成唯識論了義燈 7.55

7.55

《論》：「於二取隨眠」等。問：《華嚴》第十解十住菩薩中，云「除滅諸煩惱，永盡無有餘」等，云何此言未伏二取？答：因邪教、邪師所起不共無明伴煩惱類盡，非盡一切自分別生及俱生煩惱。又解：說佛法功力能滅煩惱，非皆已伏。又解：自分別伏，邪師等起不能伏滅。《緣起經》文，二互隨會。然《疏》下結，故此位中自分別煩惱未盡，故前解正。准發菩提心，緣力發者，名不堅固；因力發者，名為堅固。起惑亦爾，邪教起者易除，自力起者難伏。又解：經據少分有能伏義，論依全分云不能伏。初、後解正，《疏》文似脫，更勘正本。

The Treatise says: "Regarding the latencies of the twofold grasping..." Question: In the tenth chapter of the Avataṃsaka-sūtra, when explaining the bodhisattvas of the ten abodes, it says: "They eradicate all afflictions, making them forever exhausted without remainder," and so on. Why then does it say here that they have not yet subdued the twofold grasping? Answer: It refers to the eradication of the accompanying afflictions associated with the unshared ignorance that arises due to perverse teachings and perverse masters; it does not mean the exhaustion of all self-discriminated and innate afflictions. Another explanation: It speaks of the efficacy of the Buddha-Dharma being capable of eradicating afflictions, not that they are all already subdued. Another explanation: Their own discriminated ones are subdued, while those arisen from perverse teachers and the like cannot be subdued or eradicated. The text of the Pratītyasamutpāda-sūtra can be reconciled either way. However, since the commentary later concludes that in this stage one's own discriminated afflictions are not yet exhausted, the former explanation is correct. By analogy with generating bodhicitta: that which is generated through the power of conditions is called unstable, while that which is generated through the power of causes is called stable. The arising of delusions is also like this: those arising from perverse teachings are easy to eliminate, while those arising from one's own power are difficult to subdue. Another explanation: The sūtra relies on a partial sense of there being the ability to subdue, while the treatise relies on the complete sense of being unable to subdue. The first and last explanations are correct; the text of the commentary seems to have an omission, so one should check against an accurate exemplar.