For a while I'd had a question in the back of my mind: can the things written down in ancient Indian scripture be checked against modern data — satellite maps, geographic databases, the kind of elevation data NASA and other agencies have made public? Not to prove or disprove anything about faith, but to see whether some of the patterns people point to are actually there when you measure them.
I worked with an AI assistant to brainstorm which claims would make a good test — something concrete enough to put a number on. We went through a lot of ideas and narrowed to one: the Nadīstuti, a hymn in the Rigveda that names a row of rivers. For about 150 years scholars have said those rivers look like they're listed in geographic order — east to west — but they always said it as an aside, and nobody had measured it. That was exactly the kind of small, checkable claim I was looking for.
The hymn is Rigveda 10.75. Verse 5 is the one that names the rivers:
इमं मे गङ्गे यमुने सरस्वति शुतुद्रि स्तोमं सचता परुष्ण्या ।
असिक्न्या मरुद्वृधे वितस्तयार्जीकीये शृणुह्या सुषोमया ॥५॥O Ganga, Yamuna, Sarasvati, Shutudri (Sutlej), Parushni (Ravi), follow my praise! O Asikni (Chenab), Marudvridha, Vitasta (Jhelum), with the Arjikiya and Sushoma (Sohan), listen!
(Modern river names in brackets.)
What I did was simple. I found where each river sits on a modern map and asked whether the list moves steadily in one direction. The AI assistant helped me write the code, which sped up work that would otherwise have taken me a very long time. The test is one idea: shuffle the list into a random order thousands of times, and see where the real, ancient order ranks against the random ones. The shuffles stand in for chance — they show how tidy a score a random arrangement of the same rivers usually gets, so if the real order beats almost all of them, the pattern is very unlikely to be an accident.
The Rigveda order beat almost every shuffle, so it isn't luck. Then I ran the same test on four more lists — the sixteen Buddhist "great realms," the lands of the Zoroastrian Avesta, the twenty-three countries on Darius's Behistun inscription (around 520 BCE), and the Table of Nations in Genesis 10. They split. The two Indian lists follow the map. The Zoroastrian one doesn't — its order is religious. Darius's is loosely geographic, grouped into regional blocs. Genesis 10 looked unordered until the test showed why: it's stitched from family branches, and two of them are each geographic on their own, pointing opposite ways so they cancel out — a structure biblical scholars had argued for on separate grounds.
I didn't want to take the numbers on faith, so I checked them hard. Every number comes from code. A separate script re-derives all forty headline numbers from the raw data. Someone else downloaded the project fresh, deleted my results, and rebuilt them — all forty matched. I also re-ran everything with the competing modern-place identifications that scholars argue about, and again with a different map database built from satellite elevation data, including NASA's.
What I learned: whether an old list follows the map seems to depend on what the list is for — a hymn, an imperial record, a family tree — more than on its language or its century. Five lists is a hint, not a rule. And it's a measurement, not a revelation — it says nothing about who wrote these texts, or when. Mostly it was the kind of question I like: small, clear, and open for anyone to check.
Further reading — the Nadīstuti river hymn
- Nadīstuti Sūkta (Wikipedia)
- Nadistuti Sukta (Dharmapedia)
- Rivers of the Rigveda — drramakanta.com
- Rivers in the Rigveda — tri-murti.com (2003, archived)
- The Rigveda's rivers — Hare Krsna Sun
- @tskrishnan on the river hymn (X)
Code and data: github.com/ajinkyakulkarni/geographic-encoding-enumerations. The full paper is on Knowledge Commons: doi.org/10.17613/7s0j1-yzv46.