Collection of Studies on the Language and Characters of Jiajin

Author: Yu Suisheng
Publisher:
Publishing Date: 2003-01-01
Features: For a long time, the study of oracle bone inscriptions has been renowned for the scarcity of available materials. The publication of Oracles and Bones Combined (The Complete Collection of Oracle Bone Inscriptions) provided academia with a relatively complete and scientifically organized set of primary materials, largely resolving the issue of finding resources. The Comprehensive Compilation of Oracle Bone Inscriptions from Yin Ruins, edited by Yao Xiaosui and Xiao Ding, transcribed and explained the divinations from Combined and other works into modern characters, compiling them into an index of words and phrases. This was a significant achievement in organizing oracle bone documents, making them more accessible and better serving various related disciplines, and was widely welcomed by academia, especially readers outside the field of oracle bone studies. In recent years, with the advancement of computer technology, an oracle bone information processing system has been developed. Establishing and improving character, word, and corpus databases for oracle bone inscriptions, along with their retrieval systems, and making them practical and widely available to meet the needs of academic research and application has become an urgent issue. The Comprehensive Compilation and Comprehensive Classification can be seen as meaningful preparation and attempts to build a corpus of oracle bone inscriptions. This article aims to discuss several key considerations for establishing a corpus of oracle bone inscriptions, drawing on some experiences in using the Comprehensive Compilation and Comprehensive Classification. If this article, written by an outsider, seems amusing to experts, then let it be abandoned as a vessel of pottery.
### 1. On the Completeness of the Corpus
There are approximately 100,000 oracle bone inscriptions in total, with Combined41956 pieces, plus the unrecorded ones from Oracle Bones from Xiaotun Southern Site, Oracle Bones Collected by the British Museum, Oracle Bones Held by the Institute of Oriental Culture, Peking University, and Oracle Bones Collected by Whittaker, removing duplicates, there are about 50,000 in total. It can be said that almost all research-worthy oracle bone materials have been included. When establishing a corpus of oracle bone inscriptions, apart from discarding forgeries and duplicates, all inscriptions should be fully entered. Additionally, indexing items for retrieval should be marked according to common and unified principles to ensure the completeness of the corpus.
The Comprehensive Classification serves as an index of oracle bone characters and phrases, but during its compilation, some empty words and common phrases were selectively included. The compilers' intention was to make the book more concise and easier to use. Yao Xiaosui, in the preface to the book, stated: "The selection of some common empty words, as mentioned above, will undoubtedly be a source of regret for those studying ancient languages, especially grammar. However, we had no better alternatives and had to make this painful choice." The compilers' grievances are understandable, but this did indeed reduce the usability of Comprehensive Classification. The incompleteness of the corpus is regrettable, but a more serious issue is that no distinct markers were added to distinguish between selected and unselected materials in the book. It is unclear which entries are complete and which are not. The compilers mentioned "common empty words" and "common phrase examples" as being selectively included, but which words were chosen is subjective. Non-professional users, in particular, would find it difficult to determine.
For example, for the same empty word, on page 1333, the entry under "" is marked as "selected," yet on page 970, the entry under "" is marked as "selected" but only has 6 divination instances, clearly indicating it was selected. On page 1299, the entry under "" is not marked as "selected," but it contains many entries, making it unclear whether it was selected. Another example is on page 118, under the entry for ""