The Writing Revolution

by Amalia E Gnanadesikan

  Thus Champollion was able to show that the hieroglyphic script mixed logograms and consonantal phonograms, and that foreign names were not unique in using phonological spellings. He applied the sign values he had deduced so far to words that were not royal names, corroborating his results with the Coptic forms of words. And so, one sign at a time, the words of the ancient Egyptians were revived and the modern study of Egyptology could begin.

  Champollion died at only 41 in 1832, while his work was still opposed by many scholars. Others built on his work: Richard Lepsius found that phonograms could contain not just one but also two or three consonants, while the prolific Edward Hincks (who was simultaneously contributing to the decipherment of cuneiform) demonstrated that the script contained no vowels.

  Today Egyptian can be read as easily as the classical languages, despite the fact that its vowel-less character means that no one knows how to pronounce it. The ancient scribes, enjoined to preserve forever the words offered to their pharaohs, succeeded admirably. Fossilized in writing, their language long outlasted the civilization that used it.

  4

  Chinese: A Love of Paperwork

  The fields of the Chinese village of Xiao Tun in the district of Anyang yielded the occasional bonus crop. After a rainfall, or after being plowed, the earth would occasionally heave up an ancient bone. These “dragon bones” were sold to the local apothecary, who dispensed them in small pieces to be ground up into traditional Chinese medicines. Occasionally a bone would turn up inscribed with symbols that looked like writing. The apothecary carefully scraped off these scratchings before selling the pieces; dragon bones, no matter how magical, would look inauthentic with writing all over them.

  So runs the modern legend surrounding the oldest form of Chinese writing yet found. In 1899 fragments of inscribed bone came to the attention of the antiquarian Wang Yi-Jung, who realized to his astonishment that he was looking at the long-lost ancestor of the Chinese writing system. The writing dated from the latter part of the Shang dynasty (China’s second dynasty) in the period between 1200 and 1040 BC, when the Shang had their capital at Anyang (see map in the appendix, figure A.4).

  Though the art of writing was well established in Egypt and Mesopotamia long before the Shang period, Chinese writing seems to have been an entirely indigenous development. Just how and when the Chinese developed the technology is not clear, however. The climate of China is significantly different from that of Mesopotamia or Egypt. In China’s wetter climate, perishable materials duly perish, leaving much less for the archaeologist’s spade. The Shang bones thus remain the oldest surviving form of Chinese writing. Some scholars have interpreted marks incised on pottery during the Neolithic period, some as much as six thousand years old, as the forerunners of Chinese writing. But while the marks do appear to have been meaningful symbols, and while the use of symbols is an important prerequisite for writing, there is no firm evidence that the pottery symbols evolved into a true script, either the Shang script or any other.

  The “dragon” bones – actually the shoulder blades of cattle and the breast-plates of turtle shells – are the remains of the Shang system of divination. To inquire of the spirits (ancestors and deities), the diviner would hollow out a small oval in the back of a polished turtle shell or ox scapula, leaving the bone very thin inside the oval. The diviner would then apply heat to a point at the edge of the oval, and the bone would crack. The resulting T-shaped crack would convey a positive or negative answer to the diviner by some system of interpretation now forgotten. In some cases the diviner would write down what the question had been, scratching it into the bone after the divination had taken place. The questions were mostly about sacrifices and warfare, but also on such topics as rain and crops. Unfortunately, it is rare to find the answer recorded along with the question. A few bones are not divinatory but record calendars of sacrifices or reports on warfare or hunting expeditions. Despite their narrow topical focus, these oracle bone inscriptions, as they are now called, are a significant source of information about Shang society; it is always easier to understand an ancient society when we have access to their own words, in writing, than when we have only mute artifacts to work with (see plate 3).

  Despite the passage of three thousand years since the last oracle bone was used for divination, the shape of the divinatory crack is preserved in the modern character for “to divine,” (b; see figure 4.3 for a pronunciation guide). This crack shape is not the only enduring legacy of Shang Chinese. The characters on the Shang oracle bones embody the same principles that are used to write Chinese today.

  Early Chinese scribes, like other pioneers of writing, seem to have started with stylized pictograms of easily depicted things or ideas, which they used as logograms to encode complete morphemes. They sided with the Mesopotamians rather than with the Egyptians in that they chose an abstract, schematic way to make their pictograms.

  The most basic Chinese characters are thus stylized pictures of objects, such as “sun,” (rì), “moon,” (yuè), “tree,” (mù), and “horse,” (m), whose original Shang characters are shown in figure 4.1. Generally speaking, any residual pictorial element in the modern character is only visible in retrospect, after one knows the meaning of the character. Other simple characters depict a more abstract idea, such as “center,” (zhong), “above,” (shàng), “below,” (xià), and the numbers “one,” (yi), “two,” (èr), and “three” (sn). Although these characters illustrate the idea behind a word, it is the word that they represent. Thus, for example, in contexts where the word ling is used instead of èr to mean “two” (as we use the word pair sometimes, rather than two), the character is (or ), not .

