from: ROBERT KRITZER, Rebirth-and-Causation-in-the-Yogacara-Abhidharma,WIEN 1999
3. The Authorship of the Yogācārabhūmi
The other major "issue raised by the biographical sources is the authorship of the Yogācārabhūmi. Again, the traditional accounts differ. According to Paramartha, Maitreya, at Asaṅga's request, came down from Tuṣita heaven to preach the Shih ch’i ti ching (Chinese text unreadable = Yogācārabhūmi) at night, while during the day, Asaṅga, who was the only one who could approach Maitreya, explained Maitreya's speech to the rest of the audience (T. 2049: 188c14-20; Takakusu 1904: 274-275). According to Hsūan-tsang, on the other hand, Asaṅga ascended to Maitreya's palace at night and received from him the Yogācārabhūmi, the Mahāyānasūtrālamkāra, Madhyāntavibhāga [Ed.: I am sorry. The Chinese text will remain missing for the rest of this excerpt], and other texts, which he later explained to the masses (T. 2087: 896b21-24; Bea1 1983: 226).18
The Tibetan tradition is more willing to attribute the actual authorship of the Yogācārabhūmi to Asaṅga. Bu-ston first tells the story of Asaṅga's ascent to Tuṣita heaven, where Maitreya expounds the Prajñāpāramitā, the Yogācārabhūmi, and other Mahāyānasūtras. In order to elucidate these scriptures, Maitreya composes the five works of Maitreya,19 which Asaṅga brings down to earth. However, Bu-ston shortly thereafter states that Asaṅga composed "his great treatise in five divisions" (in other words, the Yogācārabhūmi) after his return to earth (Obermiller 1986: 139-140). In Tāranātha's version of the same story, Asaṅga, after arriving in Tuṣita heaven, "listened to the Mahāyāna doctrine in its entirety from Ajitanātha [Maitreyanatha] and learnt the real significance of the whole collection of sūtra-s. Then he listened to the 'Five Works of Maitreya'" (Chimpa and Chattopadhyaya 1970: 159). Unlike Bu-ston, Tāranātha does not mention the Yogācārabhūmi as one of the texts preached to Asaṅga by Maitreya. However, Tarānātha does mention the Yogācārabhūmi as one of the treatises written by Asaṅga some time after his return from Tuṣita heaven (Chimpa and Chattopadhyaya 1970: 160).
Both Bu-ston and Tarānātha, in their sections on the life of Vasubandhu, relate a story according to which Vasubandhu, hearing or reading a work or some works of Asaṅga, criticized Asaṅga:
Alas, Asaṅga, residing in the forest, Has practiced meditation for 12 years. Without having attained anything by his meditation, He has founded a system, so difficult and burdensome, That it can be carried only by an elephant!20
Although Bu-ston does not mention the Yogācārabhūmi by name, referring only to "numerous treatises" (Obermiller 1986: 143), Tarānātha says specifically that Vasubandhu "read the Five Bhūmi-s, the work of ārya Asaṅga" (Chimpa and Chattopadhyaya 1970: 168). Incidentally, Wayman, perhaps relying on Tarānātha, assumes that Bu- ston's version is, in fact, referring to the Yogācārabhūmi (1961: 40). Thus, Tarānātha unequivocally attributes the authorship of the Yogācārabhūmi to Asaṅga, while Bu-ston seems to contradict himself by first saying that Maitreya expounded it to Asaṅga and then stating that Asaṅga composed it himself.
In addition to these biographical accounts, we also have traditional attributions in the canonical collections that include the Yogācārabhūmi. According to the Taisho edition of the Yogācārabhūmi, the author is Maitreya (T. 1579: 279a5; Demieville 1978: 134), while according to the index to the Peking edition of the Bstan 'gyur, the author is Asaṅga.
Before I move on, I must again take issue with a statement of Wayman's. According to him, "There is no disagreement between Chinese and Tibetan tradition as to the common provenance of the Yogācārabhūmi, the Abhidharma-samuccaya, and the Mahāyāna-saṃgraha" (1961: 40). This is just not true. Neither Paramartha nor Hsūan-tsang mentions the Abhidharma-samuccaya or Mahāyana-saṃgraha, so we must rely on the canonical attributions for the Chinese tradition regarding these works: both texts are attributed to Asaṅga, with no mention of Maitreya (Demieville 1978: 136). The Tibetan tradition, which unlike the Chinese accepts Asaṅga as the author of the Yogācārabhūmi, agrees with the Chinese in accepting him as the author, of the Abhidharma-samuccaya and the Mahāyāna-saṃgraha. Therefore, although, as Wayman claims, the Tibetan tradition accepts the common authorship of the three texts, the Chinese tradition certainly does not. In his later work, Wayman in fact mentions the attribution of the Yogācārabhūmi to Maitreya in the Chinese canon, and he speculates that "Hsūan-tsang, believing that Maitreya had inspired Asaṅga, may have entered the name Maitreya in order to get a hearing for the work, to ensure its study in his country where the Buddhists mainly followed Sūtras, such [as] the Pure Land ones, the Avataṃsaka, the Lotus Sūtra, and so on" (Wayman 1989: 202). Wayman's reasoning seems to be that, since "Demieville has sufficiently demonstrated the absurdity of regarding Maitreya as a 'historical personage,",21 we can safely assume that the story of Asaṅga's having received the Yogācārabhūmi from Maitreya is totally fictional and that Asaṅga wrote the treatise himself. I agree with Wayman regarding the question of Maitreya, but I think that we must be more open-minded regarding the Chinese tradition that the author of the Yogācārabhūmi is not the same as the author of the Abhidharma-samuccaya and the Mahāyāna-saṃgraha.
