佛祖統紀 45.42

45.42

述曰。夫人在孩提時便知學字。師不知字故所學多非。及長又以是教人非。傳非故皆不疑。不知字而為文。文欲學古。而字不知古。何文哉。漢先君中郎尚書（後漢盧植字子幹。至磐已身為十二世。備見家諜）嘗上書。乞正五經文字。有云。古文近實而流俗。降在小學此傷之也（班固漢書列字書為小學）磐幼從師學。年五十始悟字多非。因發憤取諸字書訂正之。益知世人書字訛謬為尤多。如陳陣華花歷曆奉俸。反主俗字而違失經史之正（論語戰陳之事。諸典皆作陳。今韻略却出陣字。於下注云。亦作陳○禮記。菊有黃華。詩隰有荷華。魏曹植詩。朱華冐綠池。今韻略先出花字。於下注云。亦作華○易治歷明時。漢書律歷志皆作歷。今韻略歷曆並出○漢書餐錢奉邑。又光武詔。增百官奉。扶用反。秩祿也。今韻略先出俸字。下注云。亦作奉。此皆以俗字為主）巖岩槀[葶-丁+呆]臺䑓船舡。此造作不典。而正通俗三者之所不收。岡崗暴曝芻蒭然燃。是為山山火火重加之繆（暴音僕。孟子一日暴之。史記。暴其所長於燕。又暴露於野。並音僕。號字韻者。非曝浴）況况宜冝羈羈場塲是為增減點畫之不成字也。至於正月之音征。昭穆之音韶。異代不必諱。而述音義者。諱之何也（書正月上日音政。又音征。詩正月如字。世謂秦始皇諱政。故避者音征。今去秦遠矣。尚不聞避政事之政。何獨正月之避耶。禮記王制。三昭三穆無明音。周禮呼昭穆而進之。注上昭反。此因避晉文帝諱昭故音韶。然昭穆正取昭明之義。朱晦菴經解並用如字。此征韶二音皆陸德明釋文。不知改削之過。已亡不令之君。能與避諱。而孔聖萬世之師。不知避諱。古今儒宗不能無愧也）。佛與禪出釋典而不收。僧與梵亦釋典而置之篇韻。何也（梵語佛陀此言覺者。梵語禪那此云靜慮。此二字不收。梵語僧伽此云眾。梵者清淨離欲。今韻中收二字）是知孫強之增玉篇（梁顧野王撰玉篇。唐孫強增字。多取俗書）孫愐之增廣韻（唐孫愐撰廣韻多收俗字）與夫本朝丁度之撰集韻。皆泛引俗書雜揉正體。俾經史古意黯然不明。此字書之大阨。而有待於志士以矯正之乎。今觀歐陽序韻總。謂能正五方之訛。儒者莫能難。於私心竊有感焉。特無由見此書耳（見六一居士集）黃太史曰。魏晉人取字勢易工。不復問字之根源。如橋𭫥直[匕/且]皆不成字。思陵翰墨志有云。書字之弊無如本朝。至哉。聖學之深識也（高宗陵號永思）。

