GLOSSES ON THE BOOK OF DOCUMENTS This book is a reprint of: BY Glosses on the Book of Documents Bulletin of the Museum of Far Eastem Antiquities, No. 20. 1948. BERNHARD KARLGREN Glosses on the Book of Documents. II. BMFEA No. 21. 1949. The present work is a direct sequel to my papers Glosses on the Kuo feng Odes, 1942, Glosses on the Siao ya Odes, 1944, Glosses on theTa ya and Sung Odes 1946 (BMFEA 14, 16, 18). In order to facilitate brief references, the present glosses are numbered in sequence to those earlier ones, our first gloss here being n:o 1207. A fundamental principle in all my glosses is a constant reference to the Archaic Chinese phonology. Only with the aid of the archaic pronunciation of the words is it possible to solve the intricate problems of the so-called k i a -t s i e, phonetic loans, and the true meaning of a word can often be. best decided by etymology, i. e, through determining to which word family it belongs, what js its word stem (cf. B. Karlgren, Word Families in Chinese, :BMFEA 5, 1933). The correctness of the decisions thus reached with the aid of phonology presupposes that the phonetic system of Archaic Chinese is sufficiently known and that my reconstruction of that system (as summed up in my Grammata Serica, BMFEA 12, 1940) is in the main trustworthy. Some writers have called in question the reliability of this reconstruction or have simply sweepingly condemned it. Their reason for this judgment is principally th!s: that the language of the Odes is not homogeneous, that it is a conglomerate of a series of disparate ancient dialects, and that no reconstruction of one definite »High-Chinese>> early Chou language can be effected with the aid of the two principal sources, to which I have had recourse: the rimes of the Odes and the composition of the h i.e she n g characters (those characters which have a >>phonetiG>>). In favour of the thesis that the language of the Odes is heterogeneous, repre senting several or many ancient dialects, two arguments have been advanced. One is the classical tradition that the section Kuo feng means »The customs (moeurs) of the states>>, ~tnd that they are popular songs culled in various feudal states and simply brought together in a collection. This thesis was long ago definitely refuted by H. Maspero, and, ·as we shall see presently, it is utterly untenable. Moreover, as I have proved in Gloss 757, though fen g 'wind' can certainly mean in many contexts :a current, a custom', its meaning in the Odes Reprinted in offset is quite different. It means 'air', melody, and the Kuo feng (Pei feng, Cheng feng by etc.) are court poems composed in the Chou capital but sung with the airs of the various feudal states (>>the airs of Pei, the airs of Cheng>> etc.). ·The second argu ELANDERS BOKTRYCKERI AKTIEBOLAG ment, which has sometimes been advanced with strong emphasis and in a very GOTEBORG 1970 39 BULLETIN OF THE MUSEUM OF FAR EASTERN ANTIQUITIES B. KARLGREN: GLOSSES ON THE BOOK OF DOCUMENTS apodictic way, is that the earliest d i a I e c t dictionary, the Fang yen,. Chou time it has become more and more obsolete in the lapse of many centuries reveals that many words that occur in the Odes were d i a I e c t w o r d s, not and lingered on in middle Han time as a dialect word only, in the regions of Yen general words (what the Chinese call t 'u n g hi n g h u a): words used only in and Ts'i (just like hOs and bairn). certain provinces and quite unknown in other parts of early China. The hete If the arguments adduced for the linguistically heterogeneous nature of the rogeneous nature of the Odes seems thus to be definitely proved, since part of its Odes thus fail utterly, it does not necessarily follow that the theory as such is vocabulary can be shown to be )>dialect words)>. wrong: a true theory can sometimes be advocated with faulty arguments by This argument only goes to prove that its propounders lack even an elementary ignorant writers. Let us therefore examine the case somewhat more closely. knowledge of linguistic science, its laws and methods. Fang yen was written by The principal source used for the reconstruction of Archaic Chinese is the rimes Yang Hiung, who died in 18 A. D. It depicts the dialect vocabulary of middle of the Odes (Sill king). In my Grammata Serica (pp. 90-110) I have tabulated Han time. The Odes were composed for the most part in Western Chou time all the rimes in the Kuo feng, Siao ya, and Ta ya sections and those parts of the (1027-771 B. C.), a few of them slightly later, but none later than the 6th century Sung which have a regular riming system. There are altogether 1815 characters B. C. T h e r e i s a g a p o f 600-900 y e a r s b e t w e e n t h e 0 d e s which occur in a riming position (the words are actually more numerous, for many and t h e F a n g y e n. It is quite impossible and unallowable to draw any characters occur with two or several meanings, and sometimes with two or several conclusions whatever regarding the language of the Odes from a document like readings). They stand as rimes in 4925 stanzas. These high figures are important. the Fang yen written so4ll.any centuries later than the former. If a writer in the We have here a very extensive material indeed, and thanks to these voluminous field of Anglo-Saxon philology were to propound that a word now only occurring materials it is possible to determine with a great measure of certainty w h i c h in certain English dialects but not in High English must eo ipso have been merely r i m e s a r e r e g u I a r and which are not. In other words, it has been possible a dialect word and not a general word in the language 1100-1200 A. D., he would to establish t h e r i m i n g c a t e g o r i e s of the Odes. Such and such finals be dismissed immediately as an ignoramus. The simple fact, known to every rime regularly with. such and such other finals, through~ut the large collection of serious linguist, is that when a word becomes obsolete and dies out in a language rimes. In this way, basing myself largely on, but further developing, the results it does not do so suddenly and simultaneously in all the parts of the language field. of a number of great Ts'ing scholars, the best of them being Tuan Yii-ts'ai and, When a once frequent and general word loses ground and beeomes forgotten in the particularly, Kiang Yu-kao, I have established 26 r i :in e cIa sse s, each of them leading regions of the total field, it almost invariably lingers on, sometimes for comprising a certain number of Archaic finals. The 4th class, for instance, con many centuries, in certain parts of the field: from being a g en e r a l w o r d it tains the finals which, when supplied with their reconstructed sound values, were becomes a d i a 1 e c t word, and finally, perhaps, dies out altogether. We can as follows: an : wan : an : wan : jan : jwan : ian : iwan : an : w~n : jan : jwan. All these never conclude from