Lake & rivers

The Four Rivers of Anavatapta: What Sources Actually Say

Medieval Theravāda, Indic, Chinese, Khmer, Thai, and later reception

Several Buddhist source families connect Anavatapta with great rivers, but they do not all describe the same system. A major medieval Pāli-area dossier traces a staged southern flow that becomes five named rivers; Sanskrit and northern Buddhist material has its own four-river wording; later visual traditions can organize waters through animal outlets. The differences are the point.

four rivers AnavataptaAnotatta riversGaṅgā Anotattafive rivers Anotatta

1710 · National Diet Library, Japan, 10.11501/2542466 Nansenbushū bankoku shōka no zu — canvas 1 National Diet Library, Japan, Nansenbushū bankoku shōka no zu (1710), DOI 10.11501/2542466. Public Domain Mark · source · One canvas from a multi-canvas object, shown as an object-specific map witness.

What this page does

One source family at a time

This guide links each major assertion to a named source and a boundary. It is designed to be useful for readers, artists, teachers, and researchers without pretending that later images can repair earlier silence.

A river system is never just one fact

Sacred water moves through texts as narrative, cosmography, ritual, and geography. A river name can be a source path, a list entry, a visual channel, or an attempt to relate a cosmological system to human space. The source layer must be named before a river claim is repeated.

This site does not draw one “correct” drainage map. It records which witness supplies the route, whether that passage is direct or inherited, and what the witness does not establish.

The medieval five-river dossier

The Sāratthadīpanī gives the most elaborate audited Pāli-area southern outflow. It describes circuits around the lake, rock and aerial courses, an impact pool, a rock channel, a tunnel, Viñjha, and a later five-river result: Gaṅgā, Yamunā, Aciravatī, Sarabhū, and Mahī.

That is a remarkable witness, but it is not ordinary hydrological survey. The controlled extraction identifies its compilational character and explicit dependence, so it is presented as medieval cosmographic exposition rather than a timeless map.

What the four-river evidence does not permit

History 03 established an important grammatical correction: in S03, the statement that humans cannot cross without supernormal power governs the four rivers—Gaṅgā, Sindhu, Vakṣu, and Sītā—not automatically the lake itself.

Likewise, an attractive animal-outlet diagram belongs to particular visual or textual witnesses. It must not be attached by default to the Pāli Jātaka corpus, which does not present an animal-mouth system.

Evidence ledger

Claims you can trace

These are the precise points supporting the guide. Each includes a claim type, source, locator, and limit.

Direct sourceSāratthadīpanī

Five-river result

The Sāratthadīpanī’s staged southern flow culminates in Gaṅgā, Yamunā, Aciravatī, Sarabhū, and Mahī.

Locator: Vol. 1 p. 94; Vol. 3 p. 157

Does not establish: It remains a medieval compilational dossier, not modern hydrology.

Direct sourceIndic witnesses

Four-river correction

In S03, exceptional-crossing wording governs the four rivers rather than unambiguously the lake.

Locator: S03 claim crosswalk

Does not establish: The claim is grammatical and witness-specific.

What not to carry forward

Never convert yojanas into modern kilometres without a separate argument.
Never merge four and five river sets merely to complete an illustration.

Sources

Read the evidence layer

Primary textprobably 1165–1186, Polonnaruwa Sri Lanka

Sāratthadīpanī Gate 2 controlled extraction

Vol. 1 p. 94; Vol. 3 p. 157; bounded occurrence and claim ledgers

This medieval Vinaya subcommentary contains the largest connected Anotatta drainage dossier in the audited Pāli-area material.

Boundary: The exposition is compilational/inherited in parts and is not a surveyed watershed.

Primary textfourth–seventh centuries CE editions and witnesses

Indic and northern Buddhist Anavatapta witnesses

S01, S03, S07 in History 03 Run 1 control package

The witnesses retain their own dimensions, Gandhamādana relations, river wording, and nāga claims.

Boundary: Their measurements and spatial expressions must not be averaged into one diagram.

Primary textWestern Jin–Tang, 266–646 CE

Chinese Buddhist Anavatapta source family

T23, T24, T25, T197, T199, T2087; History 04 source manifest

Translations and Xuanzang/Bianji’s record create witness-specific Kunlun, river, mountain, and animal-direction configurations.

Boundary: Source-reported geography is not a modern field survey and does not erase other variants.

Research control record2026

Himavanta cross-history research synthesis

Cross_History_Thread_of_Truth.md; separately frozen history packages

The project comparison layer connects independently audited source families without merging their claims.

Boundary: A synthesis is not a primary witness and cannot fill a source’s silence.