Modern scholars have a variety of opinions regarding the authorship of the Yogācārabhūmi. Some, for example, Willis (1982) and Rahula (1980: xi), simply accept the traditional attribution to Asaṅga, do not acknowledge that a difference of opinion exists, and therefore make no attempt to justify their acceptance. Others recognize that there is a difference of opinion but do not seriously engage the issue. For example, Wayman states that "it is untenable to hold that certain sections of the Yogācārabhūmi were written by different persons" and supports this statement by saying, "My reading over the years in the Yogācārabhūmi has led me to conclude that it is the same person writing throughout" (1989: 201-202). At the other extreme is Nakamura, who accepts the Chinese attribution to Maitreya or Maitreyanatha and, at the end of two pages of description of the contents of the Yogācārabhūmi, merely mentions that "there is an opinion that, as the contents of the Yogācārabhūmi are substantially different to a great extent from other works ascribed to Maitreya, its author may be different from Maitreyanatha" (Nakamura 1987: 258).
Schmithausen, on the other hand, believes that the Yogācārabhūmi is a compilation (1987: 13) and, responding to Hakamaya, is unwilling to grant much credence to traditional attributions: "I for one prefer to confine myself to the statement that it is possible (perhaps even probable) that Asaṅga compiled [the] Y[ogācārabhūmi] or, as Hakamaya himself puts it in a later article, somehow participated in its compilation (i.e. that 'tradition' may, in this case, in fact contain a kernel of historical truth in our sense). But this possibility still needs verification, and for the time being I for one cannot exclude the opposite possibility" (1987: 185). In fact, in his most recent work on the Yogācārabhūmi, Alayavijñāna, Schmithausen identifies so many passages from various portions of the text that differ so substantially from passages in other portions that, to my mind, he has effectively proved his theory that the Yogācārabhūmi is a heterogeneous compilation that cannot be attributed to Asaṅga.22
The question of the authorship of the Yogācārabhūmi is obviously pertinent to the study of the Abhidharma-samuccaya, which, to the best of my knowledge, has hitherto been universally accepted (except, perhaps, by Schmithausen) as the work of Asaṅga. That the Yogācāra- bhūmi chronologically precedes the Abhidharma-samuccaya is also generally, but not universally, conceded; Willis (1982: 10) and Warder (1991: 441), who accept Asaṅga's authorship of the Yogācārabhūmi, both state that the Abhidharma-samuccaya is earlier without, as Schmithausen points out (1987: 262 n. 100), providing any evidence to support their positions.23
It is indeed difficult to imagine that the same person would have first written the Abhidharma-samuccaya, a model of brevity and organization, and afterwards produced the text to which the Tibetan tradition refers when it criticizes "Asaṅga's system" as being "so difficult and burdensome that it can be carried only by an elephant" (Obermiller 1986: 143; see above).
Perhaps a better idea of the relationship between the two texts is suggested by Rahula in the introduction to his translation of the Abhidharma-samuccaya:
"Ce que l’Abhidhammapiṭaka en pali est pour les Theravādin, le Jñānaprasthāna pour les Sarvāstivāda, le Mahāprajñāpāramitāsāstra pour les Madhyamika, le Yogācāra- bhūmisāstra l'est pour les Yogācirin” (Rahula 1980: xiv).
Although I disagree with Rahula about the authorship of the Yogācārabhūmi, I think that his characterization of the text as Yogācāra abhidharma, which is similar to Wayman's (1961: 45; see above), is accurate. It is, moreover, significant that Rahula compares the Yogācārabhūmi with texts belonging to the Theravādin and Sarvāstivādin Abhidharmapitakas proper, rather than with the later treatises.
Rahula sets aside his analogy regarding the Yogācārabhūmi and goes on to compare the question-and-answer format of the Abhidharma-samuccaya with that of certain PaIi Abhidhammapiṭaka texts, such as the Dhammasaṇgaṇi, Vibhaṅga, and Dhātukathā. But it must be stressed that if the Yogācārabhūmi is compared with the Abhidharmapitakas, the Abhidharma-samuccaya is more properly analogous to the manuals such as the Abhidharmāmṛta, which also consists largely of questions and answers similar to those found in the Abhidharma-samuccaya. In any case, the analogy should not be taken too far, since the Abhidharma-samuccaya is a much more sophisticated text than the early Sarvāstivāda manuals, let alone the works of the Pali abhidhamma.