the fact that a word is today limited to one dialect only that finals rime freely and regularly with each other. Similarly, the riming finals of it was not a general word some centuries ago. A few examples: In Ancient Nordic class 21 were: ak: wak : jak : jwak: ek: wek: juk : ag: wag: jag : jwag: eg : weg: jug. there was a general and regular frequently occurring word which meant 'head': If we now examine those rime lists more closely, we shall find that the number Icelandic hauss, Anc. Swedish and Anc. Danish hos. In Medireval Nordic texts it of cases in which a word rimes with one or several others in a reg u 1 a r way, recurs passim as a normal word for 'head'. Today it is entirely forgotten in most i. e. within the limits of the established rime classes, is 1571. Besides these 1571 parts of Sweden and Danmark; in central and northern Sweden it is entirely strictly regular rimes, there are, on the one hand, 92 rimes which might be called unknown, and in High Swedish it is so dead that a Stockholmer has no idea what •subsidiary rimes•, i. e. rimes that are not strictly regular (keeping within the a hOs means. But in some dialects in the south of Sweden and likewise in a few established classes) but are still phonetically very understandable, a slight licentia Danish dialects it still lingers; there it is still a living dialect word meaning 'head'. poetica, the phonetic similarity being still sufficiently strong to warrant a makeshift An English parallel to this is the word bairn ('child') which is now a dialect word rime.1 On the other hand there are 27 frankly bad and faulty rimes, e. g. (Scottish dialect), but which earlier was a general word (it already occurs in Beo d'iok: d'ieg; twar: gjwed; gljam: xiang; iam:tsong. wulf in the form bearn). It would be easy to adduce scores of English examples 1) Such subsidiary rimes have: l. The same final consonant, but principal vowels which do not of this kind. In precisely the same way a word which Yang Hiung (in Fang yen) rime in the strict system: m (a) which properly rimes in the m (at) class, in 14 ca8es rimes in the en in middle Han time describes as a dialect word occurring only in the provinces of (et) class; &.! and og, which in the strict system are not allowed to rime, form subsidiary rimes in Yen and' Ts'i may very well have been a general and regular t ' u n g h i n g 13 cases; 6ng ("k, &./)occasionally rime with ung (uk, ug); am a.nd/im with am. 8d with ad, an with en. 2. The same principal vowel, but different final consonants, which in the strict system do not rime Chinese ~ord 900-600 years earlier, and as such occur in the Odes. Fang yen (however, they are still consonants that are genetic.ally cognate: both dentals, or both gutturals), can prove nothing whatever to the contrary. From being a general word in early e. g. 9d with 11r; m with ar; -ag with -ang. 40 41 BULLETIN Oft' THE MUSEUM OJ? PAR EASTERN ANTIQUITIES B. KARLGREN: GLOSSES ON THE BOOK OF DOCUMENTS ------------------ ---- These figures: 1571 strictly regular rimes (often with quite a number of words The general conclusions that can be drawn from the preceding are as follows: in the riming series) and 92 subsidiary rimes as against 27 faulty and irregular l. The language of the Odes is on the whole quite homogeneous (the court rimes, afford indeed absolutely crushing evidence. Such a high degree of regularity language of Chou) and can ·safely be used as a basis for the reconstruction of its is remarkable. It reveals two important facts: Archaic phonology. 2. The rimes of the Odes are sufficiently numerous and First, there cannot be any possibility whatsoever that the Kuo feng odes are sufficiently strict in their rules to allow of a systematic reconstruction. 3. The popular songs culled in the various farmers' regions. The system is far too strict Archaic phonology thus reconstructed (as recorded in dictionary form in Grammata (the faulty rimes in the Kuo feng are indeed very few: 8 cases). No folk songs Serica) can safely be used for philological purposes, in the interpretation of difficult have ever such precision in their rimes. The Odes, even the Kuo feng, are' all texts. court poems composed by scholars who had a marvellous versificatory technique, If we now revert to the present glosses on the Book of Documents, we should quite remarkable in view of the early era of the Odes. add a few remarks. Secondly, the language of the Odes must on the whole be quite homo gene- The Shu king (Shang shu) is a collection of exceedingly difficult and, largely, o us. If it were heterogeneous, if the various odes represented different archaic very obscure texts, and it is doubtful whether it will ever be possible definitely to dialects, it would be absolutely impossible to establish those 26 riming classes, so determine its correct form and interpretation. In the cases in which we happen strictly adhered to (1571 and 92 cases against 27). The consistency, indeed the to know (through ancient quotations) the so-called Kin-wen version, we frequently very severity of the rimi.ng system in the Odes at once and irrevocably proves find that it deviates badly from the traditional Ku-wen version. This makes it that the language of the odes is one, i. e. the language of Chou, the Royal tempting to conclude that in the numerous cases, known only in the Ku-wen ver court. sion, which are obscure and difficult to construe, the text is too badly corrupted But let us for a moment suppose, for the sake of argument, that, after all, a to admit of a successful analysis. But in my opinion this should not deter us from certain number of )>dialect \-lords)>, not really belonging to the court language of endeavouring to unravel as much of its secrets as is possible by the aid of the Chou, had crept into some odes. Would they, eo ipso, by their presence in the materials available. If we can arrive at a fairly reasonable interpretation of a Odes, invalidate our system of phonetic reconstruction? By no means. In order difficult Ku-wen passage, there is always a good chance that the text is not spoiled to be )>dangerous)> they would have: l. to stand in a riming position (since the but is simply difficult because of its high age and archaic language. reconstruction is solely based on the riming functions; words in the middle of the Owing to the great difficulties of most' of the Shu chapters, the divergences lines are not pertinent to the reconstruction question); 2. to be the only repre among the Chinese scholars as to the true division of the lines and meanings of sentatives of certain reconstructed finals. As a rule, however, this never happens. the phrases have been very considerable. The Shu king literature is immense, If we examine the said 1815 characters which occur as rimes, we find that circa and it would be impossible to adduce, in every case, all the proposals for inter 1200 (two thirds, a very high figure) are such as belong