Do Koreans even exist? part 4. Stuff
4. Stuff (artifacts)
By this point you may have noticed a few things:
DNA held promise, but didn't tell a cohesive story. There is no such thing as "Korean DNA".
Language told a story, until it stopped at Shilla.
Soil took us back to Liaoning, and left us wondering what Liaoning has to do with Koreans.
To tell the remainder of the "Korean" story, we need to take out another tool from our toolbox:
Stuff.
The things people left us.
The funny thing about telling a story with Stuff is, it really has a horrible openning. Not that there's anything particularly bad about Korea, and Korean stuff, but because for most of our history we really didn't have that much stuff. No one did.
So if the history of Koreans were a movie, in the first 90% of it, barely anything happened:
People hunted game.
Gathered food.
Slept.
That's it.
Oh yeah, and as we've just seen, they moved.
They moved a lot.
They moved so much, and left so little behind, that no one is 100% sure who our ancestors were.
Our best guess is...a mixture of various people in East Asia; from the Amur to the Mongolian steppes, and down into the heart of China.
From a Stuff perspective:
They made stone tools.
They made slightly better stone tools.
They probably had baskets (though none remain).
They stayed in caves and later built primitive homes.
But that's basically it.
Though, to be honest making tools from stone is pretty damn cool.
It makes sense that they didn't have that much stuff.
If you're constantly on the move, you can't carry a lot of stuff around with you.
Of course some people didn't constantly move.
Some were lucky and lived next to an awesome food source, like the ocean.
They stayed longer, then moved.
That's why so much of Korea's really old stuff is often found near mounds of sea shells, called shell middens (패총).
Then something happened.
They stayed.
Someone either figured out, or was taught how to plant stuff, millet, to be exact.
The usual archaeological story is that millet cultivation reached Korea from northern China, probably through the Liaoning world.
Maybe.
I happen to think that anyone can plant shit and wait for it to grow.
Once people stayed, stuff exploded.
So,
- They stayed → stuff accumulated
- They farmed → needed more stuff
- Closer to our time → more stuff survived
- They produced surplus → people could start making stuff that wasn't about staying alive
Soil: why people moved.
Stuff: what happened when they stopped.
Even though Stuff is genuinely fascinating, Stuff is not the main point of our story.
Using Stuff to find the origins, if there is such a thing, of Koreans is.
So let's start digging.
- ה -
When I first began this project, my definition of "Korean" was rather simple, and in hindsight, pretty naive.
Definition #1.
Koreans =
The first human being to step onto the land we now call Korea.
But as I looked into this topic, I quickly came across two problems:
1. How on earth would we know who that person was?
and perhaps more importantly,
2. If that person, or those people, left not a single drop of blood and no legacy we can trace, in what meaningful sense are they Us?
Reasonable continuity.
When I asked my wife who she thought were the first Koreans, she said "People who settled and left a legacy for us today. People who lived on Korean soil, had children, raised families, and were buried on Korean Soil. Who left Blood, Words, Stuff that influenced us today."
Brilliant. And thus I added a third component to our definition of Us:
"I am Us."
"Yes, you are Us."
+ reasonable claim to continuity
Leading to,
Definition #2.
Koreans =
The earliest populations that left a continuing legacy,
biological, cultural, linguistic, political, or otherwise,
that we can reasonably trace into the people who eventually became Koreans.
Reconfirming the toolset we've been using so far:
Soil (Land)
Stuff (Artifacts)
Blood (DNA)
Words (Language)
Books (Written records)
- ה -
The history of Stuff
Where should we start?
Or to put things in a better frame, where should we not look?
Since we are not looking to tell a history lesson, but trace the origins of the Korean People, which increasingly looks like Peoples: multiple, not singular.
Let's start eliminating.
Modern Koreans (today) → Joseon Koreans
→ Goryeo Koreans → Unified Shilla Koreans (668 CE)
Difficult to argue with this chain.
- Modern Korean DNA is relatively established
- Koreanic Language can be relatively proven
- The Soil (territory) confirms it
- The Stuff is continuous: it can be traced upwards in time
- Historical books (introduced in depth in the next section) also record it
They can be safely removed.
668 CE is our upper limit, where we're sure they are "Korean".
How about the lower band?
~700,000 BCE: when the first humans were discovered, Home Erectus
to:
668 CE.
Roughly 700,668 years.
That's a long time, so let's continue eliminating time.
---
Homo Erectus (c. ~700,000-400,000 BCE)
"Upright man." Named this way because early archaeologists thought they were the first humand relative to walk fully upright with human-like body proportions.
While no fossils have been found (remember Korea eats its dead), there is plenty of evidence that Homo Erectus lived in what we now call the Korean Peninsula. Acheulean handaxes, heavy choppers, and primitive cleavers are found all over the place, Jeongok-ri (Yeoncheon), Geumgul Cave (Danyang), and Geomeun Moru (Sangwon) are some famous examples of their sites.
But while they left us Stuff and lived on our Soil, nothing of them remains. Korean history can claim them, but Koreans cannot. Nothing of them remain in Us today that we can decipher.
Using our definition,
Soil (Land): Yes (but migratory)
Stuff (Artifacts): Yes
Blood (DNA): No
Words (Language): No
Books (Written records): No
Are they Korean ('Us')?
→ I say No (you can disagree).
From a timeframe of 700,668 years, we are now at 400,668 years.
---
Before Homo Sapiens: Lower /Middle Paleolithic (~700,000–45,000 BCE)
After Homo erectus, East Asia did not suddenly become empty. It remained inhabited by a messy collection of archaic humans: Denisovans, robust cousins of Neanderthals, and other archaic Homo represented by Dali, Jinniushan, and “Dragon Man.”
On Korea, no DNA remains, but places like Jeongok-ri, Ryokpo, and Tokchon leave their traces.
No trace of Homo Sapiens can be found on the Korean Peninsula during this period. Homo sapiens evolved in Africa around 300,000 years ago and spread outward. By roughly 45,000 years ago, they were established in northern China and wider Northeast Asia.
But the stuff of these early Hominids exist. Homo Erectus was already extinct and Homo Sapiens hadn't appeared. So some Hominid must have existed on the Korean Peninsual. We just don't know who, or what they are.
So,
Soil (Land): Yes
Stuff (Artifacts): Yes
Blood (DNA): No
Words (Language): No
Books (Written records): No
Are they Korean ('Us')?
→ Still No (you can disagree)
From a timeframe of 700,668 years, we are now at 45,668 years.
From the original 700,668 years, we've managed to eliminate 655,000 years. That's a reduction of 93.48%! We've managed to successfully strip away over nine-tenths of the timeline we need to examine.
---
Homo Sapiens: Upper Paleolithic (~c. 40,000-10,000 BCE)
After Homo Sapiens arrived in China, because global sea levels were drastically lower, these anatomically modern humans walked freely across the dry Yellow Sea basin from mainland China and Siberia, bringing advanced stone technologies and complex social behaviors. They completely replaced all earlier archaic hominids, like Homo Erectus and the Hominids after them.
Some key discoveries:
- Seungnisan Man (Tokchon): Discovered in a cave in North Korea in 1972, these represent some of the first Homo sapiens fossils found on the peninsula. The remains, primarily a lower jaw and collarbone dating to roughly 40,000 years ago, show clear anatomically modern traits that distinguish them from the older Homo erectus populations.
- Heungsu Child (Cheongju): Found in Durubong Cave and dating to approximately 40,000 years ago, this nearly complete skeleton of a young boy is one of the most important Paleolithic discoveries in Asia. He was laid out in a deliberate burial posture, and soil analysis revealed high concentrations of pollen around the body, strongly suggesting these modern humans were conducting mortuary rituals using flowers.
- Yonggok Man (Pyongyang): Discovered in the 1980s alongside bone tools and animal fossils, these skulls provided further evidence of late-stage Paleolithic modern humans hunting across the northern half of the peninsula.
- Mandal Man (Pyongyang): Dating closer to 12,000–15,000 years ago near the end of the Ice Age, this remarkably preserved skull exhibits physical characteristics, such as a broad face and flat cheekbones, that anthropologists link directly to the ancestors of modern Northeast Asians.
Unlike Homo erectus, who relied on heavy handaxes and choppers, Homo sapiens in Korea engineered the microblade. At manufacturing sites like Suyanggae, these hunters struck razor-sharp stone flakes designed to be slotted into bone or wood to create advanced spears, darts, and knives.
They also established massive, long-distance trade networks. Obsidian tools found in southern Korea have been chemically traced to volcanic sources at Mount Paektu in the far north and Kyushu, Japan in the south. This proves these early Homo sapiens bands were not isolated; they were constantly migrating, trading, and navigating across the vast land bridges of Pleistocene Asia.
But like Homo Erectus before them, they left nothing that we can reasonably trace back to modern Koreans.
Soil (Land): Yes (but migratory)
Stuff (Artifacts): Yes
Blood (DNA): No evidence (not a flat out No)
Words (Language): No
Books (Written records): No
Are they Korean ('Us')?
→ I say No (but this time I won't laugh at you if you disagree).
From a timeframe of 45,668 years, we are now at 10,668 years.
Not as impressive as a 655,000 year reduction, but still 35,000 years is impressive.
---
Incipient Neolithic: c. 10,000–6,000 BCE
10,000 BCE, +/- 2,000 years. This is when things get interesting. Since we are using a definition of Koreans around "settled" and "left traces", this period of time is very important.
This was near the end of the last Ice age. Sea levels were 40-50 meters lower than today. Today the Yellow Sea is only 44 meters deep on average. Back then it was most likely a fertile plain with rivers flowing through it. Most of what is now the Korean Peninsula was an inhospitable mountain range to the east.
Before they discovered farming people hunted and gathered. In most places, during the winter and early spring, the only available food source was game. So they had no choice but to move with the game. And they had to stick close to sources of water.
You can live for weeks without food.
Without water you die in the days.
So they came south in the winter, and moved north in the summer. Both ancient fossils and artifacts show this. Amur people coming south. Yellow river people going east. Liaoning people moving north↔south. Korea was at best a pit stop.
