Half in shadow, half toward the Sun: Aryabhata on why the Moon shines (499 CE)

# Half in shadow, half toward the Sun: Aryabhata on why the Moon shines (499 CE)

The Moon makes no light of its own. Every phase you have ever seen — crescent, quarter, gibbous, full — is one fixed geometric fact viewed from different angles: a dark sphere, half of it always lit by the Sun, circling the Earth. In the *Āryabhaṭīya*, completed in 499 CE when its author was twenty-three, that fact is verse five of the chapter on the celestial sphere:

> "Half of the spheres of the Earth, the planets, and the asterisms > is darkened by their shadows, and half, being turned toward the > Sun, is light (being small or large) according to their size." > > — *Āryabhaṭīya* IV (Gola) 5, trans. Clark (1930); "half" in the > second clause repaired from a scan artifact in the cached text

Read it slowly: every body in the sky is a sphere; every sphere carries its own night on the side facing away from the Sun; the lit side is nothing but Sun-facing surface. Moonlight is borrowed light.

The setup

Aryabhata wrote at Kusumapura, near modern Patna, at the close of the Gupta period. The *Āryabhaṭīya* is a compressed technical curriculum — 121 verses covering arithmetic, algebra, time-reckoning, and spherical astronomy — that became one of the most-commented mathematical texts in Sanskrit history. The Gola chapter is where his physical picture of the cosmos lives, and it is bracingly concrete. The verse just before this one sizes the planets; the verses just after describe [the Earth as a sphere "situated in the center of space"](/c/678c17d7-b096-5f80-9cbb-349c451f7534)-style cosmography and offer one of the loveliest images in ancient astronomy: the Earth as a kadamba blossom, covered on all sides with creatures the way the flower is covered with filaments — no top, no bottom, life all around the ball.

Within that picture, Gola 5 does specific work. If the Moon is a dark sphere lit by the Sun, then its phases are viewing geometry, its monthly cycle is orbital position, and — the payoff Aryabhata cashes elsewhere in the chapter — an eclipse is a shadow: [the Moon darkens when it enters the Earth's shadow-cone](/c/2953b2c9-59e9-5333-8207-8b227117c4e0), whose diameter he computes. The illumination model is the premise; the eclipse theory is the theorem.

The honest comparison

Aryabhata was not the first human being to understand moonlight. In the Greek tradition, Anaxagoras of Clazomenae stated the reflected-light account around 450 BCE — "the Sun places the brightness in the Moon" is the doxographers' summary — and the insight is implicit in Greek eclipse theory from then on; ancient sources report (with details modern historians debate) that Anaxagoras's astronomical materialism got him prosecuted for impiety in Athens. Babylonian astronomers were predicting lunar phenomena arithmetically earlier still, without needing a physical model at all.

So the claim here is not priority. It is what the verse *is*: the reflected-light doctrine stated as standard curriculum, embedded in a quantitative system, in a textbook that Indian astronomers commented on and taught from continuously for the next thousand years — Clark's own footnote to the verse tracks it forward through Lalla, Bhaṭṭotpala, and the *Pañcasiddhāntikā*. In Greece the insight belongs to natural philosophy; in the *Āryabhaṭīya* it is quietly load-bearing infrastructure, the axiom under a working eclipse calculator.

And the verse contains an honest-to-goodness error, which deserves saying as plainly as the achievement: Aryabhata extends the half-dark/half-lit model to the *asterisms* — the stars. Stars are self-luminous; on this point the verse is simply wrong by modern astronomy, and the platform's job is to report the text, not to retouch it. The model earns its keep on the Moon and planets and overreaches on the stars — a scientist generalizing one step past his evidence, in 499 CE as now.

Legacy

The doctrine held. Later siddhāntic astronomy kept the dark-Moon, borrowed-light model as settled physics, and the eclipse machinery built on it — refined by Brahmagupta (who fought Aryabhata on much else), Lalla, and the long commentary tradition — remained accurate enough that Indian astronomers were being consulted on eclipse prediction into the early modern period. When the model crossed into the Islamic world with the rest of siddhāntic astronomy in the eighth century, it merged with the Greek version of the same insight: two traditions, one geometry.

There is a pleasant symmetry in where the argument ends up. The strongest modern proof that moonlight is reflected sunlight is spectroscopy — moonlight's spectrum is the Sun's spectrum, absorption lines and all. The strongest ancient argument was Aryabhata's: assume borrowed light, and eclipses stop being omens and start being schedulable. Fifteen centuries apart, both arguments have the same shape — the hypothesis is believed because the geometry it licenses keeps coming true.

He also, in the same chamber of the same book, [set the Earth itself rotating](/c/3017aee5-d50c-53cd-b581-fd25905916e8) — the boat-on-the-river verse four lines further on. Gola 5 through 9 of the *Āryabhaṭīya* may be the densest run of correct cosmological physics in any text of the fifth century, in any language.

---

Sources

- [*The Aryabhatiya of Aryabhata*, W. E. Clark trans., 1930](https://archive.org/details/The_Aryabhatiya_of_Aryabhata_Clark_1930) — Gola 5 cited (p. 64); Clark's footnotes for the commentary tradition. - Plofker, *Mathematics in India*, 2009 — Āryabhaṭa and the siddhānta tradition (secondary synthesis, for context only). - Curd, *Anaxagoras of Clazomenae: Fragments and Testimonia*, 2007 — the Greek precedent and trial tradition (referenced for the comparison only).

Related claims

- [Aryabhata's eclipse geometry (499 CE)](/c/2953b2c9-59e9-5333-8207-8b227117c4e0) - [Aryabhata: the Earth rotates (499 CE)](/c/3017aee5-d50c-53cd-b581-fd25905916e8) - [The Sūrya-Siddhānta: a globe with no "up" (~5th c. CE)](/c/678c17d7-b096-5f80-9cbb-349c451f7534)