to the most common stock pretation that have been advanced. That would have required several stout of the general language: m a 'horse', k u o 'state', tung 'east', y u 'to have', volumes. It has been necessary here to select and submit to the student the etc. The remaining circa 600 are less common, but the great majority of them opinions of a limited number of leading scholars. The authorities most frequently are such as are, after all, attested in three or four other archaic texts independent referred to are the following: of the Odes, and may therefore, with a large measure. of certainty, be expected to Ancient and mediaeval authors: have been general (t ' u n g hi n g) words. But then there are, at any rate, a Cheng Huan (2nd c. A. D;); the glosses of this author that are known today are few riming words, which are indeed exceedingly rare, some of them even quite all recorded, for instance, in the works of Kiang Sheng and Sun Sing-yen below. unknown outside the Odes. Are these not risky materials, since we have no means A few glosses of Cheng's teacher Ma Jung are likewise known. of judging whether they are general words or )>dialect words)>? No, because the Pseudo-K'ung An-kuo (3rd c. A. D.; his name here always abbreviated into; reconstruction scheme is not based on them. In ode 8 we have a rare rime word PK'ung), in the main embodying the teachings of Wang Su. His glosses are to be ~~ *kiet I kiet I k i e 'to take in the held-up flap', but the Archa~c final iet was found in the Sh'i san king chu su. not reconstructed with the aid of that hapax legomenon but of common and general K'ung Ying4a; (beginning of the 7th c. A. D.), in Sh'i san king chu su. words like 'fffl *tsiet jtsiet J t s i e and *fi kiet I kiet I k i e. In ode 28 the rime Ts'ai Ch'en (beginnirig of the 13th c. A. D.) the summarizer of the Sung school word Ji] g'ang I yang I hang is a hapax legomenon; but the Archaic final ang is interpretations. In: Shu king tsi chuan. · · based on a long series of very common words, riming in the Odes, such as JIJIJ # Ts'ing and later scholars: . • -*· ~ ~ Kiang Sheng, in: Shu tsi chu yin su (in: Huang Ts'ing king kie). 42 43 BULLETIN OF THE MUSEUM OF /t'AR EASTERN ANTIQUITIES B. KARLGREN: GLOSSES ON THE BOOK OF DOCUMENTS Sun Sing-yen, in: Shang shu kin ku wen chu su (ibid.). find that) the emperor Yao was called Fang-hiin•. (So also Legge). Cf. Shu: T'ai shi (ap. Han shu: P'ing Tang chuan): Cheng k i k u, k i en kung 16. Here the first phr. Liu Feng-lu (largely incorporating the ideas of Chuang Shu-tsu) in: Shang ends by k. i k u, just as in our y ii e j o k i k u, t i Y a o y ii e above.- B. Various shu kin ku wen tsi kie (in: Huang Ts'ing king kie su pien). Han and later authors punctuated after Y ao: Y ii e j o k i k u t i Y a o, Y ii e ]' a n g Wang Yin-chi, in: King yi shu wen (in: Huang Ts'ing king kie). h ii n. This must already.have been the opinion of Pan Ku, for in Po hu t'ung: Sheng Ch'en K'iao-tsung, in: Kin wen Shang shu king shuo k'ao (in: Huang Ts'ing jen he cites the quite analogous phr. in Shu: Kao Yao mo (Y ii e j o k i k u K a o Y a o king kie su pien). y ii e Y ii n t i etc. 17), in a way which shows that he punctuated: y ii e j o k i k u K a o Y a o, y ii e Y ii n t i: >>Examining into the ancient Kao Yao (we find that) ....> >, and he must Yu Yue, in: K'un king p'ing yi (ibid.). necessarily have understood our Yao tien phr. above in the same way. For the rest, the Sun Yi-jang, in: Shang shu p'ien chi. interp. vary. -a. Cheng Hiian: j o 6 = 18 'to accord with, to be concordant, obedient' Chang Ping-lin, in: Ku wen Shang shu shi yi. (common); k i 15 = 19 'to accord with': k u 20 = 21 'Heaven'. Thus: >>The emperor Wang Sien-k'ien, in: Shang shu K'ung chuan ts'an cheng. Yao who was j o concordant (obedient) and k i k u agreed with (>>the ancient one>>===) Yu Sing-wu, in: Shuang kien yi Shang shu sin cheng. Heaven, was called Fang-hiim. For support have been adduced: Li: Ju hing 22, on which Cheng Hiian says k i 15 = 23: »The scholar has present men with whom he associates, More occasional references are made to many more scholars, for instance, Su and ancient men with whom he accords>>; but k i 15 here may equally well be taken in Shi, Lin Chi-k'i, Lu Tsu-k'ien, Chu Hi (all Sung), Wu Ch'eng (early Yuan), Huei its ordinary sense: >>and ancient men with whom he studies>>; Hanfei: Chu tao 24 - here Tung, Wang Ming-sheng, Tuan Yu-ts'ai, K'ung Kuang-sen, Chu Tsun-sheng, Yuan k i is sure enough combined with t 'u n g 19, but it still means 'to examine': »We keep Yuan, P'i Si-juei (all. Ts'~g), Wang Kuo-wei, Ku Kie-kang. on our course and scrutinize and accord with ih>; Chou Ii: Siao tsai: >>He regulates the ·Conscriptions for war and hunts 25 according to registers and (investigations =) calcula tions>>-!-though Cheng Chrmg here says k i means 26 'counting' or 23 'bringing together', i. e. >>summing up», it frmdamentally means 'investigation', and Cheng Chung's h o 23 has here not the sense of 'to accord with'; Chuang: Siao yao yu 27 - here Lu Te-ming reads k i and Si-ma Piao says k i = 28 >>The great flow (comes to, reaches:) joins Heaven>>, Yao tien (including the Shun tien of the orthodox version). but Sii and Li read k' i as in k' i s h o u 29: >>The great flow knocks against Heaven>>, which is certainly preferable. Thus Cheng Hiian's gloss k i 15 = t ' u n g 19 lacks true 1207. Y ii e j o k i k u t i Y a o y ii e Fang h ii n 1. support; k i never really means 'to accord with'. K u 20 = 21 'Heaven' refers to_ Yi .~or y ~ e 2 there.is the variant 3 in quotation by Li Shan:comm. on Wsiian. In Wang Chou shu: Chou chu 30. But this is exceedingly scholastic. - {3. PK'ung: j o 6 = 18 YI s Luling kuang tien fu we find 4, which suggests that Wang had a Shu version with 3. (with Cheng above), k i 15 = 31 'to examine' (common), paraphrasing: 32 >>The one Yiie jo: who could accord with and scrutinize the ancient (norms) and practise them was the A. Y ii e, written 2 or 3 or 5 (all *gjwiitfNwvtj y ii e) is a common initial particle, un emperor Yam>.- I'· Ts'ai Ch'en simplifies the interpr. by taking k u as an attribute to translatabl~.; very common in the Shi etc. J o 6 is similarly often a mere initial particle, t i y·a o : >>Investigating the ancient emperor Yao>>. - A is the earliest interpr. and sup cf. Shu: Kun Shi 7, where j o does not mean 'if', since it is equal to Shu: Ta kao 8 both j o and ! ii e. being mere particles equal to wei 9 (for more ex. see Wang Yin-chi: King ported by a good Shu par. Fang-hiin: chuan shi ts'I). In our phr. above, y ii e-j o is a binominal initial particle as already A. PK'ung reads *pjwangfpiwangf fang in rising tone: >>(The one who) imitates the recognized by Ts'ai Ch'en ( = 10), though he tries to invest it with a fuller s~nse akin to merits (of the ancestors)>>.