This is when the first evidence of pottery appears. If you live in one place, pottery is amazingly useful. You can store stuff, cook food, trade it. But if you're constantly moving then it's a horrible thing to lug around. That's why most historians connect pottery with settlement.
The first proper pottery began to appear in Asia around 14,000 BCE in Amur and China (earlier versions of crude pottery was dated to ~c. 18,000 in Jianxi, but these were not settled farmers, they were highly mobile Ice Age hunters). Japan follows with 12,000 BCE. Same thing in Liaoning. But in the Korean Peninsula, pottery is strangely absent until much later in 6,000 BCE. People must have traveled through there, but no pottery, and hence no settlement. They fail our definition of "Korean".
Or do they?
Remember our hunter-gatherer survival logic:
You move because you couldn't gather enough in the winter.
There was one type of place where this was possible:
You lived near a rich river or a coastline.
Plenty of fish, shellfish, edible seaweed, marine animals to help you survive throughout the year. Perfect for people to settle.
Archaeology proves this. Ancient pottery-leaving-people across Asia lived this way.
Then why not Korea?
Because it was a hundred miles from the coast and the rivers that flowed north to south went through the Yellow Sea plains. So if people settled along those old coasts and rivers and left pottery, much of the evidence is now underwater.
Of course the surviving peninsula record looks late.
For an archaeologist,
No Pottery = No Permanent Settlement = No Neolithic era.
This suited the Japanese historians fine during the Japanese occupation.
"We have pottery dated 12,000 BCE, yours is dated 6,000 BCE. We are older. You came from us. We are united and rightfully your elder."
After independence, most Korean historians were either Japanese educated, or Japanese influenced. That thinking stuck. For the longest time, the Korean Neolithic era started in 6,000 BCE.
And then, in 1988, Gosan-ri happened. On Jeju Island, archaeologists found pottery dating back to roughly c. 10,000 BCE, along with evidence of habitation.
Exactly where our model says we should expect it:
Near the coast.
It was a crude form of pottery (low fired clay mixed with plant fibers), not the Jeulmun pottery neolithic Koreans would be known for, but finally pottery was found in Korea. And the geography supports it.
Now it becomes much harder to argue that the Korean Neolithic simply “began” around 6,000 BCE.
The transition probably began much earlier.
As for pottery on the peninsula?
Instead of digging for Stuff, we may have to swim for it.
Soil (Land): Yes (settled, but on Jeju Island)
Stuff (Artifacts): Yes
Blood (DNA): No evidence (not a flat out No)
Words (Language): No
Books (Written records): No
Are they Korean ('Us')?
→ I say Yes...Maybe???
But this raises a few questions for us:
- Can Koreans claim people who lived in the Yellow Sea or on Jeju Island as their own?
- Can Koreans claim people who left no DNA trace in modern Koreans as their own?
- Can Koreans claim people who left no trace of their language as their own?
None of them have easy answers, and they get more difficult as we progress through time. But at least we have a possible backstop to our time problem.
6000 BCE to 668 CE.
We have our first possible "Korean", depending on the definition you choose, and your comfort level. No trace left in Blood, Words, or Books. But they left Stuff, and they stayed.
- ה -
Early Neolithic: c. 6,000–3,500 BCE
By around 6,000 BCE, something had clearly changed.
First, the Bohai and Yellow Seas were oceans again. The Ice Age had ended, water levels rose, and megafauna went extinct. Korea became a peninsular, and Japan an Island.
Second, people were no longer merely passing through Korea. They were staying. Not everywhere, and not necessarily forever, but long enough to build houses, dig storage pits, make pottery too heavy and fragile to carry around, and return to the same rivers and coastlines generation after generation.
This matters to us because under our definition of "Korean", simply stepping onto Korean Soil is not enough. You have to stay, and you have to leave something behind.
Across the peninsula, that is exactly what begins appearing. Along the Han River at Amsa-dong, people built clusters of pit houses directly beside the river. Further north, communities settled around the Daedong River. On the eastern coast, people at Osan-ri built homes beside a lagoon.
Different places. Different pottery. Different ways of finding food.
But the same important change:
People were beginning to make particular places home.
Their houses make that obvious. They dug floors into the earth, drove wooden posts into the ground, built roofs over them, made hearths, and dug storage pits. These were not things you built for somewhere you expected to leave tomorrow.
And then there were the pots.
By now pottery becomes common, which makes sense. If you're constantly moving, pottery is horrible Stuff. It is heavy, fragile, and takes time to make. But once you stay, a pot becomes incredibly useful. You can cook in it, store food and water in it, leave it beside the hearth, and use it again tomorrow.
The pointed-bottom Jeulmun (즐문/빗살무늬; comb-pattern) pots found around places like Amsa-dong make the point almost comically well. They cannot even stand upright on a hard floor. But shove one into soft river sand or the ash around a hearth, and suddenly it works perfectly.
The pot was made for the place. And the place was becoming home.
But pottery is not the only Stuff they left us. There are heavy grinding stones, stone sinkers for fishing nets, axes, arrowheads, storage pits, shell ornaments, clay figurines, burials.
A Paleolithic hunter can drop a stone tool and disappear forever. A settlement leaves layers.
One house, then another. One hearth, then another. Broken pottery. Discarded tools. Food waste. Burials. People repeatedly living, eating, sleeping, raising children and dying in the same place.
And that brings us back to our actual question.
Were these people Korean?
We still have no Words, no Books, and almost no Blood. But now we have something we did not have before:
Continuity written into the Soil.
People lived here long enough to leave entire communities behind. And when their particular styles of pottery eventually disappeared, people did not disappear from the peninsula with them. New houses were built. New pottery appeared. New ways of getting food developed.
The Stuff changed.
But settlement continued.
Soil (Land): Yes (settled on the Korean Peninsula)
Stuff (Artifacts): Yes, lots of it
Blood (DNA): No evidence (not a flat out No)
Words (Language): No
Books (Written records): No
Are they Korean ('Us')?
→ Perhaps. More so than ever.
But compared with the people who merely passed through before them, it is getting much harder to say No.
If you're uncomfortable with calling the Gosanri people "Koreans", this time we have a later date for you.
3500 BCE to 668 CE.
///
Abrief conversation about Moving
I hate moving. I find it incredbily stressful.
packing
loading
driving
unpacking
arranging
address changes, finding new friends, saying good-bye to old.
And I have moving companies, cars, and all the modern aeminties that make moving easier.
Now imagine moving in 15,000 BCE.
If you were a hunter-gatherer you have no choice.
If the land no longer is able to sustain you, you have to move.
No one wants to carry evertything you and your family own on your back and trek the frozen tundra following herds of mammoth.
You didn't do it because moving was fun.
You did it because the alternative could be worse.
Stay. Die.
Move. Live
...at least until the next kill.
Now imagine you're a settled farmer in 2300 BCE.
Fields sustain you along with the land around you.
You've put generations of labor into your fields.
You know where all the good fishing spots are, where fruit and vegetables could be gathered.
You built a home for your family.
You own Stuff that you spent hours, weeks, months making.
And your entire support network is right there: uncles, aunts, grandparents, nephews, friends, kin.
Now imagine moving.
You would leave all that behind.
For what?
There has to be a damn good reason to move.
Survival is a damn good reason.
So why talk about moving now?
It's a problem that has been troubling me for some time.
People seem to have moved in waves into the Korean Peninsula.
They arrived, they had children. They raised families. Survived.
They left traces. At least, the traces that Korean soil, the Yellow Sea, and time didn't eat.
But when I look at modern Korean blood, some of the traces that I expect to find aren't there.
And some of the traces that are there, leave more questions than answers.
Why are they there, and not others?
Why did these people come here in the first place?
What happened to the people before them?
And if they disappeared, what happened to them?
Well, the answer to that question always starts with:
Someone moved.
///
Middle-Late Neolithic/Jeulmun: c. 3,500–1,500 BCE
By around 3,500 BCE, something genuinely new appears in the Korean archaeological record:
Millet.
Domesticated broomcorn and foxtail millet begin appearing at Jeulmun sites across the peninsula. The clearest evidence dates to around 3,500 BCE, although there are hints that millet may have arrived somewhat earlier.
This is important because it gives us our first experiment.
What happens when you give hunter-gatherers agriculture?
Apparently, not that much.
The Jeulmun people did not suddenly abandon fishing, shellfish, hunting, gathering, acorns and everything else they had been eating for thousands of years. They planted millet, harvested it and ate it, but millet seems to have been added to an existing broad-spectrum economy rather than replacing it. Even after cultivation became established, hunting, fishing and gathering remained enormously important.
And millet made that possible.
Millet is relatively forgiving. It grows quickly, tolerates dry conditions and poorer soils, requires relatively little infrastructure, and does not demand that an entire community reorganize itself around water management. You can plant some millet and still remain, fundamentally, a hunter-gatherer who happens to farm.
It provides food, but not enough.
Millet therefore increased the amount of food available without necessarily changing the entire structure of society.
Because later, when rice appears, everything changes.
So whatever happened next cannot simply be explained as:
Agriculture arrived.
Because, agriculture had already arrived.
The Jeulmun had been experimenting with farming for perhaps two thousand years.
And strangely, instead of producing ever larger agricultural villages, the later Jeulmun world begins to shrink.
Something had changed.
Around 2800 BCE, the warm wet conditions of the Holocene Climatic Optimum began to end in Korea.
2750 BCE (4.7 ka)
Korean pollen records show a sharp decline in trees around this time, consistent with reduced rainfall. Sea-surface temperatures in the western tropical Pacific also fell, probably weakening the transport of summer moisture onto the peninsula. In other words, Korea was becoming cooler, drier, and less predictable.
And it did not quickly bounce back.
One Korean climate reconstruction identifies the entire period from roughly 2800 to 1600 BCE as unusually dry. Within that long deterioration came a series of sharper shocks:
2250 BCE (4.2 ka)
1750 BCE (3.7 ka)
Cooling. Drying. Weaker monsoons. Failed crops.
The 4.2 ka event
Around 2250 BCE, was particularly severe and is much better known because societies across a huge part of the Old World were being battered at roughly the same time. Mesopotamia, Egypt, the Indus world and parts of China all experienced major climatic and massive social disruption.
