1. Hou-hsüan, Hu , “Pa-shih-wu nien lai chia-ku-wen ts'ai-liao chih tsai t'ung-chi” , Shih-hsüeh yüeh-k'an 1984.5, 22 Google Scholar. I am grateful to Sarah Allan, Hu Hou-hsüan, Jean A. Lefeuvre, Li Hsüeh-ch'in, Lin Yün, Stanley L. Mickel, Edward L. Shaughnessy, and Ken-ichi Takashima, for their helpful criticisms of an earlier draft of this review; its errors remain my own.

2. Ying-kuo so-ts'ang chia-ku chi; Oracle Bone Collections in Great Britain. Chung-kuo she-hui k'e-hsüeh-yüan li-shih yen-chiu-so and School of Oriental and African Studies, University of London. Li Hsüeh-Ch'in , Ch'i Wen-hsin , and Ai Lan (Sarah Allan). Vol. 1, pts. 1 and 2. Pp. 524,15. Peking: Chung-hua shu-chü, 1985. 350 RMB.

3. I have in mind (1) the annual meetings of the Chinese Paleography Association; see, e.g., the reports in Early Oiina 5 (1979-80), 113-118 (by Chou Hung-hsiang); 7 (1981-82), 114-118 (by Edward L. Shaughnessy); 8 (1982-83), 182-186 (by Chow Kwok Ching); (2) the international conferences held at An-yang in 1984 and 1987 (see, e.g., the report in Early China Neivs 1.1 (January 1988), 1, 3-4, 19-18 (by Shaughnessy). One should also signal Chinese participation in conferences held in America; notable in this regard was (3) the International Conference on Shang Civilization held in Hawaii in 1982, which led to two publications: Studies of Shang Archaeology, ed. K.C. Chang (New Haven: Yale University Press, 1986) and International Conference on Shang Civilization: Abstracts of the Papers Presented and a Summary of the Discussions (Early China Supplement 1, 1986), prepared by Lothar Von Falkenhausen.

4. See, e.g., the contributions by David S. Nivison (on the date of the Chou conquest of Shang) in Ku-wen-tzu yen-chiu 12 (1985), by Keightley (on the origins of orthodoxy historiography) and Shaughnessy (on Shang-Chou relations and the post-Wu Ting influence of the Shang royal house) in ibid. 13 (1986).

5. E.g., the “Early China Forum” articles on the Chou-yuan oracle bones by Edward L. Shaughnessy, Wang Yuxin, Li Xueqin, and Fan Yuzhou in Early China 11-12 (1985-1987), 146-194.

6. It is reassuring to have rubbings of the pieces in the Royal Scottish Museum included in Ying-ts'ang. When Paul Demieville visited the Museum in 1972, nobody could find or tell him what had become of the collection ( Lefeuvre, J.A., “Les inscriptions des Shang sur carapaces de tortue et sur os,” T'oung Pao 41 [1975], 50, n. 1)Google Scholar.

7. That the drawings of the Chin-chang inscriptions were originally reproduced in slightly reduced size made it rather difficult for scholars attempting to join them to pieces in other collections.

8. For the abbreviations by which I refer to oracle-bone collections, see Keightley, , Sources of Shang History: The Oracle-Bone Inscriptions of Bronze Age China (Berkeley: University of California Press, 1978), 229-231Google Scholar.

9. Tsung-i, Jao , Ou Mei Ya so chien chia-ku lu-ts'un (n.p., 1970), nos. 28–35Google Scholar.

10. A brief account of the history and number of bones in each collection is provided in the editors’ “Introduction” (English version), pp. 5-8; the editors will presumably identify the source of each individual piece in volume 2. More useful at present is Wen-hsin, Ch'i, “Kuan-yü Ying-ts'ang chia-ku cheng-li chung ti chi-ke wen-f i” , Shih-hsüeh yüeh-k'an 1986.3,12-22Google Scholar, which gives the statistics for the pieces held and numbers of rubbings made, as well as some account of the detective work involved in tracking down unpublished inscriptions. I have notyet seen Li Hsüeh-ch'in, “Chi Ying-kuo shou-ts'ang ti Yin-hsü chia-ku” , to appear in Chia-ku-wen yü Yin-Shang shih , ed. Hu Hou-hsuan, 3rd series.