- B. Cheng Hiian and Wang Su read *piwang in falling tone:\ th~t ~f y u e 2 'to speak': Y ii e- j o = 11 >>the saying is like this>>, i. e. mamely, enim>>. >>(The one who) spread everywhere his merits>>. -It is really futile to try to choose be !his IS vetoed by th~ va~iants ~ and 5 of the particle, and it cannot be carried through tween these attempts at the etymology of the name. m S~u: Shao kao: y u e J o I a 1 s an y ii e 12, where y ii e j o is obviously an empty 1208. K ' i n m i n g w e n s i an an 33. partiCle (cf. Gl .. 1715). - B. Various ancient comm. took j o 6 to be a verb, see For k 'in 'respectful', min g 'enlightened', wen 'accomplished' there are various ~elow. Yang Hmng had already understood it as a verb, since in his Ki Ts'in mei sin lengthy abstract definitions in the early comm., not worth while reproducing here. (U: Wsiia?) ~e says j o k u c he c h' eng Y a o Shun 13 »Among those who accorded with antiqmty, one praises Yao and Shrm>>, evidently alluding to our Shu passage. - C• •c o~vreur takes both y ii e j o in their modern sense: >>I say, if>> we examine etc.). / El ~ ~-(!;" ~ ~ g JP(t./Lt, El 3 ~ lt. ~~A$ -t ;fi ~ 1-Ji;f, .?. ~ '- ~ 7, ~ ,k ~ ·t;:t. ,_ ~ This disregards the ancient variants of y ii e. - A is confirmed by conclusive text. par. The whole sentence: J;.. !f_,l/L r.tti 10. ;_t~,g.tlJf 11. Jt ~ ~.~ ;.~:.~~¢. 3. I~ 1:J.~ <5 ~~1L~ 14. e-=##.i\!; •3 ~· The oldest, i.nterpr. ~ay be gleaned ~rom Yi Chou shu: Wu mu: Y ii e j o k i k u, Y u e. ~ h a o t I en c h 1 tao 14, which can only mean: >>Examining into antiquity 13~ .:K_ .Z.di_ IJ.A1i It lf.j~-;:fj~ /7_. E1 ;#".t,~ ;!i $-~~13 /L~J}. J1l!l,19.f<.\ ;UJ. i; t/. A.t.t.1f&;{f ~ h(wave ef tI?tds tchoamt)m tohne mweaayn oinf gB orifg 'htto Heexaavmeinn e.'.. ..T >>.h eI tp ihsr .q uisi tem eovdiedlelendt tahfatet r ko ui r 1f5a mheorues m Suhsut J....fft,j! iJ A~ 2-:J. [email protected] {~.:§ f;f ~:<.1!.~ I~ z.t;.~;t. ffM:f tc"!t t-1.};;..~-#t ft:...23 ~n~ kp~ Is ska gue, abto ~1e ,Y a an do ity isii cel eaFr athnagt -thhe iCi nh oua ntidm ein atuertphroert epdrm: c•tEuuamteidn ianfgt eirn tko ua n:t iqyu iiit ye (jw oe .tj 30 J;;..J?.. -;t; J/ 4f ..ll ~lt.J'Irt .1f ;t;-Q. ~><'1" :t.. -1j' ~ ~,~.~ tfl:t.~. j! tc Jo/. ~.:lf."A_:t~.M:-.114 x-. 44 BULLETIN OF THE MUSEUM OF PAR EASTERN ANTIQUITIES B. KARLGREN: GLOSSES ON THE BOOK OF DOCUMENTS ---- ~-·~--------··--·-- --------·---~·-----~------ A. PK'ung (Ku-wen) read as above: wen s 1 an an. Shiwen reads 34 *sjaglsil s 1 c. Si-ma Ts'ien (Wu ti pen ki) renders the line n eng min g h ii n t e 68. Possibly in falling tone, or in the ordinary way, in even tone. Cheng Hiian took it in its ordinary Si-ma had a text with h ii n t e, but that is not certain, for he frequently replaced words sense of 'to think', thus: 'thoughtful'. Ma Jung on the other hand says 35 >>the virtue in the Shu text by interpretation synonyms (e. g. here n eng for k 'o). It is therefore being pure and complete is called s i>>. This is based on Yi Chou shu: S h i fa 36. Some more probable that he thought the 53 *ts-jwan or 66 *siwiJn of the text was a loan char. Ts'ing scholars imagine that Ma took 34 *sjag to be loan char. for 37 *sgk 'sincere' of B for 59 (or, perhaps, for 60 *(l'iwiJn f dz'iuen Ish u n 'docile, obedient', which he then below, but this is not convincing. His definition, that of Yi Chou shu, may just as well glossed by 69). This 59 *xiwan I xiuiJn I h ii n is etym. id. w. 61 (same sound). The be a free paraphrase of s i 34: 'thoughtful = wise, sage'. There is nothing to prove that stem means 'to teach to train' and also '(taught, trained=) docile, obedient, concordant'. the Chou shu definition refers to our Shu passage. PK'ung took k ' in m i n g w e n Sii Kuang (early 5th e.) briefly says 59= 61. But Si-ma's line has been understood in s i as subject and an an as verb: >>By his reverence etc. he tranquillized (the world)>>. two ways:-a. Si-ma Cheng (8th c.): 59= 60 (cf. above): »He was a?le to make bri~ht his Ts'ai Ch'en more nq,turally takes an an as an adjective coordinated with the preceding: (docile:) concordant virtue>>. -fl. Sun Sing-yen: 59= 61 as an active verb, referrmg to >v>eHres iwona s( Kreisnp-ewcetnfu) l,a pe.n Slihgahntegn sehdu, akc'acoom lipnlgis hyeod r, etahdo uwghetnfu ls aen dy epenac efyule>>n. -38: B•.R Aesnpeoctthfuelr,· Cheng Chung's comm. on Chouli: T'u hiin (h ii n = 62 'to guide'), thus: l>He was able ~o· enlightened, accomplished, sincere, peaceful (mild)». For 39 *sakjsgkl s e 'sincere' see Gl. 73, make bright his (training =) instructing virtue>>, which is more far-fetched. - There IS no reason to abandon the earliest interpr. (A). with text par. Whether the orig. Shu had 40 *·an I ·an Ian or 41 *·an I ·an I yen, the meaning is 'peaceful, mild', with Ts'ai Ch'en above. -We have s e 39 in precisely this kind 1211. Y i t s ' i n k i u t s u 63. of enumeration of virtues in the Shi (see Gl. 73), and version B therefore seems pre-ferable. >>And so he made affectionate the nine branches of the family>>. There have been two 1209. K u an g p e i s i pi a o 42. expl. of the term k i u t s u. PK'ung (Ku-wen versionf'read like this; the Han shu: Wang Mang chuan (Kin-wen A. Ma Jung and Cheng Hiian, foll. by PK'ung and Ts'ai Ch'en: the 9 branches were from version) had 43. great-great-grandfather down to great-great-grandson, all of the sa~e ~in g 64 fam~ly A. Cheng Hiian (foll. by Ts'ai Ch'en), disregarding the Kin-wen, ,takes k u an g 44 name. Pan Ku (Han shu: Kao tiki) records that in 200 B. C. >>one mst~tuted the office with its ordinary reading and sense: »His brightness (glory) (covered=) reached to the of t s u n g c h e n g 65 governor of the clan in order to regulate k I u t s u t~e 9 four extreme points>>. - B. PK'ung, saying k u an g 44 = 45 'fully (extensively)' branches of the family>>; here k i u t s u can only mean those who had the same s 1 n g takes 44 as short-form for 46 (Erya 46 = 46), which is a variant of 47 (both *kwdng, and clan temple. -B. The schools of Ou-yang and Hia-hou (ap. K'ung Ying-~a's comm. falling time), thus considering the Ku-wen form k u an g 44 as a variant for the k u an g on Tso: Huan 6): the 9 branches were collateral branches on both the fath~r s and the 47 of the Kin-wen: »He extensively (covered =) possessed the four extreme points». This is mother's side and those relatives by marriage for whom one wore mournmg. Po hu logically best and confirmed by many text par., see Gl. 899. PK'ung thinks that the t'ung: Tsung tsu (compiled by the same Pan Ku quoted un~er A above) _enumerates ,t~en; subject of the clause is the virtue told before: >>(His virtuous fame) (fully =) extensively in detail, adding the amusing speculation that though k I u 66 certainly ~eans mne (covered ==) spread out over the four extreme points>>, but the parallels adduced in Gl. branches at the same time it means k i u 67 'exhaustively'' all those who are tiCd together 899 show that the subject is Yao himself. The following phr.: (Ku-wen:) k o y ii shan g by bonds' of t s 'in kinship affection.