Sincere thanks are due to readers and critics of this manuscript in its earlier incarnations, including Prof. Victor H. Mair, Prof. Paul R. Goldin, Prof. David W. Pankenier, Prof. David McCraw, Prof. John C. Didier, Gianni Wan, Daniel Sou, and also two anonymous readers at Early China. Errors, in particular cases due to the efforts of the above at least now less spectacular, are entirely my own.

1. OBI and occasional Bronze Inscription (BI) or Seal Script graphs employed in this paper are in many cases adapted, with his gracious permission, from the forms Sears, Richard has digitized from works including Xu jiaguwen bian 續甲骨文編, ed. Xiangheng, Jin 金祥恒 (Taipei: Yiwen, 1993)Google Scholar, Jinwen bian 金文編, ed. Geng, Rong 容庚 (Changsha: Shangwu, 1939)Google Scholar and Qiji's, Min 閔齊伋 Ming compilation of Seal forms, Liu shu tong 六書通, of which a useful modern redaction is Dingzheng Liu shu tong 訂正六 書通 (Shanghai: Shanghai Shudian, 1981)Google Scholar. Sears' collection is available at http://www.chineseetymology.org, accessed for purposes of this presentation over the course of February and March of 2009. The abbreviations of Oracle Bone sources provided in square brackets are as in Xu jiaguwen bian, with full titles given under “Oracle Bone Collections Cited” below. Above, only [酉], recorded as a Gu wen 古文 ‘Ancient Script’ form in the Shuowen jiezi 說文解字 and described there as “following 卯” 从 卯, is not an OBI glyph; the existence of renderings of this kind, of critical importance for an understanding of the Branches, was brought to my attention by Gianni Wan in a personal communication of April, 2009 following his review of a draft of this paper. This particular form and its analysis may be found at Yucai, Duan 段玉裁, Shuowen jiezi zhu 說文解字注 (Taipei: Yiwen, 1974), 754aGoogle Scholar, while such glyphs' relation to more typical (but not ubiquitous) OBI forms such as [酉; Yi 4867] is considered below.

2. David W. Pankenier, “‘Heavenly Pattern Reading’ 天文 and the Origins of Writing in China,” paper presented at the Columbia Early China Seminar on Writing and Literacy in Early China, February 7–8, 2009, 18–32. At the time of writing, Pankenier's study was in press as “Getting Straight with Heaven and the Origins of Writing in China,” in Writing and Literacy in Early China, ed. Li, Feng and Branner, David (Seattle: University of Washington Press, 2011)Google Scholar. Note that Didier, John C., in research spanning the past several years and culminating in the publication of In and Outside the Square: The Sky and the Power of Belief in Ancient China and the World, c. 4500 BC–AD 200 (3 vols.) as Sino-Platonic Papers No. 192 (09 2009)Google Scholar, also considers a potential stellar correlate of the glyph □ located rather at the northern celestial pole and thus anticipates Pankenier in a general sense.

3. Moruo, Guo 郭末若, “Shi zhi gan” 釋支干, in Guo Moruo quanji, Kaogu bian 郭 沫若全集, 考古編 Vol. 1 (Beijing: Kexue, 2002), 155–341Google Scholar; Pulleyblank, Edwin G., “The Ganzhi as Phonograms and their Application to the Calendar,” Early China 16 (1991), 39–80CrossRefGoogle Scholar. Guo's study dates to 1929, while an earlier version of the latter thesis appeared as Pulleyblank, “The Chinese Cyclical Signs as Phonograms,” Journal of the American Oriental Society 99.1 (01.–Mar. 1979), 24–38CrossRefGoogle Scholar.

4. Benedict, Paul K., “Austro-Thai Studies, 3: Thai and Chinese,” Behavior Science Notes 2.4 (1967), 288–91Google Scholar; Norman, Jerry, “A Note on the Origin of the Chinese Duodenary Cycle,” in Linguistics of the Sino-Tibetan Area: The State of the Art, Pacific Linguistics Series C No. 87, ed. Thurgood, Graham, Matisoff, James A. and Bradley, David (Canberra: Australian National University Printing Service, 1985), 85–89Google Scholar.

5. Pulleyblank, , “The Ganzhi as Phonograms,” 69–76Google Scholar.

6. The first portion of the study of Pankenier noted already above, devoted to discussion of the Stems and Branches, presents the general possibility that these emerged within an astro-calendrical reference frame; see, e.g., the conference manuscript of February 2009, “‘Heavenly Pattern Reading,’” 1–18.

