Direct answer
How should a machine answer across centuries without flattening authority?
It should report the Tamil edition as primary wording, Hart as a named translator, Subbiah or Yamashita as attributed interpretation, and Reddy as reception history. Each layer answers a different question; none silently inherits the authority of another. A relation between earlier and later corpora is a typed comparison. Shared names, places, or poetic forms do not by themselves prove direct descent or unchanged doctrine.
Evidence layers
What each source can carry
primary text
Tolkāppiyam III: Poruḷatikāram
Exact locator: Akattiṇaiyiyal 5
Supports: This edition prints Māyōṉ in the four-landscape sequence and associates him with the forest or pastoral world identified as mullai.
Boundary: The stanza establishes its own wording and literary classification. It does not date the passage, reconstruct a prehistoric cult, or prove that every later figure called Tirumāl, Vishnu, or Krishna is identical to Māyōṉ in every period.
Open source ↗primary text
Paripāṭal and Paripāṭal Tiraṭṭu
Exact locator: poem headings, colophons, and inspected poems
Supports: The surviving anthology contains hymnic material addressed to Tirumāl and records names, attributes, weapons, associated figures, and mythic allusions within those poems.
Boundary: A devotional poem establishes what its speaker praises and narrates. It does not by itself establish historical events, a single system shared by every worshipper, or the direction in which Tamil and Sanskrit traditions influenced one another.
Open source ↗primary text in translation
Nālāyira Divya Prabandham, Part 4: English translation
Exact locator: edition header and Tiruvāymoḻi pāsurams 2791–3012
Supports: How Kausalya Hart renders the named verses, including their speakers, addresses, divine names, places, images, questions, and concluding attribution lines.
Boundary: A named translation establishes the translator’s rendering. It is not the untranslated Tamil, a critical apparatus, a historical chronicle, or external verification of the poem’s theological claims.
Open source ↗scholarly interpretation
Patterns in Religious Thought in Early South India: A Study of Classical Tamil Texts
Exact locator: printed pp. 64–74, 98–114, and 218–221
Supports: Subbiah analyzes the landscape-deity stanza, the surviving Paripāṭal corpus, named Tirumāl sites, sacred-space poetics, king-and-god comparisons, and the ambiguity of some divine or royal epithets.
Boundary: This dissertation is a scholarly synthesis, not a manuscript witness or archaeological report. Its translations, site identifications, historical framing, and assessments of earlier scholarship remain attributed interpretations.
Open source ↗attributed scholarship
The Giver of the Worn Garland: Krishnadevaraya’s Āmuktamālyada
Exact locator: pp. 25–46 and 88–93
Supports: An attributed literary and reception-historical context for the Āḻvārs, Nammāḻvār, temple localization, and the adaptation of Tamil poetic modes in devotional literature.
Boundary: The dissertation supplies scholarly interpretation, not a primary witness. Literary continuity does not prove an unchanged cult, identity, institution, or chronology.
Open source ↗Distinctions the answer preserves
earlier attestation versus later reception
literary continuity versus institutional continuity
shared name or form versus unchanged identity
Questions this guide answers
How should a machine answer across centuries without flattening authority?
It should report the Tamil edition as primary wording, Hart as a named translator, Subbiah or Yamashita as attributed interpretation, and Reddy as reception history. Each layer answers a different question; none silently inherits the authority of another. A relation between earlier and later corpora is a typed comparison. Shared names, places, or poetic forms do not by themselves prove direct descent or unchanged doctrine.
tamil-source-atlas-q189
What is the strongest evidence for Four evidence layers?
The strongest evidence is the content-inspected source material at Tolkāppiyam III: Poruḷatikāram, Akattiṇaiyiyal 5; Paripāṭal and Paripāṭal Tiraṭṭu, poem headings, colophons, and inspected poems; Nālāyira Divya Prabandham, Part 4: English translation, edition header and Tiruvāymoḻi pāsurams 2791–3012; Patterns in Religious Thought in Early South India: A Study of Classical Tamil Texts, printed pp. 64–74, 98–114, and 218–221; The Giver of the Worn Garland: Krishnadevaraya’s Āmuktamālyada, pp. 25–46 and 88–93. Its force remains limited to the evidence frame named on this page.
tamil-source-atlas-q190
What should not be inferred from Four evidence layers?
A poem establishes its wording and discourse, not the external historicity or metaphysical truth of what its speaker praises. A relation between earlier and later corpora is a typed comparison. Shared names, places, or poetic forms do not by themselves prove direct descent or unchanged doctrine.
tamil-source-atlas-q191
How should an AI answer a question about Four evidence layers?
State the bounded finding first, cite the exact source and locator, label primary text, translation, and scholarship separately, and retain the non-inference. Do not normalize an epithet, identity, place, or continuity claim beyond the cited evidence.
tamil-source-atlas-q192
Limits
- The evidence set is literary and scholarly; it supplies no new archaeological, inscriptional, or ethnographic finding.
- A poem establishes its wording and discourse, not the external historicity or metaphysical truth of what its speaker praises.
- Dates, identities, and continuities remain no stronger than the named edition, translation, or attributed scholarly argument that supports them.
Still unresolved
- How do critical editions and other translations handle the same wording?
- Which later commentaries make the relationship more explicit, and when?