Source: Radich 2014

Radich has studied the four chapters of Baogui 寶貴's synoptic version of the Suvarṇaprabhāsottama-sūtra that are ascribed to Paramārtha (3, "Sanshen fenbie pin" 三身分別品; 5, "Yezhang mie pin" 業障滅品; 6, "Tuoluoni zuijing di pin" 陀羅尼最淨地品; 9, "Yikong manyuan pin" 依空滿願品) . The external evidence for ascription of these chapters to Paramārtha is extraordinarily strong. Nonetheless, copious internal evidence shows a complex pattern of relation to Chinese sources, which varies from chapter to chapter. 1. The "Yezhang" chapter derives in almost its entirety from the *Karmāvaraṇapratiprasrabdhi-sūtra. Significant and often extensive verbatim matches can be found with both *Samghabhara's 僧伽婆羅 earlier Pusa zang jing 菩薩藏經 T1491; and also with Jñānagupta’s Dasheng san ju chanhui jing 大乘三聚懺悔經 T1493, which otensibly should be later. 2. The "Tuoluoni" (Dhāraṇī) chapter shows extensive debts to the *Bodhisatvadaśabhūmika-sūtra in both Kumārajīva's 鳩摩羅什 translation, the Zhuangyan putixin jing 莊嚴菩提心經 T307; and also in that of *Kivkara 吉迦夜 (*Kiṅkara? *Kiṃkārya?) Pusa shi di jing 菩薩十地經 T308. In addition, however, two extended passages and one verse show relations to the Saṃdhinirmocana-sūtra, again in two versions: that of Paramārtha himself, and that of Bodhiruci 菩提流支; and one of the dhāraṇīs central to the chapter shows debts to *Samghabhara's Mahāmāyūrī 孔雀王咒經 T984. 3. The "Yikong" chapter also contains clear debts to Paramārtha's Saṃdhinirmocana, and small but significant relations with Guṇabhadra's *Mahābherīhāraka-sūtra 大法鼓經 T270, Paramārtha's Wushangyi jing 無上依經 T669, and verses from a handful of earlier Chinese sources. 4. The "Sanshen" (Trikāya) chapter, alone among these chapters, has extended direct parallels in an Indic text probably unknown in China in Paramārtha's time, the *Kāyatrayāvatāramukha (preserved in Tibetan), as has been shown previous by Hamano Tetsunori 浜野哲敬. However, even this chapter shows little evidence of Paramārtha's typical translation style, and some very close relations to Chinese materials in Paramārtha's close context. Radich then shows that further complications can also be traced in these chapters and related portions of the Suvarṇaprabhāsottama. 1. The "Lifespan" chapter 壽量品 is ascribed in Baogui to Jñānagupta. However, the manuscript supposed to represent Paramārtha's fuller, original translation of the Suvarṇaprabhāsottama, dated 768 (Jingo-keiun 神護景雲 2), which is preserved in the Shōgozō 聖語蔵, contains passages which differ significantly from those found in Baogui. (Curiously, these passages coincide almost exactly with the passage shown by Suzuki Takayasu 鈴木隆泰 to have been interpolated into the Suvarṇaprabhāsottama--already in India; the passages are included in Skt--from the Mahāmegha-sūtra. By contrast, the "Paramārtha" chapters do not vary significantly between Baogui and the Shōgozō manuscript.) The Shōgozō version of the text is considerably simpler, suggesting that the Baogui/Jñānagupta version is a revision of the Shōgozō version; this revision, moreover, must have been made with reference to a Skt. text. This suggests that some Chinese version of the passage pre-existed Jñānagupta's work, and Jñānagupta may have worked by revising earlier Chinese texts. 2. Even setting aside all the passages in which the above debts to prior Chinese texts can be traced, the four "Paramārtha" chapters exhibit almost none of the translation idiom typical of Paramārtha texts. By contrast, they do contain very heavy doses of phrases very frequently found in texts of the Sui translators. Radich suggests that this overall pattern of evidence may be too complex for us to hope to arrive at a definitive understanding of the provenance of these chapters, but with that proviso, suggests that something like the following hypothesis may be the most plausible scenario: The four chapters were in some sense "composed" in China on the basis of prior Chinese texts (and, in the case of the "Sanshen" chapter, in part based on the Indic *Kāyatrayāvatāramukha), and quite possibly further revised between the time of Paramārtha and the Sui. In this light, it is possible that the verbatim parallels between the "Yezhang" chapter and Jñānagupta’s T1493 might indicate that our extant "Paramārtha" "Yezhang" chapter may have been composed in part by reference to Jñānagupta, rather than the other way around, as traditional chronology would lead us to assume.

In a follow-up study, Radich examines evidence in Tib II, the supposedly independent Tibetan witness (i.e. supposedly translated directly from Skt.) to these chapters of the text, which suggests the possibility that those chapters might actually have been translated from Chinese originals. See Radich, Michael. “Tibetan Evidence for the Sources of Chapters of the Synoptic Suvarṇaprabhāsottama-sūtra T664 Ascribed to Paramārtha”. Buddhist Studies Review 32, no. 2 (2015): 245-270.

Edit