh and How-tseih, and descending through the prosperous eras of Yin and Chow to the times of decadence under kings Yëw and Le, he selected in all 305 pieces, which he sang over to his lute, to bring them into accordance with the musical style of the Shaou, the Woo, the Ya, and the Sung." This is the first notice which we have of any compilation of the ancient poems by Confucius, and from it mainly are derived all the subsequent statements on the subject.

In the History of the Classical Books in the Records of the Suy dynasty (A.D. 589--618), it is said:—"When odes ceased to be made and collected, Che, the Grand Music-master of Loo, arranged in order those which were existing, and made a copy of them. Then Confucius expurgated them; and going up to the Shang dynasty, and coming down to the State of Loo, he compiled altogether or 300 pieces."

Gow-yang Sëw (A.D. 1006-1071) endeavours to state particularly what the work of expurgation performed by Confucius was. "Not only," says he, "did the sage reject whole poems, but from others he rejected one or more stanzas; from stanzas he rejected one or more lines; and from lines he rejected one or more characters."

Choo He (A.D. 1130-1200), whose own classical Work on the Book of Poetry appeared in A.D. 1178, declined to express himself positively on the question of the expurgation of the odes, but summed up his view of what Confucius did for them in the following words:—"Poems had ceased to be made and collected, and those which were extant were full of errors and wanting in arrangement. When Confucius returned from Wei to Loo, he brought with him the odes which he had gotten in other States and digested them, along with those which were to be found in Loo, into a collection of 300 pieces."

These statements not supported by evidence. The view of the author.I have not been able to find evidence sustaining these representations, and propose now to submit to the reader the considerations which prevent me from concurring in them, and have brought me to the conclusions that, before the birth of Confucius, the Book of Poetry existed substantially the same as it was at his death, and that, while he may have somewhat altered the arrangement of its Books and odes, the principal service which he rendered to it was not that of compilation, but the impulse to the study of it which he communicated to his disciples. The discrepancy in the number of the odes as given in the above statements will be touched on in a note.

The groundlessness of the above representations.2. If we place Ts‘ëen’s composition of the memoir of Confucius in B.C. 100, nearly four hundred years will thus have elapsed between the death of the sage and any statement to the effect that he expurgated a collection of poems, or compiled previous that which we now have, consisting of a few over 300 pieces; and no writer in the interval, so far as we know, had affirmed or implied any such facts. But independently of this consideration, there is ample evidence to prove, first, that the poems current before Confucius were not by any means so numerous as Sze-ma Ts‘ëen says, and, secondly, that the collection of 300 pieces or thereabouts, digested under the same divisions as in the present Classic, existed before the sage's time.

The old poems were not numerous.3. [i.] It would not be surprising, if, floating about and current among the people of China, in the 6th century before Christ, there had been even more than 3000 pieces of poetry. The marvel is that such was not the case. But in the "Narratives of the States," a Work attributed by some to Tso K‘ëw-ming, 1 there occur quotations from 31 poems, made by statesmen and others, all anterior to Confucius; and of those poems it cannot be pleaded that more than two are not in the present Classic, while of those two one is an ode of it quoted under another name. Further, in the Tso Chuen, certainly the work of Tso K‘ëw-ming, and a most valuable supplement to Confucius’ own Work of the Chun Ts‘ëw, we have quotations from not fewer than 219 poems; and of these only thirteen are not found in the Classic. Thus of 250 poems current in China before the supposed compilation of the Book of Poetry, 236 are found in it, and only 14 are absent. To use the words of Chaou Yih, a scholar of the present dynasty, of the period K‘ëen-lung (A.D. 1736-1795), "If the poems existing in Confucius’ time had been more than 3000, the quotations found in these two Books of poems now lost should have been ten times as numerous as the quotations from the 305 pieces said to have been preserved by him, whereas they are only between a twenty-first and twenty-second part of those from the existing pieces. This is sufficient to show that Ts‘ëen’s statement is not worthy of credit." I have made the widest possible induction from all existing Records in which there are quotations of poems made anterior to Confucius, and the conclusion to which I have been brought is altogether confirmatory of that deduced from the Works of Tso K‘ëw-ming. If Confucius did make any compilation of poems, he had no such work of rejection and expurgation to do as is commonly imagined.