-It might seem th~t, since Yao. hi~self personally, hi a 48 (Kin-wen: k i a y ii shan g hi a 49 - the words 50 *klfik I kvk I k o and through his virtue, >>made affectionate the 9 branches>>, this can~ot, w1~h mterpr. A, have 51 *kd I ka I k i a were synonymous) should be explained thus: >>He reached to (Heaven) in view 9 generations alive simultaneously. But the expressiOn k 1 u .t s u may? of above and (J1Jarth) below>>, i. e. he obtained the favour both of God in Heaven and the course, have been generalized from the literal and technical sense of >>the mne ge~erati?ns people on earth. - We compare: Shu: Kao Yao mo (now Yi Tsi): T i k u an g t' i en of the clam into a more general and vague expr. for >>the whole clam. In archa1c China c h 1 hi a c h 1 y ii h a i y ii 70. Here again (with Kiang Sheng and Sun Sing-yen) it was certainly the branches which had the same s i n g 68 th~t were felt to be _the funda k u an g should be equal to 46, thus: »Oh emperor, extensively, all under Heaven, even to mental unit of the family, and those who had a different s In g 69 - relat~ves on the the corners of the seas .•. » female side and relatives by marriage - were much more outsiders. Hence mterpr. A 1210. K ' o m i n g t s ii n t e 52. in the general sense of >>the whole clam is more plausible than B. The word *tsjwiJn I tsjuen 1t s ii n 53 means 'great, eminent', and in the Odes it is often 1212. Pien chang posing 71. e wr. 54 (same reading), see Gl. 561. Li: Ta hiie quotes our line k' o min g s ii n t iSiS, It has long been established that p 'in g 72 (small seal form 73) is a scribe's error the 66 *s-jwiJn I s-juen Is ii n meaning 'high, lofty', a cognate word. It is impossible to for p i en 7 4. (small seal 7 5) = pie n 76 'to distinguish' (S'i-ma Ts'ien rendering it by ~he tell whether the orig. text had *tsfwiJn or *siwrm. Probably the Chou graph was simply loan char. pi en 77, and Cheng Hiian reading pien chang 78), see Gl. 716 With 57, and the Han scholars have supplied either rad. 'man', taking it to be *tsjwJn 'great', or 'mountain', taking it as *siwgn 'lofty'. A. The oldest interp ..i s that of the Ta.hiie author, who clearly expounds the meaning as: »He was able to make bright his lofty (great) virtue>~. Wang Ch'ung (Lun heng: Ch'eng JV~&,-\3 ~-17. ~-:38. Jt"k~J~ ~ 37 ~'to ~ 4; ~ 1-.i .J'L :f:Jt \'9 ~ 43.-$~ fJJl# ;It, H £ ~,.f7\':_, itsn' aei)o mfomllo. wosn tShihsu i nsateyrsp rt. ,s aiin nd Ct hee nmg eHaniisa n'm inen c oofm emm. inoenn Lt iv: iTrtau eh'i iaen. d- PKB'.u nCgh eenxgp oHuiniadns 47:1ii ~;s.M-1 J: 1-1t1 1~J... t J:. -r:: so. ~s1.1n.>.t. ;tr.J)J 1ft~.~,;-J~.!sJ. 7ttJJl ~.t..~:.'f~~w£co ~ further: >>He was able to (make bright =) proem ote those of eminent virtue>>. (Both Legge 5i~ti..t!Ji t&H~Jr.~J'\ ~J'ltl'l~'l•x $-;lt.-.n;<#~.rH1. .0:-..~. iE•c:/La ~""' AJ:':lcr: jtltl, and Chavannes MH I, 41 take t s ii n and t separately: >>the t s ii n eminent, able and the t e virtuous>>, but that was certainly not Cheng's idea). - 70. $ fL.kz 1- .?i.r ~r.% rl..f-31 B ft?:i Sf 73. <5 ·~ffi rJ.llf R:fit. rrcD!?SmftH iJj.!!..W~ 46 47 BULLETIN OF THE MUSEUM OF FAR EASTERN ANTIQUITIES B. KARLGREN: GLOSSES ON THE BOOK OF DOCUMENTS ---- Shl P.arallels. When ~~'ung foll. bY: Ts'.ai Ch'en takes 72 in its ordinary reading p 'in g it by y an g 94 'to nourish'. Just as in the preceding line p i e n c h a n g 1 is a com and sense: >>He tranqmllized ...> > he IS wide of the mark. P o s i n g »the hundred clans>> bination of two analogous words: 'to distinguish' and 'to (give marks of distinction to =) in spite of the fact that it often means »the people>> generally (passim in the Shu) here' honour', so here we have y ii -fan, two analogous words: y ii 84 ( = 96 = 95) 'to with Cheng Hiian and PK'ung, undoubtedly means >>the gentry>> (those who held ~fficiai give plenty of food to' and fan 90 'plenty, ample, prosperous'. Thus the B text line positions), a~ shown by the context: Yao's i~~luence regulates first k i u t s u his family, means: •The numerous people were amply-nourished and prosperous, and then became then p o sIng the gentry (the great families), then wan pang »the myriad feudal concordant». The short-form 84 for 96 has nothing astonishing about it. In the same st~tes» anf finally t~e l i min ?m~ti~ude of the (common) people>>. Sun Sing-yen way the phr. y ii y i 2 (Ch'u: Kiu pien) recurs vi'ith the short-form y ii y i 3 in a Wei pomts out that we fmd the same distmctwn in Kyii: Chou yii: p o s i n g c h a o m i n time stone inscription.-C. A Han-time stone inscr. in the Li shi has a third text version: 79 >>th~ hundred clans and the multitude of the (common) people>>, on which Wei Chao: 4. The Ts'ing scholars discuss whether pie n 5 is at loan char. for the 91 of A or the 90 p o sIng = 80 »the hundred officers>>. For Ts'ai Ch'en's idea that p o sing here of B. Chang Ping-lin would take this pi e n 5 = 6 = Shuowen 7) to mean 'to rejoice', would mean >>the common people of the Royal domain proper>> there is no support whatever. since Mao Heng defines it as such (then loan char. for 8) in Ode 197; but that is not admis Thus our line means: •He distinguished and (gave marks of distinction to =) honoured the hundred sible, see Gl. 591. There is thus no safe text corroboration for a meaning 'to rejoice'. In the orthodox A text the char. 84 constitutes an almost insuperable difficulty, whether clans (the gentry) •· with interpr. a or fl. No such difficulty mars the version B, which is attested early (W. Han 1213. L i min y ii pie n s hi. y u n g 81. time) and makes very good sense. On the other hand, it can hardly be claimed that the -~o~ l i m i ~ : not >>the ?Iack~haired people>> but >>the numerous people>> see Gl. 430. pie n 91 (*pljan) of A was a loan char. for the fan 90 (*pjwan) of B. We have, in fact, Y~ S~g-wu believes th~t li 82 IS equal to 1 i 83 'to benefit'; but the standing phr. I i two different and irreconcilable versions, of which B is decidedly preferable. m I n IS very common m contexts where 83 would make no sense. 1214. N a i m i n g H i H o 9. A. The ort~odox version 'reads as above. PK'ung says nothing of the awkward y ii In my paper Legends and Cults in ancient China (BMI!'EA 18) I have extensively studied 84 and explams: >>The numerous people were (changed:) transformed and then became this and the following paragraphs and arrived at the conclusion that the Yao tien is here concordant>>. Y ii 84 has been explained in two ways.