The twenty-two are among the oldest characters still in daily use anywhere on
earth, and the way to feel that is to notice what they were doing before they had
names. The
Shang kings at Anyang, around 1200 BC, cracked ox shoulder blades and turtle shells
to ask their ancestors questions, and every question was dated. The dates are
stem-branch pairs. In the translations that Columbia's Sources of Chinese
Tradition prints, a diviner named Gu makes his cracks "on yiwei (day 32)," a
diviner named Zheng "on renzi (day 49)" and again "on guichou (day 50)," and the
questions they ask are whether "Father Yi" or "Grandfather Ding" is the ancestor
harming the king. The sixty-day cycle is not being invented on those bones. It is
being used as casually as a Tuesday, and Zhang He, in a 2020 comparative study of
the Chinese and Mesoamerican cycles, notes that a complete run of all sixty pairs
appears "neatly aligned" in the inscriptions by 1200 BC.
Look at those ancestors' names again. Father Yi. Grandfather Ding. The Shang royal
house named its dead by stems: Pan Geng, who moved the capital to Anyang; Wu Ding,
the king whose reign the oracle bones begin in; Di Xin, the last of the line. Zhang
He's list runs to seven kings and the observation that "only the names of the
heavenly stems were chosen to bestow upon people." Not one Shang king is named for
a branch. The ten were the half that belonged to people, and the reason is
probably the plainest one: ten stems made a ten-day week. The Shuowen
entry for that week is three characters long, "ten days make a xun," and one of
the standard divinations on the bones, cracked on the last day of one xun about
the next, is three characters that David Keightley renders "In the next ten days
there will be no disasters." Long before anyone put a stem on a branch, the stems
were the days you lived through.
The names came a thousand years later. "The system was not clearly named until
about 200 BCE," Zhang He writes, when people began calling the ten gan and
the twelve zhi. The gloss that stuck is Cai Yong's, from his
second-century handbook Duduan, which the Ming manual
Sanming Tonghui quotes at the head of its chapter on origins: gan is
a trunk, there are ten of them and they are also called the ten mothers, today's
jia through gui; zhi is a branch, there are twelve of them and they are
also called the twelve children. One tree, a trunk
and its limbs; and behind it Sima Qian's older and stranger image of mothers and
children, which the Ming author keeps, because the ten do generate the sixty and
the twelve are what they generate it on.
Where the twelve came from in the first place nobody knows, and it is worth saying
so plainly rather than picking a pretty story. Guo Moruo, the great
twentieth-century reader of the oracle bones, suggested that the ten stems were
simply ten fingers and that the twelve branches had been borrowed from the
Babylonian zodiac; Zhang He points out that the Babylonian constellations were
not formalized at twelve stations until 700 BC, five centuries after the branches
appear complete on the bones, and files the idea under speculation. So will we.