11. Hu Hou-hsüan provides an account of the way in which drawings, rubbings, and photographs have been used in various combinations in different oracle-bone collections. He expresses the hope that photographs of the British inscriptions can also eventually be published so that scholars may compare all three forms of reproduction for any particular bone (Ying-ts'ang, “Preface,” 3-4). I gather, in fact, that the photographs will be published as Ying-kuo so-ts'angchia-ku wen-tzu ts'un-chen (conversation with Hu Hou-hsüan, 7 October 1988).

12. I obtain these figures from Ch'i Wen-hsin, “Kuan-yü Ying-ts'ang,” 14.

13. Kuo Mo-jo , chief ed.; Hu Hou-hsüan, ed.-in-chief, Chia-ku-wen ho-chi (N.p.: Chung-hua shu-chü, 1978-1982) 13 vols. For my review of this fundamental new reference, see Keightley, , “Sources of Shang History: Two Major Oracle-Bone Collections Published in the People's Republic of China,” Journal of the American Oriental Society 110.1 (1990)CrossRefGoogle Scholar, in press.

14. For an introduction to Tung's periodization, see Keightley, Sources of Shang History, 92-94. Hsüeh-ch'in, Li, “Kuan-yü Ying-kuo so ts'ang chia-ku ti chi-chien shih” , Shu-p'in 1987.2, 15-16Google Scholar, explains that he employed Tung's old system because it has endured for fifty years and is in general use; he did not think it proper to impose his own system, still in dispute, involving two major traditions and nine groups, in a resource intended for the scholarly community as a whole.

15. The Ying-ts'ang categories, for example, make no reference to Ho-chi's “Slaves and Commoners” or “Slave-owner Lords.”

16. On the way in which the range of Shang divinatoiy concerns grew more restricted in the later periods, see Keightley, Sources of Shang History, 122; see also Keightley, , “Shang Divination and Metaphysics,” Philosophy East and West 38.4 (October 1988), 378-383CrossRefGoogle Scholar.

17. For an introduction to the disputes, see Shaughnessy, Edward L., “Recent Approaches to Oracle-Bone Periodization: A Review,” Early China 8 (1982-83), 1-13CrossRefGoogle Scholar; see too the discussions of periodization in Keightley, “Sources of Shang History.”

18. The character for ch'i in the phrase ch'i jih , “on the seventh day,” is virtually invisible in the Ying-ts'ang rubbing; Chalfanfs original drawing, here as in other cases, helps us read it.

19. These records are important for they imply that systematic sacrifices, offered on the kan -day of the recipient's temple name, were being offered in the reign of Wu Ting, before the regular divinatory evidence for their existence appears in the reign of Tsu Chia . For a pioneering discussion of the evidence, see Li Hsüeh-ch'in, “Lun Pin-tsu chia-ku ti chi-chung chi-shih k'e-tz'u” , paper presented at the International Conference on China's Yin-Shang Culture, Anyang, 9-16 September 1987; among the inscriptions he cites in that paper— a version of which will be published in Ying-ts'ang, vol. 2 (see n. 37 below)—are Ho-chi 1262,1263,1770,10410, and 12333.