  Figure 4.1 At left, a few of the Shang characters that have identifiable modern descendants. At right, the stages of evolution of two characters from oracle bones to modern standard and cursive scripts. About 2,200 characters received simplified forms in the People’s Republic of China in the mid-twentieth century. Some of these characters, such as that for “turtle,” seem to have been begging for simplification.

  Any aspiring developer of a writing system soon realizes the limitations of pictography, not to mention the limitations of human memory that come into play if one abandons pictography and attempts to devise arbitrary symbols for each of the language’s thousands of morphemes. These limitations would have led the Chinese to discover the rebus principle, by which the sign for one word was used to indicate a different word of the same or similar sound. A few simple rebus characters still exist, such as (wàn) for “ten thousand,” derived from a Shang character that meant “scorpion”. To resolve ambiguities caused by rebus writing, the Chinese, like their predecessors in the Middle East, added clues to meaning in the form of semantic determinatives.

  But here is where the Chinese took a crucial step differently than other peoples. They put the determinatives and the rebus signs together into a single compound character. As a result, most Chinese characters are compound – sometimes multiply compounded. One part of the character is the rebus, or phonological element, which contains another character (possibly itself compound) whose pronunciation is similar to the one being written. The other part of the character is the semantic determinative, which identifies the semantic category of the morpheme being written. For example, is the traditional (i.e. not simplified in mid-twentieth-century reforms in the People’s Republic of China) character for “horse,” m. The character for “woman” is , n. The word for “mother,” m, belongs to the same semantic class as “woman,” but it sounds like “horse,” though pronounced with a different pitch of the voice. The character for “mother” is thus incorporating the meaning “woman” on the left with the sound [ma] on the right. As can be seen from the way the parts are squeezed together, it is a single, compound character, not a sequence of two characters. The compound character is the same size as either of the original two characters. No matter how many times a character is compounded (the practical limit seems to be about six), all the parts must be
squeezed into the same-sized box.

  The position of the parts within a character is not fixed; so, for example, the character originally meaning “tree” and now “wood,” , mù, serves as the semantic determinative in various types of trees or wooden objects. When added to , róng, “face,” it yields , “banyan tree,” also pronounced róng, but with the “tree” determinative added on the left. Róng can also mean “hibiscus,” but in that case it has the “grass” determinative, adapted from the character , co, added to the top, indicating that a hibiscus is, like grass, a smaller plant than a tree: . The “tree” determinative itself can occur in other positions within a character, sometimes on the bottom, as in (lì, “chestnut tree”), and occasionally on the top, as in (xìng, “apricot”). So it is not always obvious which part of a character is the semantic determinative and which is the phonological component.

  Over 80 percent of modern Chinese characters are semantic–phonological composites of this type. Since the size, location, and precise shape of the components of a compound character will vary, each character must be treated as a unified entity by writers, typesetters, and word processors.

  This practice of squeezing phonological and semantic elements into the same character is part of why the Chinese script has remained logographic and not become logosyllabic like cuneiform or logo-consonantal like Egyptian hieroglyphs. As in all scripts, phonological components are extensively used, but in Chinese they are not used as separate characters. The use of rebus signs without semantic determinatives did not become well established, so the Chinese writing system did not develop purely phonological signs. Chinese characters refer directly to morphemes.

  Another reason why the Chinese script resisted a move toward a logosyllabic system lies in the nature of Chinese morphology. Unlike Sumerian, which expressed grammatical relations by adding affixes to root morphemes, the Chinese language does not inflect its words, and uses very few affixes. The need for phonological signs to spell out affixes was therefore absent in Chinese.

  Chinese morphemes are usually a single syllable long. Therefore, one logogram stands for a single morpheme, but simultaneously for a single syllable. The most common source of two-syllable words is compounds, which are made up of two smaller words, just as English doghouse and Scotland are. Chinese compound words are written with two characters (or occasionally three, if three morphemes are compounded), but the characters within a word will be spaced no more closely together than characters that belong to separate words. Each written unit on the page is thus one character long, which is rather misleading to Westerners used to defining a “word” as something that is written with a space after it.

  When words of two or more syllables are borrowed into Chinese from other languages, they will be written with as many characters as syllables, in keeping with Chinese speakers’ general sense that each syllable is a separate morpheme and each morpheme has its own character. What keeps the Chinese system from being a syllabary, however, is that each character is vested with a meaning. Exceptions do occur in the writing of foreign words (such as the Sanskrit terms of Buddhism), and foreign names, but even in such cases the individual syllables tend to acquire meaning; marketers of foreign brands must be careful what characters they use for their products’ names, so as to avoid any unsavory connotations. The symbols are not used as pure sounds dissociated from meaning as they would be in a true syllabary.

  The Chinese script thus remains true to the logographic principles of its Shang ancestor. Yet despite the clear relationship between Shang writing and the modern Chinese script, reading Shang Chinese is no easy task. Of the roughly 4,500 identified characters, about a third have recognizable modern descendants or have been deciphered, while the rest of the characters are still not well understood.

