Thesis / main argument
Apple reconstructs Atiśa’s (982–1054) Madhyamaka thought and its transmission to Tibet through recently recovered Kadampa manuscripts (the Collected Works of the Kadampas, published 2006–2015). He demonstrates that Atiśa taught a “pure” (undifferentiated) Madhyamaka lineage descending from Nāgārjuna through Candrakīrti, mediated by his Indian teachers Vidyākokila, Avadhūtipa, and Bodhibhadra. This lineage was contemplative rather than scholastic, faith-based rather than epistemological, and significantly different from both the Yogācāra-Madhyamaka synthesis of Śāntarakṣita/Kamalaśīla that Tibetans had received during the Imperial Era and the later Gelukpa systematisation of Tsongkhapa.
Key claims
- Atiśa’s Madhyamaka was structured around the two realities (satyadvaya) as articulated in his Satyadvayāvatāra (28 verses) and Madhyamakopadeśa — conventional reality is a false projection of ignorance; ultimate reality is one, undifferentiated, and beyond conceptuality (SDA vv. 1–9)
- Atiśa rejected valid cognition (pramāṇa) as a means to realise ultimate reality — direct perception and inference are useful only for refuting opponents, not for realising emptiness (SDA vv. 10–13)
- A buddha has no continuum of wisdom (jñāna) whatsoever — buddhahood is a nondual fusion with the dharmadhātu bereft of all mind and mental factors. This was Atiśa’s most controversial position, vehemently opposed by his Yogācāra contemporaries (Jñānaśrīmitra, Ratnākaraśānti) and not fully understood even by some of his own Tibetan disciples
- Conventional realities are “mere appearances” (snang ba tsam) — dependent designations without any real basis (avastuka), unlike the Yogācāra-Madhyamaka position which grants conventional reality a real basis in mental elements. Atiśa explicitly classified the Tibetans’ pebbles-and-sticks illusion analogy as applicable only to “Mind Only and below”
- Atiśa did not make propositional assertions (pratijñā), did not concede to the principle of common establishment (ubhayasiddhatva), and used only consequences (prasaṅga) and other-acknowledged inferences — following Candrakīrti against Śāntarakṣita/Kamalaśīla
- Atiśa described his system as “Great Madhyamaka” (dbu ma chen po) — an undifferentiated tradition. He did not use the Prāsaṅgika/Svātantrika classification, nor did the earliest layer of Kadampa commentaries. These terms are later Tibetan neologisms
- The early Kadampa commentaries show Atiśa synthesising Bhāviveka and Candrakīrti under a unified Madhyamaka — using Bhāviveka pedagogically and Candrakīrti for advanced instruction
- Atiśa taught Madhyamaka publicly through the works of Bhāviveka (Tarkajvālā) while giving his Candrakīrti-lineage instructions (Madhyamakopadeśa, Satyadvayāvatāra) privately to close disciples
- The early Kadampa community at Radreng was contemplative in nature — the founding members were titled “yogis,” and the Madhyamaka taught there emphasised meditation over debate
- The recovered manuscripts show Atiśa’s Madhyamaka is “totally opposite” from post-16th-century Gelukpa presentations — directly challenging the assumption that Gelukpa scholasticism represents the Kadampa inheritance
Methodology
Historical-philological: translation and analysis of recently recovered Kadampa manuscripts (unknown since the 17th century, confiscated by the Fifth Dalai Lama), supplemented by biographical sources and comparative analysis with Indian Buddhist contemporaries. Apple reads Atiśa’s thought through Atiśa’s own words and those of his immediate followers rather than through later Tibetan systematisations.
Notable quote
“These texts illustrate how unique the Gelukpa presentation of Madhyamaka is from the standpoint of Atiśa and the majority of early Kadampa thinkers.”
