開元釋教錄 10.4.7

10.4.7

求。或憑時代。或寄參譯。各紀一隅務存所
見。斯並當時稽古。識量修明而綴撰筆削。
不至詳審者。非為才不足而智不周也（云
云）上集群目取訊僧傳等文勘閱詳定。便參
祐房等錄祐錄徵據文義可觀。然大小儡
同。三藏糅雜。抄集參正。傳記亂經。考括始
終莫能通決。房錄後出該瞻前聞。然三寶
共部偽真淆亂。自餘諸錄胡可勝言。今余
所撰望革前弊（撰錄者曰。宣公所撰類例明審實有可觀作者之風見於茲矣。然少有差
雜未能盡善。述作之事誠謂難哉。今略敘數科以詳厥誤。非欲指陳臧否。實惟甄異是非。只如人代存亡錄中
新小品等六經並云大品之同本者不然。義理雖通據會全異。而言同本。一誤。菩薩淨行經與寶髻菩薩經俱云
竺法護譯。此是一經兩名。存其二本。二誤。弘道廣顯三昧經與阿耨達龍王經。亦是一經兩名。二本俱存三誤。須彌
藏經二卷此是高齊耶舍所翻。前後重上。四誤。攝大乘論真諦譯者有十二卷本有十五成者二本俱存。五誤。般
若鈔經大乘錄中及集傳內二處俱載。六誤。又如分乘入藏錄中。深密解脫經前後重上。前與菩薩處胎經同帙。後
與解深密經同帙。七誤。以舊十輪經為單本。新十輪經為重譯。八誤。以起世經為單本。樓炭經為重譯。九誤。此類
非一不能備舉。如舉要轉讀錄中信力入印法門經。雖是華嚴支類。大部中無同於度世漸備等經。攝歸大部不為
轉讀之數一不然也。諸部般若唯舉大品一經。放光等九部云重沓罕尋舉前以統大義斯盡玉華後譯大般若者
明佛一化十有六會得存供養難用常行今謂不然豈可以凡愚淺智而堰截法海乎。人性不同所樂各異。豈以
自情好略令他同己見耶。般若大經轉讀極眾。佛記弘闡在東北方而言難用常行竊為未可。又小品大品據本
全殊一廢一興二不然也。如須彌藏經等雖是大集別分大部中無攝而不行理為未當。又須彌藏經前則攝歸大
集。後即別舉流行。前後差殊三不然也。如樓炭經等攝入長含起世一經別令轉讀起世之與樓炭梵本何殊。一廢
一興四不然也。如安般守意與大安般無殊。此乃何緣二俱轉讀五不然也。如十八部論與部異執梵本不殊此復
何緣俱令轉讀。六不然也。又云錄外有外道金七十論破外道涅槃論破外道四宗論未暇故闕者。如金七十
論外道所撰非是佛法除之可然。涅槃四宗同彼刪削將為未可。此是內論破於外宗。一例刪除。七不然也。又如人
代存亡錄及舉要轉讀錄大乘錄中三藏備具入藏錄中大乘無律。豈可前後俱有。中間獨無。自為矛盾。八不然也。
如上所列非無乖舛而云革弊。或所未然差錯極多。卒難陳委。明達之輩幸自詳焉）。
續大唐內典錄一卷（開元庚午歲西崇福寺沙門智昇撰）
歷代眾經傳譯所從錄（從麟德元年甲子至開元十八年庚午前錄未載今
故續之）。
大唐古今譯經圖紀四卷（大慈恩寺翻經沙門靖邁撰）
大慈恩寺翻經堂內。壁畫古今翻譯圖變。
靖邁因撰題之于壁。但略費長房錄。翻經
之者紀之。餘撰集者不錄。逮至皇朝總成
四卷。房所錯者此亦同然。更欲廣陳恐繁
故止。
續古今譯經圖紀一卷（開元庚午歲西崇福寺沙門智昇撰）
從奘法師後至輸波迦羅。前紀未載今故
續之。
大周刊定眾經目錄十五卷（天后天冊萬歲元年乙未勅東都佛授記寺
沙門明佺等撰定）
大小乘經律論及賢聖集傳（合三千六百一十六部八千六百四十
一卷其見定入藏流行部卷不在此數）
大乘單譯經目卷第一（二百八十三部五百二十八卷）
大乘重譯經目之一卷第二（一百七十八部一千三百九十四
卷）
大乘重譯經目之二卷第三（一百六十八部三百三十八卷）
大乘重譯經目之三卷第四（一百八十部四百二十一卷）
大乘重譯經目之四卷第五（一百六十八部三百六十五卷）
大乘律大乘論目卷第六（一百六十二部七百一十六卷）
小乘單譯經目卷第七（一百二十八部三百一十卷）

seeking. Some relied on chronological eras; some depended on participating in translations. Each recorded a single corner, striving to preserve what they had observed. These were all investigations of antiquity in their own time, composed and edited by people of refined understanding and intellect. That they did not achieve thorough precision was not because their talent was insufficient or their wisdom incomplete […]’ The above brings together the titles of catalogues, consulting texts such as the Biographies of Eminent Monks to verify, collate, and determine them in detail. It then drew upon the catalogues of Sengyou, Fei Changfang, and others. The citations and evidence in Sengyou's catalogue have commendable phrasing and meaning, yet Mahāyāna and Hīnayāna are lumped together, the three piṭakas are intermingled, extracted compilations are mixed with canonical texts, and biographical records are confused with sūtras; when examining them from beginning to end, none can resolve these ambiguities. Fei Changfang's catalogue appeared later and was comprehensive of earlier lore, yet the three treasures share the same section, and the spurious and the authentic are commingled. As for the rest of the various catalogues, how could they all be mentioned? What I have now compiled aspires to reform earlier shortcomings. (The compiler of this catalogue [Zhisheng] says: What Duke Xuan compiled is clear and rigorous in its classification; it is truly worth examining, and the style of a master author is evident here. However, with a few inconsistencies and confusions, it is not yet entirely perfect. Authorship is indeed a difficult undertaking! I now briefly present several points to examine its errors, not wishing to pronounce on merits and demerits, but truly only to distinguish truth from falsehood. Just as in the Catalogue of Historical Translations, Extant and Lost, with Translators and Dynasties, the six sūtras such as the newly translated Aṣṭasāhasrikā are all said to be the same text as the Pañcaviṃśatisāhasrikā; this is not so. Although the doctrine is coherent, the assemblies are entirely different; to call them the same text is the first error. The Bodhisattva-caryā-viśuddhi-sūtra and the Ratnacūḍa-bodhisattva-sūtra are both said to have been translated by