Method
How the evidence works.
How a connection is typed, how a date is classified, how geographic uncertainty is shown, what the lanes mean — and every source cited anywhere on this map.
The reference page. Everything about how a claim is typed, dated, placed and sourced is settled here once, so the other pages can point at it instead of re-explaining it. The complete bibliography is at the foot.
↩ Back to the AtlasThe history is only as good as the line behind it.
This project keeps two kinds of truth visible at once: what a tradition says about itself, and what manuscripts, inscriptions, historical records and modern scholarship can establish. The first is not dismissed as “mere belief”; the second is not bent to protect a lineage claim. They are shown side by side.
Composition, surviving manuscript, inscription, archaeology, traditional attribution, hagiography and modern documentation are not treated as the same kind of date.
Documented transmission is visually distinct from supported contact, plausible contact, comparison-only, contested connection and traditional lineage claim.
Where scholarship is divided—Gauḍapāda and Buddhism, early Dzogchen, Sikh nonduality, Udāsī identity, Nahua “monism”—the dispute is part of the record rather than something to smooth away.
Where each argument is made
Five things this site says more than once.
Each of these recurs, because each is load-bearing in more than one place. Each is argued in exactly one place, listed here. Everywhere else it appears, a pointer says so.
How to read the evidence atlas
Four things are encoded at once: when something happened, where, what structure of non-separation it proposes, and how well any of that is actually attested.
Lanes are cultural and geographic, not national. Modern borders are not imposed on ancient intellectual history. Gandhāra, Uḍḍiyāna, Kashmir and the Punjab are separate lanes because they were separate intellectual worlds, whatever state now administers them.
Dating is layered. A mark's evidence type says whether its date comes from composition, redaction, a surviving manuscript, an inscription, archaeology, a traditional attribution, later hagiography, an attested lifespan, a publication, or modern documentation. For traditions that transmitted orally, traditional antiquity, probable historical antiquity, earliest textual documentation, earliest manuscript evidence, and modern ethnographic documentation are five different things, and the map never treats them as interchangeable.
Lines carry their own evidence. Every connection is one of six kinds: documented transmission; strongly supported contact with uncertain doctrinal transmission; plausible contact seriously discussed by scholars; philosophical parallel with no demonstrated transmission; contested connection; and traditional lineage claim. The last two are the most common places where popular accounts of this history overreach.
Structure is provisional. The eleven categories used here are a working instrument, not a result. Several of them failed on contact with the sources, and where they failed the map says so in the entry rather than forcing a fit. The category devotional non-separation was added specifically to test whether some Sant, Sikh, Bhakti and Sufi material works through a structure in which lover and Beloved remain relationally distinguishable while separation is overcome — different from both Śaṅkara's identity claim and Dzogchen's empty-luminous knowing.
Absence of a line is information. Two traditions sitting in the same century in adjacent lanes are not thereby connected. Co-occurrence in time is not evidence of contact, and this map draws nothing it cannot source.
Contested is not a euphemism. Where scholarship is genuinely divided — over Buddhist influence on Gauḍapāda, over Śaiva–Buddhist tantric borrowing, over Chan and early Dzogchen, over whether Buddha-nature is a disguised self, over whether Guru Nanak's teaching is monotheist, monist, or neither — both sides are named in the entry and the dispute is left open.
Bibliography
Every source cited anywhere on this map, deduplicated.
The evidence behind this page
Every claim above opens into the atlas, and the atlas is loaded here — the same timeline, the same map, the same entries. Open it in place, or go to the Atlas itself.