Proofs of the existence of the Book of Poetry before the time of Confucius.[ii.] But I believe myself that he did no work at all to which the name of compilation can properly be applied, but simply adopted an existing collection of poems consisting of 305, or at most of 311 pieces. Of the existence

of the She, or Book of Poetry, before Confucius, digested under four divisions, and much in the same order as at present, there may be advanced the following proofs:—

First, in the "Official Book of Chow," we are told that it belonged to the grand-master " to teach the six classes of poems,—the Fung, with their descriptive, metaphorical, and allusive pieces, the Ya, and the Sung." Mr. Wylie says that the question of the genuineness of the Official Book may be considered as set at rest since the inquiry into it by Choo He, and that it is to be accepted as a work of the duke of Chow, or some other sage of the Chow dynasty. 2 Without committing myself to any opinion on this point, as I find the passage just quoted in the Preface to the She (of which I shall treat in the next chapter), I cannot but accept it as having been current before Confucius; and thus we have a distinct reference to a collection of poems, earlier than his time, with the same division into Parts, and the same classification of the pieces in those Parts.

Second, in Part II. of the She, Book vi., Ode IX.,—an ode assigned to the time of king Yëw, B.C. 780-770, we have the words,

"They sing the Ya and the Nan,
Dancing to their flutes without error."

So early then as the 8th century before our era, there was a collection of poems, of which some bore the name of the Nan, which there is nothing to forbid our supposing to have been the Chow-nan and the Shaou-nan, forming the first two Books of the first Part of the present classic, often spoken of together as the Nan; and of which others bore the name of the Ya, being probably the earlier pieces which now compose a large portion of the second and third Parts.

Third, in the narratives of Tso K‘ëw-ming, under the 29th year of duke Sëang, B.C. 543, when Confucius was only 8 or 9 years old, we have an account of a visit to the court of Loo by an envoy from Woo, an eminent states man of the time, and of great learning. We are told that, as he wished to hear the music of Chow, which he could do better in Loo than in any other State, they sang to him the odes of the Chow-nan and the Shaou-nan; those of P'ei, Yung, and Wei; of the Royal domain; of Ch'ing; of Ts‘e; of Pin; of Ts‘in; of Wei; of Tang; of Ch'in; of Kwei; and of Ts’aou. They sang to him also the odes of the Minor Ya and the Greater Ya; and they sang finally the pieces of the Sung. We have here existing in the boyhood of Confucius, before he had set his mind on learning, 3 what we may call the present Book of Poetry, with its Fung, its Ya, and its Sung. The odes of the Fung were in 15 Books as now, with merely some slight differences in the order of their arrangement;—the odes of Pin forming the 9th Book instead of the 15th, those of Ts‘in the 10th instead of the 11th, those of Wei the 11th instead of the 9th, and those of T'ang the 12th instead of the 10th. In other respects the She, existing in Loo when Confucius was a mere boy, appears to have been the same as that of which the compilation has been ascribed to him.

Fourth, in this matter we may appeal to the words of Confucius himself. Twice in the Analects he speaks of the odes as a collection consisting of 300 pieces, 4 That Work not being made on any principle of chronological order, we cannot positively assign those sayings to any particular periods of Confucius’ life; but it is, I may say, the unanimous opinion of the critics that they were spoken before the time to which Sze-ma Ts‘ëen and Choo He refer his special labour on the Book of Poetry. The reader may be left, with the evidence which has been set before him, to form his own opinion on the questions discussed. To my own mind that evidence is decisive on the points.

The Book of Poetry, arranged very much as we now have it, was current in China long before the sage; and its pieces were in the mouths of statesmen and scholars, constantly quoted by them on festive and other occasions. Poems not included in it there doubtless were, but they were comparatively few. Confucius may have made a copy for the use of himself and his disciples; but it does not appear that he rejected any pieces which had been previously received, or admitted any which had not previously found a place in the collection.


Further errors in the statements in the first Paragraph.4. Having come to the above conclusions, it seems superfluous to make any further observations on the statements adduced in the first paragraph. If Confucius expurgated no previous Book, it is vain to try and specify the nature of his ex purgation as Gow-yang Sëw did. 5 From Sze-ma Ts‘ëen we should suppose that there were no odes in the She later than the time of king Le, whereas there are 12 of the time of king Hwuy, 13 of that of king Sëang, and 2 of the time of king Ting. Even the Sung of Lee which are referred to by the Suy writer and Choo He are not the latest pieces in the Book. The statement of the former that the odes were arranged in order and copied by Che, the music-master of Loo, 6 rests on no authority but his own;—more than a thousand years after the time of Confucius. I shall refer to it again, however, in the next chapter.