Dharmarakṣa; this is one sūtra with two titles, yet he kept both texts; this is the second error. The *Mārgavipula-prakaṭana-samādhi-sūtra and the Anavatapta-nāgarāja-sūtra are also one sūtra with two titles, yet both texts are kept; this is the third error. The Sumerugarbha-sūtra in two fascicles was translated by Narendrayaśas of the Northern Qi, yet it is entered twice, earlier and later; this is the fourth error. For the Mahāyānasaṃgraha translated by Paramārtha, there is a twelve-fascicle text and a fifteen-fascicle completion, yet both texts are kept; this is the fifth error. The Prajñā-cāo-jing is recorded in two places, in both the Mahāyāna catalogue and the compilations and biographies; this is the sixth error. Furthermore, as in the Catalogue of Historical Sūtras by Vehicle Entered into the Canon, the Sandhinirmocana-sūtra is entered twice, earlier and later: first in the same case with the Bodhisattva-garbhastha-sūtra, and later in the same case with the Jie shenmi jing; this is the seventh error. Taking the old Daśacakra-sūtra as a sole translation and the new Daśacakra-sūtra as a retranslation is the eighth error. Taking the Qishi jing as a sole translation and the Loutan jing as a retranslation is the ninth error. Such cases are not singular, and cannot all be enumerated. As in the Catalogue of Essential Readings for Chanting, although the Xinli ruyin famen jing is an extracted branch of the Avataṃsaka, it is not present in the large work, being similar to sūtras such as Dushi jianbei jing; subsuming it back into the large work and not counting it for chanting is the first improper point. Among the various Prajñā texts, only the single sūtra of the Pañcaviṃśatisāhasrikā is cited, claiming that the nine works such as the Fangguang jing are repetitious and rarely investigated, and that citing the former subsumes them, so that the broad meaning is exhausted; regarding the Mahāprajñāpāramitā-sūtra later translated at Yuhua Palace, it explains that it clarifies the Buddha's lifelong teaching in sixteen assemblies and may be preserved for veneration, but is difficult to apply for regular recitation. I now say this is not so. How could one dam up the sea of Dharma with mundane foolishness and shallow intellect? Human natures are not the same, and what people delight in differs; how could one impose one's own preference for brevity onto others to match one's own view? The great Prajñā sūtra is chanted by very many people, and the Buddha prophesied that its broad dissemination would be in the northeast; to say that it is difficult to apply for regular practice is, in my humble opinion, unacceptable. Moreover, the Aṣṭasāhasrikā and the Pañcaviṃśatisāhasrikā are completely different based on their Indian texts; to discard one and promote the other is the second improper point. Works such as the Sumerugarbha-sūtra, although an extracted section of the Mahāsaṃnipāta, are not found in the large work; subsuming it and not allowing it to circulate is unreasonable. Furthermore, the Sumerugarbha-sūtra was earlier subsumed into the Mahāsaṃnipāta, and later cited separately for circulation; this discrepancy between earlier and later is the third improper point. Works such as the Loutan jing were subsumed into the Dīrghāgama, while the Qishi jing was individually commanded to be chanted; how does the Sanskrit text of the Qishi jing differ from that of the Loutan jing? Discarding one and promoting the other is the fourth improper point. The Anban shouyi jing does not differ from the Da anban; on what basis are both commanded to be chanted? This is the fifth improper point. The Samayabhedoparacanacakra-śāstra does not differ in Sanskrit text from the Buyizhi lun; on what basis are both commanded to be chanted? This is the sixth improper