20. Correcting the “2525” of the English “Introduction,” p. 10, line 16.

21. The editors seem to disagree about the authenticity of this piece. Ch'i Wen-hsin (“Kuan-yü Ying-ts'ang,” p. 19), as part of her larger discussion of fakes in the collection, argues, primarily on the basis of the derivative or unique graph forms and the careless and weak writing style, that the genealogy on Ying-ts'ang 2674 is a fake inscription added to a used oracle-bone. Li Hsüeh-ch'in, by contrast, believes it is genuine; see his talk, “A Discussion of a Shang Dynasty Genealogy Inscribed on a Scapula Formerly in the Chalfant Collection” (School of Oriental and African Studies, 2 February 1982), as reported in Early China 7 (1981-82), 119; see too his “Kuan-yü Ying-kuo so ts'ang,” pp. 16-17. Allan's “On the Engraving of Oracle Bone Inscriptions,” which will appear in Ying-ts'ang, Vol. 2 (see n. 37 below), concludes, on the basis of a microscopic study of groove depth, cleanness of groove edge, stroke order, and fissures in grooves, that the inscription is genuine; she tentatively dates it to the reign of Wu Ting. I find her contribution, which she has been kind enough to let me see in manuscript, impressive. My own examination of the bone on 28 September 1978 led me—though with no firm conviction—to a position close to that of Ch'i Wen-hsin; I noted, in particular, that the graphs lapse into a style that I associate with a number of evident fakes held in Paris collections (see Keightley, Sources of Shang History, 144, n. 40). In the light of Allan's more rigorously defined criteria, I would wish to reexamine the scapulas in both London and Paris.

22. “Kuan-yü Ying-ts'ang,” 14-15.

23. “Kuan-yü Ying-ts'ang,” 15-17, figs. 1-4.

24. Che-mao, Ts'ai, “Tu Ying-kuo so-ts'ang chia-ku-chi shang-pien” Ta-lu tsa-chih 74.5 (1987), 201 Google Scholar. He notes that two of Ch'i Wen-hsin's four rejoinings had already been made: one (reproduced as Ying-ts'ang 2119) by Lu, Hsia , “Hsüeh-hsi ku-wen-tzu sui-chi erh-tse” , Ku-wen-tzu yen-chiu 6 (1981), 178-179Google Scholar; the other by Po Yü-cheng , “Chia-ku chui-ho lu hsiao” , Chung-kuo wen-tzu 3 (1981), 210, 228. To be fair to Ch'i Wen-hsin, however, Po had only provided an exploded rejoining of K'u-fang 1844 and Chin-chang 739; Ch'i, in her article, supplies the missing fragment, K'u-fang 1556. (Chin-chang 739 appears by itself, however, as Ying-ts'ang 1813; the Ying-ts'ang editors have not reproduced this rejoining in the plates.) Ts'ai (preferring the placement originally proposed by Tseng Yi-kung ) also queries the way in which Ch'i has rejoined Chin-chang 535 and 477 to form Ying-ts'ang 593 (he illustrates the two possible rejoinings with his figs. 5 and 6 on p. 202 of “Tu Ying-kuo”).

25. One of these, involving Ying-ts'ang 2259 (K'u-fang 1768 = Ho-chi 41312) + Ying- ts'ang 2261 (Chin-chang 363 = Ho-chi 41312) is particularly important, because it provides additional evidence for the view that the ancestral title chung-tsung could refer only to Tsu Yi and not to Tsu Ting (“Tu Ying-kuo,” 201, 204, fig. 4). Ts'ai's discussion of this term should be read in conjunction with the Shih-wen for no. 2281 in Hsiao-t'un nan-ti chia-ku , vol. 2, sect. 1 (Shanghai: Chung-hua shu-chü, 1983), pp. 996-997, (This collection, containing 4,589 rubbings, mainly of oracle bones excavated in 1973, is the second collection reviewed in Keightley, “Sources of Shang History.” Ts'ai also notes that the accuracy and true scale of the Ying-ts'ang rubbings permitted him to retract a rejoining he had earlier proposed (“Tu Ying-kuo,” 200).

26. See Ch'en Meng-chia, Yin-lisü pu-tz'u tsung-shu (Peking: K'e- hsüeh ch'u-pan she, 1956), 652-653. Ch'i Wen-hsin (“Kuan-yü Ying-ts'ang,” 18-20) adds three more cases to Ch'en's list of bones that were total fakes, for example, and finds that seventeen of his total fakes are in fact only partial. She found far fewer fakes in Chin-chang: of the seven pieces marked “spurious” in the original book, she found four to be only partly so. Of four others she had originally suspected of being fake, three turned out to be partly so and one totally so.