Did Confucius then do anything for the Book of Poetry?5. The question arises now of what Confucius really did for the Book of Poetry, if, indeed, he did anything at all. The only thing from which we can hazard the slightest opinion on the point we have from his own lips. In the Analects, IX. xiv., he tells us:—"I returned from Wei to Loo, and then the music was reformed, and the pieces in the Ya and the Sung all found their proper places." The return from Wei to Loo took place when the sage was in his 69th year, only five years before his death. He ceased from that time to take active part in political affairs, and solaced himself with music, the study of the Classics, the writing of the Ch'un Ts‘ëw, and familiar intercourse with those of his disciples who still kept about him. He reformed the music,—that to which the poems were sung; but wherein the reformation consisted we cannot tell. And he gave to the pieces of the Ya and the Sung their proper places. The present order of the Books in the Fung, slightly differing, we have seen, from that which was common in his boyhood, may also have now been determined by him.

As to the arrangement of the odes in the other Parts of the Work, we cannot say of what extent it was. What are now called the correct Ya precede the pieces called the Ya of a changed character or of a degenerate age; but there is no chronological order in their following one an other, and it will be seen, from the notes on the separate odes, that there are not a few of the latter class, which are illustrations of a good reign and of the observance of propriety, as much as any of the former. In the Books of the Sung again, the occurrence of the Praise songs of Loo between the sacrificial odes of Chow and Shang is an anomaly for which we try in vain to discover a reasonable explanation.

Confucius’ service to the She was in the impulse which he gave to the study of it.6. While we cannot discover, therefore, any peculiar labours of Confucius on the Book of Poetry, and we have it now, as will be shown in the next section, substantially as he found it already compiled to his hand, the subsequent preservation of it may reasonably be attributed to the admiration which he expressed for it, and the enthusiasm for it with which he sought to inspire his disciples. It was one of the themes on which he delighted to converse with them. 7 He taught that it is from the odes that the mind receives its best stimulus. 8 A man ignorant of them was, in his opinion, like one who stands with his face to wards a wall, limited In his views, and unable to advance. 9 Of the two things which his son could specify as particularly enjoined on him by the sage, the first was that be should learn the odes. 10 In this way Confucius, probably, contributed largely to the subsequent preservation of the Book of Poetry;—the preservation of the tablets on which the odes were inscribed, and the preservation of it in the memories of all who venerated his authority, and looked up to him as their master.



From Confucius to the dynasty of Ts‘in.1. Of the attention paid to the study of the Book of Poetry from the death of Confucius to the rise of the Ts‘in dynasty, we have abundant evidence in the writings of his grandson Tsze-sze, of Mencius, and of Seun K‘ing. One of the acknowledged distinctions of Mencius is his acquaintance with the odes, of which his canon for the study of them pre fixed to my larger volumes is a proof; and Seun K‘ing survived the extinction of the Chow dynasty, and lived on into the times of Ts‘in.

The Poems were all recovered after the fires of Ts‘in.2. The Poems shared in the calamity which all the other classical Works, excepting the Yih, suffered, when the tyrant of Ts‘in issued his edict for their destruction. But I have shown, in the prolegomena to vol. I., that only a few years elapsed between the execution of his decree and the establishment of the Han dynasty, which distinguished itself by its labours to restore the monuments of ancient literature. The odes were all, or very nearly all, recovered; 11 and the reason assigned for this is, that their preservation depended on the memory of scholars more than on their inscription upon tablets and silk. We shall find reason to accept this statement.

Three different texts.3. Three different texts of the odes made their appearance early in the Han dynasty, known as the She of Loo, of Ts‘e, and of Han; that is, the Book of Poetry was recovered from three different quarters.

