For more than two thousand years, people have mapped the stars. But a new study claims one of the world’s oldest surviving star catalogs — the Star Manual of Master Shi — might be more ancient than we thought. It might rewrite the history of astronomy itself.

Using a machine learning technique, Chinese researchers argue that this star calendar is centuries older than the one created by the famous Greek Hipparchus. However, not everyone is convinced.

A Time Machine in Ink

Star calendars track the predictable appearance of stars in the night sky to signal seasonal changes. It’s essentially a system that uses stars and constellations to mark time. These calendars have practical purposes (for agriculture and navigation) but are also important for religious purposes. Ancient civilizations across the globe — from the Babylonians and Egyptians to the Maya and Chinese — had their own star calendars.

In China, astronomers developed a complex system aligned with constellations to track time and guide imperial rituals. The Star Manual of Master Shi, or Shi Shi Xing Jing (石氏星经) is the earliest known Chinese star calendar. It’s a meticulous record of 120 stars and 28 constellations, originating in China. Historians knew it was old. But how old, exactly?

Some experts placed it around the 1st century BCE. Others argued for a later date — maybe even the Tang Dynasty, 700 years later. But Boliang He and Yongheng Zhao from the Chinese National Astronomical Observatories weren’t satisfied with guesses. So, they built a model to calculate more precisely.

Using simulations and a modern astronomical reference catalog, they crunched ancient angular distances and compared them against how the stars would have appeared at different times. Essentially, they backtraced how the stars moved over time to see when they had the configuration presented in the calendar.

The Earth wobbles like a top, completing one full cycle every 26,000 years. Because of this motion, the North Star changes over millennia, and constellations subtly shift across centuries. By modeling this motion backward in time, astronomers can estimate when a given star chart was drawn.

The researchers used a technique called the Generalized Hough Transform that uses AI to find and mitigate significant errors between similar images.

What the Analysis Found

Two clear time stamps emerged from the noise. One from 355 BCE. The other from 125 CE.

Researchers say this means the calendar wasn’t a “one and done” affair. Instead, it was born in 355 BCE — perhaps in the lifetime of Shi Shen, the chart’s reputed author — and then updated centuries later. The second round of observations, they suggest, may have been carried out centuries later.

For centuries, Western scientific tradition has placed the dawn of systematic astronomy in Greece, with Hipparchus and Ptolemy leading the charge. Hipparchus’s lost catalog (dated to about 130 BCE) and Ptolemy’s Almagest (2nd century CE) were considered the oldest structured records of the heavens. But if Master Shi’s catalog really does date to 355 BCE — and includes mathematical coordinate systems and precession-aware tracking — it would beat Hipparchus by over 200 years.

This comes with some historical bragging rights. But it also forces scholars to reassess whether knowledge flowed between these ancient civilizations. It’s possible that the Chinese and Greeks independently developed stellar catalogs, but it’s also possible that some knowledge flowed down the Silk Road.

It’s a provocative idea. And, as you might guess, not everyone agrees with the study’s conclusions.

Not So Fast, Say the Skeptics

Daniel Morgan, a historian at the French Centre for Research on East Asian Civilizations, has a different take. The discrepancies in the catalog, he suggests, might not indicate multiple epochs. Instead, they could be due to a simple calibration error. He told Live Science that this could all be explained if the original instrument was off by about one degree. When you account for this mechanical slip, the whole catalog could plausibly line up with a single observational period: around 103 BCE.

That’s still old, but it loses the headline-grabbing status of “world’s oldest.”

There’s also a sort of nationalism or political pride in countries trying to claim various achievements, but while this study paints an intriguing picture, it’s not entirely conclusive. The study is still in preprint; it hasn’t been peer-reviewed yet. Even if it passes review, there’s still a reasonable doubt about what it means.

But there’s a much more tangible, practical angle: as AI and digital imaging methods evolve, they offer powerful new ways to examine old materials. What we’re seeing with the Shi catalog could be just the start. Expect similar breakthroughs in ancient medicine, engineering, and mathematics.

So Is This the Oldest Star Chart?

The short answer is maybe. Don’t you just hate it when we can’t draw a clear conclusion?

But the real takeaway is bigger than any one civilization. It says something about human nature.

For thousands of years, humans have watched the sky, trying to understand our place in it. Some used bamboo slips. Others used marble tablets. Now we use algorithms. But the impulse is the same.

To look up, wonder, and write it down.