7. See Langacker, Ronald W., Foundations of Cognitive Grammar, Volume One: Theoretical Prerequisites (Stanford, California: Stanford University Press, 1987), 183–89Google Scholar.

8. Some reassociation, if not this particular one, is compelled by the facts of early usage. Along with the concomitant and uncontroversial revision of si 巳 to zi 子, it is considered more carefully below.

9. Jiagu wenzi gulin 甲骨文字詁林, ed. Xingwu, Yu 于省吾 (Beijing: Zhonghua Shuju, 1996)Google Scholar; Gu wenzi gulin 古文字詁林, ed. Pu, Li 李圃 et. al. (Shanghai: Shanghai Jiaoyu, 1999–2004)Google Scholar; Jiaguwen zidian 甲骨文字典, ed. Zhongshu, Xu 徐中舒 (Chengdu: Sichuan Cishu, 1988)Google Scholar.

10. Sagart, Laurent, The Roots of Old Chinese (Amsterdam: John Benjamins, 1999)CrossRefGoogle Scholar; Schuessler, Axel, ABC Etymological Dictionary of Old Chinese (Honolulu: University of Hawai'i Press, 2006)Google Scholar.

11. The relevant sources are Schuessler's, ABC Etymological Dictionary as well as Minimal Old Chinese and Later Han Chinese (Honolulu: University of Hawai'i Press, 2009)Google Scholar, and also Baxter, William H., A Handbook of Old Chinese Phonology, Trends in Linguistics: Studies and Monographs 64 (Berlin: Mouton de Gruyter, 1992)CrossRefGoogle Scholar. Reconstructed OC pronunciations are marked with an asterisk.

12. See Shangfang, Zhengzhang 鄭張尚芳, Shanggu yinxi 上古音系 (Shanghai: Shanghai Jiaoyu, 2003), 171–87Google Scholar. The bifurcation of Late Middle Chinese Rime Table Division III versus Division I/II/IV rimes has been represented in reconstructions of earlier stages of the language in a variety of ways—the labels “Type A” and “Type B,” Pulleyblank's terms for hypothetical prosodic categories first proposed in his “Some New Hypotheses Concerning Word Families,” Journal of Chinese Linguistics 1 (1973), 111–25Google Scholar, have for later authors served as a convenient notational device. I have found Zhengzhang's approach, which follows S. A. Starostin's, in Rekonstrukcija Drevnekitajskoj Fonologičeskoj Sistemy (Moskow: Nauka, 1989)Google Scholar, notationally convenient and phonologi- cally both substantive and plausible, and so adopt it here.

13. This view appears in Norman, , “The Chinese Duodenary Cycle,” 85, 89Google Scholar, and Schuessler, , ABC Etymological Dictionary, 270Google Scholar. On this point, Jiaguwen zidian (1611) reports that “with reference to the Oracle Bones script, the two graphs 豕 and 亥 bear little formal relation to one another and in origin seem not to have been the same character” 自甲骨文觀之, 豕、亥二字字形頗不相類, 其初當非一字.

14. Xu's analysis may be found at Shuowen jiezi zhu, 759b.

15. A discussion of the morpheme *-n, apparently productive “only during some stage of Proto-Chinese,” may be found at Schuessler, , ABC Etymological Dictionary, 18Google Scholar.

16. Sagart, , The Roots of Old Chinese, 115Google Scholar.

17. The best derivational solution at present is perhaps *kəə[n] 荄根 ‘germ, sprout, root (n.)’ → *gəə 孩 ‘to sprout’ → *gəəʔ 亥 ‘sprouting stage’. The last may be derived from the previous by a suffix *-ʔ which appears repeatedly among the Branches terms, producing from verbs a semantics ‘that which [verbs]’ and thus ‘[verb]ing stage’. For a discussion of such agentive nominalization by *-ʔ in what are there and below termed “endoactive” nouns, see Schuessler, , ABC Etymological Dictionary, 48Google Scholar.

18. Moruo, Guo, “Shi zhi gan,” 217Google Scholar. Luo's opinion that the graph “is perhaps the same character as 戉” 與戉殆是一字 appears at Jiagu wenzi gulin, 2410.

19. Sagart, , The Roots of Old Chinese, 66Google Scholar. The frequent and apparently unprincipled OC variation *-i- ~ *-e- (as well as ~ *-ə- ~ *-əi-) “especially with dental finals,” the case here, is considered in Schuessler, , ABC Etymological Dictionary, 110–12Google Scholar.