The Text of Loo.[i.] Lëw Hin’s catalogue 12 of the Works in the imperial library of the earlier Han dynasty commences, on the She King, with a Collection of the three Texts in 28 chapters, which is followed by two Works of commentary on the Text of Loo. The former of them was by a Shin P‘ei, of whom we have some account in the Literary Biographies of Han. He was a native of Loo, and had received his own knowledge of the odes from a scholar of Ts‘e, called Fow K‘ëw-pih. He was resorted to by many disciples, whom he taught to repeat the odes, but without entering into discussion with them on their interpretation. When the first emperor of the Han dynasty was passing through Loo, Shin followed him to the capital of that State, and had an interview with him. The emperor Woo, in the beginning of his reign (B.C. 139), sent for him to court when he was more than 80 years old; and he appears to have survived a considerable number of years beyond that advanced age. The Dames of ten of his disciples are given, all men of eminence, and among them K‘ung Gan-kwoh. A little later, the most noted adherent of the school of Loo was a Wei Hëen, who arrived at the dignity of prime minister, and published "the She of Loo in Stanzas and Lines." Up and down in the Books of Han and Wei are to be found quotations of the odes, which must have been taken from the professors of the Loo recension; but neither the text nor the writings on it long survived. They are said to have perished during the Tsin dynasty (A.D. 265-419). When the catalogue of the Suy library was made, none of them were existing.

The Text of Ts‘e.[ii.] The Han catalogue mentions five different works on the She of Ts‘e. This text was from a Yuen Koo, a native of Ts‘e, about whom we learn, from the same chapter of Literary Biographies, that he was one of the Great scholars of the court in the time of he emperor King (B.C. 155-142), a favourite with him, and specially distinguished for his knowledge of the odes and his advocacy of orthodox Confucian doctrine. He died in the next reign of Woo, more than 90 years old; and we are told that all the scholars of Ts‘e who got a name in those days for their acquaintance with the She sprang from his school. Among his disciples is the well known name of Hëa-how Ch‘e-ch‘ang, who communicated his acquisitions to How Ts‘ang, a native of the present Shan-tung province, and author of two of the Works in the Han catalogue. How had three disciples of eminence,—Yih Fung, Sëaou Wang-che, and K‘wang Hang. From them the Text of Ts‘e was transmitted to others, whose names, with quotations from their writings, are scattered through the Books of Han. Neither text nor commentaries, however, had a better fate than the She of Loo. There is no mention of them in the catalogue of Suy. They are said to have perished even before the rise of the Tsin dynasty.

The Text of Han Ying.[iii.] The Text of Han was somewhat more fortunate. The Han catalogue contains the titles of four works, all by Han Ying, whose surname is thus perpetuated in the text of the She which emanated from him. His biography follows that of How Ts'ang. He was a native, we are told, of the province of Yen, and a "Great scholar" in the time of the emperor Wan (B.C. 178-156), and on into the reigns of King and Woo. "He laboured," it is said, "to unfold the meaning of the odes, and published an 'Explanation of the Text' and 'Illustrations of the She,' containing several myriads of characters. His text was somewhat different from the texts of the She of Loo and Ts‘e, but substantially of the same meaning." Of course Han founded a school; but while almost all the writings of his followers soon perished, both the Works just mentioned continued on through the various dynasties to the time of Sung. The Suy catalogue contains the titles of his text and two Works on it; the Pang those of his text and his Illustrations; but when we come to the catalogue of Sung, published in the time of the Yuen dynasty, we find only the Illustrations, in 10 Books or chapters; and Gow-yang Sëw tells us that in his time this was all of Han that remained. It continues, entire or nearly so, to the present day.

A fourth Text; that of Maou.4. But while these three different recensions of the She all disappeared, with the exception of a single fragment, their unhappy fate was owing not more to the convulsions by which the empire was often rent, and the consequent destruction of literary monuments, such as we have witnessed in our own day in China, than to the appearance of a fourth Text which displaced them by its superior correctness, and the ability with which it was advocated and commented on. This was what is called the "Text of Maou. It came into the field later than the others; but the Han catalogue contains the She of Maou in 29 chapters, and a commentary on the text in 30. According to Ch‘ing K‘ang-shing, the author of this commentary was a native of Lee, known as Maou Hang or the Greater Maou, who was a disciple, we are told by Luh Tih-ming, of Seun K‘ing. The Work is lost. He had communicated his knowledge of the She, however, to another Maou,—Maou Chang, or the Lesser Maou,—who was "a Great scholar" at the court of king Hëen of Ho-këen. 13 This king Hëen was one of the most diligent labourers in the recovery of the ancient Books, and presented Maou's text and the Work of Hang at the court of the emperor King,—probably in B.C. 129. Chang himself published his "Explanations of the She," in 29 chapters, which still remain; but it was not till the reign of the emperor P‘ing (A.D. 1-5) that Maou's recension was received into the imperial college, and took its place along with those of Lee, Ts‘e, and Han.