Let's focus a little on China, because it matters to our Korea story.
To the west, the Longshan world (c. 2600–1900 BCE) was breaking apart. In Shandong it was followed by the poorer Yueshi Culture (c. 1900–1500 BCE).
Farther inland, something very different was happening.
Erlitou (c. 1900–1500 BCE)
↓
Erligang (c. 1600–1400 BCE)
↓
Shang (c. 1600–1046 BCE)
Cities grew larger. Bronze technology became increasingly sophisticated. Political organization became more complex.
But northeast of them, another world was developing.
By around 2000 BCE, Liaoxi and the West Liao River region were already developing their own fortified settlements and early bronze traditions, distinct from the Central Plains, in what archaeologists call the Lower Xiajiadian world (c. 2000–1400 BCE).
Across the Liao River in Liaodong, the picture was different again. Southern Liaodong belonged largely to the Shuangtuozi cultural world, while farther north the Machengzi tradition was developing.
That matters because Korea sat directly east of both worlds.
And the easiest land route between those worlds ran through Liaodong and the Yalu.
DNA shows some of this movement too. People carrying ancestry related to the farming populations of Shandong and the Yellow River moved northeast and mixed with populations of the West Liao region.
We can't know exactly what happened in Korea because reliable Korean DNA from this period is almost nonexistent. But a later Bronze Age Korean individual, and modern Koreans, carry substantial ancestry related to this same mixed West Liao population world.
Korea did not escape the 4.2 ka event either
Climate did not suddenly kill the Jeulmun.
Their decline had already begun.
Instead, imagine the 4.7 ka event as the beginning of a long squeeze.
A bad harvest is survivable.
Several bad harvests are harder.
A century in which rainfall is less reliable changes where you live, what you eat, how often you move, and how many people the land can support.
Do that for generations and an environmental problem becomes a demographic problem.
Settlements became smaller. Inland villages declined. Communities fragmented. Coastal and island settlements held on better, but by the Late Jeulmun the large settled world we saw earlier was disappearing.
And now our little conversation about Moving matters.
Since time immemorial,
When food becomes unreliable, people move.
When crops fail, people move.
When the land can no longer support everyone living on it, people move.
Sometimes they move peacefully.
Sometimes they move onto land occupied by someone else.
Sometimes the people already there have to move instead.
And sometimes they do not move at all. They lose.
Land gets taken. Communities break apart. Men die or stop leaving descendants. Women are incorporated into the population that comes next.
Then, around 1500 BCE, something new arrived.
Unlike millet,
Rice changed everything.
So:
Soil (Land): Yes, absolutely
Stuff (Artifacts): Yes, lots of it
Blood (DNA): No evidence (not a flat out No)
Words (Language): No
Books (Written records): No
Are they Korean ('Us')?
→ Still Perhaps. But definitely leaning towards yes.
- ה -
On writing history purely based on evidence (Stuff)
I originally wanted to write what comes chronologically:
Middle Neolithic (still Jeulmun) → Late Neolithic (Mumun) → Bronze (Late Mumun/early polities).
Pushing the time window forward, one archaeological timeframe at a time.
But I realized that something was missing. I talk about Stuff, but the Stuff don't tell stories.
We know a little about "What happened", we don't know "Why it happened".
Basing a story purely on academically peer-reviewed material exposes this crucial flaw. You tend to say only what can be proven. But footsteps on the sand do nothing about explaining who that person was:
Why was he there?
Why did he come?
Did he find some truth about himself?
Footsteps do not equal the incredible stories a person represents.
Stuff, especially stuff in rigorously researched, tested, and peer-reviewed academic literature, leaves out the best stories we care about. No wonder historians can't find jobs.
Let me be clear.
There is nothing wrong with writing history based on evidence. But I'm not an academic, and this Blog is not a research paper waiting to be published. I write to learn new things, formalize my thoughts, to tell stories.
Evidenced stuff tells the story of Stuff that survived time.
You need conjecture and reasoned theories to figure out what plausibly happened, including where direct evidence simply does not survive.
At some point you need to grow some balls and take a leap of faith.
So here are my balls:
1. Large populations and cultural systems do not simply evaporate.
If a widespread culture disappears, something happened to the people: continuity, migration, absorption, displacement, demographic collapse, replacement, or some combination.
2. DNA can disappear from a later gene pool, but that disappearance itself requires a demographic explanation.
Lineages can go extinct and ancestry can become diluted beyond detection. But because DNA is copied and recombined rather than simply blended away, populations usually leave genetic fragments in their descendants. People do not genetically vanish by magic.
3. Do not assume “assimilation” was peaceful.
If paternal lineages disappear much more strongly than maternal ones, sex-biased replacement becomes an obvious mechanism to investigate. Violence is one possibility, but so are confiscation of land, exclusion, reproductive inequality, enslavement, political subordination, patrilocal marriage, and anything else that prevents one set of men from leaving descendants while incorporating women into the succeeding population.
4. A small, one-off migrant population normally cannot transform a much larger resident population.
It will usually disappear, be absorbed, or remain marginal unless it receives repeated reinforcement, remains reproductively isolated, or gains an enormous reproductive advantage. Mumun (rice farmers) and later Yayoi (the same rice farmers moving to Japan) therefore requires migration at scale or sustained migration over time.
5. A migration carrying leaders, craftsmen, ritual specialists, farmers, and an entire technical package is organized migration.
Those people are embedded in communities and institutions. Their movement requires networks, planning, sponsorship, political fracture, catastrophe, or some other strong incentive.
6. Do not assume prehistoric people were peaceful. Hunter-gatherers and early farmers were human beings.
Conflict, raiding, coercion, conquest, exclusion, and organized violence belong on the list of explanations whenever the archaeological and demographic pattern calls for them.
Using this I wanted to revist my original questions:
1. Do "Koreans" exist
⇒
"Korean" or any other ethnic identity is a made-up construct. At it's core it requires:
"I am Korean"
"Yes, you are Korean"
+ reasonable evidence to claim
2. Where do Koreans come from?
⇒
Top → Down (upper limit established at): Unified Shilla, 668 CE (for now).
Bottom → Up (lower limit established at), depending on your definition you feel comfortable with:
- Incipient Neolithic, 6000 BCE (if your definition is "some form of settlement")
- Early Jeulmun/Early Neolithic, 3500 BCE (if your definition is "permanent settlement")
- Middle Jeulmun/Middle Neolithic, 2000 BCE (if your definition is "farming")
- Early Mumun/Early Bronze, 800 BCE (if your definition is "bronze appears")
- Gojoseon/Late Mumun/Middle-Late Bronze (1500-300 BCE) + Early Iron (300 BCE-), 108 BCE (if your definition includes "polity" or "state")
Soil (Land): Yes (all these people lived in the Korean peninsula)
Stuff (Artifacts): Yes (just different types of artifacts across time)
Blood (DNA): No, until Mumun arrive, then Yes, finally a plausible Yes
Words (Language): No, until Mumun arrive, then a strange Yes
Books (Written records): No, until Gojoseon is mentioned in the Guanzi (管子; ~c. 7th Century)
This leaves to some additional questions:
1. What happened to ancient hunter-gatherer (perhaps proto-Jomon) DNA, after Jeulmun settlers came carrying millet?
⇒
Before Jeulmun, the population seems to be a mixture of many people:
- Yokchido (욕지도; Tongyeong, South Gyeongsang Province):
95% Jomon-related - Yeondaedo (연내도; Tongyeong, South Gyeongsang Province) and Changhang (Gadeokdo Island, Busan):
Mixture of Jōmon and Northeast + Hongshan - Ando (안도; Yeosu, South Jeolla Province):
Northeast + Hongshan - Gaya (가야; 4-5th C Nakdong River Basin):
7-34% Jomon
2. What happened to pre-Mumun DNA (especially Y-DNA)?
⇒
Around 1500-1300 BCE: Jeulmun dramatically disappear, replaced with homogenous Mumun. But why rice vs millet. Was rice that much more "violent" in its disruption? Rice alone does not explain this.
Perhaps:
- Climate change?
- War and aggression?
- One wave, or two waves, or multiple waves?
If Jeulmun were "absorbed", why the collapse in Y-DNA, while mtDNA remain relatively intact?
Need to look into conflict in and around 800 BCE:
walled villages, weapons, signs of conflict, Songgukri.
3. What happened to the proto-Japonic language, Jeulmun/some branch of Mumun, after Bronze wielding later Mumun farmers came (assuming 2+ waves)?
⇒
- Around 1500-1300 BCE first Mumun pottery and rice arrives and spreads, while Jeulmun drastically disappears.
- Around 1300-1100 BCE small groups of Mumun/proto-Yayoi appear on Kyushu.
- Around 1100-900 BCE major migration to Kyushu happens, rice appears in Japan. Proto-Yayoi, speaking Japonic. Why did they decide to move, then? And why what look like elites: pottery makers, leaders, craftsmen etc.? In significant numbers to be not swalloed by the larger Jomon people.
- Around 800 BCE major push south of Bronze-wielding wet-rice farming Mumun arrive.
- Around 300 CE a second migration happens, starting the Kofun period: bronze and iron, different burial customs, statehood. Who were they? Gaya after Goguryeo kicked their ass? And why were Wa in Gaya attacking Shilla? What relationship do the Gaya have with Wa, and vice-versa?
In other words, what the fuck happened around 1300-800 BCE???
To answer this question, instead of moving into Mumun (Late Neolithic/Early Bronze), I'm going to go back to Shilla (668 CE), and move backwards. We've shaved roughly 699,000 years off our starting point.
Maybe we can shave a few centuries off the other end too.
- ה -
North-South Kingdom Period: c. 669-926 CE
OK. What the hell is this North-South Kingdoms bullshit all of a sudden?
Silla was one kingdom on the Korean Peninsula, but it was not the only Korean state of the period. To its north was Balhae, controlling a vast territory stretching from northern Korea across much of Manchuria and into what is now the Russian Far East. At its greatest extent, it also reached west into Liaodong.
Balhae was founded by people from the former Goguryeo world, alongside large Mohe populations, after Goguryeo collapsed in 668.
So if we want to understand this period, we cannot tell the story as Unified Silla alone.
