摘要
The Tangut dictionaries "Ocean of Characters" and "Ocean of characters, Mixed Classes" divide characters into their component parts without indicating their function. However, this kind of information is essential for an understanding of the structure of characters. Without the knowledge of the functions of the component parts of a character, we can not decide whether it is a phonetic compound (形聲字) or a semantic compound (會意字).
As Tangut characters make extensive use of semantic compounds, the study of the types of compounds should play an important part in understanding Tangut script. It is through the description of semantic structure that the graphic structure can be linked with the meaning of word.
The component parts of a character are mostly abbreviated forms of other characters. Since there are different ways of abbreviation, the classification of characters according to radicals becomes extremely difficult. It is, therefore, suggested that a distinction be made between full forms and abbreviated forms and that the former be used as the basis for classification.
For the above-mentioned purposes all Tangut characters are divided into two groups, simple characters and compound characters. The latter is further divided into twelve subgroups depending on whether they contain phonetics or not, whether their components are abbreviated forms or not, how many components a character is composed of, and so on.