The commentary says: When people are in childhood, they learn characters; but teachers do not understand characters, so much of what is learned is wrong. When they grow up, they teach others these errors; because they transmit errors, none suspect it. Not understanding characters, they compose texts; they wish their texts to emulate antiquity, yet they do not know antiquity in their characters—what kind of literature is that? My ancestor of the Han dynasty, the Palace Gentleman and Master of Writing (Lu Zhi of the Later Han, courtesy name Zijian; Zhipan is his twelfth-generation descendant, as fully recorded in the family genealogy), once submitted a memorial requesting the correction of the characters of the Five Classics, which stated: 'Ancient script is close to reality, yet following vulgar usage, it has declined into primary philology; this has harmed it' (Ban Gu's Book of Han categorized character dictionaries under primary philology). I, Zhipan, studied under teachers in my youth, and only at the age of fifty did I realize that many characters were wrong. Thereupon, I roused myself, gathered various character dictionaries, and collated and corrected them. I came to know even more clearly that worldly people write characters with excessive corruptions and errors. For example, for chen / zhen (陳/陣), hua / hua (華/花), li / li (歷/曆), and feng / feng (奉/俸), they conversely make vulgar characters primary and deviate from the orthodox forms of the classics and histories (in the Analects, 'matters of battle arrays', all the canons write chen 陳; the present Yunlue instead produces the character zhen 陣, noting beneath it: 'also written as chen 陳'; in the Book of Rites, 'chrysanthemums have yellow blossoms [hua 華]', and in the Odes, 'the lowlands have lotus blossoms [hua 華]', and in Cao Zhi of the Wei's poem, 'vermilion blossoms [hua 華] cover the green pool'; the present Yunlue first produces the character hua 花, noting beneath it: 'also written as hua 華'; in the Book of Changes, 'order the calendar and clarify the seasons', and in the Book of Han, the Treatise on the Pitchpipes and Calendar, all write li 歷; the present Yunlue produces both li 歷 and li 曆; in the Book of Han, 'board money and salary towns [fengyi 奉邑]', and furthermore Emperor Guangwu's edict, 'increase the emoluments [feng 奉] of the hundred officials', pronounced fu-yong fan, meaning official stipend; the present Yunlue first produces the character feng 俸, noting beneath it: 'also written as feng 奉'; these all take vulgar characters as primary). Yan / yan (巖/岩), gao / gao (槀/[葶-丁+呆]), tai / tai (臺/䑓), chuan / chuan (船/舡)—these are uncanonical fabrications not collected by the three categories of correct, common, and vulgar. Gang / gang (岡/崗), bao / pu (暴/曝), chu / chu (芻/蒭), ran / ran (然/燃)—these are errors of redundantly adding mountains upon mountains or fire upon fire (bao 暴 is pronounced pu. In Mencius, 'if one warms it for one day [pu 暴 zhi]'; in the Records of the Grand Historian, 'exposing his strengths [pu 暴 qi suo chang] to Yan'; and 'exposed in the wilderness [baolu 暴露 yu ye]', all are pronounced pu. Those who rhyme it with hao are wrong; it is not bao-yu 曝浴). Kuang / kuang (況/况), yi / yi (宜/冝), ji / ji (羈/羁), chang / chang (場/塲)—these are non-characters formed by adding or subtracting dots and strokes. As for the pronunciation of zhengyue (正月) as zheng, and zhaomu (昭穆) as shao, one need not observe taboos for different dynasties; why then do those who compile sounds and meanings taboo them? (In the Book of Documents, 'on the first day of the first month [zhengyue 正月]', it is pronounced zheng 政, and also pronounced zheng 征. In the Odes, zhengyue is read according to its standard character. The world says that the First Emperor of Qin tabooed Zheng 政, so those who avoided it pronounced it zheng 征. Now we are far removed from Qin; we do not even hear of avoiding the zheng of zhengshi 政事 [state affairs], so why avoid it only in zhengyue? In the Book of Rites, the Royal Regulations, 'three zhao and three mu' has no clear phonological note. In the Rites of Zhou, 'calling upon the zhao and mu to advance', the note says: shang zhao fan. This was because of avoiding the taboo name Zhao of Emperor Wen of Jin, so it was pronounced shao 韶. However, zhaomu precisely takes the meaning of bright and clear [zhaoming 昭明]. Zhu Hui'an's commentaries on the classics all use the standard pronunciation. These two pronunciations of zheng and shao are both Lu Deming's Jingdian shiwen; they are errors of not knowing to revise and delete them. That they could observe taboos for bad rulers who have already perished, while for the Sage Confucius, the teacher of myriad generations, they do not know to observe taboos—the Confucian masters past and present cannot be without shame). 'Fo' (佛) and 'Chan' (禪) come from Buddhist canons and are not collected, while 'Seng' (僧) and 'Fan' (梵) also come from Buddhist canons yet are placed in the dictionaries and rhyme books; why is this? (In Sanskrit, 'Buddha', here translated as 'Awakened One'; in Sanskrit, 'Dhyāna', here translated as 'quiet contemplation'; these two characters are not collected. In Sanskrit, 'Saṅgha', here translated as 'assembly'; 'Brahma' means pure and free from desire; these two characters are now collected in rhyme books). From this we know that Sun Qiang's additions to the Yupian (Gu Yewang of the Liang compiled the Yupian; Sun Qiang of the Tang added characters, mostly taking vulgar writings), Sun Mian's additions to the Guangyun (Sun Mian of the Tang compiled the Guangyun, which largely collected vulgar characters), and the compilation of the Jiyun by Ding Du of our dynasty all indiscriminately cite vulgar writings and intermix them with orthodox forms, causing the ancient meanings of the classics and histories to be obscured and unclear. This is a great impediment of character dictionaries; does it not await dedicated scholars to correct it? Now observing Ouyang's preface to the Yunzong, which states that he was able to correct the discrepancies of the five regions such that Confucian scholars could not confound him, in my private thoughts I am deeply moved; it is merely that I have had no means to see this book (seen in the Collected Works of Retired Scholar Liuyi). Senior Academician Huang said: 'People of the Wei and Jin dynasties adopted character forms that were easy to execute skillfully, without questioning the roots and sources of characters. Characters such as qiao 𭫥 and zhi [匕/且] do not form characters at all.' In the Hanmo zhi (Record of Brush and Ink) of Sileng, it is said: 'The defects of written characters have never been worse than in our dynasty.' How supreme! Such profound insight of sagely learning! (Gaozong's mausoleum is titled Yongsi).