South of the divide was Silla.
North of it was Balhae.
Together, they make up what Korean historians call the North-South Kingdoms Period.
---
Unified Shilla: c. 668-935 CE
We've already covered Unified Shilla so I won't go into the weeds.
Just to recap:
- Shilla allies with Tang to destroy Baekje in 660, then Goguryeo in 668.
- Shilla then fights Tang itself and drives Tang forces out of most of the peninsula.
- Unified Shilla lasts until 935.
- Its final monarch, King Gyeongsun (경순왕), peacefully surrendered to Wang Geon (왕건), founder of the Goryeo (高麗), whose name was deliberately inherited from Goguryeo (高句麗).
No one in their right mind would think Unified Shilla could not be claimed by modern Koreans as "Korean."
///
Because Unified Shilla has already been settled as "Korean", I originally didn't plan to cover their Stuff much. But while doing research I discovered two things that completely blew my mind, and I had to share it with you.
1. The Silla Village Register (신라촌락문서; 新羅村落文書)
In 1933, conservators at Tōdai-ji's Shōsōin in Nara, Japan were repairing the wrapper of a Buddhist text and discovered old Silla administrative paper pasted inside it as reinforcement. In other words, an obsolete government record appears to have become scrap paper, crossed to Japan, got glued into a book covering, and survived there for more than a thousand years.
What survived concerns four villages around Seowon-gyeong (西原京), modern Cheongju. And this is where it gets ridiculous.
The government recorded:
- households and their tax/administrative grade
- men and women, divided by age categories
- births, deaths and migration during the previous three years
- enslaved people
- horses
- cattle
- rice paddies
- dry fields
- hemp fields
- mulberry trees
- pine-nut trees
- walnut trees
- and changes in these things since the previous survey.
Not “this village grows mulberries.”
They counted the fucking trees.
Take Sahaejeom-chon (沙害漸村), one of the four villages. The surviving register gives:
- 11 registered household units
- 147 people
- 25 horses
- 22 cattle
- 1,004 mulberry trees
- 120 pine-nut trees
- 112 walnut trees
And then it says, essentially:
- Of those 1,004 mulberry trees, 90 were planted during the last three years; 914 were already there.
- Likewise, 34 of the pine-nut trees and 38 of the walnut trees had been added during that three-year interval.
That is nuts.
Some Silla clerk counted 1,004 mulberry trees in a village and wrote the number down.
And somebody came back three years later and updated the spreadsheet.
That last part matters enormously. This wasn't a one-time royal survey. The evidence indicates a three-year reporting cycle. The state was maintaining an administrative model of these communities over time.
There is another thing here. Silla did not merely count people. It classified households into nine grades. Exactly what determined the grade remains debated, but labor capacity, population and/or economic resources appear central. Those grades could then be converted into calculated administrative units used to assess what the village could provide to the state.
丁 / 丁女 (jeong): 29 / 42
助子 / 助女子 (jo): 7 /11
追子 / 追女子 (chu): 13* / 9
小子 / 小女子 (so): 16* / 16
除公 / 除母 (je): 1 / 2
老公 / 老母 (no): 0 / 1
* These totals include two people separately recorded as having moved into the village during the previous three years: one 追子 and one 小子.
And it gets even better: the document identifies enslaved people inside those categories. There were at least 9 enslaved people among the 147: 2 males and 7 females.
The categories were basically administrative age/labor brackets. The exact age boundaries are debated, but 丁 was the crucial working/tribute-bearing category; the younger groups stepped downward from it, while 除 and 老 marked older people with reduced or removed labor obligations.
But there is a second grading system. The 11 households themselves were ranked:
中下 (middle-lower): 4 households
下上 (lower-upper): 2 households
下下 (lower-lower): 5 households
So every single household in Sahaejeom-chon fell in the lower half of Silla's nine-grade household hierarchy.
The full 9 grade hierarchy is:
上上 (upper-upper)
上中 (upper-middle)
上下 (upper-lower)
中上 (middle-upper)
中中 (middle-middle)
中下 (middle-lower)
下上 (lower-upper)
下中 (lower-middle)
下下 (lower-lower)
So when we say:
Unified Silla was a state.
This document lets us see what that actually means.
There is somebody in Cheongju, hundreds of kilometers from Gyeongju, who knows that a particular village has 147 people, 22 cattle, 25 horses and 1,004 mulberry trees.
That information gets aggregated upward through an administrative hierarchy and apparently reaches a higher regional or central office. The National Institute of Korean History argues that the surviving version was probably compiled above the village/county level and ultimately reported toward the central administration.
That is state capacity in a very concrete sense.
2. Hwangnamdaechong Gold crown (황남대총 금관)
No doubt it's a fabulous crown. Blah, blah.
But what's fascinating about this crown are two things:
(1) The Jade 곡옥 (the green things dangling on the crown)
Called gok-ok, literally "curved jade", they are a comma-shaped piece of jade jewelry
What's interesting is that curved jade ornaments have a very deep history to the north. More than 3,000 years before this crown, the Hongshan culture of Liaoning was already producing distinctive curved and coiled jade ornaments for elite burials.
That's far too large a gap to claim that Silla 곡옥 came directly from Hongshan.
But the recurring combination of jade + curved form + elite burial is hard not to notice.
Stuff connecting people across enormous stretches of time.
(2) The gold antlers (녹각형)
If you look at the Hwangnamdaechong crown above, it has three central tree-like uprights and two lateral uprights, convetionally identified as deer antlers.
Across the Eurasian steppe, deer and antlers had been important ritual imagery for centuries. At Pazyryk in the Altai, horses were actually buried wearing elaborate masks that transformed them visually into antlered deer.
But we don't even have to go that far.
In Inner Mongolia and Liaoning, archaeologists have found Xianbei-associated elite headdress ornaments from roughly the 3rd–5th centuries CE combining animal or deer forms, large branching antlers or trees, and thin gold leaves suspended from them that shimmered as the wearer moved.
Stuff connecting people across enormous stretches of time.
I didn't want to say this again, but
Liaoning, again.
///
Soil (Land): Yes
Stuff (Artifacts): Yes
Blood (DNA): Yes
Words (Language): Yes
Books (Written records): Yes
Are they Korean ('Us')?
→ Yes
c. 668 upper band reconfirmed.
---
Balhae: 698-926 CE
Goguryeo fell in 668 CE.
But Goguryeo did not disappear.
Its people were scattered across the north. Some remained in former Goguryeo territory. Some fled south. Others were forcibly relocated west into Liaoxi. Mohe communities that had lived inside, beside, and under Goguryeo were scattered with them.
Then, in 696, a Khitan rebellion exploded around Yingzhou, near modern Chaoyang.
In the chaos, groups of Goguryeo refugees and Mohe broke away and began moving east.
Two leaders emerged: a refugee Goguryeo general, Geolgeol Jungsang (걸걸중상; 乞乞仲象), and the leader of the Baishan Mohe, Geolsa Biu (걸사비우; 乞四比羽). As a Korean, these are very strange sounding names.
Tang forces pursued them.
Geolsa Biu was killed. Geolgeol Jungsang died soon afterward.
Leadership passed to Geolgeol Jungsang’s son:
Dae Jo-yeong (대조영; 大祚榮).
(apparently Geol means Big, and the Chinese word for Big is Dae (大)...so Geolgeol became Dae)
In 698, Dae Jo-yeong gathered the surviving Goguryeo and Mohe forces and met the pursuing army of Li Kaigu (이가고; 李楷固) at Tianmenling (천문령; 天門嶺).
He destroyed it.
That victory gave the refugees what they needed most:
Space.
Dae Jo-yeong continued east into the mountain basins around Dongmo Mountain (동모산; 東牟山), near modern Dunhua, and founded a new state called Jin (진; 震).
Within fifteen years, that state would be known as:
Balhae.
And this was not some tiny refugee kingdom hiding in the mountains.
It expanded.
Fast.
Goguryeo communities joined it. Mohe communities joined it. Other peoples were conquered, absorbed, allied with, or incorporated as Balhae pushed outward across northeast Asia.
By the eighth century, Balhae controlled a vast territory stretching across northern Korea, Manchuria, and into what is now the Russian Far East.
Eventually it would administer that space through:
5 capitals
15 prefectures
62 provinces
making it the largest territorial state in Korean history.
And Balhae was very clear about where it thought it came from.
In 727, King Mu sent an embassy to Japan.
復高麗之舊居，有扶餘之遺俗
His letter described Balhae as having:
“recovered the old territory of Goguryeo”
and preserving:
“the customs of Buyeo.”
That is not a claim made by modern Korea.
That is Balhae describing itself.
King Mu
Dae Jo-yeong died in 719 and was succeeded by his son Dae Mu-ye (대무예; 大武藝), remembered as King Mu (무왕; 武王).
And King Mu was aggressive as hell.
To Balhae’s northeast lived the Heishui Mohe (흑수말갈; 黑水靺鞨) around the lower Songhua and Amur systems. Unlike many other Mohe groups, the Heishui remained independent.
When their relationship with Tang grew closer, King Mu saw a problem immediately. A powerful outside state was establishing influence among independent tribes sitting directly behind Balhae’s northeastern frontier.
So he decided to attack them.
His younger brother Dae Mun-ye (대문예; 大門藝) argued against the campaign, warning that it could trigger a much larger war.
Mu ordered it anyway.
Mun-ye defected.
The family argument now became an international crisis.
Then, in 732, Balhae did something rather spectacular.
King Mu sent Jang Mun-hyu (장문휴; 張文休) across the Bohai Sea and attacked Dengzhou (등주; 登州) on the Shandong Peninsula. The major port connecting northern China to the northeast Asian maritime world.
Balhae captured it. The governor, Wei Jun (위준; 韋俊), was killed.
At roughly the same time, Balhae was fighting farther west in Liaoxi and maintaining relationships with Khitan and other northern powers.
The response came from multiple directions.
Tang Forces attacked from the west.
Shilla sent an army north in 733.
Both expeditions failed.
Balhae survived and prospered.
Only thirty-five years after Tianmenling, the refugees hiding around Dongmo Mountain had become strong enough to raid Shandong, fight across Liaoxi, resist attacks from the south, and contest control of the northeast.