point. Furthermore, it is said that outside the catalogue are the non-Buddhist Sāṃkhya-kārikā-bhāṣya (*Suvarṇasaptati), the Treatise Refuting the Non-Buddhist Nirvāṇa, and the Treatise Refuting the Four Non-Buddhist Tenets, which were omitted due to lack of leisure. A work such as the *Suvarṇasaptati was composed by non-Buddhists and is not the Buddha's Dharma; removing it is acceptable. But to excise the Nirvāṇa and Four Tenets texts alongside it is unacceptable: these are Buddhist treatises refuting non-Buddhist traditions, and deleting them uniformly is the seventh improper point. Furthermore, in the Catalogue of Historical Translations, Extant and Lost, with Translators and Dynasties and the Catalogue of Essential Readings for Chanting, the three piṭakas are fully present in the Mahāyāna catalogue, yet in the Catalogue of Works Entered into the Canon there is no Mahāyāna vinaya. How could it be present before and after, yet uniquely absent in the middle, creating a self-contradiction? This is the eighth improper point. What has been listed above is not without errors, and as for claiming to reform shortcomings, perhaps that is not the case; the errors are extremely numerous and difficult to recount in full quickly. May the discerning examine them for themselves.)
Continuation of the Great Tang Catalogue of Inner Canonical Texts, one fascicle (compiled in the gengwu year of Kaiyuan by the śramaṇa Zhisheng of the West Chongfu Temple)
Catalogue of the Provenance of Canonical Translations across the Dynasties (from the jiazi year, the first year of Linde, to the gengwu year, the eighteenth year of Kaiyuan; not recorded in the earlier catalogue, now continued).
Pictorial Records of Ancient and Modern Sūtra Translations of the Great Tang, four fascicles (compiled by the translation śramaṇa Jingmai of Da Cien Temple)
Inside the translation hall of Da Cien Temple, there are mural illustrations and tableaux of ancient and modern translations.
Jingmai thereupon wrote inscriptions for them on the wall. He merely abridged Fei Changfang's catalogue, recording those who translated sūtras and omitting other authors and compilers. Down to the present imperial dynasty, it formed four fascicles in total. Where Changfang erred, this catalogue follows suit. Wishing to elaborate further, I feared it would be too verbose and therefore stopped.
Continuation of the Pictorial Records of Ancient and Modern Sūtra Translations, one fascicle (compiled in the gengwu year of Kaiyuan by the śramaṇa Zhisheng of the West Chongfu Temple)
From Master Xuanzang down to Śubhakarasiṃha; not recorded in the earlier record, now continued.
Great Zhou Catalogue of Canonical Scriptures Authorized upon Revision, fifteen fascicles (compiled and fixed by imperial order in the yiwei year, the first year of Tiancedanwan, by the śramaṇa Mingquan and others of the Fo Shouji Temple in the Eastern Capital)
Mahāyāna and Hīnayāna sūtras, vinayas, and śāstras, and compilations and biographies of sages (totaling 3,616 works in 8,641 fascicles; the works and fascicles currently established as entering the canon and circulating are not included in this count)
Catalogue of Mahāyāna Sole Translations of Sūtras, Fascicle 1 (283 works, 528 fascicles)
Catalogue of Mahāyāna Retranslations of Sūtras, Part 1, Fascicle 2 (178 works, 1,394 fascicles)
Catalogue of Mahāyāna Retranslations of Sūtras, Part 2, Fascicle 3 (168 works, 338 fascicles)
Catalogue of Mahāyāna Retranslations of Sūtras, Part 3, Fascicle 4 (180 works, 421 fascicles)
Catalogue of Mahāyāna Retranslations of Sūtras, Part 4, Fascicle 5 (168 works, 365 fascicles)
Catalogue of Mahāyāna Vinayas and Mahāyāna Śāstras, Fascicle 6 (162 works, 716 fascicles)
Catalogue of Hīnayāna Sole Translations of Sūtras, Fascicle 7 (128 works, 310 fascicles)