-a. Ying Shao (2nd c. A. D.) in composed of two different parts. The primary part consists of the first 18 words: N a i comm. on Han shu (c f. below) says y ii 84 = y ii s h i 85 'thereupon'. It is true that m i n g H i H o k ' i n j o h a o t ' i e n, l i s i a n g j i y ii e s i n g c h ' e n, k i n g b?th Ery~ and Mao's comm. on Shi say y ii an 86 ('thereupon') = y ii 84 and Sun s h o u j en s hi >>And then he charged Hi Ho (i. e. one person, the cult-master of the Smg-yen m consequence ~ys that y ii pie n is equal to y ii an pie n 87 >>thereupon sun) reverently to follow the august Heavem etc. followed by the calendar passages: j i was. transfor~ed>>: But m fact there are no text par. whatever with y ii 84 alone and c h .e n g s i n g n i a o, y i y i n g c h u n g c h ' u n, k ii e m i n s i, n i a o s h o u by Its~~ se~mg m the s~nse of y ii s hi 85 = 86. Moreover, in the Shu the regular t s i w e i; j i y u n g s i n g h u o, y i c h e n g c h u n g h i a, k ii e m i n y i n, prepositiOn IS y ii 88 (*g1wo), not yii 84 (* ·jo); in a few instances the present orthodox n i a o s h o u h i k o; s i a o c h u n g s i n g s ii, y i y i n c h u n g t s ' i u, k ii e text has 84, ?ut then there are always early text variants with the 88 that is typical of m i n y i, n i a o s h o u m a o s i e n; j i t u a n s i n g m a o, y i c h e n g c h u n g the _S~u and IS hence correct, and the 84 is due to errors of scribes. - fl. Ts'ai Ch'en, t u n g, k ii e m in y ii, n i a o s h o u j u n g m a o. This primary passage has been rea~zmg_tha:t 84 cannot ~e the preposition y ii, therefore reads the char. 84 *·oj-uo 1w u, embellished and tampered with by the editor in Western Chou time. He wanted to have an mterJeCti~n common m the Shu and the Shi, thus: »The people - lo! - were trans the four seasons 0f the calendar specially represented each by one cult-master, and so he formed>>. This, however, is equally impossible, for this interjection must always commence took the Hi Ho above to be two men: Hi and Ho, and invented younger brothers for them, the.~lause (passim in Sh~ and ~hi), ~nd the line should then have run: w u, l i min pie n who were sent to the east (Hi Chung), the south (Hi Shu), the west (Ho Chung) and the s ~I :fun g. - B. An Imperial edwt of 23 B. C. (Han shu: Ch'eng ti ki) quotes the line: north (Ho Shu). He therefore added a secondary part, four passages inserted at the head of l(cio mmmI.n ~.b Yt: di.i) ftaakne s s fhai ny u9 0n (g* b ~1~-wva nT hI ibs, ~~wasv nb eIefn a dni)f fearse nltolayn e xcphlaari.n efod:r -thae .9 1Y i(n*gp lSiahna o1 eYa cahn s ge a-sko nu ,o f ythien orci ghi n' aul caj lie, ndpa r': inF ge nc hmi i nt gu nH g i t Cs ho u. .n. g; ts sh ee nY imi -iy n i g y Hii ei pjan I pie n) of the A version, which is plainly impossible.-{3. Wei Chao (comm. ibid.) S h u etc. That this is so follows clearly from the fact that in the primary text it is said: says f a n 90 = 92 'much, plenty', which is the ordinary meaning of the word ('luxuriant, N a i m i n g H i H o, whereas, if these were originally meant to be two and the elder ample, prosper??.s') .. Again, Wang Fu in Ts'ien fu lun: K'ao tsi (2nd c. A. D.) expounds brothers of the rest, the passage would necessarily have run: N a i m i n g H i P o H o Y a o y a n g li m I n e r c h i s h i y u n g 93 as >>Yao nourished the numerous people Po (corresponding to Hi Chung, Hi Shu, Ho Chung, Ho Shu). The rea,sons for concluding and caused them to be concordant». Sun Sing-yen thinks that by yang 94 'to nourish' that Hi Ho was primarily one person are fully detailed in my paper cited above. But Wang e:cpounded the word fan 90, but that is not correct. Thew. 95 *·jo 1'i .wo 1y ii to the editor of the Yao tien, such as we have it in the Shu, the phrase certainly meant: ~tha~ this word had no final consonant in Archaic Chinese is. proved by a rime in Ode 164) >>And then he charged Hi and Hm>. e t? give plenty of fo?d, to satiate' had a variant 96. Tso: Siang 26 k i a s h an t s y ii s I. ~7 >>When they mcreased the number of their dishes, they gave abundantly (to their m_mister~)>> (:ru Yii: y ii = 98, :to satiate'f This expr. y ii s i 99 recurs written y ii ·~I 100 to give plenty of food m Hou Han shu: Ma Jung chuan. The variant 96 for 95 R_ to B it .r;$_tt ~a.;fii£52~ 8'3.4:.1 t~t:I:R !l5.ffl ~a i._K1.ffi~ 1l ~9'8'. -f' f't~ 1,(77'\'~ IS not known from any text earlier than this, but Wang Fu's explanation of our Shu line I:F.'t~i.fo{fi '!d~t:t1Y n #'..~~ ~w£>t...a~~9'1-l_?r~::t'.~:;7./Jc~~·1~~f$9.r~n~ above . (y a. n g 94) reveals that it was current in Hari time and that Wang took the y ii a8r4c h(*a i·c~o w I o·r~dw o9 I5 y( *i i·)io Io f· jwthoe I tye ixi)t t'too bgeiv ea pslheonrtty- foofr fmoo odf'; 9th6,i sa i sv tahrei arenat soofn t wheh yw heell -gklnososwedn ;m~~~ %...~ .t..lfo~ 3.77:' ~ '1. ;r, -t-a'fJfir~ ~. ft 7. "# 8~.1..9. /7 ~,li-a;[email protected] 49 4 ~8 BULLETIN OF THE MUSEUM OF FAR EASTERN ANTIQUITIES JJ.- KARLGREN: GLOSSES ON THE BOOK OF DOCUMENTS 1215. L i siang j i y ii e sing c h' en 10. /34 of Ma Jung's versions (D) and translate: >>Causing a regulation of the works of the East». A. Cheng Hiian: s i n g -c h ' e n is a binome: >>To calculate and delineate the sun, _,As we shall see immediately below, however, the gloss pie n -24 = s hi 35 is hardly the moon and (the other) heavenly bodies (i.e. stars and constellations)>>. For this general ;~:~.cceptable. -:- All the rea<lings pie n (A, B), p ' in g (C) and p ' eng (D) make good meaning 'heavenly body' of the word c h 'en 11; cf. Kyii: Lu yii: >>The emperor K'u ~ense; in such a .c ase we should follow the earliest attested version: p i e n (Shang shu could 12 determine the (order =)courses of the 3 heavenly bodies (sc; sun, moon, stars)>>. ta chuan, Shi ki). ---., We compare: · Evidently Cheng thought that the phr. sing c h' en referred to the four constellations Shu: Tsiu kao: W u p i en n a i s i min mien y ii t s i u 39. A. PK'ung says: in the following calendar (n i a o, h u o, h ii, mao). - B. The same Cheng Hiian, pie n 24 = s hi 35 'to cause' (Kuangya has the same definition), and takes w u 40 in comm. on .Chouli: Ta tsung po, same phr. j i y ii e sing c h ' en 13 says sing as the ordinary imperative negation, s i m i n as 'those who govern the people' = offi means 14 'the five planets' and c h' en means 15 'the .conjunctions of sun and moon', cials: »Do not cause your officers to plunge into wine». For this one slender support has or, with Legge, »the zodiacal spaces>> in which the firmament is divided by the.places of been adduced. In the Shu sii (Preface) we find wang p e i J u n g p o t so 41 »The conjunctions. Thus: »To calculate and to delineate the sun, the moon, the planets and king caused the prince of Jung to compose>> etc. ap.d Shiwen says: Ma Jung's verskm for (regions of) conjunction». (Chavannes on the other hand translates: »Le soleil, la lune -p e i 42 'to cause' had p i e n 24. This, of course, cannot prove that p i e n was syno sing les constellations, ch'en les syzygies de conjonction»). For ch'en this is based aymous with p e i and meant 'to cause'. In Ma's version the clause means: >>The king on Tso: Chao 7, phr. 16 »(The regions of) the conjunctions of sun and moon, those are the (distinguished =) selected the prince of Jung to compose>>. - B. Ts'ai Ch'en takes ~ u c h ' e n». - Both meanings (A and B) of c h ' e n are thus attested in