This was no longer a refugee state.
It was becoming an empire.
King Mun
King Mu died in 737 and his son King Mun (문왕; 文王) inherited a very different Balhae.
Mu had secured its survival.
Mun spent more than fifty years turning it into a state capable of governing what Mu had conquered.
The administration expanded.
Cities grew.
The capital moved.
Balhae developed a system of five capitals, connected by roads across an enormous territory. Government was organized into ministries, provincial administrations, military commands, and local jurisdictions capable of holding together populations spread from Liaodong to the Sea of Japan.
The great capital at Shangjing (상경; 上京) eventually became one of the largest planned cities in northeast Asia.
King Mun also opened Balhae much more aggressively to diplomacy and trade.
Relations with Tang stabilized.
Balhae students travelled abroad to study.
Books, Buddhism, administrative ideas, art, technology, textiles and luxury goods flowed into Balhae.
But Balhae was not simply looking west.
It was also looking east across the sea.
Balhae and Japan
According to the (whatever Japanese book) Balhae had first contacted Japan in 727, and that relationship became one of the most persistent diplomatic connections in its history.
Between 727 and Balhae’s collapse, Balhae sent 34 recorded embassies to Japan. Japan sent 13 in return.
This was diplomacy.
It was also trade.
Balhae ships carried furs, hides, ginseng, horses and other northeastern products across the Sea of Japan. Japanese goods returned in the opposite direction. By the ninth century, Balhae envoys were trading not only with the Japanese court but directly with aristocrats and merchants.
And politics occasionally made the relationship wonderfully stupid.
In the late 750s, Japan wanted to attack Shilla.
Again.
Japan tried to bring Balhae into a joint invasion, sending repeated missions between 758 and 764.
Balhae never committed.
The invasion never happened.
Then came the family argument.
In 771, King Mun sent Japan a diplomatic letter describing the relationship between the two rulers not as brothers, but as uncle and nephew.
Japan objected.
According to the Japanese court, Goguryeo and Japan had previously addressed each other as brothers. Balhae had changed the relationship.
Japan demanded the wording be changed.
The Balhae delegation eventually rewrote the letter, everyone stopped arguing about imaginary family relationships, and diplomacy continued.
Except these weren't really imaginary family relationships.
To us, calling another king your brother, uncle, or nephew sounds like ceremonial nonsense.
It wasn't.
In a world organized heavily through family and Confucian ideas of Li (禮), family relationships described order. Father and son. Older brother and younger brother. Uncle and nephew. Each relationship carried different expectations about seniority, responsibility, respect, and obligation.
Change the relationship and you change where everyone stands.
The Japanese response actually says exactly that:
於禮失矣
(“This violates proper order.”)
Japan was angry enough to stop the formal reception of Balhae's embassy until the letter was corrected.
And buried inside that argument is something much more important.
Japan wasn't merely calling Balhae “Goguryeo.”
It expected Balhae to inherit Goguryeo's relationship with Japan.
Japanese records referred to Balhae as Goguryeo, its envoys as Goguryeo envoys, and its rulers as Kings of Goguryeo. When Balhae tried to redefine the relationship, Japan reached backward past 668 and said, essentially:
No. We already know what our relationship is. We had it with Goguryeo.
That tells us considerably more than the name written at the top of some old Japanese chronicle.
"I am Goguryeo."
"Yes, you are Goguryeo."
+ reasonable evidence to claim: The Shoku Nihongi; 續日本紀 book 32
King Seon
After King Mun died in 793, Balhae went through twenty-five years of instability.
Six rulers came and went.
Then, in 818, King Seon (선왕; 宣王) took the throne.
Under Seon, Balhae expanded again.
Its armies pushed outward across the northeast, bringing additional Mohe territories under Balhae control and extending the state westward into Liaodong and eastward toward the coast.
By the ninth century, Balhae stretched from the northern Korean Peninsula through Jilin and Heilongjiang, across the Changbai Mountains, through the Mudanjiang and Tumen basins, and into Primorye.
Different peoples lived inside it.
Goguryeo descendants.
Mohe.
Khitan.
Shiwei.
Turkic populations.
Communities that had previously belonged to entirely different political worlds were now subjects of the same state.
This is also when Balhae became one of the great commercial crossroads of northeast Asia.
Routes ran:
- west toward Liaodong and the continental interior,
- northwest toward Khitan and the steppe,
- south toward Shilla,
- east through Hunchun and Primorye,
- and across the sea to Japan.
Balhae was not sitting at the end of Asia.
It was sitting in the middle of several different Asias.
Its wealth, cities, administration and cultural life became famous enough that later histories remembered it as:
해동성국 (海東盛國)
"The Flourishing State of the East."
And then it disappeared.
926
In late 925, the Khitan ruler Yelü Abaoji (야율아보기; 耶律阿保機) invaded Balhae.
He moved quickly.
Balhae’s western defenses collapsed.
The Khitan army advanced on the capital.
In early 926, King Dae In-seon (대인선; 大諲譔) surrendered.
Balhae was gone.
A state that had existed for more than two centuries and ruled an enormous part of northeast Asia collapsed in a matter of weeks.
Why?
We don't really know.
There had been political instability.
Khitan power had grown enormously.
Balhae may have been internally fragmented.
Its western defenses may simply have failed catastrophically.
The famous theory that Mt. Baekdu destroyed Balhae does not work very well. The enormous Millennium Eruption appears to have occurred after Balhae had already fallen.
So we know who delivered the final blow: Khitan.
We understand much less clearly why the empire suddenly proved so vulnerable.
But again:
the state disappeared.
The people didn't.
Balhae refugees began moving south into Goryeo.
The most famous migration came in 934, when Dae Gwang-hyeon (대광현; 大光顯), a Balhae prince, entered Goryeo with a huge body of followers.
The King of Goryeo, Wang Geon, welcomed them. Dae Gwang-hyeon was given the royal surname Wang, entered into Goryeo’s ruling order, and his followers were given land, homes and offices.
Balhae did not simply vanish from Korean history in 926.
Part of it walked south.
And part of it remained in the north.
Which creates one last wonderfully inconvenient fact.
When the Jurchens later rose against the Khitan, the founder of the Jin dynasty, Wanyan Aguda, appealed to Balhae people by declaring:
“Jurchen and Balhae were originally one family.”
So Balhae could be remembered as Goguryeo by Japan.
Absorbed as kin by Goryeo.
And claimed as family by the Jurchens.
That is where the easy story ends.
Because now we finally have enough Balhae on the table to ask the question that actually matters:
What the hell does it mean to call Balhae "Korean"?
Not because the claim is unreasonable.
Quite the opposite.
The evidence for a Goguryeo connection is substantial.
But Balhae is the first state in this story that forces Soil, Blood and Words apart.
Most of the empire lay outside the Korean Peninsula.
Much of its population was Mohe.
Many of those people almost certainly did not speak Koreanic.
And suddenly our story has moved decisively into three things we can no longer avoid:
Manchuria.
Mt. Baekdu.
The Mohe.
So,
Soil (Land): Yes
Stuff (Artifacts): Yes
Blood (DNA): Yes
Words (Language): Yes
Books (Written records): Yes
Are they Korean ('Us')?
→ Yes
That part is easy.
But Balhae is also:
Mohe.
Jurchen inheritance.
Chinese history.
Some weird familial relationship with Japan.
and because of modern borders, Russian Far Eastern history.
None of those claims cancel the others.
From a timing perspective, nothing changes. We are still at the upper limit of 668 CE.
And oh yeah, this one is so obvious:
Liaoning, again.
---
The Three Kingdoms Period: ~c. 57 BCE-668 CE
Before Shilla allied with Tang to unify most of the Korean Peninsula, there was a long stretch of time when the Korean world was divided among multiple states and polities.
There were others, but three eventually dominated:
Goguryeo. Baekje. Shilla.
And for much of its early history, Shilla was by far the weakest of the three.
Historians call this the Three Kingdoms Period (삼국시대; 三國時代).
It is also where much of what modern Koreans recognize as ancient Korea really begins to take shape: its kingdoms, heroes, wars, legends, royal lineages, Buddhism, tombs, art, and founding stories.
In that sense, the Three Kingdoms Period is the spiritual home of Korean history.
The sequence in which we introduce each state is not important, since they all pretty much started around the same period (plus the starting dates are all mythical and involve magial turtles, glowing eggs, and quite a few angel-like beings).
Let's start with Gaya, because it (they?) tell the most interesting story when it comes to talking about "Koreans."
---
Gaya: ~42 CE-562 CE
If Korean high school history were a word cloud, with font size representing relative importance, it would probably look something like this:
Shilla. Goguryeo. Baekje. Balhae. Gaya. Okjeo. Dongye.
Goguryeo and Shilla were chapters.
Baekje was a section.
Balhae got a few pages.
Gaya got half a page.
Okjeo and Dongye were footnotes.
So before researching Gaya for this section, I knew practically nothing about it beyond the names and roughly where it existed.
Calling Gaya a kingdom is probably wrong.
Calling it a state is not quite right either.
But these were not a bunch of hunter-gatherers fishing along the Nakdong River.
Gaya was a collection of wealthy, stratified polities with rulers, armies, monumental tombs, long-distance diplomacy, specialized craftsmen, and an enormous international trade network. What it never managed to do was centralize into a single state the way Goguryeo, Baekje, and Shilla eventually did. Even the archaeology reflects this. The great Gaya cemeteries share recognizable burial traditions and material culture, while each regional center maintains its own distinctive style.
Before Gaya, this region was called Byeonhan (변한; 弁韓).
The Sanguozhi describes Byeonhan as twelve small polities. Over the following centuries, some of those polities grew into the larger political centers we now call Gaya.
There was no morning when everyone woke up and said:
Cool. We're Gaya now.
The transition was gradual.
But somewhere near the center of that transformation was one thing:
Iron.
One of the twelve Byeonhan polities was Guya-guk (구야국; 狗邪國), centered around modern Gimhae.
Guya-guk eventually became Geumgwan Gaya (금관가야; 金官伽倻).
And Geumgwan Gaya got rich.
Filthy rich.
