On the act and representation of reading in medieval China.

• Author: Chen, Jack W.

I.

Books, throughout the medieval period in China, were precious objects, handed down as treasures from generation to generation in families wealthy or fortunate enough to possess them. It is therefore not surprising that the sixth-century scholar Yan Zhitui [TEXT NOT REPRODUCIBLE IN ASCII] (531-591), in his remarkable Yanshi Jiaxan [TEXT NOT REPRODUCIBLE IN ASCII], sternly emphasizes how the care of books is a serious matter and that to allow books to be carelessly strewn about one's study is injurious to moral character. Moreover, he also feels it necessary to point out that,

Whenever I read the books of the sages, never once do I not treat the books with solemn respect. If old paper has the words or explanations of the Five Classics, or even the names of worthies and the perfected, I would not dare use it for something foul. [TEXT NOT REPRODUCIBLE IN ASCII](1)

Yan reads and reveres the canonical books of the tradition, even extending that reverence to the waste paper that otherwise would be used to wipe oneself in the bathroom. There is a fetishistic quality here, one that elevates any words relating to the Classics to the same canonical status as that of the Classics, not to be defiled either by careless treatment in the study or worse. Citations, names, and glosses relating to these sacred books themselves become sacred, as these objects of textuality serve as a link between the reader and the sagely referent.

Given the importance ascribed to books and canonical texts, it is thus curious that how one should read--that is, the actual practice of reading--is not something that Yan actually describes. Indeed, much of our understanding of reading practices in traditional China is derived from later sources, one of the most famous of which is the "Method of Reading" ("Dushu fa" [TEXT NOT REPRODUCIBLE IN ASCII]) by the Neo-Confucian thinker Zhu Xi [TEXT NOT REPRODUCIBLE IN ASCII] (1130-1200). As Daniel K. Gardner has noted, the reason for such a prescriptive approach to reading may be found in the proliferation of books in the age of printing. Gardner writes, "Thus there developed among readers, hoping to cover as many texts as possible, a tendency to jump from one to another," with the result that "readers absorbed little or nothing of what they read." (2) A certain level of anxiety towards uncareful reading may be said to permeate Zhu Xi's thoughts on moral pedagogy, an anxiety that does not seem to be felt in the earlier age of the manuscript, when texts available for reading were presumably fewer.

What, however, can be said about reading in the period prior to the Song dynasty, which is to say, in the medieval period? The French scholar Jean-Pierre Drege notes that information about reading practices from the period of manuscripts is "essentially limited to chapters, in various works, that are devoted to exhortations to study or to biographies contained in the dynastic histories where literary education is sometimes briefly mentioned." (3) The most common textual representations of reading are found in biographies where the person is depicted as a child prodigy who could read and recite various texts from a young age. A typical example can be found in the opening lines of the biography of the poet Cao Zhi [TEXT NOT REPRODUCIBLE IN ASCII] (192-232), which notes that "when he was just a little over ten years of age, he could read and recite poems, essays, and rhapsodies, a total of several hundred thousand words" [TEXT NOT REPRODUCIBLE IN ASCII]. (4) However, this kind of statement belongs to biographical or hagiographic typological conventions, and it tells us more about the construction of Cao Zhi than about the act of reading itself.

To understand what it meant to read in medieval China, one must first understand the terminology employed in definitions and accounts of reading. The most common term for reading in classical Chinese is du [TEXT NOT REPRODUCIBLE IN ASCII], which is defined by Xu Shen [TEXT NOT REPRODUCIBLE IN ASCII] (c. 58-c. 147) in the Shuowen jiezi [TEXT NOT REPRODUCIBLE IN ASCII] as follows: "The character du means 'to read a document aloud" [TEXT NOT REPRODUCIBLE IN ASCII](5) Xu Shen's use of song[TEXT NOT REPRODUCIBLE IN ASCII] points to the oral nature of reading, though song is not so technically a term for reading itself as it is a term that denotes a particular mode of textual voicing. (6) As Wolfgang Behr and Bernhard Fuhrer note, the term song was often used in performative contexts or as part of pedagogical recitation practices. (7) The latter is most likely the case in the Cao Zhi biographical excerpt, which conjoins the term song with du to form the compound songdu [TEXT NOT REPRODUCIBLE IN ASCII] describing the young Cao Zhi's ability to read hundreds of thousands of characters.

