The purpose of studying manuscript aims not at revealing secrets, finding something which has hitherto escape detection, discovering the truth or essence concerning the works or the writers themselves. We do not unearth the hidden truth buried beneath the surface. What we do is to turn to a detour, to turn away from dialectical cliches. How do we escape the sterotype? Everytime we encounter the issue of creativity, we instinctively shy away from it. In order to delve into the question of the genetic process, I intend to discuss the traditional Chinese concept of revision and then deal with Wang Wen-hsing's usage of the word in his manuscripts. In an interview, Wang stated that it is impossible for him to revise. If a revision is made anyway, there seems to be no difference between what has been revised and what has remained unchaged. To put it another way, the fault remains even if the revision is done. ＂To try to change a single word does not work, unless you strike all the words which comes after that word,＂ according to Wang. The statement raises a couple of questions. Why can't a writer rewrite his own writing? Does revision has something to do with language? Or does the act concern with language? Does this involve the tradition from whence the writer comes? Or is this an unique phenomenon belonging to Wang alone? Though Wang has emphasized that in the process of transferring from the final draft to the fair copy, not a single word is changed, in actuality his manuscripts are full of revisions and deletions. After the completion of Jiabien [Family Catastrophe], he mentioned that he intended to rewrite everything. Did he or did he not do that? Is there in existence some paradox?