Bone of contention

Has the record of writing in ancient China been skewed because the more durable bone and bronze used for communication with the dead survived, while the bamboo and wood that would have been used for everyday writing perished? William Boltz examines the evidence

The earliest Chinese writing known is no earlier than the late Shang period and takes the form of "oracle bone inscriptions". These are found both on bones and on turtle plastrons - the underside of the turtle's shell. Like all writing systems, the Chinese script is a graphic means for representing a language, in this case the language of the Shang state in the last two centuries of the second millennium BC. With very few exceptions, each character of the script stands for a word in the language, and carries meaning only by virtue of that representation. No character stands for a meaning independently of representing a word, nor does any character stand in the abstract, divorced from the living tissue of the Chinese language, representing an idea directly. The script on the Shang oracle bone inscriptions is the direct precursor of the modern Chinese script.

The oracle bone inscriptions contain the record of ritual divination procedures carried out by the Shang royal house. Those rituals were direct appeals to communication with the ancestors (the dead) through the medium of the bone itself. The diviners applied fire to hollowed out spots on the bone, causing cracks. These were taken as meaningful signs from the ancestors in response to the substance of the divination topic. It was the king's responsibility to interpret the meaning of the cracks. The inscriptions do not themselves represent an effort to communicate with the ancestors; that was the purpose of the pyromantic cracking of the bone or shell in the first place. For the purposes of Shang divination as we understand it at present, the ancestors need not have been literate.

With the Zhou conquest of the Shang, in about 1045BC, written records take a different form. There are lengthy inscriptions cast in bronze vessels. The Shang inscribed bronze vessels, but those inscriptions were typically short, often only a few characters designating the name of the clan who had the bronze made. Almost immediately after the Zhou conquest of the Shang we find long, historically detailed inscriptions on bronze vessels, commemorating military victories or events of comparable importance to some member of the ruling class.

A bronze vessel cast with such a commemorative inscription was used ceremonially in a way that would, on the one hand announce an important event to the ancestors and on the other preserve its glory for generations yet to come. The way in which the ancestors' attention was drawn to the substance of the inscription was probably through the vessel's use in ancestral ritual sacrifices. Inscriptions on bronze vessels typically occur on the inside surface of the vessel, in a position where they are least likely to be seen when the vessel is not in use, and completely unseen when it is filled with food, as we suppose it was when used ritually. The fact that inscriptions tend to be placed on the inside surfaces suggests that the message was intended primarily for the ancestors; whether the mortal participants in the sacrifices could read it was of less moment. Writing for the Zhou, unlike for the Shang, was a means of communicating with the ancestors - who must have been presumed to be literate.

Given the specialised nature of both bone and bronze inscriptions, coupled with the fact that there are no other extant written texts from the Shang and early Zhou, it is often asked whether the writing of this period was more mundane and quotidian than the archival and ritual purposes with which we are familiar. A popular theory is that the extant corpus of Shang and early Western Zhou writing is an accident of the archaeological record. Thanks to the durability of the materials on which the oracle bone and bronze inscriptions were written, these texts have survived where other more everyday writing on perishable material such as bamboo and wood has not. The archaeological record is said to be skewed in a way that deprives us of an accurate picture of the true variety of writing in ancient China.

This view seems to be bolstered by the fact that the character normally used in later stages of the language to write the word ce, meaning "bamboo or wood strips used for writing", is found to occur in Shang inscriptions. It follows that if the Chinese in the Shang period had a word for it, viz, bamboo or wood strips used for writing, they must have had the thing itself and these have not survived because of the perishable nature of the wood/bamboo.

