Fuzzy Word Identification: A Case Study from the Oracle Bone Inscriptions
In this paper, a new method of “fuzzy identification” is proposed for circumstances in which an exact match of an epigraphic written word with later attested forms is not possible (for example, because the word has been lost from the language). Based on our increasingly sophisticated understanding of early Chinese morphological patterns and word families, it is sometimes possible to achieve an approximate understanding of pronunciation and meaning in the absence of a precise identification. As an illustration of this approach, I consider the oracle-bone graph as it appears in a famous eclipse inscription. This graph has been identified as 斲 zhuó and 剅 dōu (among others). I argue that any such identification is overly precise. A fuzzy identification, as a member of the word family based on root *tok with meaning ‘cut, chop’, is a more accurate reflection of the state of our knowledge and provides greater insight into the possible pronunciations and range of meaning and function of the word.