The Chinese critics have carefully traced the line of scholars who had charge of Maou's text and explanations down to the reign of P'ing;—Kwan Ch‘ang-k‘ing, Hëae Yen-nëen, and Seu Gaou. To Seu Gaou succeeded Ch‘in Këah, who was in office at the court of the usurper Wang Mang (A.D. 9-22). He transmitted his treasures to Sëay Man-k‘ing, who himself commented on the She; and from him they passed to the well-known Wei King-chung or Wei Hwang, of whom I shall have to speak in the next chapter. From this time the most famous scholars addicted themselves to Maou's text. Këa Kwei (A.D. 25-101) published a Work on the "Meaning and Difficulties of Maou's She," having previously compiled a digest of the differences between its text and those of the other three recensions, at the command of the emperor Ming (A.D. 58-75). Ma Yung (A.D. 69-165) followed with another commentary;—and we arrive at Ch‘ing Heuen, or Ching K‘ang-shing, who wrote his "Supplementary Commentary to the She of Maou," and his "Chronological Introduction to the She." The former of these two Works complete, and portions of the latter, are still extant. That the former has great defects as well as great merits, there can be no question; but it took possession of the literary world of China, and after the time of Ch'ing the other three texts were little heard of, while the names of the commentators on Maou's text and his explanations of it speedily become very numerous. Maou's grave is still shown near the village of Tsun-fuh, in the departmental district of Ho-këen.

The different texts guarantee the integrity of the recovered She.5. Returning now to what I said in the 2nd paragraph, it will be granted that the appearance of three different and independent texts, immediately after the rise of the Han dynasty, affords the most satisfactory evidence of the recovery of the Book of Poetry, as it had continued from the time of Confucius. Unfortunately only fragments of them remain now; but we have seen that they were diligently compared by competent scholars with one another, and with the fourth text of Maou, which subsequently got the field to itself. In the body of the larger Work attention is called to many of their peculiar readings; and it is clear to me that their variations from one another and from Maou's text arose from the alleged fact that the preservation of the odes was owing to their being transmitted by recitation.The texts were all taken down at, from recitation. The rhyme helped the memory to retain them, and while wood, bamboo, and silk were all consumed by the flames of Ts‘in, when the time of repression ceased scholars would be eager to rehearse their stores. it was inevitable that the same sounds, when taken down by different writers, should in many cases be represented by different characters. Accepting the text as it exists, we have no reason to doubt that it is a near approximation to that which was current in the time of Confucius.…


1 "Wylie's Notes on Chinese Literature, p. 6. Tso K‘ëw-ming was not far removed from the era of Confucius.

2 Notes on Chinese Literature, p. 4.

3 Confucian Analects, II. iv, 1.

4 Confucian Analects, II. ii.; XIII. v.

5 Every instance pleaded by Sëw in support of his expurgation of stanzas, lines, and characters has been disposed of by various scholars.

6 When this Che lived is much disputed. From the references to him in Ana. VIII. xv., XVIII. ix., we naturally suppose him to have been a contemporary of Confucius.

7 Analects, VII. xvii.

8 Ana., VIII. viii.; XVII. ix.

9 Ana., XVII. x.

10 Ana., XVI. xiii.

11 In the last section reference was made to the number of the odes, given by Confucius himself as 300. He might mention the round number, not thinking it worth while to say that they were 305 or 311. The Classic now contains the text of 305 pieces, and the titles of other 6. It is contended by Choo and many other scholars, that in Confucius’ time the text of those six was already lost, or rather that the titles were names of tunes only. More likely is the view that the text of these pieces was lost after Confucius’ death.

12 Proleg., Vol. I. p. 4.

13 The petty kingdom of Ho-këen embraced three of the districts in the present department of the same name in Chih-le, and one of the two districts of Shin Chow. King Hëen's name was Tih.

Next: Contents