§10.1 The problem of textual conflict
The corpus is enormous and it contains statements that, taken at face value, conflict. A method that denied this would be useless. The tradition does not deny it; it ranks.
§10.2 The Prasthāna-trayī filter
All secondary literature must pass through the triple canonical filter:
- Śruti-prasthāna — the principal Upaniṣads.
- Nyāya-prasthāna — the Vedānta-sūtra of Bādarāyaṇa Vyāsa.
- Smṛti-prasthāna — the Bhagavad-gītā.
Every ācārya who founded a school commented on all three; that is what founding a school consisted of. Within the Brahma-Mādhva-Gauḍīya line the Śrīmad-Bhāgavatam is held to state the same conclusion in expanded form, a position argued at length by Jīva Gosvāmī in the Tattva-sandarbha.
§10.3 The canonical ranking hierarchy
- T1 · Śruti — Upaniṣads and Saṁhitā. Highest. A direct statement here is not overturned by anything below it.
- T2 · Vedānta-sūtra. The reasoning instrument. Determines how T1 is to be read when T1 appears to conflict with itself.
- T3 · Bhagavad-gītā; Viṣṇu-sahasranāma. Smṛti of the highest order; treated as settling questions T1 leaves open.
- T4 · Śrīmad-Bhāgavatam. In this lineage, the natural commentary on T2 and the expansion of T1–T3. Where it appears to conflict with T1, the conflict is a signal to re-read, not to choose.
- T5 · Other Purāṇa and Itihāsa. Real evidence, ranked by the guṇa classification (§10.4) and by agreement with T1–T4.
- T6 · Āgama, Tantra, Nīti, and the technical śāstras. Authoritative within their own domain; not authoritative on tattva against T1–T4. The Bharadvāja Saṁhitā of §7.3 sits here.
- T7 · Later works, regional literature, oral tradition. Valuable, frequently beautiful, never load-bearing for a contested tattva question.
§10.4 The guṇa classification of the Purāṇas
The Purāṇas are traditionally sorted by predominant mode — sāttvika, rājasika, tāmasika — and the sorting is standardly derived from statements in the Padma Purāṇa (Uttara-khaṇḍa 236.18–21 in the circulating vulgate; Grade C pending collation, register entry D-24). This device is used constantly and abused nearly as often, so its correct scope must be stated.
The related doctrine of asura-vimohana — that the Lord arranges bewildering presentations for those disposed to be bewildered — is subject to exactly the same restriction and for the same reason. It explains why a body of literature exists. It is not a licence to attribute bewilderment to any text one dislikes.
§10.5 The conflict protocol
When two passages appear to contradict, run this in order. Stop at the first step that resolves.
- Verify both texts. A startling number of apparent contradictions are between a real passage and a misquotation. Check the locator in a printed edition first. This step resolves more disputes than the remaining seven combined.
- Establish subject by the six marks (§8.2). Often the passages are not about the same thing and there is no conflict to resolve.
- Check for arthavāda. Is one passage praise embedded in a section on another subject? Praise does not transfer subject-hood.
- Check register (§5.2). A statement true ādhidaivika and a statement true pāramārthika are not in conflict.
- Apply the evidence hierarchy (§8.3). Direct statement outranks contextual inference outranks name.
- Apply the corpus ranking (§10.3). Tier governs where the marks are silent.
- Consider audience and mode (§10.4) — as explanation of emphasis, subject to Rule C-02, never as sole dismissal.
- If unresolved, record it as unresolved. Enter it in the open register (Appendix D) with both readings stated. Do not manufacture a resolution. An honest unresolved entry is worth more to this platform than a confident wrong one, because the first is repairable by someone else and the second is not. Manufacturing a resolution is failure mode M18.