Khitan Small Script
Khitan Language
The Khitan, descended from the Donghu, were one of the northern peoples of ancient China. At the end of the Tang Dynasty, the Central Plains were in turmoil and divided into various Fanzhen. Yelü Abaoji, the leader of the Khitan aristocracy, seized the opportunity to unite the various clans and became the emperor in 907 A.D. In 916, he established a prominent dynasty in Chinese history that was called the Khitan Dynasty or Liao Dynasty. After the establishment of the Liao Dynasty, in order to meet the needs of political, military, economic and cultural development, prompted by ethnic consciousness, two scripts called the Khitan Large Script and the Khitan Small Script were created.
When the Khitan scripts were in use, books were written and translated in them. Probably due to the strict control of books by the Liao Dynasty and the subsequent wars, not a single book written in Khitan scripts has been found to date. Only in the books of the Song Dynasty are Khitan signs and banners found, written by the people of that time. In addition, although there is the Langjun Xingji is engraved on the wordless stele in front of Wu Zetian’s tomb, no one knew that it was engraved with Khitan characters until the twentieth century. In the twentieth century, with the discovery of numerous inscribed tombstones and wall paintings, the study of Khitan scripts became vigorous.
Khitan Small Script
Khitan Small Script is an functionally phonetic script, in which the meanings are weakly involved in the correspondence between language and script. The Khitan words are written phonetically with yuanzi, which represent consonants, vowels or diphthongs, or syllables. The following analysis is based on Chinggeltei (2005).
When recording Han loan words, a yuanzi usually corresponds to a syllable. There are also cases where two or more yuanzi correspond to a single syllable, which makes it possible to analyze their function as corresponding to initials (consonants) and finals (vowels or diphthongs). Other opinions include the idea that these yuanzi contain an inherited vowel and thus still correspond to syllables, while syllable merging occurs during yuanzi combination.
When recording native words, a yuanzi usually corresponds to a syllable as well, and is usually not specialized for words with specific meanings. The idea that yuanzi contains an inherited vowel is partially supported by the fact that in the cases where native words are recorded, some of the yuanzi thought to correspond to consonants are found to be used separately and thus to correspond to syllables, while on the other hand it is not certain that all yuanzi can form their own syllables.
The yuanzi in Khitan Small Script are clearly derived from Chinese characters, but it is not clear whether Khitan Small Script is also cognitively ideographic. Some studies have shown through a few examples that the glyphs of yuanzi are more or less related to Chinese characters, but none of these analyses are generalizable.
Khitan Small Script Character Set
The character set design of Khitan Small Script consists of two parts, namely, yuanzi collation and shaping structure design.
In April 2010, China National Body (Sun Bojun, Jing Yongshi and Li Yang) submitted the first proposal for the encoding of the Khitan Small Script to WG2 (WG2 N3820 and WG2 N3918), in which the code charts summarized the repertoires from Kittan Shōji Shakudoku Mondai, “The font style norm of original characters of the Khitan Small Script and complete table for original characters” and The Kitan Language and Script, and add a column for a unified repertoire.
In April 2016, Andrew West, Viacheslav Zaytsev and Michael Everson (WG2 N4725R) further organized the above proposals of the China Nation Body with thirteen documents, to form a repertoire of 472 yuanzi. Apart from the three documents mentioned above, “Kittan shōji-no onka suitei oyobi sōkan mondai”, Qìdān Xiǎozì Yánjiū, Qìdān Xiǎozì Shìdú Wèntí (Xiūdìngběn), Mílín Wènjìng, Mítián Gēngyún, “A collection of inscriptions on steles excavated from the Supreme Capital of the Liao dynasty”, Qìdān Wénzì Yánjiū Lèibiān, Qìdān Xiǎozì Cíhuì Suǒyǐn, Kittango no Kenkyū and New Materials on the Khitan Small Script: A Critical Edition of Xiao Dilu and Yelü Xiangwen are considered.
