interpretation. "By asserting," said he, "that the current editions read so and so, Tai implies that the Yung-lo ta-tien text supplies in each case the correct reading." This weighted interpretation has been unquestioningly accepted by all of Tai's critics for a hundred years. When they discovered that the Yung-lo copy in many cases contained comparable errors, they forthwith concluded that Tai purloined his improvements from other sources, and then attributed all credit to the Yung-lo text in order to win the favor of his Imperial patron, one of whose purposes in creating the Ssŭ-k'u ch'üan-shu was the recovery of lost works from sources like the Yung-lo ta-tien.
But such an interpretation of Tai's editorial procedure is as unjustified as it is untrue. He did not imply that his corrections were based on the Yung-lo copy, for in fact this was only one of numerous texts and works of reference which he consulted in a lifetime of research in this field. A survey of his editorial notes shows that he refers by name to forty-two works of various types. But we know with certainty that there were many other works which he employed but did not regard it necessary to cite in support of his corrections. He rarely thought it worthwhile to state the grounds for reconstructions that could be confirmed by works of reference familiar to every investigator in the field; and for many other proposed readings there were neither older texts, nor works of reference, which would lend authority to them. At many points he found all available texts equally corrupt, and the necessary reference material often quite unreliable. In an unusually long footnote in chüan 25, he points out that, in making the necessary corrections, he found all existing texts—including the Yung-lo copy—equally faulty; and that the one indispensable source for comparison was a forty-word passage in the Han Dynastic History which itself contains eight errors! In a letter to Ts'ao Hsüeh-min 曹學閔 (1720–1788), written in 1770, he again points out that a single passage quoted by Ts'ao from the Yüan-ho chün-hsien chih—one of the indispensable reference works for the study of the Shui-ching chu—contains six grave errors of fact!
For these reasons, all successful students of the Shui-ching chu have been compelled to go beyond the available texts and works of reference. The most notable instance is furnished by chüan 19, in which both Tai and Chao made transpositions amounting to many thousand words, without the benefit of an authoritative source. And in all his separations of the commingled ching and chu, Chao I-ch'ing adopted precisely the same procedure as Tai, merely noting in each case that such and such a passage had been mistakenly placed in the ching or the chu. In all these cases Chao, as well as Tai, held himself responsible for the changes or corrections he made. No sinister motives or intentions can possibly be deduced from this method. For a long line of intelligent scholars to build up a case against Tai Chiên on some such alleged intention is as unfair as it is absurd—unfair because the person accused is long dead and cannot defend himself, and absurd because there is no evidence to support it. The charges made against Tai are hereby thrown out of court as unworthy of serious consideration.
These, then, are my conclusions: (1) That there is absolutely no evidence to show that Tai Chên saw or utilized the work of Chao I-ch'ing on the Shui-ching chu, before he had completed his own text for inclusion in the Imperial Manuscript Library. (2) That during the years 1786–94 the printed edition of Chao's work had the benefit of slight editorial improvements by Liang Li-shêng, but that no
