Our Deep History Our Deep History A museum of unexplained ancient sites and relics
South Korea rooms

Gyeongju, South Korea

Cheomseongdae

Reign of Seondeok, 632 to 647 (observatory, altar, or both)

Cheomseongdae in Gyeongju. The square opening sits in the middle courses. The top is a square frame, and the base is a square platform. Bernard Gagnon, CC0 1.0. Shown unedited.

In Gyeongju a granite tower about 9.17 metres high has been called the oldest surviving observatory in East Asia. Its name means a platform for gazing at stars. No instrument and no observing log have come down with it.

Why this is extraordinary

Cheomseongdae stands in Gyeongju, a granite bottle on a square base, 9.17 metres high. The base is about 5.7 metres across. The body is 27 courses of fan-shaped blocks. A square window cuts the 13th, 14th, and 15th courses and looks a little east of south. The top is a square frame. Stones project so a ladder could be hung on. Below the window the shaft is full of rubble. Above it, the tower is hollow. The reign given for the work is Queen Seondeok’s, 632 to 647. She is counted as the 27th ruler of Silla, and the 27 courses are read as that number.

The name means a platform for gazing at stars. The body is counted at 365 stones, a year. With the base the courses come to 28, the lunar mansions. With the top they come to 30, a lunar month. Twelve courses each side of the window are the months, and the two twelves are the 24 solar terms, the farming divisions of the East Asian year. An old line says a person went in through the middle and climbed. The catch is that no sighting tool and no log survived, and the thematic study will not say exactly how an observation was made.

At this latitude the noon sun lives in the south, so the window faces the sun’s midday half of the sky, not the sunrise and not the pole star. One account has the equinox sun reaching the floor inside. The study that prints the direction does not print that measurement. What this changes: if the tower was climbed, Silla kept a stone observatory in the capital. If it was an altar or a monument, the calendar is still built into the courses, and the stars are the subject of the building whether or not a watcher stood on the frame. Other readings, an altar or a memorial, are printed beside the observatory one. The Gyeongju historic areas, this tower among them, are a World Heritage property.

Why it matters

If it was an observatory, a Silla court built a stone instrument for the year that fed the farms and the omens. If it was an altar or a monument, the same counts still spell a calendar in masonry, and the stars were the subject even if nobody climbed up to measure them.

The tower stands in the Gyeongju historic areas, a World Heritage property. It has three parts. A square base, about 5.7 metres on a side, is one course of stones. A curved body is 27 courses. A square top, seen from above, is read as the Chinese character for a well. A square window occupies the 13th to 15th courses and faces a little east of south. Below the window the interior is packed with rubble. Above it the shaft is empty. Stones stick out at the top and on two higher pairs of courses, which would take a ladder. An old note says people entered through the middle and climbed.

Photo tour

2 photographs, unedited. On a phone, swipe the row. Choose one to look closely.

Cheomseongdae in Gyeongju. The square opening sits in the middle courses. The top is a square frame, and the base is a square platform. Bernard Gagnon, CC0 1.0
Another side. The square top and the projecting stones are where a ladder could be set. The window is the only opening in the body. Bernard Gagnon, CC0 1.0

What is still unexplained

These sit beside the consensus account, with the same visual weight. None of them is marked solved.

Open Counts that match the year The body is counted at 365 stones, read as the days of a year. Twenty-seven courses are read as Seondeok’s place as the 27th ruler of Silla. Add the base and the count is 28, the lunar mansions of the East Asian sky. Add the two courses of the top and the count is 30, a lunar month. Twelve courses below the window and twelve above it are read as the twelve months, and together as the 24 solar terms that divide the farming year. These are fits between a building and a calendar. A fit can be the design, or it can be a count that works once you decide to count that way. The thematic study of astronomical heritage prints the fits and the word “probably.”
Open A window a little east of south The opening faces slightly east of due south. At about 36 degrees north the noon sun is always in the southern sky, so a south-facing window looks toward the sun’s midday path, not toward sunrise and not toward the pole. One educational account says sunlight falls on the interior floor at the spring and autumn equinoxes. The thematic study prints the direction and does not print that equinox measurement. How you would observe is simple if the shaft were open. You would watch a beam cross the floor, or you would stand on the top with a ladder and an instrument and look at the sky. The rubble fill stops the view through the lower half, so any sight from inside starts at the window or above it. What the opening meant in Silla is tied, in the same study, to farming decisions and to astrology as politics. Both are possible meanings. Neither is a logbook.