The weak point in this argument lies in its assumption that the character in question stands for the same word ce in the Shang language that we know it to stand for in the later language. This unsafe assumption rests on the interpretation of the Shang form of the character as pictographically representative of several bamboo strips laced together with a thong of some kind, a physical form near to that which we know "books" to have had centuries later. As with all pictographic analyses of characters, it is an impressionistic argument. According to most oracle bone scholars, the character stands either for a word meaning "to announce something (ritually to the ancestors)" or for a word meaning "to sacrifice by butchering (in some ceremonial way)". If we did not already know what form bamboo strip "books" took in the Eastern Zhou and Warring States eras some centuries later, I think it unlikely we would see with any confidence a depiction of that physical object in the character as it appears in Shang texts.

When we assume that there were documents in the Shang and early Western Zhou that have been lost due to the perishability of the strips on which they were written, the implication is that such lost writings would have included all kinds of texts, not just the specialised kind we find in the bones and bronzes of this period. To speculate in this way is to speculate on the extent of literacy in these periods. The question can be asked in two ways. What was the extent of writing in the Shang and early Zhou apart from the oracle bone and bronze inscriptions? Or what segments of the population were literate and in what ways did they use their literacy.

Assessing the extent of literacy in Shang and early Zhou China is not easy. Is literacy the ability to read or to write? And does a person who can only write his name qualify as literate? What is meant by the ability to read is straightforward; but the ability to write may mean, for the Shang engraver, simply commanding the skill to execute copies of characters on to the hard surfaces of shells and bones. There is evidence that some Shang inscriptions were prepared on what has been called a "mass-production basis" whereby the characters were not carved one at a time as we would expect in normal writing, but rather all of the horizontal and slanting strokes for all of the characters in the inscription were carved at once, and then the bone or shell was rotated through 90 degrees and the rest of the strokes were carved. While this does not prove that the engravers were illiterate, it does suggest a mechanistic quality to the engraver's task that gainsays the image of the Shang court scribe as a "man of letters".

It is often argued that the evidence of brush-written characters, found to occur on a few scattered pottery fragments and on some oracle bones, shows that the technique of writing with a brush was known as early as the late Shang period. Because brush writing would have been the natural form of writing we would expect to find on bamboo or wood strips, this evidence is coupled with the evidence for bamboo and wood writing strips and the whole argument is expanded to suggest that brush written bamboo or wood strips were common. The indefensible assumption here is that the evidence of brush writing on a few pottery pieces and on some oracle bones can be extrapolated to mean that brush writing was also used on putative bamboo or wood writing strips for any number of unidentified purposes. Brush writing existed but it may have had as limited a role in the overall scope of Shang and early Zhou literacy as bone inscriptions had. It may have served no function other than as a way to prepare a "draft copy" of the text to be incised on the bone or shell, or cast in the bronze vessel.

It has often been proposed that the extent of literacy in a society is directly affected by the nature of that society's writing system. For instance it is suggested that it was the Greek invention of the alphabet that led to widespread literacy in the Aegean and ultimately for the "success" of European civilisation throughout the Mediterranean. Conversely, by the same proposition, those ancient societies outside the European fold that did not enjoy the benefits of an alphabetic writing system were doomed to harbour large parts of their population forever illiterate. This is a thesis that holds more appeal for the Hellenist than for the Sinologist, but in either case its validity is not borne out by any real evidence. A recent consideration of literacy among the pre-Columbian Maya concludes that it is not script that promotes or thwarts literacy, but the society itself. The Athenians enjoyed an effective democracy, the success of which depended on the participation of an informed populus, which depended in its turn on literacy. The Spartans, whose political system was undemocratic and which did not entail the active participation of the populus, appear to have been far less literate.

The converse of this proposition holds with equal force. Where there are no social conditions that would either depend upon or be enhanced by writing, there is no reason to favour the assumption of widespread literacy. This applies to ancient China, While ancient Chinese society in no overt way prevented literacy, there were no social conditions that would seem to have promoted or required the spread of literacy beyond the restricted classes and specialised uses already described. It is, in other words, not "obvious" that because writing was used extensively in the Shang royal court, for very specialised purposes, and mutatis mutandis, in the early Western Zhou, it was also widespread among the populus.