early texts. But 10 as an ordinary negation ( = 43), which is co~on in early texts, and punctuates after A suits the context better; since sing -c h ' en reasonably (with A) alludes to the fol s i: »If · yo~ do not properly arrange your officers, the people will steep itslllf in wine •· Shuowen lowing four constellations (this is evidently why Chavannes translates sing by 'con-- say~ p i en 24 = c h i 44 'to regulate, arrange', properly 'to distinguish, distribute, stellation', not 'star'). - arrange', as y.J:>pye. - C. Kiang Sheng: p i e n 24 = p i e n 45 'all round'. In several 1216. Tse Yii yi 17. passages in Yili (IJiang yin tsiu li, Yen li), where the current text has 24; Cheng Hiian The Ku-wen reads 18 (*d'iiJc I a'vk Its e). The Kin-wen version lap. the Stone classics says the Kin-wen version had 45 'all round', and later comm. all say that 24 *b'jan is loan of 175 A. D.) read 19 (*d'ag I d'uo It u), both meaning 'to dwell, to reside'. For 19- char. for 45 *pian. That is hardly necessary; there is simply an extension of meaning: in this sense see Gl. 794. For Yii-yi there are a great number of variants.· In the present .24 '*b'ian 'to distribute, dis~ributively, seriatim, all round'. So far Kia:q.g's theory is glosses I shall refrain entirely from discussing geographical names and their-localization .. admissible. But he explains,: >>If (in your admo:pishments) you do not go all round to For my reasons for this see BMFEA 18, p. 208. your officers, the· people will plunge into wine?>, which is very forced. - B, which takes 1217. Pien (p'ing1) chi tung tso 20. .the word p i e n in ~ts normal sense and which divides the line in a rhythmically satis A. In Gl. 1212 above we saw that, in a phr. there, .p ' in g 21 was simply a scribe's. factory Wl).y: w u ·P i e n n a i s i, m i n m i e n y ii t s i u, is surely best. error for p i e n 22. The present case is much more complicated, but some early sources. N i a o s h o u t s i w e i 46. evidently had p i e n 22 = 23 or 24 'to distinguish, distribute, arrange', for Cheng Hiian Si-ma Ts'ien renders the line 47 .. The tw~. t s i are interchangeable and mean 'to (comm. on Chouli: Feng siang shi) quotes pie n chi- 25, and Kia Kung-yen (ibid.) breed'. W e i 48 and 49 .were b~th *mjwar (though in different tones) and can serve as says Shang shu ta chuan had 26 and S i -m a Cheng says it had 25. Thus: •To arrange, loan char. for each other (a person was called 50, in Lun and 51 in- Ts'e). and regulate the ·works of the Eash. (For another interpr. of tung t so: »the [heavenly] A. , Shuowen as quoted by P'ei Yin says wei 48 ('tail') == 52 'to copulate', evidently actions of the east>> see Gl. 1226 A below). -B. Si-ma Ts'ien (Wu ti pen ki) evidently alluding to our Shu passage. The present Shuowen text says 48 = '49 'a small thing' likewise read pie n, for, as in the case Gl. 1212, he replaces the char. by pie n 2r fan appendix', pefinition by sound similarity, COJllii10n in Shuowen; h\)re certainly the (loan-char.). But he reads the next word differently: pie n c h' eng tung t so 28: author was influenced by hjs knowledge of Si-~:q_a's reading). But since 49 'a small thing' >>To arrange and set the tasks for the works of the East>> (Chavannes translates c h '.eng· is his definition of the qommon char. for 'tail', it is evident that Hii (or whoever form by 'promulguer', which is certainly not correct). Probably, however, the c h 'eng of Ulated this version of the Shuowen) t!)ok Si-m,a's 49 'tlie small thing' to be a variant ~or Si-ma's does not indicate a variant reading but is his gloss word for c h i 29: 'to regulate'" !1:8 'tail' and our line to mean: "Birds and be&~~ts breed and (•taih =) copulate». PK'ung taken in -the sense of settling the c h ' e n g 30 'measures, quota' of the daily work. - and Ts'ai Ch'en (with Legge) followed this. - B. Kiang Sheng insists that 48 'tail' is C. There were, however, early versions which actually had p ' in g 21, not· pie n 22' a loan char. for 49 'small', and thl)-t Si-ma meant: »Th~ birds and beasts breed and (are (cf. D below). Shuowen quotes 31 (32 is a variant for 29, same reading and meaning); small =) there are small ones>> (accepted by Chavannes). He refers to Kyii: Lu yii shang, »To equalize and regulate the works of the East>>. Both Chao K'i (2nd c. A. D.) 1n comm. where n i a o s h o u ·y ii n 53 >>birds and beasts breed>> (spring time) is placed in on Meng: Wan Chang and PK'ung read p ' i n g c h i 33, and this· has been accepted as opposition to n. ia o s li o.u c :h' eng 54 >>birds and beasts are groww> (summer time), the orthodox version. - D. Ma Jting (ap. Shiwen) says p ' eng 34 =·35. This shows tph 'aet ning th3e4 cish aar aschteorr theer wreaayl loyf hwardi,t inags 3p6h o(n*ept'eicn,g Ip p' 'ei nng g I p2 1'e ann dg) ,n ovta rp. 3i 7e, ni n 2O2d. e H25i7s : ~ ::i.ffi.: 11.~ 1-t.fr ~fo.-.1'3 t1 f-j ~ttl¥. li:fff;.>. a 'lil z ~ lt. a .B z -'t ~~ _)tt,,._ t:.lllo~~ = 'to cause'; the same word as 38 (same sound) 'to cause' in Shu: Lo k~¥> and Li cheng 16. ;t ./'i)If.j.o, 3f~ {~ 21 :f-li )I( .tllJt.t't1M-.2.>~~~7:{J!o.1~ ~ {~.29:~~1. (cf. Gl. 973). Thus: >>Causing a regulation of the works of the East>>. -E. It would appear tha.t there is one more possibility. PK'ung and Kuangya have a gloss pie n 24 = 35 'to· 31.-~ ~Jil~ .:lt._t~h7:f~CI~-:¥- ~{jt_~!f 1?-i?f-~:IS-ff-~:;>~#1'~ Of) ~j@ ~ ~\{7 ~~It; _L cause'. H that could be duly substantiated, we could abide by the p i e n 22 = 24 of the early versions but interpret this pie n as meaning 'to cause', synon. w. the p ' eng 1~ 3fti s {~ ~ ~ 4.1 lfj 9'1 1 ;;.fJ -¥'. -"~' *t --~ l-(~~?7.. .\ ~ i!'~ #~ 1-1~.w.f;Yf(1E:.. ~ 37. ~~~ 50 51 BULLETIN OF THE MUSEUM OF FAR EASTERN ANTIQUITIES B. KARLGREN: GLOSSES ON THE BOOK OF DOCUMENTS giving the same idea of spring as being the time when the animals have small ones. Since of astronomical observations, and since c hi j i 76 is a technical term attested in two there is no text par. of wei 'tail' meaning 'to copulate', it is tempting to accept B. ancient sources, B after all seems the most convincing. But we have no reason for believing that Si-ma really had a Shu text reading 49. His 1221. K ii e m in yin 77. 49 may be, as often, his own gloss word. In accepting B we should have to assume a This phr. is one of the four referring to the occupations of the people during the seasons. k i a t s i e loan char. 48 for 49 in the Shu text, whereas A takes the actual word of the Spring: the people s i 78 disperse (leave the houses and go out into the fields); summer: text: 48 'tail' with a perfectly natural extension of meaning. A therefore seems safest. the people yin 79, as above; autumn: the people y i 80 are at ease (common meaning e T s Nan k i a o, see Chavannes MH I, p. 45. _ Si-ma paraphrases it by y i 70); winter: the people y ii 81 keep in the warmth of 1219. Pie n c h 1 n an n go 55. their houses. A. PK'ung: yin 79 = yin t s i u 82 (after Shuowen 79 = 8.1), expound For variants and expl. of the first two words, see Gl. 1217 above. ing: »The old and weak yin t s i u (lean upon, associate themselves with =) join A. The line above is PK'ung's orthodox version. 