馬王堆帛書《刑德》、《陰陽五行》諸篇曆法與神煞再考

Reexamining the Calendar System and Calendrical Spirits in Xingde and Yinyang Wuxing from Mawangdui

秦至漢初部分數術文獻的編纂使用曆法歲實取三百六十六日，以甲子年正月甲子朔立春為曆元，馬王堆帛書《陰陽五行》甲、乙篇與《刑德》乙篇的部分章節就是應用這種曆法的實例。在此曆法中，立春日的干支每年都在子、午之間循環，因此帛書〈太陰刑德大游圖〉中特定子、午干支後的「位春」當讀為「立春」。立春是太陰的遷徙日，藉此可以解決文獻中三處看似矛盾的紀年問題，確定從秦末至太初改曆這一段時間中不存在太歲超辰的現象。帛書中的「上朔」即年首之上，立春之前日，這樣得出的干支與傳世文獻所載完全相合，年內其他上朔仿照其原理，設定在三、五、七、九、十一月節的前一天，可以稱其為「節上朔」。由於編纂使用曆法與當時實際行用的顓頊曆每年有四分之三日的差距，所以基於此曆法形成的數術模型無法長期應用，數術家嘗試通過修改具體規則的方式繼續使用其中的部分神煞，例如《刑德》甲篇的編者將上朔的確定方法調整為實際立春之前最近的亥日或巳日，年內其他上朔直接重複其干支。如何在理想與現實之間取捨，是數術家必須面對的問題，馬王堆帛書中的諸篇數術文獻生動地再現了二者的糾葛過程。

Some of the occult texts compiled from the Qin to early Han dynasties adopt a calendar that has 366 days in a tropical year and begins on jiazi甲子(the first day) of the first month, also known as lichun立春(“Beginning of Spring”), in the year jiazi. Actual examples of this form of calendar can be found within certain sections in Yinyang wuxing陰陽五行(lit. yin and yang and the five phases or elements) A and B and Xingde刑德(lit. punishment-virtue) B from the Mawangdui archaeological site. In such a calendar, the lichun of every year rotates between zi子and wu午; therefore, in the manuscript“Diagram of the Great Travelling of the Greater Yin and Punishment and Virtue”太陰刑德大游圖, the phrase weichun位春that appears after certain zi and wu stem-branch terms should be read as lichun (namely, the solar term“Beginning of Spring”). Since lichun is the day when the Greater Yin migrates, we can use this to resolve three seemingly contradictory year notations in early texts and determine that there was no occurrence of Tai Sui overrunning its supposed branch from the late Qin to the Taichu calendar reform in the early Han. The phrase shangshuo上朔found within the manuscripts means“before the beginning of a year,”namely, the day before lichun, and interpreted as such, its stem and branch completely conform with those recorded in transmitted texts. Accordingly, the other shangshuo days within the year are set to be the days prior to the solar terms in the third, fifth, seventh, ninth, and eleventh months, and thereby should be regarded as the“solar term shangshuo.”Since the calendar used in these texts differed from the Zhuanxu calendar that was in actual use at the time by three quarters of a day, the occult model based on this calendar could not be applied for long periods of time. To continue using some of the calendrical spirits, occultists modified particular rules, exemplified by the editor of Xingde A changing the method for determining the first shangshuo by setting it to the day hai亥or si巳immediately before the actual lichun as well as then making the other shangshuo days in the year directly repeat its stem and branch. The problem that these occultists were thus faced with was how to choose between the ideal and real, and the Mawangdui occult texts bear witness to their struggle between the two.