The iron trade was important enough that some random dude sitting in China actually bothered to write it down in the third-century Sanguozhi:
國出鐵，韓、濊、倭皆從取之。又以供給二郡。
(“The country produced iron; Han, Ye, and Wa all came to obtain it. It was also supplied to the two commanderies.”)
Han (韓) here means the Samhan, not Han Chinese (漢). Confusing, I know.
Ye (濊) means the Ye people farther north along the eastern peninsula.
Wa (倭) means the peoples and polities of the Japanese archipelago, increasingly connected to the political world that would become Yamato.
And the two commanderies were Lelang (낙랑; 樂浪) and Daifang (대방; 帶方), the remaining Han Chinese (漢) commanderies on the peninsula.
Gaya iron was traded so heavily that standardized iron ingots appear to have functioned partly as money. Archaeologists keep digging them out of Gaya tombs, while Gaya iron and ironworking technology turn up across the Korea Strait in Japan.
But Geumgwan Gaya's real advantage was not merely that somebody found iron nearby.
Look at Gimhae.
It had:
- access to the Nakdong River,
- a protected bay,
- easy maritime access to Tsushima and Kyushu,
- iron flowing down from the interior,
- and access westward along the entire southern coast.
It sat almost perfectly between the inland production zones and the sea.
Geumgwan Gaya became the middleman.
And middlemen get rich.
And because this is the Stuff section, Gaya left us some pretty fabulous stuff.
- Iron swords.
- Iron spearheads.
- Iron armor.
- Iron helmets.
- Horse armor.
- Horse helmets.
- Horse trappings.
- Agricultural tools.
- And piles of standardized iron ingots.
But Gaya was not just iron.
Its pottery is beautiful and incredibly distinctive.
High-footed dishes, long-necked jars, enormous vessel stands, boat-shaped pottery, wagon-shaped pottery, and increasingly elaborate ceremonial objects appear across the Gaya world.
And the pottery actually lets us see the political structure.
Ara Gaya, around Haman, loved high-footed dishes with distinctive flame-shaped perforations.
Daegaya, around Goryeong, developed its own pottery, weapons, and elaborate metalwork.
Other Gaya polities developed their own regional styles.
Same cultural family.
Different political communities.
Then there are the tombs.
- Daeseong-dong in Gimhae.
- Marisan in Haman.
- Okjeon in Hapcheon.
- Jisan-dong in Goryeong.
- And others scattered across the south.
They are huge elite cemeteries filled with weapons, armor, horse equipment, pottery, jewelry, imported luxury goods, and evidence of long-distance exchange.
And buried among the Gaya stuff is everybody else's stuff.
- Chinese objects.
- Northeastern continental objects.
- Wa pottery.
- Wa weapons.
- Wa horse equipment.
Meanwhile, across the sea in Japan, archaeologists find Gaya pottery, iron ingots, weapons, metalworking technology, crowns, jewelry, and horse gear. Gaya ceramic technology also contributed heavily to the development of Japanese Sue ware.
This was not occasional trade.
This was a system.
And then around 400 CE, Geumgwan Gaya apparently got a little too confident.
By this point, Geumgwan Gaya and Wa forces were putting serious military pressure on Shilla around the lower Nakdong region.
Shilla panicked and called Goguryeo.
King Gwanggaeto was only too happy to help.
In 400 CE, Goguryeo sent 50,000 infantry and cavalry south.
They drove the Wa forces out of Shilla and chased them deep into Imna-Gara (任那加羅), generally associated with the Geumgwan Gaya world.
Whatever Geumgwan Gaya thought was about to happen, that wasn't it.
The old Geumgwan-centered order never really recovered.
Its dominance collapsed and the political center of the Gaya world began shifting elsewhere.
First toward:
Ara Gaya, around Haman.
And eventually:
Daegaya, around Goryeong.
Gaya did not disappear.
It reorganized.
But the window in which Geumgwan Gaya might have become the dominant state of southeastern Korea had closed.
In 532, Geumgwan Gaya itself was absorbed into Shilla.
Other Gaya states survived.
Daegaya became the strongest of them.
Then, in 562, Shilla conquered Daegaya.
Gaya disappeared from the map.
But again:
States disappear.
People don't.
Gaya's ruling families, soldiers, craftsmen, and population were absorbed into Shilla.
The ultimate irony?
Kim Yoo-shin (김유신; 金庾信), the Shilla general who later helped destroy Baekje and Goguryeo, was descended from the royal family of Geumgwan Gaya.
Talk about an epic revenge story.
But unfortunately for the iron, the tombs, and all that fabulous stuff, none of that is the main reason Gaya matters to this series.
Gaya matters because it may represent one of the last places where we can still glimpse something that has completely disappeared from Korea today:
Japonic.
1. The last bastion of Japonic?
Remember when we talked about how place names can survive long after the people who named them disappear?
This is where that becomes interesting.
We have almost nothing written in the Gaya language.
Almost nothing.
But one actual Gaya word survives in the Samguk Sagi:
加羅語謂門爲梁云
(“In the Gaya language, ‘gate’ was called 梁.”)
The exact ancient pronunciation is difficult to reconstruct, but linguists have compared the word with Old Japanese to, meaning:
door / gate.
Then there are the older Byeonhan place names preserved phonetically in Chinese.
One of them is:
彌烏邪馬
usually reconstructed approximately as Mioyama.
That final -yama looks suspiciously like Proto-Japonic:
yama
meaning:
mountain.
Which is still Japanese yama (山) today.
And this is not occurring randomly somewhere in Manchuria.
It is sitting in exactly the southern peninsula where our earlier linguistic and archaeological reconstruction predicted Japonic-speaking populations should have survived.
Is that enough to say:
Gaya spoke Japonic?
No.
We simply do not have enough surviving language.
But it is enough that I am not comfortable pretending the question doesn't exist.
If Gaya was entirely Koreanic by this period, then somebody still needs to explain why Japonic-looking linguistic debris survives precisely here.
And then the linguistic evidence runs straight into something considerably harder to ignore.
2. Gaya and Wa
Gaya and Wa were close.
Not:
“Hey, we're trading partners.”
close.
Really fucking close.
- Tons of Gaya stuff appears in Japan.
- Tons of Wa stuff appears in Gaya.
- Finished objects moved.
- Iron moved.
- Pottery moved.
- Armor moved.
- Horse equipment moved.
And most importantly:
concrete evidence that people moved.
And not just small family. An entire system of people.
Recent work on Gaya-Wa iron exchange finds not merely Gaya iron objects in Japan, but the transfer of Gaya craftsmen and manufacturing technologies into the archipelago. Japanese-made weapons and armor traveled the other way.
By the sixth century, Wa representatives were even living for extended periods in Ara Gaya, participating in the absurdly complicated diplomacy among Gaya, Baekje, Shilla, and Wa. Those resident envoys are part of the historical residue that the Nihon Shoki would later inflate into the famous Mimana Nihonfu (任那日本府).
Japanese imperial historians later turned that into:
Japan ruled southern Korea.
That colonial theory is bullshit.
But the people weren't invented.
Wa representatives really were living in Gaya.
Gaya and Wa really were conducting diplomacy through one another.
People really were crossing the strait constantly.
The lie was turning an extraordinarily entangled relationship into a Japanese colonial government.
And this takes us back much farther.
Earlier in this series, we talked about the large movement from the Mumun world of southern Korea into Japan around the beginning of the Yayoi period.
People crossed.
Rice crossed.
Pottery crossed.
Technology crossed.
DNA crossed.
That was the first great movement:
Mumun → Yayoi
Then, centuries later, during the Kofun period, it happens again.
Not the same people.
Not the same circumstances.
But another enormous stream of people and technology moves from the Korean Peninsula into Japan.
This time the archaeological fingerprints are increasingly:
Gaya.
Baekje.
Ironworkers.
Potters.
Horse culture.
Armor.
Weapons.
Elite families.
Craft specialists.
And this movement accelerates during exactly the centuries when the southern Korean political world is being ripped apart by Goguryeo expansion.
But the fourth and fifth centuries bring another enormous pulse of continental people and technology into the archipelago.
And suddenly the 400 CE destruction of Geumgwan Gaya's old political order becomes interesting for reasons that have nothing to do with who won the battle.
Where did the losers go?
There was a very rich, extremely maritime population sitting directly across from Japan, already connected to Wa by trade, family, craftsmen, diplomacy, and war.
Then Goguryeo smashed its political center.
People don't disappear.
3. Blood
And then there is DNA. Which is where this gets weird.
We actually have ancient genomes from fourth- and fifth-century Gaya.
Eight individuals have been sequenced from Gimhae, including people buried at the great Daeseong-dong cemetery, the royal burial complex of Geumgwan Gaya.
Genetically, they show strong continuity with modern Koreans.
No surprise there.
But they were not genetically uniform.
Some carried substantial Jōmon-related ancestry, the ancestry associated with the older populations of the Japanese archipelago.
Meanwhile, the genetic evidence from Japan points in the opposite direction.
A recent whole-genome study of Yayoi, Kofun, and modern Japanese ancestry found that the majority of continental migrants entering Japan from the Yayoi through Kofun periods came primarily from populations related to those of the Korean Peninsula. The ancient Gimhae Gaya samples are themselves genetically unusually close to the Japanese cluster, partly because of that Jōmon-related ancestry.
That does not mean I can point at one skeleton in Gimhae and say:
His grandmother was from Kyushu.
But look at the whole picture.
The Chinese records tell us they traded.
The tombs contain each other's stuff.
Their craftsmen moved.
Their diplomats lived among each other.
Their armies fought together.
Their languages may have been related.
And now their dead contain genetic traces pointing across the same strait.
People were moving in both directions.
The iron already told us that.
The pottery told us that.
The tombs told us that.
Now the dead are telling us too.
And that is why this little confederacy I got half a page about in high school suddenly matters enormously to the question we are actually trying to answer.
Because Gaya sits exactly where several stories collide:
Mumun → Yayoi
Byeonhan → Gaya
Gaya ↔ Wa
Gaya/Baekje → Kofun
and finally:
Shilla → Koreanic
By 562, Gaya was gone.
Eventually Japonic was gone from the Korean Peninsula too.