56 was *ngwafnguaf n g o. This company with the strong men in the fields and help in the husbandry work». It is, of graph is interchangeable with 57, same sound. Si-ma Ts'ien's Shi ki in the version of course, impossible to read this into the short Shu phr., and Sun Sing-yen simplifies: >>The Chang Shou-tsie (foil. by the Sung Po na ed.) had this variant, reading 58 (Chang reads people (lean on each other =) ally themselves with each other>>, sc. in the field work. 57 *gwiafjwif1./ wei, but Ts'ie yiin, correctly identifying it with 56, reads *ngwafnguaf Cf. Tso: Chuang 19: »They made an insurrection and 84 (leaned on =) allied themselves f n go). Shang shu ta chuan and Cheng Hiian in gl. on Chouli: Feng siang shi both with the Su clam. This is better and would be quite good if the Shu phr. had run: k ii e likewise quote 59. This *ngwa (56 or 57), however, has three different meanings in the m i n s i a n g yin 85. Without the reciprocal s i an g the expl. is hardly admissible. ancient texts: a. synon. w. t so 60 'to act, to do, to move, to work', in this sefl:SC _ B. Ts'ai Ch'en: yin 79 = 86 >>they disperse and even more disperse>>, i. e. y i n sometimes also wr. 61, see Gl. 508, with Shi examples. b. 'false', see Gl. 306, with Shi: = 'to follow up, to go on, to continue', referring it to the s i 'disperse' of the preceding examples. c. 'to change' f= 62) e. g. in Ode 191. This has given rise to different interpr.: paragraph. Thus: »The people continue>> sc. to disperse. Y i n in the sense of 'to base a. PK'ung (foil. by Ts'ai Ch'en, Kiang Sheng and Legge) n go= 62 'to transform', oneself on something as a point of departure and follow it up, carry on' is common, e. g. thus: >>To arrange and regulate the transformations of the South». {Cf. also Cheng Hiian in Li: Li k'i 87 »The Hia created (the rites), the Yin (followed them up =) took them over Gl. 1226 A below). - {J. Chang Shou-tsie: 57 means 'to do, to act, to work', thus: •To arrang& and continued them, carried them om. It is, however, a curious idea that the »dispersing•> and regulate the works of the South•, forming an exact par. to the preceding tung t so 63 of spring time was not complete and had to be »followed up» by a still greater dispersion >>works of the East>>. - B. The Shi ki in the version of Si-ma Cheng (now foil. by most cur in summer. It would then be better to interpret in a more general way: >>The people rent editions) read 64, and Si-ma Cheng insists that this 67 should be read in its ordinary follow up (carry on, sc. the work of the year)>>. Even so, the interpr. is strained. - way *gwia I jwif1. I wei 'to do, to act' (Han shu: Wang Mang chuan writes 65-this 66 C. Kiang Sheng therefore reverts to Shuowen's definition yin 79 = t s i u 83, but goes *gwia 1j wif1./ wei 'to fake, to concoct' is etym. the same word as 67 'to make'). The on: Shu6wen defines t s i u 83 as = 88 'go to high places' (a typical script etymology meaning comes to the same as A fJ (curiously enough Sun Sing-yen quotes Huai: T'ien of Hii's, the graph 83 containing 89 'height'), thus: >>The people go to high places>>. A wen 68 »When the year is dry the grain is not [made =] achieved>>, i. e. does not ripen, frightful speculation. The w. yin of course has no such sense. -D. Yii Yiie points out and suggests that nan wei means >>the achievement [of the crops] of the South•>). that yin 79 is the original graph of 90 'a mat' (they are etymologically the same word: -Whether Si-ma Ts'ien wrote 57 (*ngwa) or 67 (*gwia) is immaterial; the meaning comes the fundamental sense of the stem * ·ien, even tone, is 'to lean on, rest on, rely on' and to the same thing. The Shu text obviously had the former 56 v.. 57 *ngwa, and quite *·jen 'mat' means 'that on which one rests'), and this yin 90 is defined in Shuowen as = as obviously in the sense of 'to do, to act', with A {J, as proved by the parallellism 91 'double mat in a carriage'. Hence, .Yii says, yin 79 = yin 90 has the sense of with' 63 (so also Chavannes). It should be added, however, that the expl. of 56 = 'to. 'double', and the Shu phr. means: >>The people have double covers>> (as protection against change, transform' in A a above is influenced by the corresponding winter paragraph the summer heat). Shuowen's definition, however, has very little value. Yin 90 means 69, which PK'ung thought meant >>the transformations>> (Cheng Hiian even believed that 'mat, mattress', certainly not 'double'. If yin 90 were a coarse and plain straw mat, tung t so 63 etc. referred to celestial phenomena, see Gl. 1226 below) and hence it would have been tempting to think of the custom that the farmers during the busiest conceived of as being a case of parallelism: n go 56 v.. y i 70. But, as we shall see in summer time did not go home but slept in grass huts in the fields, thus: >>The people Gl. 1226, y i 70 as well means simply >>work», like t so 60, n go 56, c h' eng 54 (straw-mat =) sleep on straw-mats>> (in the field huts). But yin 90, which is common (>>achievements>>), so that all through the parallelism is in favour of A fl. (Li, Yili, Hanfei etc.), always means a fine mat or even a mattress, which makes this 1220. King chi 71. interpr. unlikely. - E. Another interpr. Yin 79 as an independent word often means A. PK'ung connects this with the preceding: »and respectfully bring about>> sc. the 'to avail oneself of the opportunity, to take advantage of the suitable time, make the things pertaining to -summer. - B. Ts'ai Ch'en (foll. by the best Ts'ing authorities, best of the occasion'. Lii: Shun shuo 92 >>By availing himself of the opportunity (Kao Kiang Sheng, Sun Sing-yen and others): c h 1 72 means »the summer solstice>>, thus: •And pay careful attention to the (summer) solstice». This is based on Chouli: Fang siang ~;Z~n{j'. .>-Y.A'>-.YY:f-.#. ~~t,clt.J7~SY.{\t*-.1~-~\ili:i~ca{i. r:1 "i't..a1'~tJffi.{"li •<t shi 73 >>In winter and summer he brings to its utmost point the sun>>, i. e. he determines the solstices (as explained in detail by Cheng Hiian). Similarly ·in Tso: Huan 17 it is {JeA;.ii~ A,,.;.r.~1~a-ik,c7A cs.A* *--1--#.% 49 if'jj $i ra ~ 7,/ ~_£(:r.tifti'J. :1--{*.13 7/r)t{ said that the astronomer chi j 1 74, more correctly 75 (variant of 72, see Gl. 820). Chavannes, who accepts A, brushes this expl. aside as being far-fetched, but in fact it 1:3 7f~ 11i£tt3 71 ffi: {;(. 1~ ?8.-#7'1 \B 5t>. ~ rr1 ~ft. ~~~.1~ r1F8~~ JY@(tVf@ 16:1 &:.fiT is not. The difficulty is that there is no parallel phr. in the passages relating to the other seasons which could help us to decide. Since, however, it is here throughout a question il'ij X,;.fff-s1 l.~t.trtt t5: it~ t'l. ieto 1'8 r1 1- i ~ 1.2 18 g~ t?{ ~ ...r J.-AJ~7l ;t it 93. ~J~ Ji.Jf<. 52 53