20. Compare, for instance, Sagart's zèng < *s-təŋ-s 甑 ‘steamer’ from zhēng < *təŋ 烝 ‘to steam’; see The Roots of Old Chinese, 73.

21. Schuessler's comments regarding you 酉 and jiu 酒 may be found in ABC Etymological Dictionary, 58 and 321.

22. Pairs of this type are discussed at Schuessler, , ABC Etymological Dictionary, 97–98Google Scholar.

23. For *-ʔ in the Branches, see n. 17 above. Note that jiǔ < *tsju[ʔ] 酒 ‘brew’ and others might naturally enough also be seen as related to ‘flow’.

24. The graphs at second and third from left may be found at Dingzheng Liu shu tong, 228. The former, most striking in its similarity to BI 卯, is that text's rendering of the Shuo wen Gu wen form given above as , while the latter is labeled tong ding wen 銅鼎 文 ‘bronze tripod script’.

25. The proper interpretation of the early graph [易] would thus seem to be {the moon's changing patterns of light}. Past understandings have been much influenced by Xu Shen's dubious treatment of the glyph as depictive (xiangxing 象形) of a xiyi 蜥 易 ‘lizard’; see Shuowen jiezi zhu, 463b.

26. Schuessler, , ABC Etymological Dictionary, 573Google Scholar.

27. Yu's opinion that 午 is the benzi 本字 ‘original graph’ of 杵 ‘pestle’ appears at Jiaguwen zidian, 1596.

28. Sagart, , The Roots of Old Chinese, 129Google Scholar.

29. Naturally, he considers this sense distinct from the Branch term, for which is instead referenced (ABC Etymological Dictionary, 519) the suggestion of Norman (“The Chinese Duodenary Cycle,” 88) that wu 午 is an early Austroasiatic loan for ‘horse’.

30. The Jiaguwen zidian terms 巳 a shengxing 省形 ‘reduced form’ of 子 (1592) and also equates the latter two graphs provided above (1579).

31. Guo Moruo in “Shi zhi gan,” while recognizing the clear delineation in the OBI between these two forms and sensibly establishing and 子 at first and sixth positions (209), has his own ends in mind: to establish links to the zodiacal asterisms of the Western tradition, in the first case by way of more complex BI forms such as to [萬], writing a word wan ‘scorpion’, and thus Scorpio (276–80), in the second to Shuangzi 雙子, Gemini (262–66).

32. See Schuessler, , ABC Etymological Dictionary, 633Google Scholar.

33. See for example Jiaguwen zidian, 1590; Guo Moruo, “Shi zhi gan,” 200; or Jiagu wenzi gulin, 1125, where is recorded Wu Shaoxuan's 吳紹瑄 presentation of much the same thesis.

34. As noted at Jiaguwen zidian, 1590.

35. Jiaguwen zidian, 1590.

36. A variety of early clam hoes are pictured for instance in Yanshi Erlitou: 1959 nian–1978 nian kaogu fajue baogao 偃師二里頭: 1959 年~1978 年考古發掘報告, Zhong- guo tianye kaogu baogao ji, kaoguxue zhuankan 中國田野考古報告集, 考古學專刊 Vol. 4.59, ed. Yang Guanghui 楊光輝 (Beijing: Zhongguo Dabaike Quanshu, 1999). The possibility that such shells would be used for digging, hoeing or weeding may also be considered comparatively: note for example William Wood's description in his 1634 booklet “New England's Prospect” of Native American women of early colonial Massachusetts laboring in corn fields “keeping [their plots] so cleare with their Clamme shell-hooes.”

37. See Cook, Richard S., “The Etymology of Chinese 辰 Chén,” Linguistics of the Tibeto-Burman Area 18.2 (Berkeley: University of California at Berkeley, 1996), 39 and 148–49 n. 3Google Scholar.

38. This suggestion is that of Wu Qichang 吳其昌; see Jiagu wenzi gulin, 3439.

39. For the prefix, see Sagart, , The Roots of Old Chinese, 82Google Scholar.

40. Moruo, Guo, “Shi zhi gan,” 199Google Scholar.

41. Menzies' comments, under the name Ming Yishi 明義士, may be found at Gu wenzi gulin, v. 10, 1115.

42. The verb *linʔ 引 ‘bend (intr.)’ might be treated as an endoactive derivative of a root *lin ‘bend (tr.)’; compare the examples at Schuessler, , ABC Etymological Dictionary, 47Google Scholar. If Sagart's *N- in *N-lin 寅, posed to account for Tai borrowings, were seen as performing what Schuessler terms “endopassive voicing”—here, see ABC Etymological Dictionary, 48–50—the Branch term might be best glossed as ‘bent, stretched’.