Tenpa’s critical notes
This source is directly relevant to the paper’s thesis. Atiśa operates fully within the Mahāyāna hermeneutical framework — Two Truths structure, graduated pedagogy, Candrakīrti lineage — while arriving at a position quite different from Tsongkhapa’s later Gelukpa systematisation. This demonstrates the paper’s core point: the framework generates diverse, sophisticated interpretations rather than a single dogmatic reading. The fact that Atiśa’s “pure Madhyamaka” predates and differs from later Gelukpa Prāsaṅgika adds important historical depth.
However, as Tenpa anticipated, there is heavy Gelukpa influence — not in Atiśa’s own views, but in the framing. The Gelukpa tradition claims Kadampa lineage descent, yet Apple shows the recovered manuscripts reveal a Madhyamaka significantly different from Gelukpa scholasticism. This tension — between what Atiśa actually taught and what Tsongkhapa later systematised — is itself evidence for the paper’s argument about how the hermeneutical framework evolves productively.
Atiśa’s rejection of pramāṇa for realising emptiness places him closer to the Karmapa and Gorampa than to Tsongkhapa on this point. His undifferentiated Madhyamaka and refusal of the Svātantrika-Prāsaṅgika distinction align with Mipham’s, the Karmapa’s, and Shakya Chokden’s dissolution of this hierarchy — though from a historically earlier vantage.
The “no wisdom continuum for buddhas” position is striking: it was controversial in the 11th century and continues to divide interpreters. Tsongkhapa would reject this position (accepting a buddha’s omniscient wisdom), while Atiśa’s Kadampas insisted on it based on Candrakīrti. This is a significant data point for framework-internal diversity.
Connections
- Directly relevant to Tsongkhapa: Tsongkhapa claimed Kadampa lineage but systematised Madhyamaka differently — incorporating pramāṇa into the realisation of emptiness, which Atiśa explicitly rejected
- Complements shantarakshita-madhyamakalankara: Atiśa’s Madhyamaka is presented as a direct alternative to Śāntarakṣita/Kamalaśīla’s Yogācāra-Madhyamaka that the Tibetans had inherited
- Connects to karmapa-feast-fortunate-1578: the Karmapa’s rejection of pramāṇa for ultimate realisation echoes Atiśa; both emphasise contemplative over scholastic Madhyamaka
- Connects to gorampa-distinguishing-views-1469: Gorampa’s charge that Tsongkhapa over-systematises parallels Apple’s evidence that early Kadampa Madhyamaka was simpler and more contemplative
- Directly opposes kalupahana-mmk-1986: Atiśa exemplifies an Indian master who reads MMK within the full Mahāyāna framework — Two Truths, Candrakīrti, graduated path — producing a coherent, non-deflationary reading
- Connects to ruegg-svat-pras-2006: Ruegg’s point that the Prāsaṅgika label is a later Tibetan construction is strongly supported by Apple’s evidence that Atiśa and the earliest Kadampas did not use it
Relevance to paper
- Section 3.3 (Candrakīrti): Atiśa’s lineage from Candrakīrti demonstrates a contemplative transmission of Candrakīrti’s Madhyamaka distinct from later scholastic reception
- Section 4.1 (early Tibetan Madhyamaka): Atiśa and the Kadampas should be discussed as a precursor to the Tibetan commentarial traditions — they represent a missing link between Indian and Tibetan Madhyamaka
- Section 4.2 (Tsongkhapa): Apple’s evidence that Tsongkhapa’s Gelukpa Madhyamaka is significantly different from what Atiśa actually taught complicates the Gelukpa claim of Kadampa lineage descent
- Section 6.1 (framework necessity): Atiśa’s two realities structure and graduated teaching (Bhāviveka publicly, Candrakīrti privately) exemplify the framework in action at the moment of Madhyamaka’s Tibetan transmission
- Section 6.3 (framework present but disputed): Atiśa’s differences from Śāntarakṣita/Kamalaśīla (on pramāṇa, conventional reality, illusion analogy) show that Indian Madhyamaka diversity was already robust before the Tibetan commentarial tradition amplified it