picture of two boats joined together). But among the other nine was fang, meaning “street” or “locality,” in such common use that it became necessary to have some means of writing it. Instead of inventing an altogether new character, as they might have done, the Chinese took 方 “square” and used it also in the sense of “locality.” This was a simple expedient, no doubt, but one that, applied on a large scale, could not but lead to confusion. The corresponding difficulty which presented itself in speech was overcome, as we saw, by many devices, one of which consisted in prefixing to the word in question another which served to determine its special meaning. A native does not say fang simply when he wishes to speak of a place, but li-fang “earth-place.” Exactly the same device was now adopted in writing the character. To fang “square” was added another part meaning “earth,” in order to show that the fang in question had to do with location on the earth’s surface. The whole character thus appeared as 坊. Once this phonetic principle had been introduced, all was smooth sailing, and writing progressed by leaps and bounds. Nothing was easier now than to provide signs for the other words pronounced fang. “A room” was 房 door-fang; “to spin” was 紡 silk-fang; “fragrant” was 芳 herbs-fang; “to inquire” was 訪 words-fang; “an embankment,” and hence “to guard against,” was 防 mound-fang; “to hinder” was 妨 woman-fang. This last example may seem a little strange until we remember that man must have played the principal part in the development of writing, and that from the masculine point of view there is something essentially obstructive and unmanageable in woman’s nature. It may be remarked, by the way, that the element “woman” is often the determinative in characters that stand for unamiable qualities, e.g. 妒 “jealous,” 奸 “treacherous,” 妄 “false” and 妖 “uncanny.” This class of characters, which constitutes at least nine-tenths of the language, has received the convenient name of phonograms. It must be added that the formation of the phonogram or phonetic compound did not always proceed along such simple lines as in the examples given above, where both parts are pictorial characters, one the “phonetic,” representing the sound, and the other, commonly known as the “radical,” giving a clue to the sense. In the first place, most of the phonetics now existing are not simple pictograms, but themselves more or less complex characters made up in a variety of ways. On analysing, for instance, the word 遜 hsün, “to withdraw,” we find it is composed of the phonetic 孫 combined with the radical 辶, an abbreviated form of 辵 “to walk.” But 孫 sun means “grandson,” and is itself a suggestive compound made up of the two characters 子 “a son” and 系 “connect.” The former character is a simple pictogram, but the latter is again resolvable into the two elements 丿 “a down stroke to the left” and 糸 “a strand of silk,” which is here understood to be the radical and appears in its ancient form as , a picture of cocoons spun by the silkworm. Again, the sound is in most cases given by no means exactly by the so-called phonetic, a fact chiefly due to the pronunciation having undergone changes which the written character was incapable of recording. Thus, we have just seen that the phonetic of 遜 is not hsün but sun. There are extreme cases in which a phonetic provides hardly any clue at all as to the sound of its derivatives. The character 欠, for example, which by itself is pronounced ch‘ien, appears in combination as the modern phonetic of 坎 k‘an, 軟 juan, 飮 yin and 吹 ch‘ui; though in the last instance it was not originally the phonetic but the radical of a character which was analysed as 欠 ch‘ien, “to emit breath” from 口 “the mouth,” the whole character being a suggestive compound rather than an illustration of radical and phonetic combined. In general, however, it may be said that the “final” or rhyme is pretty accurately indicated, while in not a few cases the phonetic does give the exact sound for all its derivatives. Thus, the characters in which the element 僉 enters are pronounced chien, ch‘ien, hsien and lien; but 意 and its derivatives are all i. A considerable number of phonetics are nearly or entirely obsolete as separate characters, although their family of derivatives may be a very large one. 臤, for instance, is never seen by itself, yet 堅, 緊, and 賢 are among the most important characters in the language. Objections have been raised in some quarters to this account of the phonetic development of Chinese. It is argued that the primitives and sub-primitives, whereby is meant any character which is capable of entering into combination with another, have really had some influence on the meaning, and do not merely possess a phonetic value. But insufficient evidence has hitherto been advanced in support of this view.
The whole body of Chinese characters, then, may conveniently be divided up, for philological purposes, into pictograms, ideograms and phonograms. The first are pictures of objects, the second are composite symbols standing for abstract ideas, the third are compound characters of which the more important element simply represents a spoken sound. Of course, in a strict sense, even the first two classes do not directly represent either objects or ideas, but rather stand for sounds by which these objects and ideas have previously been expressed. It may, in fact, be said that Chinese characters are “nothing but a number of more or less ingenious devices for suggesting spoken words to a reader.” This definition exposes the inaccuracy of the popular notion that Chinese is a language of ideographs, a mistake which even the compilers of the Oxford English Dictionary have not avoided. Considering that all the earliest characters are pictorial, and that the vast majority of the remainder are constructed on phonetic principles, it is absurd to speak of Chinese characters as “symbolizing the idea of a thing, without expressing the name of it.”
The Chinese themselves have always been diligent students of their written language, and at a very early date (probably many centuries B.C.) evolved a sixfold classification of characters, The “Six Scripts.” the so-called 六 書 liu shu, very inaccurately translated by the Six Scripts, which may be briefly noticed:—
1. 指 事 chih shih, indicative or self-explanatory characters. This is a very small class, including only the simplest numerals and a few others such as 上 “above” and 下 “below.”
2. 象 形 hsiang hsing, pictographic characters.
3. 形 聲 hsing shêng or 諧 聲 hsieh shêng, phonetic compounds.
4. 會 意 hui i, suggestive compounds based on a natural association of ideas. To this class alone can the term “ideographs” be properly applied.
5. 轉 注 chuan chu. The meaning of the name has been much disputed, some saying that it means “turned round”; e.g. mu “eye” is now written 目. Others understand it as comprising a few groups of characters nearly related in sense, each character consisting of an element common to the group, together with a specific and detachable part; e.g. 老, 考, and 耉, all of which have the meaning “old.” This class may be ignored altogether, seeing that it is concerned not with the origin of characters but only with peculiarities in their use.
6. 假 借 chia chieh, borrowed characters, as explained above, that is, characters adopted for different words simply because of the identity of sound.
The order of this native classification is not to be taken as in any sense chronological. Roughly, it may be said that the development of writing followed the course previously traced—that is, beginning with indicative signs, and going on with pictograms and ideograms, until finally the discovery of the phonetic principle did away with all necessity for other devices in enlarging the written language. But we have no direct evidence that this was so. There can be little doubt that phonetic compounds made their appearance at a very early date, probably prior to the invention of a large number of suggestive compounds, and perhaps even before the whole existing stock of pictograms had been fashioned. It is significant that numerous words of daily occurrence, which must have had a place in the earliest