of locusts could be caused by shortchanging an ancestor when sacrifices were made. Therefore, many sacrifices were intended to keep the ancestors from punishing the living or to make up for previous slights that caused the ancestors to punish the living. This custom survives today, as among the Akha minority of the far south: “a man who had a stroke made the connection that his stroke was caused by not offering the correctly colored chicken at his ancestral offerings” (Tooker 2012: 38). In Shang, this practice not only provided excuses for state-building rituals but also trapped Shang in a round of destruction. Humans and wealth goods were sacrificed at an appalling rate.
The vast majority of written records of Shang are questions to the gods and ancestors, carved on scapular bones and turtle shells. The bones and shells were cracked by heat, the cracks being read as answers to the questions. Often the answers were then carved on the bones.
These oracle bones can provide a whole ethnography of Shang (Flad 2008; Keightley 2000, 2006). They show, most obviously, consumption of sheep, pigs, turtles, deer, and so on. The inscriptions are somewhat less clear. There is continuing controversy as to what they are “really” about. If all that mattered was the forecast, it would have been easier to write it down with a brush, as indeed the scribes sometimes did (Keightley 2006); why go to incredible effort and expense to carve it? Evidently something about state power and authority is involved. Showing off expensive evidences of ritual divination may have been the goal. Presumably there was a validating religious idea that only carved oracles were truly effective. The Shang gods, like so many gods worldwide, probably demanded that the worshipers show seriousness by diligent hard work.
Writing seems to have been invented—or just possibly diffused as a concept from the Near East—around the beginning of Shang. By the end of Shang, writing was highly developed, with ancestral forms of modern characters well standardized and widely used. Early characters were largely pictographic, but eventually someone got the bright idea of using a pictograph to write various nonpicturable words that sound the same. The linguist David Prager Branner (2011: 107f–117) has described what followed. A picture of a person swimming in a stream was used to write “to swim” (now pronounced yong) and then was also used to write “eternally” (also yong—evidently they were already homonyms in Zhou times). Very soon, such duplications became terribly confusing, and people began to write small classifying particles, now called radicals, next to the phonetic symbol. These radicals indicated broad classes of meaning. Thus the word “to swim” added some dots that look like drips—the radical for “water”—while the original “swim” character, without a radical, became the word for “eternally.” This sort of marking reversal—with the original pictograph acquiring a radical while a derived word did not—was very common. It seemed logical; it is easier to mark “swimming” with “water” than to mark “eternally” with anything. Thus, similarly, the word for “ancestor,” a picture of an ancestral tablet (not—in spite of a rampantly viral folktale—a phallus), came to mean a particle indicating “good” (among other things). Both are pronounced zu, but “good” remains a simple picture of a tablet, whereas the word for “ancestor” has acquired a radical used to mark terms referring to gods and the supernatural. Then the word “butcher’s block,” which is pronounced the same way, was written with the same tablet plus a little picture of two bamboo plants, the radical for “bamboo” objects (Branner 2011). The original picture of a tablet has become a “phonetic”—a graphic device that merely marks the sound. Through such extension by meaning and sound, any word can be written, and foreign words can be transliterated.
The vast majority of modern Chinese characters consist of a radical and a phonetic. Usually the phonetic supplies nothing but the pronunciation and has nothing to do with the meaning. The cute stories that Chinese love to tell about the appropriateness of the phonetic are mostly fiction—delightful, but fiction. However, there are some important exceptions: cases in which characters are genuine ideographs. The commonest is the character for “good” (hao), which shows a woman (nu) with a son (zi—no sound correspondence there). Another important case is the extension of ren “human person” to mean “humaneness.” Originally they were the same word and character. Scribes added the character for “two” to the character for “person” when it meant “humaneness,” because it was the way two people should act toward each other. “Person” is both the “phonetic” and the “radical” in this case, “two” being inserted purely to differentiate meaning. A still different type of evolution concerns the familiar yin and yang. These terms originally referred to the shady and sunny sides of a hill (the north and south slopes). Both use the “hill” radical. Yin combines it with the word for “shadow”—used as phonetic but clearly with an eye to the meaning. Yang similarly uses an old character for “sun” as a meaningful phonetic.
Possibly the most interesting of all the Shang graphs is a cross, each arm tipped with a bar. This is the original form of the character wu, meaning a shaman or spirit medium. In Shang times it meant the divine powers associated with the directions (Keightley 2000: 73ff.). Divinization of the directions is a concept shared with Native Americans and many other cultures. There may also be some graphic relationship with the turtle, whose shell and four limbs may have been a world symbol (Keightley 2000: 93); the cosmic turtle is another concept known from India to South America. Wu, anciently pronounced something like *mag, may be cognate with “magus”—deriving from an ancient Indo-Iranian word for a magician, and, indeed, is the root of our English word (see Mallory and Mair 2000). However, its presence in Shang documents makes the link tentative, since there is not much evidence of contact at the time and since the root meaning was quite different.
Shang religion involved concern for weather and geographic features. The Li Ji, that thoughtful Han Dynasty political work masquerading as an ancient ceremonial text, expressed concisely some 3,000 years of Chinese reverence for nature: “Mountains, forests, streams, valleys, hillocks, mounds, can emit clouds and produce wind and rain, and [make one] see strange things, and are all called divine beings” (see Keightley 2000: 124, but I have retranslated). By Han, the old beliefs were subject to agnostic speculation, and seeing gods in the mists was not the literal Revelation it apparently was in Shang.
Keightley thinks the elite “may have been nature worshippers—or, more precisely, worshippers of certain Powers in nature—but they were unlikely to have been nature lovers” (2000: 116). I doubt this. Fear and awe of nature not only can accompany love for it; they can also even be the reason for love. Nature’s powers are truly awesome, in the old sense of the word, and demand respect and reverence. Humans have their own powers and can mutually love and respect other beings, a common experience in cultures everywhere that are dependent on direct interaction with nonhuman forces. Even the emperors of Shang were tiny and helpless in the face of the overpowering natural forces around them, and they knew it. They probably loved the smiling faces of heavenly and chthonic forces as much as they feared their wrath. China’s famous love for the natural environment was surely well initiated by this time.
One bit of evidence is the Shang kings’ love of hunting, abundantly attested in the oracle-bone records. Hunting and love of nature go together (paradoxical as this may sound) and certainly goes along with the love of animals. Those Many Dogs Officers prove that it was clearly as true of Shang as it was of Tudor England: “There is a saying among hunters that he cannot be a gentleman which loveth not hawking and hunting, which I have heard old woodmen well allow as an approved sentence among them. The like saying is that he cannot be a gentleman which loveth not a dog” (The Institucion of a Gentleman, Humfrey Braham, 1568, quoted in Almond 2003: 33).
Very few animal species are shown in Shang art; about half of them are mythical, mostly various types of dragons. Of real animals, sheep and water buffaloes are notably important. Many birds are shown; most are highly stylized and unidentifiable (Keightley 2000: 109–11). The pig, by far the most common animal in archaeological finds, was apparently too plebeian to show. Pigs are shown in both Neolithic and (rarely!) Zhou Dynasty art but not in Shang. An interesting sidelight, however, is the character jia (household), a picture of a pig under a roof. David Keightley (2000: 111) believes that, far from showing typical living conditions, this probably started as a sacrificial pig under a temple roof. This is another