  Besides the oracle bones and short texts in an elaborated version of the script on commemorative bronze vessels, no other examples of Shang writing have survived. Bones have a durability convenient to archaeologists, but the practice of divination with oracle bones was discontinued soon after the Shang were conquered by the Zhou, leaving a silent period in the archaeological record of Chinese writing.

  The Zhou (whose dynasty lasted in some form or other from about 1122 to 221 BC) seem to have learned to write from the Shang shortly before they conquered the Shang region and established themselves as rulers of the Chinese heartland. The Zhou took to writing with the zeal of the converted. Aristocrats and courtiers were expected to be literate, and bureaucratic paperwork blossomed. Orders were written down, apparently just for the sake of the impressive flourish with which they could be read out to underlings. The first Chinese histories were written.

  Yet very little original Zhou material remains. Early Chinese books were written on bamboo strips which were strung together to make a book. Bamboo, unlike clay tablets or stone inscriptions, is highly perishable. Some examples of bamboo strips do survive from the very end of the Zhou dynasty (dating from the Warring States period of 403–221 BC), but these are about six hundred years younger than the latest Shang inscriptions. The surviving literature from the Zhou period has been copied and recopied, so that the early forms of the characters are no longer preserved. As a result, the history of many Shang characters cannot be traced into Zhou times, with the result that reading Shang Chinese today is to a large extent a matter of deciphering an unknown script.

  During the Warring States period, China was divided into seven separate pieces. It is perhaps not surprising that when China was unified under the First Emperor, Qin Shi Huangdi (221–210 BC) of the short-lived but dynamic Qin dynasty (221–207 BC) that gave China its Western name, he found that his realm had no standard way of writing. The original Shang characters had spread and developed regional and idiosyncratic forms in the intervening Zhou period. Unchecked, the writing system might have gone on to produce mutually illegible descendant scripts used for more and more widely diverging local languages. But the First Emperor was both shrewd and literate, and he recognized the power of the written word. He had his minister of justice, L S, create an official list of characters in an official style. The resulting 3,300-character list was the first Chinese dictionary, and the style of writing was known as the “small seal” script, based on the “large seal” script, one of the styles in use during the Zhou period. (In a less kindly recognition of the power of writing, Shi Huangdi and L S burned a great many of the books extant at the time. The target was free-thinking philosophers. Spared were books owned by members of the Academy of Learned Scholars, books of official history, and books on practical subjects such as agriculture.)

  The small seal script did not entirely win the day. New writing materials (the camel’s-hair brush, and silk cloth as a writing surface) allowed characters to be written with a new ease and fluidity of motion – a development that both spurred the art of calligraphy and affected the forms of characters written for more everyday purposes. Writing that is put to practical use is subject to evolutionary pressure: the clerks and bureaucrats who actually lived by the written word were more interested in getting things written down quickly and accurately than in what style was official. A style known as “clerical” script became popular, and it is this clerical style that eventually became the new standard and underlies the forms of characters usually used in modern times (see figure 4.1, right-hand side).

  There seems to have been some attempt to move away from strict logography in the centuries following L S. Characters were sometimes used just for their phonological value, without a semantic component. This flirtation with phonological writing was quashed by X Shèn, who in AD 121 published an impressive 15-volume dictionary of about 10,000 characters complete with definitions, analysis, and classification. X Shèn introduced a system of sorting characters by semantic determinatives and described the various legitimate ways of composing a character. Characters were to remain logographic, with phonological information restricted to only part of a character.

  The original function of semantic deter
minatives was to provide a clue to the general meaning of a word. With X Shèn’s standardization, however, the determinatives were pressed into service as a classification scheme for characters. The entire corpus of characters was divided into 540 groups, according to the 540 determinatives that Xä Shèn identified. Later the determinatives were reclassified (and sometimes oversimplified) into a total of 214, then raised to 227 in modern times. These determinatives or semantic classifiers are often called “radicals” in English – a term that can be misleading. The classifiers are not the “roots” or cores of characters, though as determinatives they do suggest something about the meaning of characters.

  What does it mean to classify characters? In our alphabetic system of writing, words can be easily sorted into ordered lists based on the alphabetical position of the letters they contain. As a result, all kinds of information can be stored for easy retrieval. One type of information that is typically stored in alphabetical order is that found in dictionaries. If you know the spelling – near enough – of a word, you can look it up in a dictionary and find its meaning. In a language like English, where the phoneme-to-letter correspondence is relatively poor, you may also need to look up the pronunciation. Dictionaries, or their antecedents, word lists, have served as technical manuals to literacy since the times of the Sumerians and Akkadians.

  But how do you find a word in a Chinese dictionary? From the days of X Shèn until modern times the answer has been that first you identify the character’s radical – which may not be easy, as it may be at the top, bottom, left, or right of the character. The characters are listed in the dictionary under their respective radicals, which are themselves ordered according to their complexity of shape – measured in terms of how many strokes of the brush or pen are required to write them. Then you count the number of strokes that the character contains in addition to the radical. The character will be found after characters requiring fewer strokes and before characters requiring more strokes.