The 30-juan long Jingdian shiwen is the result of a year-long collection of commentaries on the Classics and some ancient "masters and philosophers". Lu Deming consulted more than 130 books including phonetic commentaries written between the Han 漢 (206 BCE-220 CE) and the Sui 隋 (581-618) period.
The first fascicle includes an introduction in which Lu Deming explains the intention of the collection, the method of selecting and reviewing texts and the order in which they were put together, as well as the sources from which the texts were extracted. The Classics and their commentaries are arranged in the traditional order Zhouyi 周易, Guwen shangshu 古文尚書, Maoshi 毛詩, Zhouli 周禮, Yili 儀禮, Liji 禮記, Chunqiu-Zuozhuan 春秋左傳, Gongyangzhuan 公羊傳, Guliangzhuan 穀梁傳, Xiaojing 孝經 and Lunyu 論語.
These are followed by the Daoist writings Laozi 老子 and Zhuangzi 莊子 and the classical gloss book Erya 爾雅. It is very interesting that a collection intending to assemble commentaries to the Confucian Classics includes Daoist texts, but not the book Mengzi 孟子. This can only be explained by the low position that the Mengzi text occupied before it was elevated as one of the Four Books 四書 during the Northern Song period 北宋 (960-1126).
Moreover, in the early Tang period Daoist texts were venerated as writings of an imaginary ancestor of the imperial family, namely Master Laozi, called Li Er 李耳 (Li 李 being the surname of the Tang dynasty). The text of the Laozi was at that time transmitted in a very bad state, including many errors, and it is therefore quoted in full in the Jingdian shiwei, while the other texts are only quoted in so far as to identify the sentence that is to be commented on.
The commentaries were mainly phonetic, which is the result of a very important tradition of explaining Confucian texts. Yet during Lu Deming's time there were numerous inconsistent and simply wrong phonetic comments that were wholly unreliable. Lu therefore took the burden to rectify all these phonetic comments with the help of ancient dictionaries and gloss books. It was his intention to compare ancient and contemporary pronunciation of words (yinqie 音切) and therefore had to create a systematic phonetic system to render the sounds of Chinese characters. This was not an easy task because the pronunciation of characters had changed between the Han and the Tang periods, and the pronunciation in southern China was not always identical to that of the north.
The result of Lu's research is that a lot of words are pronounced in two or more different ways. Lu Deming analysed 16,121 characters. His phonetic preference reflects his origin from southeast China, the capital region of the Liang 梁 (502-557) and Chen 陳 (557-589) dynasties. His book Jingdian shiwen is a very important source for the study of the pronunciation of early middle Chinese. It is, furthermore, one of the earliest, qualitatively high-standing, text-critical editions of the Confucian Classics and was therefore later often consulted by Confucian scholars. It is a very detailed commentary on many passages of the Classics, not only as a collection of earlier writings, but also as a source providing Lu Deming's own interpretations.
In the original manuscript, phonetic commentaries had been written in black ink, while commentaries on the meaning of the texts in vermillion ink. When the first prints were produced, the commentaries were separated from the text and placed at the end of each Classic, but during the Yuan period 元 (1279-1368) they were again integrated into the main text. These restructurings had an impact on the quality of the text, so that the Qing-period 清 (1644-1911) scholar Xu Qianxue 徐乾學 (1631-1694) revised this text according to an ancient manuscript copy and had it printed in by the Tongzhi Studio 通志堂 (see Tongzhitang jingjie 通志堂經解).
Yet this version still suffered from many errors. Lu Wenchao 盧文弨 (1717-1795) therefore used a Song-period 宋 (960-1279) copy to revise this version. His corrected version is included in the series Baojingtang congshu 抱經堂叢書 and has later been integrated in the Congshu jicheng 叢書集成. This version is enriched with a critical commentary with a length of 30 juan, written by Lu Wenchao.
Recently a fragment of an ancient manuscript from the Tang period has been discovered in Dunhuang 敦煌. A critical study of this "original" by Luo Changpei 羅常培 (1899-1958) has revealed many differences to the transmitted text. His Tang xieben Jingdian shiwen canjuan si zhong ba 唐寫本經典釋文殘卷四種跋 is included in the 1941(2) issue of the journal Qinghua xuebao 清華學報.
A Song- or Yuan-period copy of the Jingdian shiwen has been preserved in the Imperial Household 清內府 library of the Qing. The Beijing Library 北京圖書館 was able to purchase this copy on the antiques book market of Liulichang Lane 琉璃廠. It was republished in 1980 by the Shanghai Guji Press 上海古籍出版社 as a thread-bound facsimile.