\tThe Shuowen jiezi \u8bf4\u6587\u89e3\u5b57 "Explaining simple and analyzing compound characters", short Shuowen \u8bf4\u6587, is the oldest and one of the most important character dictionaries of ancient China. It was compiled by the Later Han period \u540e\u6c49 (25-220 CE) scholar Xu Shen \u8bb8\u614e. The book was finished in 100 CE but was only submitted to the court in 121 by the author's son, Xu Chong \u8bb8\u51b2. The characters are arranged in 540 so-called radicals (bushou \u90e8\u9996) in 14 chapters, and one chapter including a list of the radicals and Xu Shen’s own postface (xu \u53d9).<\/p>\n
\tThe initial point of Xu's dictionary was the fact that during the Former Han period \u524d\u6c49 (206 BCE-8 CE) a lot of different Confucian books had come to light, written in different styles of script, from the modern "chancery script" lishu \u96b6\u4e66 (the so-called "modern script classics" jinwenjing \u4eca\u6587\u7ecf) to the old "seal script" zhuanshu \u7bc6\u4e66 (the so-called "old script classics" guwenjing \u53e4\u6587\u7ecf). In order to provide a tool for a study of these texts, especially the old text classics, which began to dominate Confucian scholarship at the beginning of the Later Han period, Xu Shen provided a dictionary which analysed the seal script characters and their meaning. The allegedly more original old script versions seemed to be more reliable than the new script texts.<\/p>\n
\tThe lemmata heads are written in small seal script (xiaozhuan \u5c0f\u7bc6), while the analytic and explanatory text is written in contemporary chancery script. From the Qing period \u6e05 (1644-1911) on editions of the Shuowen also added transcriptions of the seal script characters, the large seal script characters (zhouwen \u7c40\u6587, also known as dazhuan \u5927\u7bc6), the old characters (guwen \u53e4\u6587) and popular variants (suti \u4fd7\u4f53), which have been provided by Xu Shen to some of the standard small seal script characters.<\/p>\n
\tIn his postface (xu) to the Shuowen, Xu Shen gives an account on the development of the Chinese script. It is said to have been invented by Cang Jie \u4ed3\u9889, a minister of the mythological Yellow Emperor \u9ec4\u5e1d, after he had seen the traces of bird feet on the soil. The simple characters he created are mainly illustrations of objects and ideas, simple in appearance and therefore called "patterns" (wen \u6587). In a later stage the characters or ideographs were combined from an ideographic part (xing \u5f62 "shape") and a phonetic part (sheng \u58f0). This type of compound characters is called zi \u5b57. Today both terms are combined to the word wenzi \u6587\u5b57, meaning "Chinese character" or "Chinese script". Xu Shen discerns six theoretical types of characters, the liushu \u516d\u4e66 "six types of script":<\/p>\n
\t• The simplest form are pictograms (xiangxing \u8c61\u5f62 "illustration of a shape"), pictures of optically perceivable or imaginable things, like \u6728 "tree", \u5c71 "mountain", different animals and plants (\u9a6c "horse", \u7f8a "sheep", \u7af9 "bamboo", \u7c73 "grain"), \u624b "hand", \u7709 "eyebrow", \u6c14 "breath", or various objects (\u6208 "halberd", \u9f0e "tripod"). This group also includes symbols of figurative meaning, like \u4ea4 "exchange" (a picture of crossed legs).
\t• The second type of characters are ideograms of simple relationships (zhishi \u6307\u4e8b "pointing at things"), often derived from a pictogram. The relationship to the pictogram is indicated with a stroke, like \u4e0a "above", \u4e0b "below", \u5203 "blade" of a knife, \u672c "root" or \u672b "branch" of a tree. Turned characters also belong to this type, like \u4e4f deficient" (opposite of \u6b63 "correct"), or \u5315 "change", a turned \u4eba "man".
\t• The third type (huiyi \u4f1a\u610f "assembled meanings") is a combination of two pictograms, like \u6b66 "war" from \u6208 "halberd" and \u6b62 "base"; \u4fe1 "trust" from \u4eba "man" and \u8a00 "spech"; \u4e27 "funeral" from \u54ed "weeping" and \u4ea1 "gone, dead"; \u65e6 "dawn" from \u65e5 "sun" and the horizon; or \u516c "public" from \u516b "to separate" and \u3119 "private". There are a lot of characters from this type, but only in a few cases Xu Shen explicitly mentions the word huiyi.
\t• The fourth type (xingsheng \u5f62\u58f0 "shape and sound"), which applies to about 90 percent of all Chinese characters, is a combination of pictogram and a character of which the sound is used, like shang \u8d4f "to grant a reward", from \u8d1d "shell, i. e. money", and the phonetic shang \u5c1a. The same phonetic part \u5c1a is used, for instance, in the characters tang \u5802 "hall" (phonetic \u5c1a and radical \u571f "pounded earth") or shang \u88f3 "garment" (phonetic \u5c1a and radical \u8863 "clothing")
\t• The fifth type (zhuanzhu \u8f6c\u6ce8 "comment by turning") is a rarely understood type, because it is not sufficiently explained by Xu Shen. In his preface, he gives the examples kao \u8003 and lao \u8001. It seems to be that because both have a similar meaning ("old") and similar pronunciation, the characters have been conciously designed in a very similar way, but with one part mirrored horizontally. Yet in the explanation to two lemmata themselves, Xu Shen derives the character kao \u8003 from an abbreviated \u8001 "old" as a radical and the phonetic part kao \u4e02. The following characters also might belong to this group: fan \u8fd4 "give back" and huan \u8fd8 "turn back", or biao \u6807 "tip of a branch" and miao \u676a "end of a stalk"
\t• The sixth type of character (jiajie \u5047\u501f "wrongly borrowed") are loan-characters borrowed for a word similarly pronounced but with a different meaning, like ling \u4ee4 "order" from ming \u4ee4 "command" (later written \u547d) and zhang \u957f "headperson", from chang \u957f "long hair". Many grammatical particles are of this type. The ancient Chinese simply borrow another character with the same or a similar pronuncition for these words, like nai \u4e43 "breast" for nai "therefore", qi \u5176 "basket" for qi "his, her, its", zhi \u4e4b "to go" for a genetive particle and object pronoun, or ye \u4e5f "uterus" for an equalizing particle. In some cases, new characters were created for the original words, like \u5976 for "breast, milk", and \u7b95 for "basket".<\/p>\n
\tXu Shen has developed a special syntax for his analysis. Huiyi characters are generally analysed with the sentence cong A, B \u4ece\u7532\u4e59, or cong A, cong B \u4ece\u7532\u4ece\u4e59 "from A and B". Xingsheng characters are analyzed with the sentence cong A, B sheng \u4ece\u7532\u4e59\u58f0 "from A and the sound of B". One part of the huiyi characters is in many cases also used phonetically, in which case Xu Shen writes cong A, cong B, B yi sheng \u4ece\u7532\u4ece\u4e59\uff0c\u4e59\u4ea6\u58f0 "from an and B, B is also used phonetically". In a lot of characters the phonetic part is abbreviated, a phenomenon which in huiyi type characters also occasionally occurs. Xu Shen's formula for this phenomenon is cong B sheng sheng \u4ece\u4e59\u7701\u58f0 "from abbreviated B, used phonetically".<\/p>\n