In 499 CE, a twenty-three-year-old Indian mathematician wrote a treatise that for centuries enabled the calculation of solar and lunar eclipses accurate to the minute—all while relying on a model of the cosmos where the Sun dangles four million kilometers from Earth and the Moon hovers beyond it. This is a case of the perfect alibi assembled from false evidence.
Aryabhata was born in 476 CE in Kusumapura—a city that would later become Patna, capital of the Indian state of Bihar. In an era when most astronomers explained eclipses as the hungry demon Rahu devouring the celestial bodies, he assembled 108 verses and 13 introductory stanzas into the treatise "Aryabhatiya". Four padas—Gittikakanda, Ganitapada, Kalakriyapada, and Golapada—contained mathematical apparatus that would have been the envy of European universities eight centuries later.
His sine system worked on half-chords, formulas for sums of squares and cubes fell into tables without glitches. The "kuttaka" method solved Diophantine equations—problems where integer solutions hide in algebraic labyrinths. He calculated the number π as 3.1416, with an error of fractions of a percent. The sidereal rotation period of Earth—23 hours 56 minutes 4.1 seconds—differs from the modern value by seconds. All this packaged in Sanskrit verse, where each metrical foot carried a digit or constant.
But under the hood of this mathematical machine lay a geocentric cosmology with epicycles, where the Sun circles Earth at a distance of 4.3 million kilometers—thirty-five times closer than reality. And the Moon, according to his scheme, hangs beyond the Sun, not in front of it.
Aryabhata was the first in Indian astronomy to discard mythology and declare eclipses the result of shadow geometry. A solar eclipse occurs when the Moon stands between Earth and Sun, casting a cone of darkness on Earth's surface. A lunar eclipse—when Earth blocks the Sun from the Moon, and Earth's shadow crawls across the lunar disk. No demons, just optics and mechanics.
He calculated the diameter of Earth's shadow at the Moon's orbit, derived relationships between the sizes of luminaries and their angular diameters. His tables allowed prediction of the beginning, peak, and end of an eclipse with minute accuracy—astrologers and priests across India used "Aryabhatiya" for ritual calendars for years. The Moon and planets, he wrote, shine by reflected sunlight—a claim European science would confirm a thousand years later.
The paradox is that his cosmological model contradicted his own observations. If the Moon lies beyond the Sun, it cannot cast a shadow on Earth—shadow geometry requires the occulting body to be closer to the observer. But Aryabhata circumvented this collapse by introducing epicycles—small circles along which the Moon moves relative to its own orbital center. Epicycles allowed the Moon to periodically appear in front of the Sun relative to the terrestrial observer, without changing the basic hierarchy of spheres.
Epicycles functioned as an algebraic crutch—they compensated for the model's structural error, transforming it into a predictive instrument. The evidence pointed one way, the crime map another, but the final version of events converged.
Aryabhata asserted that Earth is spherical and rotates on its axis. The apparent motion of stars is an illusion caused by the planet's own rotation. In the treatise he wrote that a person on a ship sees stationary trees on shore moving backward—likewise an observer on Earth perceives the rotation of the celestial sphere.
This claim was a millennium ahead of its time. European astronomers up to Copernicus clung to a stationary Earth, because planetary rotation required explaining why clouds don't tear away into space and birds don't lag behind the surface. Aryabhata gave no answer to these questions—he had no Newtonian mechanics—but he recorded the fact of rotation itself and calculated its period with accuracy that held in Indian calendars for seven hundred years.
Yet he left the Sun revolving around Earth. The result was a chimera: the planet spins, stars are stationary relative to the cosmos, but the Sun and planets move in geocentric orbits. This construction resembled a machine where half the parts are connected to electricity and half to a steam boiler, but somehow the entire mechanism outputs the correct time.
"Aryabhatiya" reached Baghdad in the 8th century, when the Abbasid Caliphate was collecting scientific texts from all corners of the known world. Arab translators adapted his methods for calculating eclipses, sine tables, and calendar algorithms. In the 11th century through Toledo—a city where Muslim, Christian, and Jewish scholarship intersected—Indian calculations seeped into Latin Europe.
The Toledo Tables, compiled by the astronomer al-Zarqali, included methods for predicting eclipses tracing back to Aryabhata. European mathematicians used them for calendar reforms and astrological forecasts, not knowing that at their foundation lay a model with the Sun at a distance of four million kilometers. The accuracy of predictions troubled no one—eclipses arrived on time, and that was enough.
When Kepler in the 17th century replaced circular orbits with elliptical ones and finally dismantled geocentrism, Aryabhata's computational apparatus had already operated for twelve centuries. His formulas for calculating sines and shadow diameters remained in textbooks—only the cosmological foundation beneath them was replaced with a heliocentric one.
Aryabhata created the first documented case in the history of science where an instrumental model yielded accurate results with a completely false ontology. His eclipses calculated correctly not because the Moon truly lay beyond the Sun, but because epicycles and trigonometry compensated for the scheme's structural defect. The mathematical apparatus proved flexible enough to adapt to any geometry—as long as angular sizes and periods matched.
This paradox manifests in modern science as well. Newtonian mechanics works for engineering calculations, though we know spacetime is curved. Quantum mechanics gives the most precise predictions, but its interpretation remains a matter of debate. A model is not a map of the territory, but an algorithm that produces the right answer for a specific class of questions.
Aryabhata proved that one can predict eclipses without understanding the cosmos. His treatise is a case where the guilty party sat in shadow, but all evidence led to conviction. Except the conviction turned out correct, and the guilty party was never found—because they never existed. Mathematics won a trial in which physical reality was merely a witness whose testimony could be ignored.