After discussion, based on WG2 N4725R, yuanzi #151 should be normalized to yuanzi #246, and yuanzi #349 should be analyzed as a combination of yuanzi #348 and yuanzi #467. With two yuanzi deleted, the total number of yuanzi becomes 470, and the 470 yuanzi (U+18B00..U+18CD5) thus forms the Khitan Small Script character set in Unicode 13.0.
From the text of Khitan Small Script, it can be seen that Khitan Small Script has a two-layered structure of zhengzi (整字) and yuanzi (原字). The text of Khitan Small Script can be divided into a number of vertical lines by obvious line spacing, and each vertical line can be divided into a number of segments by obvious word spacing, and the text element corresponding to each segment is called the main text element, i.e., zhengzi. Each zhengzi often has an internal structure in which the text elements with independent linguistic function are called secondary text elements, i.e. yuanzi. In a zhengzi, yuanzi are stacked from left to right and from top to bottom in a pattern of up to two per horizontal line. Some zhengzi have two yuanzi in the first horizontal line (called Type A Cluster), while others have only one yuanzi in the first horizontal line (called Type B Cluster), as described in WG2 N4943R.
The use of space (U+0020) for spacing between zhengzi is not questioned; the structure between yuanzi must be fully considered. Due to the limited stacking pattern of Khitan Small Script, there is no need to introduce unnecessary flexibility, and it is only necessary to consider how to differentiate between Type A Cluster and Type B Cluster, resulting in the fact that if a control character is inserted between the first yuanzi and the second yuanzi of a zhengzi, it means that the zhengzi belongs to Type B Cluster, and otherwise it belongs to Type A Cluster, as described in WG2 N4943R. This control character was recognized as a Khitan Small Script-specific control character (U+16FE4) in UTC #156 meeting (L2/18-183R2), and is encoded in the Ideographic Symbols and Punctuation block in Unicode 13.0.
Typefaces
Before the Khitan Small Script was encoded, the font developed by Menksoft was quite common, with yuanzi mapped to PUA codepoints and Microsoft field codes to form zhengzi.
Below are a Khitan Small Script typefaces that I was involved in developing, which is in Fangsong style.
Noto Fangsong Khitan Small Script
This typeface is the product of the redesign project for Noto Serif Tangut. Since the initial version of Noto Serif Tangut maintained the same stroke characteristics as Noto Serif CJK SC, redesign did not consider changing the style.
Stylistic Design. In order to take into account the characteristics of Kaishu and the needs of modern typography, the Fangsong style, which combines the structure of Kaishu and the stroke form of Song style, came into the vision of the design team. Juzhen Fangsong is its initial and representative. It is built on the foundation of continuous traditional Chinese culture, has the style of Ouyang Xun style Kaishu, and has both traditional aesthetic spirit and modern technical aesthetics, so Fangsong style became the final choice of the design team.
Metric Design. In the original Khitan Small Script documents, the strokes of the yuanzi are interspersed, so that the zhengzi forms an organic whole, and the height is not an integer multiple of that of Chinese characters. However, in typesetting and digital documents, when Khitan Small Script and Chinese characters are mixed, the body box of yuanzi is the same as that of Chinese characters, and the close stacking between the body boxes makes the whole character one to four times the height of Chinese characters. Therefore, in type design, adopting the body boxes of typesetting and digital documents is convenient to reduce the workload of outline design and shape design, and keep the Yuanzi in a relatively stable appearance, which is easy to recognize.
Multi-Scenario Layout. In the original documents, the zhengzi is written top-to-bottome and right-to-left. In typesetting literature, when mixed with Chinese characters, zhengzi and Chinese characters are set either left-to-right and top-to-bottom, or top-to-bottom and right-to-left. In digital literature, in addition to the same horizontal layout and vertical layout rows as in typesetting literature, some documents do not produce stacking of yuanzi, but are arranged from left to right, known as discrete layout; when presenting the content of Khitan Small Script in a web page, some browsers do not recognize the Khitan Small Script as U (upright) characters but as R (rotated) characters due to the shaping engine version, and it is necessary to produce fonts in rotated layout.