The earliest book on oracle bone inscriptions was Liu E's 劉鶚 (1857-1909) Tieyun canggui 鐵雲藏龜. This catalogue pioneered the study of Shang-period 商 (17th-11th cent. BCE) texts, yet the beginnings were very difficult because the shape of characters differed widely from that of the ancient seal script style (zhouwen 籀文, dazhuan 大篆) and that of the Warring States period 戰國 (5th cent.-221 BCE). While E's book had the function of a catalogue presenting the appearance of texts incised on "dragon" bones (longwen 龜文), the following generation of scholars began to decipher and interprete them. Sun Yirang belonged to the group of Chinese scholars who seriously attempted to read oracle bone inscriptions.
His book Qiewen juli used the bones shown in Liu E's book for a first study of the inscriptions. It is divided into ten chapters discussing the shape of certain characters or topics, namely dates (Ri-yue 日月), words for prognostication (Zhenbu 貞卜), the progress of divination (Bushi 卜事), spirits and deities (Guishen 鬼神), diviners (Buren 卜人), court officials (Guanshi 官氏), regional states (Fangguo 方國), rites and ceremonies (Dianli 典禮), selected characters (Wenzi 文字), and rules of writings and composition (Zali 雜例). Sun mainly used bronze inscriptions from the Shang period and the etymological explanations in the Han-period 漢 (206 BCE-220 CE) character dictionary Shuowen jiezi 說文解字 for his interpretations.
First page of the Qiewen juli 契文舉例, with the author's observations of oracle script characters in comparison to such on bronze vessels and the seal script. Exemplary translation see below.
Sun found out that Liu's interpretation of a certain character as 問 "to ask" was wrong, and put it right as the character 貝, as an abbreviation of 貞 "to prognosticate". He was also able to to identify the character 且 as an abbreviation of 祖 "ancestor".
He counted the names of twenty different diviners mentioned in the oracle texts collected by Liu E. Even if the number of characters of words discussed and the textual basis are quite small (not to speak of the bad reproduction quality in the Tieyun canggui), and some interpretations are wrong (王 wrongly read as 立, 獸 wrongly read as 獲, 止 wrongly read as 正, wrong distinction of 母 and 女), the Qiewen juli was the basis for a rich area of research that realized in the following decades.
The book was finished in 1904, but was only printed in 1917 in Luo Zhenyu's 羅振玉 (1866-1940) series Jishi'an congshu 吉石盦叢書. In 1927, the Yinyin Study 蟫隱廬 in Shanghai produced a reprint.
|龜甲文簡略，多紀某日卜故。今存殘字亦日名最多十榦。惟乙己二字與小篆同，餘則多差異。||Characters [incised] on tortoise plastrons are very simple. Many [texts] record oracles on certain days, and among surviving words, most [refer] to the Ten Celestial Stems. [Among these], only the characters for 乙 and 己 are equal to the Small Seal Script variants, the others differ.|
|如甲字皆作「十」, （一之四。凡 文字恒見者唯摒。始見及文義略完萌者舉證一二不悉摒也。）金文母甲觶甲字正如是作。||For instance, the word 甲 is written 十. (1/4. N.B. I only provide reference to oft-seen words. Reference for first appearance and words with unclear meaning is only given occasionally). The word 甲 in the inscription of the bronze vessel Mu Jia zhi 母甲觶 is just written like this (十).|
|丙字皆作𠔿(四之一)。金文魚父丙𣡫(爵)、父丙𣡫(爵)並略同。||The word 丙 is always written 𠔿 (4/1), just like in the bronze inscriptions Yu Fu Bing jue 魚父丙爵 and Fu Bing jue 父丙爵.|
|丁字皆作口（一之二）。金文父丁𣡫(爵)亦同。||The word 丁 is always written 口 (1/2), like in the bronze inscription Fu Ding jue 父丁爵.|
|戊字多作（十七之四）或作（四十二之四)。金文子孫父戊觚作,父戊舟𣡫(爵)作，與此略同。||The word 戊 is in many cases written (17/4) or (42/4). In the bronze inscription Zisun Fu Wu gu 子孫父戊觚, it is written , and in the Fu Wu Zhou jue 父戊舟爵 written - quite similar to this [the oracle shapes].|