Nevertheless, at the same time, the compound songdu can also be used to describe more personal, nonperformative modes of reading. In the Shi ji [TEXT NOT REPRODUCIBLE IN ASCII] the military strategist Zhang Liang [TEXT NOT REPRODUCIBLE IN ASCII] (d. 189 B.C.) encounters an old man who gives him a secret book to read and then departs. The account goes on to relate: "The next day he looked at the book; it was The Grand Duke's Art of War. Liang accordingly thought it remarkable, and he often made a practice of reading it" [TEXT NOT REPRODUCIBLE IN ASCII] (8) The text that serves as Zhang Liang's fount of secret military knowledge is none other than the Six Secret Strategems (Liutao [TEXT NOT REPRODUCIBLE IN ASCII]), known also as the Taigong bingfa [TEXT NOT REPRODUCIBLE IN ASCII] (9) Since the Taigong bingfa is represented as an esoteric transmission, Zhang's reading--again, songdu--is almost certainly not declamative or recitative reading, but rather a more private form of reading, perhaps speaking the text's words aloud to himself or sotto voce.

One of the earliest examples of reading that makes explicit its oral context is also the earliest philosophical comment about the act of reading itself. This is the famous passage from the Zhuangzi [TEXT NOT REPRODUCIBLE IN ASCII].

Duke Huan was reading above in the hall. Wheelwright Pian was cutting a wheel down below in the hall; he put down his mallet and chisel and ascended the hall. He asked Duke Huan, saying, "I venture to ask, what kind of words are those that the Duke is reading?" The Duke said, "The words of the sages," Pian asked, "Are the sages still present?" The Duke said, "They are already dead." Pian said, "If that is the case, then that which my lord is reading are merely the lees and dregs of the ancients." [TEXT NOT REPRODUCIBLE IN ASCII]

[TEXT NOT REPRODUCIBLE IN ASCII](10)

When Wheelwright Pian asks Duke Huan of Qi what the duke is reading, he is doing so because he can hear the sound of the duke's voice, reading in the hall above. What is interesting about this passage, however, is how Pian's questioning leads to a more complicated set of issues concerning the nature of text and language. In asking "what kind of words" that the duke is reading, Pian uses the term yan [TEXT NOT REPRODUCIBLE IN ASCII] "spoken words" or "speech," rather than any of the terms that would denote writing, such as wen [TEXT NOT REPRODUCIBLE IN ASCII]. To the duke's reply that he is reading the words (yan) of the sages, Pian then asks if the sages are still "present" (Zai [TEXT NOT REPRODUCIBLE IN ASCII]). When the duke responds that they are long dead, Pian comments that if the sages are themselves dead, then what the duke is reading are simply "lees and dregs" (zaopo ). The death of the sages negates the metaphysical authenticity of their speech--indeed, without the full presence and immediacy of the sages, the words preserved by books are merely impoverished traces, vestiges that can only gesture back to a now-absent essence or truth. (11)

II.

The relationship between speech and reading is of great importance in premodern literary cultures, whether in China or in the West, as the written record clearly shows that vocalized reading was the dominant mode. Zhuangzi's ironic dismissal of reading aside, the voice of the reader was crucial to the decoding of the text. Paul Saenger has pointed out how Western medieval texts were presented in unpunctuated and unspaced lines of writing--what was known as scriptura continua. This format compelled the reader to parse the words of unfamiliar texts by reading them aloud in an often time-costly, laborious manner. Reading scriptura continua, as Saenger further notes, forced the eyes to move across the page at an uneven rate, in a series of progressive and regressive jumps known as "saccades." (12) Because the complex grammatical constructions of classical texts made it necessary for the ancient reader to parse not only the morphemic boundaries, but also the sentence constructions, reading aloud "helped the reader to hold in short-term memory the fraction of a word or phrase that already had been decoded phonetically while the cognitive task of morpheme and word recognition, necessary for understanding the sense of the initial fragment, proceeded through the decoding of a subsequent section of text." (13)

For classical Chinese, of course, morphemic boundaries mostly coincided with character boundaries, so that the problems posed by scriptura continua were not as pronounced as one would find in alphabetic scripts. Nevertheless, in the absence of widespread conventions of punctuation, the mastery of reading classical Chinese still required sustained and directed study, generally with a teacher who would show the beginning student how correctly to identify sentences and rhetorical periods. A passage in the Li ji [TEXT NOT REPRODUCIBLE IN ASCII] describes the goal of the first year of education as follows: "After one year, one can see whether [the student] can parse the classics and discern what the writer had in mind"--[TEXT NOT REPRODUCIBLE IN ASCII]. The Eastern Han commentator Zheng Xuan [TEXT NOT REPRODUCIBLE IN...