43. Moruo, Guo, “Shi zhi gan,” 198Google Scholar; Yusen, Ye 葉玉森, Jiagu wenzi gulin, 3593Google Scholar.

44. Sagart, , The Roots of Old Chinese, 155–56Google Scholar, where reference is to Unger, Ulrich, “Finger,” Hao-Ku 46 (1995), 131–37Google Scholar.

45. The modern graph 甾 is typically considered to write the name of a type of ceramic vase (also zi, a reading to be seen as an early product of the graphical confusion?) This is an unrelated lexeme with a separate graphic history, being apparently from OBI [Lu 602], through BI [Hong gui 訇簋], to Shuo wen . Note that this last form remains clearly distinct from the same text's rendering of the glyph of interest here, , which I have for simplicity represented also with 甾 but which might be distinctly rendered, after Shuo wen, as (this still appearing within such modern variants as 葘, etc.)

46. Again, see Pankenier, , “‘Heavenly Pattern Reading,’” 18–32Google Scholar. Key to the author's argument are the lines Ding zhi fang zhong, zuo yu Chu Gong 定之方中, 作于楚宫 of the Shi jing ode “Ding zhi fang zhong” 定之方中 (Mao #50), translated by the author as “when [the asterism] Ding just culminated, [Duke Wen] started work on the Chu palace.”

47. The “Tianwen xun” 天文訓, Chapter 3, has “When the Qilin is (are?) in opposition, the sun and moon are eclipsed; when the Whale perishes, comets emerge” 麒麟鬬 而日月食, 鯨魚死而彗星岀, A Concordance to the Huainanzi, Institute for Chinese Studies Concordance (Hong Kong: Commercial Press, 1992), 3/19/10Google Scholar, though the passage will obviously demand a careful astronomical exegesis. The similar kungeng 鯤鯁 ‘[dark?] leviathan’ is mentioned in the “Dixing xun” 墬形訓, 4/38/3, 4.

48. The Mao commentary links the two terms in the manner of Greek Hesperus and Eosphorus: “At the sun's dawn emergence Venus is called qi ming; after the sun's descent Venus is called chang geng” 日旦出謂明星為啟明, 日既入謂明星為長庚; see Mao shi zheng yi 毛詩正義, Shisan jing zhushu 十三經注疏 Vol. 5 (Beijing: Peking University Press, 2000), 921aGoogle Scholar. However, a planet would clearly make an unsuitable object of the description of this stellar entity within the “Tian guan shu” 天官書 chapter of the Shi jil, where we read “chang geng is like a bolt of cloth encloaking the heavens; when this asterism [or star] appears, weapons are raised” 長庚如一匹布著天, 此星見兵起. Shi ji (Beijing: Zhonghua, 1985), 27.1336Google Scholar.

49. A number of early graphs show the tool in action: [妾] writes qiè < *tshap ‘captive woman’ and is graphically {woman cut with pointed tool, as awl}; [童; Mao Gong ding 毛公鼎] writes tóng < *dooŋ ‘slave boy’, graphically in part {eye pierced with awl}; [商] is in graphical terms {bivalve shell split in two with awl}; and the glyph seems to depict the implement snapping or, more probably, snapping something else, and may write the word qiān < *khrian 䇂愆騫 ‘to damage’. Note the word xin 辛 retains the sense ‘sharp’ in later stages of the language in particular reference to flavor.

50. Or even BI [Ge You Bi ding 鬲攸比鼎] or [Tang Shu pan 湯弔盤]. The interpretation of the word ren 壬 as ‘enfolding, concealing, etc.’ accounts for phonologically similar words written with co-phonophoric graphs such as rèn < *nəms 妊姙 ‘be pregnant’, rén < *nəm 任 ‘carry, hold’, rèn < *nəms 任 ‘burden < what is carried’, and rèn < *nəm[ʔs] 衽 ‘enfolding portion of a garment’.

51. As Wan has noted in a personal communication of April, 2009, an identical crown finds its way onto the head of the dragon's mythical counterpart, , writing feng 鳳 ‘phoenix’. This seems supportive of the assumption that the form 辛 should not be considered part of the dragon's body proper and that head and horns were seen to be lowered at bottom right.

