Chinese villages ‘probably were not systematically planned’ (AS2007:349). This kind of diversity of interpretation is reflected in the individual entries in this book, many of which give alternative etymologies for the same graph. Thanks to the cumulative efforts of scholars both premodern and modern, a vast amount of knowledge on historical stages of both the Chinese language and the script has been built up, but even so, much of that knowledge is as yet tentative and incomplete, and further work is needed.
The above summary of word-families in Chinese in relation to the script has been presented in as straightforward a way as possible, but this is a complicated area involving significant linguistic technicalities and interpretation of material. Scholars often diverge in their analysis of the phonetic component of graphs, hence frequently several different interpretations are made. In the analysis of the graphs in this book, for word-families reliance has been placed on the published work of established scholars, particularly Tōdō, Mizukami, and Katō.
4 Fluctuation in the Script: Variant Forms
In handwritten texts, even in an alphabetical script with a modest number of different symbols, some degree of variation in the shape of individual signs is common. The potential for shape variation naturally increases in the case of a script that uses hundreds if not thousands of separate symbols or graphs. In Chinese and Japanese texts, variation in graph shape from the earliest period onwards is often seen. This situation is normally handled by taking one form of a graph as representative, and other forms as variants. The earliest major dictionary of Chinese characters compiled in China, the Shuowen jiezi, sets out over 9,000 characters. Each entry begins with a character in seal form which served as the standard, but in some cases an alternative form is noted. Thus, for instance, the Shuowen has 鬽 as the entry heading—and hence main form—for a word meaning ‘malevolent spirit, demon’, and notes 魅 (the standard form in modern Japanese usage) as an alternative form. This illustrates the point that the standard form of a graph in one period can change to be regarded as the variant form in another period. An early example of a dictionary focused on character shapes, prescribing which ones were to be used in official texts as opposed to corresponding variants, is 干禄字書 Ganlu zishu (Character Dictionary for Seeking a [Government] Stipend; J.: Kanroku jisho), a work compiled around the beginning of the eighth century by 顔元孫 Yan Yuansun (? – 714 AD). What is a variant form in one country or jurisdiction can occupy the position of standard in another. Variant forms are in fact very common in all historical periods, though possibly less prominent today due to such factors as widespread education, official script simplification in countries/jurisdictions such as Japan and mainland China, and the regularizing influence of computerization of text with its relatively limited support of variant forms. Other examples of variant forms, selected here at random, are 畄 for 留 ‘stop’, and
5 Early and Pre-Modern Character Dictionaries
A number of character dictionaries are referred to in the explanations for individual characters listed in this book, and so these are described briefly below for the convenience of readers. Titles in Chinese characters are followed by the Chinese romanization of the title together with an English translation and the corresponding Japanese romanization.
説⽂解字 Shuowen jiezi (Explanation of Indivisible Characters and Analysis of Compound Characters; J.: Setsumon kaiji)
In discussing the beginnings of character dictionaries in early China, mention is sometimes made of a few short texts of characters for learners such as 急就篇 Jijiupian (Rapid Access Text; J.: Kyūshūhen), compiled in the 1st century BC. Whether these warrant being regarded as actual dictionaries, though, is questionable, and raises the question of how the word ‘dictionary’ is defined.
The first Chinese character dictionary proper to have survived from early times is the Shuowen jiezi. Completed ca.100AD, this was a large-scale work and one which was highly innovative in that it organized characters according to a system of formal recurrent elements (radicals or determinatives, 540 in total). The compiler 許慎 Xu Shen (J.: Kyo Shin) (born sometime in the period 58–75 AD, died ca. 147–149 AD), set out the small seal forms for some 9,400 characters together with a number of older and variant forms. Xu Shen analyzed the meanings of characters, dividing them into indivisible characters (文) and compound (i.e., divisible) characters (字). Through his work he provided valuable insights for later generations of scholars. Xu Shen’s analyses cannot be relied on in all cases, as he was limited by not having access to very early bronzes or oracle bone characters. However, in assessing and understanding his analyses, modern scholarship is in a position to make allowance for this. Unfortunately, the surviving manuscripts of Shuowen jiezi are all late, with the earliest dating from about 850 years after the time the work was compiled.
For the present book, reliance has been placed to some extent on the treatment and recognition of individual seal forms by scholars such as Mizukami Shizuo, and to some extent on the variorum text of Shuowen jiezi with its parallel modern Chinese translation (published 2009, Wanjuan Publishing Co.).
⽟篇 Yupian (Jade Chapters; J.: Gyokuhen or Gokuhen)
Dating from the early 6th century AD, this work compiled by 顧野王 Gu Yewang is a dictionary in which 12,158 characters are arranged according to pronunciation. As a general trend, the number of characters listed in dictionaries increased with the passage of time. For the present book, quoted excerpts from Yupian appearing in scholarly works have been used.
集韻 Jiyun (Collected Rhymes; J.: Shūin)
A very voluminous dictionary completed in 1039AD; the chief editor was 丁度 Ding Du. Arranged according to pronunciation, it lists 53,525 characters. For the present book, quoted excerpts from Jiyun appearing in scholarly works have been used.
字彙 Zihui (Character Collection; J.: Jii)
This was the first character dictionary to be arranged according to the system of 214 determinatives (as opposed to the 540 in Shuowen jiezi) which has been used until modern times in China and continuously down to the present in Japan. Dealing with 33,179 characters and compiled by 梅膺祚 Mei Dingzuo, Zihui was published in 1615 AD. While clearly a very significant work that was used in the compilation of later dictionaries, Zihui is noted here primarily for its significance in devising the 214 determinative system; it has not been referred to directly in the process of compiling this book.
正字通 Zhengzitong (Mastery of Orthodox Characters; J.: Seijitsū)
Originally compiled by 張自烈 Zhang Zilie, the manuscript of Zhengzitong was purchased by 寥文英 Liao Wenying, who supplemented and published it in about 1671AD. Zhengzitong lists around 33,000 characters arranged according to the 214 determinatives.
Note: in compiling the present book, Zhengzitong was used occasionally to supplement Kangxi zidian (see below), using a 1996 facsimile edition (Guoji Wenhua Publishing Co.). This is because for technical reasons the electronic version of Kangxi zidian referred to immediately below was not always available to consult.
康煕字典 Kangxi zidian (The Kangxi Dictionary; J.: Kōki jiten)
This very prestigious dictionary was compiled at the command of Emperor Kangxi (康煕) (1654–1722) by a group of scholars which included 張玉書 Zhang Yushu and 陳廷敬 Chen Tingjing. Completed and published in 1716, the aim of this work was to provide an authoritative character dictionary to remedy the shortcomings of Zihui and the verbosity of Zhengzitong. The Kangxi zidian, which lists 47,035 characters, served as a standard for matters relating to Chinese characters from the time of compilation down into the 20th century. This is not to say that the Kangxi zidian is totally error free. The existence of errors—probably inevitable in any case in a work of such scale—appears to be due in part to the need to complete compilation within a time-frame that was set by Emperor Kangxi.
Note: for compiling the present book, two editions of Kangxi zidian were referred to: one was the edition published in 2009 by Wanjuan Publishing Co.; the other one—used to ascertain actual character shapes promoted in Kangxi zidian entries—was the electronic