27. Chalfant, e.g., copied only the fake characters on K'u-fang 1531 (= Ying-ts'ang 2568) and omitted the nine genuine ones, which are important for the study of the Period II ritual cycle.

28. “Kuan-yü Ying-ts'ang,” 14-15, 20-21.

29. Since the Ho-chi editors were not able to reprint the Ying-ts'ang rubbings in their basic corpus (referred to in n. 13 above), they had to be content with reproducing Chalfant's drawings, which, together with drawings from other collections, they placed in volume 13. It is necessary, accordingly to be aware that Chalfanf s errors are preserved in Ho-chi. In this case, for example, the three erroneous dots are reproduced in Ho-chi 39621.

30. E.g., Tung Tso-pin, Yin-li p'u (Nan-ch'i, Szechwan: Academia Sinica, 1945), pt. 2, ch. 9, 31b, no. 144; Chang Tsung-tung, Der Kult der Shang-Dyrmstie im Spiegel der Orakelinschriften: Eine paläographische Studie zur Religion im archaischen China (Wiesbaden: Harrassowitz, 1970), 104, no. 6.26; Paul L-M Serruys duplicates the mistake in his review of Chang's book, “The Language of the Shang Oracle Inscriptions,” T'oung Pao 60.1-3 (1974): 70. Ho-chi 39861 reproduces the mistake.

31. Ch'i Wen-hsin, “Kuan-yü Ying-ts'ang,” 21, also refers to this case of misdrawing.

32. Ho-chi 41704 fails to correct the erroneous “3” or to indicate the rejoining.

33. See Ch'ü Wan-li's commentary to Chia-pien 297; the rejoining is reproduced as Chia-t'u 016. Ch'ü gives the correct reading of “second moon” in his commentary and drawing. So does Tung Tso-pin, Yin-li p'u, pt. 2, ch. 2, “Ssu-p'u 2,” 6b.

34. Chalfanfs errors are reproduced by Ho-chi 40867. For an accurate transcription, and for the rejoinings, see Wan-li, Ch'ü, “Pa Li Yen-chai hsien-sheng chui-ho ti liang pan ‘yung hou fun’ niu-ku pu-tz'u” , Ta-lu tsa-chih 31.3 (1965), 2 Google Scholar.

35. “Kuan-yü Ying-ts'ang,” 21-22. Some clarification may be needed in some of these cases. Ch'i, for example, claims that nineteen characters are missing from Ying- ts'ang 1015, originally Chin-chang 695; these are certainly the same bone, but both the drawing and the rubbing contain the same eight characters (plus some fractional characters). Nothing appears to have been lost and Chin-chang 695, at least as it appears in the published drawing, never hád nineteen characters to lose. The matter will presumably be clarified when vol. 2 of Ying-ts'ang (see below) is published.

36. The editors conclude, “Although examples … in which the drawings are inaccurate or insufficient are countless, [Chalfant's drawings] are generally accurate representations of the original bones. In many cases in which the drawings have been suspected of inaccuracies, an examination proves that they are correct after all” (“Introduction,” 10-11).

37. The authors and their essays are (using the p'in-yin romanization they will employ): Li Xueqin, “On Some Bin-Diviner Group Record Inscriptions Engraved on Ox Scapulae”; Qi Wenxin, “The Case of Yi Yin and Huang Yin Being Two Persons”; Sarah Allan, “On the Engraving of Oracle-Bone Inscriptions” (see n. 21 above); and E.N. Arnold, “Identification of an Oracle Bone in Cambridge University Library.”

38. Twelve finding lists will permit scholars to move between the constituent collections, Ho-chi, and Ying-ts'ang.

39. I am grateful to Sarah Allan for supplying this information (letter of 1 November 1988); see too, Li Hsüeh-ch'in, “Kuan-yü Ying-kuo,” 17.