For roughly 2,300–2,500 years — about 100 human generations — the Rigveda existed only as sound passing from one mouth to the next ear. Composition c. 1500–1200 BCE is an inference; the oldest reported manuscript is c. 1040 CE.
“I praise Agni, the household priest, divine minister of the sacrifice, the invoker, greatest bestower of treasure.” — Rigveda 1.1.1
agním īḷe puróhitaṃ yajñásya devám ṛtvíjam · hótāraṃ ratnadhā́tamam
Four interlocking recitation modes: corrupt one syllable and the modes disagree — the error is caught.
c. 1380 BCE — the Mitanni treaty. A Hittite–Mitanni treaty from the Hattusa archive invokes Mitra, Varuṇa, Indra, Nāsatyā as divine witnesses — the first time anyone outside the chain wrote down anything from inside it: four god-names, in someone else's script, far from the Punjab. The lone external anchor. note →
~2,300–2,500 years · zero manuscripts. Everything before the manuscript is a bridge built of inference — and this is every plank it has.
Australian Aboriginal coastal-flooding traditions: stories at 21 coastal sites describe land now under the sea, and the matching sea-level events date to 7,250–13,070 cal BP (Nunn & Reid 2016).
No pāṭha. No treaty plank. No manuscript wall. The only bridge offered is the landscape itself — story geography matching the drowned-land contours — and whether that bridge holds is exactly what is contested. This chain renders dim because the vault rates the time-depth claim low: the geological events are real (class 1); the transmission chain is undated and unverified.
Henige · Hiscock — the steelman against deep time
- The dating is circular: it assumes the very fidelity it claims to prove. There is no independent clock on oral tradition.
- Every well-documented case — the same tradition recorded at two dates — shows narrative change within decades to a few centuries of contact.
- Communities were not isolated: language replacement, migration, and inter-group contact are corruption vectors over millennia.
- Flood and coastal-change stories arise globally — convergence needs no deep memory.
- Post-hoc sampling: stories were selected because they match the geology.
“Impossible to disprove yet impossible to believe.” — Henige 2009, the standing critique (it predates Nunn & Reid 2016). source record →
The Vedic counterexample
- Four interlocking recitation modes — pada, krama, jaṭā, ghana — make undetected insertion or deletion extremely difficult.
- Phonological preservation is empirically demonstrable: regional schools separated for centuries recite with near-identical phonology.
- The text came down “almost entirely without corruptions.”
- So high-fidelity deep transmission is possible — but it required a dedicated, engineered mnemonic institution.
- And even this chain cannot date its own origin: the safeguards are post-composition innovations, semantic drift is not directly verifiable, and composition c. 1500–1200 BCE remains inference (3-reconstruction), not attestation.
Rigveda gap (06_dharmic) →Rigveda gap (04_indo_european) →Aboriginal time-depth claim →proposal →
Every date and claim above comes from the linked vault notes; the drift lane is a labeled, deterministic simulation (the honest control condition). The verse is Rigveda 1.1.1 per the whisper-chain proposal. Reduced-motion preferences disable the flicker and plank wobble.