It survived across the Korea Strait and became Japanese.
So the question Gaya leaves us with is not really:
What happened to Gaya?
We know what happened to Gaya.
Shilla ate it.
The much harder question is:
What happened to the people who were living there before everyone became Shilla?
Soil (Land): Yes
Stuff (Artifacts): Yes
Blood (DNA): Yes
Words (Language): Maybe not?
Books (Written records): Yes
Are they Korean ('Us')?
→ Yes
---
Shilla: ~c. 57 BCE-668 CE
Shilla is boring.
Until it suddenly isn't.
Born out of Jinhan (진한; 辰韓), one of the Samhan confederacies, Shilla began as Saro-guk (사로국; 斯盧國), one small polity among twelve clustered around modern Gyeongju in the southeastern corner of Korea.
Its traditional founding date is 57 BCE.
For the next several centuries, it was the runt of the pack.
Baekje was richer.
Gaya had iron and Wa.
Goguryeo was becoming an empire.
Shilla had mountains.
Then, around 400 CE, Gaya and Wa forces began putting serious pressure on Shilla.
Shilla called Goguryeo.
King Gwanggaeto sent 50,000 infantry and cavalry, smashed the invading forces, and chased them all the way into the Gaya world.
Then Goguryeo stayed.
For a while, Shilla came damn close to being swallowed up by Goguryeo.
Goguryeo troops were stationed inside Shilla.
Goguryeo interfered directly in Shilla politics.
When Silseong returned from years as a hostage in Goguryeo in 401, he became king with Goguryeo support. The struggle between Silseong and the future King Nulji became tangled up with Goguryeo politics as well.
Goguryeo was not merely Shilla's friendly neighbor.
It had become deeply involved in deciding what happened inside the Shilla court.
And then Shilla started pushing back.
Under King Nulji (r. 417–458), Shilla gradually drove Goguryeo troops out and began slipping away from its control.
In 433, Shilla allied with its old enemy Baekje against Goguryeo.
By the middle of the fifth century, the relationship had deteriorated badly. In 450, a Goguryeo commander was killed in Shilla territory. Fighting followed. Goguryeo and Shilla were no longer patron and dependent.
They were enemies.
But Goguryeo was still much stronger.
So for another century, Shilla waited.
And grew.
It centralized.
It built fortresses.
It absorbed neighboring Jinhan territory.
It expanded north along the east coast.
It crossed the Sobaek Mountains.
And then, in the middle of the sixth century, Goguryeo finally gave Shilla the opening it had been waiting for.
Goguryeo was having a horrible time.
Its aristocrats were fighting one another in a succession crisis.
Thousands reportedly died in fighting around the capital.
At almost exactly the same time, Göktürks (돌궐; 突厥) appeared on Goguryeo's northwestern frontier and attacked its fortresses around Fushun and Liaoyang.
Goguryeo suddenly had problems in every direction.
So in 551, Baekje and Shilla attacked.
Baekje took the lower Han River.
Shilla took the upper Han.
Goguryeo lost the central peninsula it had dominated since Jangsu conquered Hanseong in 475.
Then Shilla looked at its ally Baekje and apparently thought:
Thanks. I'll take yours too.
In 553, Shilla seized the lower Han River from Baekje.
Baekje King Seong retaliated the following year.
He was killed.
Shilla now controlled the entire Han River basin, including the old Baekje heartland that Goguryeo had taken eighty years earlier.
And suddenly the boring little mountain state had something it had never possessed before:
direct access to the Yellow Sea.
Which meant:
China.
That changed everything.
Nine years later, in 562, Shilla finished absorbing Gaya.
A century after living under Goguryeo's shadow, Shilla now controlled most of southeastern and central Korea and had direct diplomatic access to the continental powers across the Yellow Sea.
From there we get the story we already know:
Shilla allied with Tang.
Baekje fell in 660.
Goguryeo fell in 668.
Then Shilla discovered that Tang had no intention of politely handing everything over and going home.
Shilla turned on Tang.
And won.
For the purposes of this series, that is really all the Shilla history we need.
They were piddly.
They got stronger.
They escaped Goguryeo.
They betrayed Baekje.
They ate Gaya.
They allied with Tang and destroyed Baekje and Goguryeo.
Then they turned around and fought Tang too.
Fucking survivors.
But none of that is what makes Shilla interesting to me.
What makes Shilla interesting is:
Words.
Why the hell is Shilla speaking Koreanic?
If the linguistic story we have built so far is even approximately right, southern Korea had once been heavily Japonic-speaking.
We just saw the remnants of that world in Gaya.
Across the Korea Strait, Japonic survives spectacularly.
It becomes Japanese.
But right beside Gaya, tucked into the southeastern corner of the peninsula, is Saro.
And Saro becomes Shilla.
Shilla speaks Koreanic.
That is weird.
Baekje gives us an obvious geographical story.
Northern populations move down the western side of the peninsula through the Han River world. They even say they come from Buyeo and Goguryeo.
Fine.
But Shilla is on the other side of the mountains.
It is nowhere near that route.
So how did Koreanic get to Gyeongju?
Look north from Gyeongju.
Not west. Follow the coast.
Shilla ← Dongye ← Okjeo ← Goguryeo / Buyeo world
Suddenly there is another route.
Not down the western plains.
Down the east coast, through the long chain of valleys between the Taebaek Mountains and the sea.
And unlike the western route, we already have ancient Chinese observers telling us that Okjeo spoke much like Goguryeo, while the broader Ye world was closely connected to the same northern population sphere.
That does not prove that a bunch of Okjeo refugees marched into Gyeongju one afternoon and invented Shilla.
But it gives us a perfectly good route for Koreanic to reach the southeast without ever crossing the central peninsula.
Now the language map starts doing something fascinating.
On the southern coast:
Gaya → Wa → Japonic survives in Japan.
Along the eastern coast:
Okjeo → Dongye → Shilla → Koreanic survives in Shilla.
By the end of the Three Kingdoms period, one of those languages has disappeared from the peninsula.
The other has conquered it.
And suddenly boring little Shilla becomes very interesting indeed.
As to their Stuff?
For such a piddly little kingdom, Shilla left behind some ridiculously impressive shit.
And now that we have a northern route into Shilla, some of their Stuff starts looking a little less random.
Take the famous Shilla gold crowns.
Trees.
Antlers.
Gold.
Their closest conceptual relatives are not sitting in Gaya or Kyushu. They point north, into a much larger northern Eurasian world.
Soil (Land): Yes
Stuff (Artifacts): Yes
Blood (DNA): Yes
Words (Language): Yes
Books (Written records): Yes
Are they Korean ('Us')?
→ Yes
Baekje: ~c. 18 BCE-660 CE
Glorious Baekje.
I don't say this lightly, because what they left behind is truly glorious. A version of Korea that speaks directly to my soul: Cosmopolitan. Accepting. Sharing. Expanding. Multi-cultural.
They easily pass the "Are they Korean, 'Us' test", with flying colors:
Soil: Yes
Located in the southwest Korean Peninsula, but its soul is in the Han River basin, modern Seoul.
Even their name for their capital was Hanseong, Seoul's old name.
Stuff: Yes
Left some of the greatest artifacts in Korean history.
Blood: Yes
Not that "Korean" DNA exists, but its pattern is undeniably Korean.
Words: Yes
Koreanic. With a Japonic underbelly.
Books: Yes
It's own records, plus rich records in both China and Japan
Are they Korean ('Us')?
→ Yes
Yet, Baekje is possibly the most ignored "Korean" in all of Korean history. Not because of who they were, but perhaps more because of where they ended up.
They were simply, inconvenient.
Let me explain:
1. They lost
Baekje collapsed to a joint Tang-Shilla army in 660 CE. They continued resisting until 663 CE, until they were finally defeated at the Battle of Baekgang. It's land absorbed, its people scattered: into Shilla, deported to Tang, and escaped to Wa.
Losers don't get to write their own story.
2. Their story is told by those outside Baekje
- Shilla unified the Korean Peninsula and ruled for a thousand years.
- Goguryeo fought valiantly against all odds, and maintained one of the largest empires in Korean history.
- Baekje made stuff and traded.
Not exactly the stuff made for legends.
Joseon claimed Shilla.
Goryeo claimed Goguryeo.
Modern Korean politicians claimed Shilla.
Modern Korean nationalists claimed Goguryeo.
No one claimed Baekje.
3. They're complicated
- A Buyeo founding story.
- Born inside Mahan.
- A trading powerhouse that connected the East Asian world.
- Diplomatically brilliant.
- A Koreanic-speaking elite governing a Koreanic/Japonic-speaking mixed citizenry.
Projecting beyond the Korean Peninsula.
When Baekje collapsed, the Wa even brought one of the surviving Baekje princes back to Korea, with tens of thousands of troops, in the hopes of reviving Baekje.
And as we saw in the Balhae section, after those revival attempts failed, they went all the way north to Balhae, hoping to enlist their help in attacking Shilla (Balhae refused).
4. They make everyone uncomfortable: Koreans, Chinese, and Japanese
Look at this ridiculous map (you can decide whether it's truly "ridiculous" for yourself).
- Goguryeo is a little larger than comfort,
- Buyeo is massive,
- and apparently Gaya controls a huge swath of Kyushu.
- Southwest Korea, Yes
- Most of Japan, Yes?
- Majority of China's east coast, Yes??
Chinese Records
Chinese records make some extraordinary claims about Baekje.
The Song Shu (宋書), completed in the fifth century, says quite plainly that after Goguryeo expanded into Liaodong:
Goguryeo took Liaodong, and Baekje took Liaoxi.
It then says that the place governed by Baekje was called Jinping Commandery, Jinping County (晉平郡 晉平縣).
The Liang Shu (梁書) later gets even more explicit. It says that during the Jin period, while Goguryeo occupied Liaodong,
Baekje:
Occupied the lands of Liaoxi and Jinping commanderies
and established its own Baekje Commandery (百濟郡) there.
That's not the Samguk Sagi.
That's not Shin Chae-ho.
That's not some dude on YouTube drawing Greater Baekje maps.
Those are Chinese dynastic histories saying it.