Location and dating

Location
Inwang-dong, Gyeongju, North Gyeongsang. The tower is about 44 metres above sea level, near the tomb mound of an earlier Silla king.
Period label
Current consensus — open to revision The reign of Queen Seondeok, 632 to 647, inside the Silla kingdom. The tower has kept its shape for about thirteen centuries and leans slightly to the north-east.
How this dating was done
The date is the historical reign, not a radiocarbon sample from the fill. The sky reading is the stone counts and the window’s direction, set against the East Asian calendar.
What could change it
A measured path of sunlight through the window across a year, published against the equinoxes and the solar terms. Or a foundation deposit that dates the first course apart from the reign.
Published ranges, drawn as bars. A bar is a claim about a span, not a point.
Seondeok’s reign
632 CE640 CE647 CE
  • Seondeok’s reign. The reign given for the construction. Not a laboratory date for the first stone.

What is actually there

Open questions

Question Status
Did anyone observe from the top, or do the counts carry the calendar without a watcher? Open
Does the window admit a deliberate equinox beam, or only the southern sky in general? Open. The direction is published. An equinox measurement is not in the thematic study.

The current narrative

Where publication stands today. It is the best-supported reading, and it is open to revision. It is not a verdict.

The current narrative — open to revision

A tower for watching the sky

Cheomseongdae was built under Seondeok as a place to observe the heavens. The name, the ladder stones, the old note about climbing in through the middle, and the calendar counts all belong to that use. Seasonal change guided farming, and astrology guided politics, so a court tower for the sky is the straight reading.

What supports this

  • The name means star-gazing platform, and the reign is 632 to 647.
  • Projecting stones and an empty upper shaft fit a ladder and a platform.
  • The 365 stones, 28 mansions, 12 months, and 24 solar terms are a calendar written in the courses.

What pushes back

  • No instrument and no observing record survive, and the thematic study says the exact method is unknown.
  • The lower interior is rubble, so it is not an open tube to the sky.
  • The same counts work if the tower is a symbol and nobody measured a star from it.

What would settle it

It is the centre because the name, the date, and the counts point the same way. It loosens if a year of measured light through the window fails to pick out the equinoxes or the solar terms.

Emerging narratives

Other published readings. Each may be right. None is adopted here unless the evidence carries it.

An emerging narrative

An altar or a monument that carries a calendar

The tower is an altar, a religious sign, or a monument to the reign, and the sky is in the numbers rather than in a nightly watch. The window and the well-shaped top are architecture. They are not, on this reading, a sighting instrument.

What supports this

  • Official accounts of the monument list an altar, a religious symbol, or a commemorative monument beside the observatory reading.
  • Nothing you can hold, a sighting tool or a log, has been passed down.
  • Rubble below the window limits what an observer inside could see.

What pushes back

  • The name and the climbing note are odd furniture for a closed altar.
  • A monument can still be used on the nights that mattered. Symbol and observatory are not enemies.

What would settle it

It keeps the gap honest. The counts stay astronomical either way. What it doubts is the nightly practice, not the calendar in the stone.

Should the current view be reevaluated?

Open to revision

The tower, the reign, and the calendar counts hold. Whether anyone stood on it to measure the sky is the open part, because the instruments are gone and the window’s equinox claim is not in the thematic study.

What drives this

  • A dated reign, 632 to 647, and a name that means star-gazing.
  • 365 stones, 27 and 28 and 30 courses, 12 and 24 as months and solar terms.
  • A window slightly east of south, and no surviving instrument.

This is the museum’s own weighing of the evidence. It is not a visitor vote, and it is separate from the slider.

Weigh it yourself

Anonymous. Nothing here is stored except on this browser. Your lean does not change the assessment beside it.

A tower for watching the sky Unsure An altar or a monument that carries a calendar

If the contested reading is right

What would follow if a reading that is not the consensus turned out to be true, and what could settle it. This explains the stakes. It does not adopt the reading.

The tower was a symbol, not a working observatory

Gyeongju would still have a calendar in granite. Guides would stop saying that observations were taken from the top, and would say the year was built into the courses.

Why it is still worth a test

A year of light through the window, and a search of Silla records for an observation said to come from this tower.

Sources

  1. Cheomseongdae observatory, case study UNESCO Portal to the Heritage of Astronomy, thematic study
  2. Cheomseongdae, heritage of astronomy UNESCO Portal to the Heritage of Astronomy
  3. Gyeongju Historic Areas UNESCO World Heritage Centre
  4. Mohs scale of mineral hardness Wikipedia

Image credits

← South Korea World map