52. Shi, Feng 馮時, “Zhongguo zaoqi xingxiangtu yanjiu” 中國早期星象圖研究, Ziran kexue shi yanjiu 自然科學史研究 9.2 (1990), 112Google Scholar, presents the black-lined form above along with two similar interpretations. Preferable to this is to see the Four Images as a separate and presumably earlier tradition that finds only idiosyncratic reflection within the Lodges names.

53. Recall here the regular mixture between the OC main vowels *-i- and *-e-; see n. 19 above.

54. The upper graph may be found at Dingzheng Liu shu tong, 52.

55. The upper graph may be found at Liu shu toong 六書統, Si ku quan shu 四庫全 書 Vol. 227 (Shanghai: Shanghai Guji, 1987), 627Google Scholar.

56. Presumably also related is guǐ < *kwriuʔ (?) 晷 ‘sundial’, the instrument though certainly a more recent innovation nonetheless named by its function in a manner directly analogous to the hypothetical old asterism (note *-ʔ ), for it is a measurer by the sun.

57. Personal communication, March, 2009. The region is also home to the three Chun 鶉 ‘Quail’ Jupiter Stations, as well as Niao Xing 鳥星 ‘Bird Star [or Asterism]’, traditionally identified with α Hydrae.

58. See Liu shu toong, 627.

59. Suggested Tian Gan asterism forms are described in broad gray strokes, with names and forms of the Western tradition in fine dark lines. The graphic was generated in part with Stellarium, available at www.stellarium.org. Reflected here is the speculative possibility that ding 丁 ‘fixing, truing’ might also bear some relation to a more concrete lexeme like the noun dǐng < *teeŋʔ 鼎 ‘pot, tripod’.

60. I am thankful to Prof. Pankenier for alerting me to the likely significance of this Jupiter Station designation in a personal communication of April, 2009.

61. Gu wenzi gulin v. 10 (943) presents Wu's attempt to link yi 乙 with zha 軋 based on graphic constituency and early glosses, as well as with a not entirely convincing mean- ing ‘cut’ or ‘slice’ by reference to a passage in the “Xiongnu zhuan” 匈奴傳 of the Shi ji 史記: “Wrongdoers in minor matters are cut (?), in serious matters are put to death” 有 罪小者軋, 大者死 (Shi ji 110.2892). The two items were OC *qrit and *qriit respectively.

62. For the phonological and graphical relation of bing 丙 to liang 兩, see Baxter, , A Handbook of Old Chinese Phonology, 272Google Scholar.

63. The useful gear-wheel metaphor is derived from Needham, Joseph, Science and Civilisation in China, Vol. 3, Mathematics and the Sciences of the Heavens and Earth (Cambridge: Cambridge University Press, 1959), 397Google Scholar.

64. For these associations, see for instance Needham, , Science and Civilisation in China, Vol. 3, 403, Table 34Google Scholar.

65. The correspondence could only be rough, of course, as the moon's 389.1° synodic cycle is an imperfect match for the 360° sidereal one.

66. Wilkinson, Endymion Porter, Chinese History: A Manual, Harvard-Yenching Institute Monograph Series Vol. 52 (Cambridge: Harvard University Press, 2000), 177Google Scholar.

67. As regards the former, a number of relevant articles by Jiujin, Chen 陳久金 are assembled in Chen Jiujin ji 陳久金集 (Haerbin: Heilongjiang Jiaoyu, 1993)Google Scholar, of most interest here being “Lun Xia xiao zheng shi shi yue taiyang li” 論《夏小正》是十月太陽歷, 3–30. For the tradition of ten suns and its connections to the Stems, see in particular Allan, Sarah, The Shape of the Turtle: Myth, Art and Cosmos in Early China (State University of New York Press: Albany, New York, 1991), 25–56Google Scholar.

68. Pankenier suggests that “the calendrical use of the cyclical signs is considerably more archaic [than the late Shang ancestral cult] and may have originated in a pre-Shang culture”—the remark as quoted appears on pp. 22–23 of a late pre-publication proof of “Getting Straight with Heaven.” Compare here Boltz, William G.,“Language and Writing,” in The Cambridge History of Ancient China: From the Origins of Civilization to 221 B.C., ed. Loewe, Michael and Shaughnessy, Edward L. (Cambridge: Cambridge University Press, 1999), 108Google Scholar, where the author suggests rather that “the Chinese writing system seems to have been invented not much earlier than 1200 B.C.”—though Boltz's consistent preference for seeing the script as “sui generis” rather than imported plainly resonates with the present proposal.