Now, whether Baekje actually possessed a chunk of Liaoxi is a completely different question. There is no nice archaeological Baekje governor's office waiting for us in Liaoxi, and the northern Chinese histories don't give us the same clean corroboration. So historians have variously interpreted this as real occupation, temporary military activity, a maritime enclave, displaced communities, confused geography, or simply an erroneous tradition that one southern history copied from another.
But the record exists.
And then it gets stranger.
The Nan Qi Shu (南齊書) preserves petitions from Baekje's King Dongseong in which Baekje officials carry, or are nominated for, Chinese territorial titles including:
Guangyang 太守
Daifang 太守
Guangling 太守
Qinghe 太守
Lelang 太守
Chengyang 太守
Joseon 太守
And the Southern Qi emperor actually confirms several of them.
Now, this does not prove Baekje governed all those places. Chinese courts handed out nominal and honorary territorial offices all the time.
But Guangling was around modern Jiangsu. Chengyang was associated with Shandong. Qinghe belonged to the northern Chinese geographical world.
So what the fuck is a Baekje king doing handing his officials titles like Governor of Guangling and asking a Chinese emperor to recognize them?
That deserves an explanation.
And it gives your ridiculous map its first little seed of reality.
Then there is southern China.
The Old Book of Tang describes Baekje geographically like this:
東北至新羅，西渡海至越州，南渡海至倭國，北渡海至高麗
(Northeast, one reaches Silla; west, crossing the sea, one reaches Yuezhou; south, crossing the sea, Wa; north, crossing the sea, Goguryeo)
Yuezhou was around modern Shaoxing, Zhejiang.
The New Book of Tang repeats essentially the same geography.
Important distinction: that does not say Zhejiang belonged to Baekje. It says Zhejiang lay across the sea from Baekje.
But then, centuries later, Choe Chiwon writes something wonderfully inconvenient. Describing Goguryeo and Baekje at the height of their power, he says that they:
南侵吳越 北撓幽燕齊魯
(invaded Wu and Yue to the south, and harassed You, Yan, Qi and Lu to the north)
In modern geography, that's a sweep from the lower Yangtze coast toward Shandong and the northeast. It's late, rhetorical, and he annoyingly bundles Goguryeo and Baekje together, so you cannot turn it directly into a map. But once again, the idea of Baekje operating across the Yellow Sea didn't originate with modern Korean nationalists.
Did Baekje rule the Chinese coast?
Almost certainly not in the ridiculous continuous-empire sense shown on the map.
Did Chinese and Korean sources preserve repeated memories of Baekje operating, claiming offices, fighting, trading, and perhaps holding enclaves across that coast?
Yes. And that's much harder to dismiss.
Japanese records
The Shoku Nihongi (續日本紀), completed in 797, records that Takano no Niigasa, mother of Emperor Kammu, descended from Prince Junda (純陀太子), a son of Baekje's King Muryeong. The Nihon Shoki independently records the death of a Baekje Crown Prince Junda in Japan in 513.
Kammu wasn't some peripheral emperor either. He reigned from 781 to 806 and founded the political order associated with the move to Heian-kyō, Kyoto.
Back in 1988, when I was a college freshman, I had the opportunity to represent Korea in a multi-national youth conference held in Fukuoka. During that trip I happened to sit next to then crown prince Akihito and have some drinks. As the evening progressed, and we all had a little bit too much sake, he turned to me with a smile and said, "you know...we're related." And then he winked. At that time I had no fucking idea what we was talking about. Related? You're the fucking next emporer of Japan, descended from some fucking sun god, unbroken bloodline. WTF?
And then years later Akihito becomes emperor.
In 2001, Emporer Akihito did something extraordinary.
At his birthday press conference, talking about Korea, he said:
“I ... feel a certain kinship with Korea”
and explicitly explained why: Shoku Nihongi records Emperor Kammu's mother as belonging to the line of King Muryeong of Baekje. He then talked about Baekje scholars bringing the Chinese classics to Japan and King Seong's role in the transmission of Buddhism. This isn't a paraphrase from the Korean press. It's sitting on the Imperial Household Agency's own website.
After Baekje fell, members of its royal house settled permanently in Japan. The descendants of Seon'gwang / Zenkō (善光), traditionally identified as a son of King Uija, became the Kudara no Konikishi (百濟王氏), literally the “Baekje King” clan. They entered the Japanese aristocratic and bureaucratic order while retaining an explicit Baekje royal identity. Under Kammu, Baekje-descended families received particular attention precisely because Kammu himself claimed Baekje royal descent through his mother.
So “deep blood ties with Yamato” isn't just:
some Koreans moved to Japan.
It is:
members of the Baekje royal house became Japanese aristocrats, and one documented Baekje-derived maternal line entered the Japanese imperial genealogy itself.
And then there are tombs.
In the Yeongsan River basin, especially modern Jeolla and Gwangju, archaeologists have found roughly sixteen keyhole-shaped burial mounds (전방후원형고분) closely related in form to the zenpō-kōen-fun (前方後円墳) of Kofun-period Japan. They appear primarily in the late fifth and early sixth centuries, exactly while the region was being drawn increasingly into Baekje's political sphere. Examples include the tombs at Wolgye-dong in Gwangju, Chilam-ri in Gochang, Shindeok in Hampyeong, and Yongdu-ri in Haenam.
And nobody can agree on who the fuck is buried in them.
Wa immigrants?
Mercenaries?
Yamato representatives?
Local Mahan elites deliberately adopting Wa symbols?
Local rulers serving Baekje while maintaining independent links with Kyushu?
Some combination of all of the above?
Recent scholarship leans heavily toward a mixed frontier world, local Yeongsan elites interacting simultaneously with Baekje, Yamato, and northern Kyushu, rather than a simple “Japan conquered Korea” or “Korea copied Japan” story.
And then cross the sea.
Japanese archaeology contains the reverse phenomenon.
The Takaidayama Kofun in Kashiwara, Osaka, belongs to a type of passage-grave tradition that Japanese archaeological scholarship has explicitly described as Baekje-derived. The Takaidayama-type chamber appears early in the development of horizontal stone chambers in the Kinai region and subsequently feeds into elite Japanese burial architecture.
And Japanese tombs contain objects that appear to have been made in Baekje itself.
The Eda Funayama and Daibō tombs in Kumamoto, for example, yielded gilt-bronze crowns, shoes, earrings and weapons with close Baekje parallels. Detailed technological analysis argues that some were more likely produced in Baekje central workshops than copied locally in Japan.
More broadly, archaeology finds substantial commonalities between the tomb architecture and funerary practices of Ungjin/Sabi Baekje and Asuka/Nara Japan.
In southwestern Korea, people were being buried in Japanese-style tombs.
In Japan, people were being buried with Baekje objects and in Baekje-influenced tombs.
A Baekje prince dies in Japan.
A Baekje royal clan becomes Japanese nobility.
A descendant of King Muryeong becomes the mother of a Japanese emperor.
And 1,200 years later, the Emperor of Japan publicly says: Yeah. I know.
And then we go into unsubstantiated territory. Reasoned conjecture territory. Something that makes Japanese blood boil.
Ōjin is the 15th emperor in the traditional list, but his chronology is legendary and probably displaced by decades. His mother is the semi-mythical Empress Jingū, whose pregnancy story is already bizarre: her husband dies, she supposedly conquers Korea while pregnant, and Ōjin is born after an impossibly prolonged pregnancy. That is not a genealogy you would treat like a birth certificate.
And then, during Ōjin’s reign, the chronicles suddenly become saturated with Baekje connections: Baekje immigrants, literate specialists, Achiki, Wani, royal princes, horses, craftsmen, scribes, continental technology. He sits at roughly the period when Yamato itself becomes archaeologically much more formidable.
- Egami Namio’s famous Horse-Rider Theory suggested that a continental/Buyeo-associated warrior elite entered Japan and established a new ruling order.
- Gari Ledyard later proposed a Buyeo-derived group moving through the Korean peninsula and into Japan.
- Hong Wontack went much further: Ōjin himself was a Baekje prince who crossed to Japan and established a new Yamato dynasty.
- Chizuko Allen’s survey of these theories explicitly records Hong’s argument that Ōjin was sent by Baekje’s Geunchogo/Geungusu-era ruling house,
- Ch’oe Chae-seok went still further and envisioned sustained Baekje supervision of Yamato.
Mainstream archaeology does not accept “Ōjin was a Baekje conqueror” as established history. There is too much continuity in Kofun culture for a clean foreign-conquest event, and the genealogy is far too murky.
But is there some truth there? Or to put it more bluntly, can modern archaeology theory explain the discrepancies better?
No.
Ōjin is exactly where you would look if you were trying to hide a dynastic intrusion inside a later sacred genealogy.
There is even an additional Korean-thread problem: the Kojiki genealogy gives Jingū an ancestor, Amenohiboko, who is explicitly a prince from Silla. So even the official mythology does not present Ōjin’s maternal ancestry as hermetically “Japanese.” That is Silla, not Baekje, but it tells you something important about how porous these early ruling genealogies actually were.
If Ōjin is mythically weird, Keitai is genealogically weird.
Emperor Buretsu dies without an heir.
And then the court goes:
Uh...there's this guy in Echizen.
Keitai arrives from the Ōmi/Echizen region, outside the Yamato political center, and is declared to be a fifth-generation descendant of Emperor Ōjin.
Fifth generation.
Conveniently distant.
The detailed genealogy that was supposed to establish this connection does not survive intact. Later sources preserve fragments, but the very weakness of the connection spawned a serious Japanese scholarly theory after WWII that Keitai represented an actual change of dynasty.
And now comes the Baekje part.
Keitai takes the throne in 507.
King Muryeong rules Baekje 501–523.
These two courts become extraordinarily close.
A National Museum of Korea archaeological study puts it unusually strongly:
Baekje immigrants played a role in establishing the Keitai dynasty, and Keitai’s government followed a strongly pro-Baekje diplomatic policy.
Then in 513, during Keitai’s reign, the Nihon Shoki casually records:
百濟太子淳陀薨
(Baekje Crown Prince Junda died)
The same Junda whom the later Shoku Nihongi identifies as a son of King Muryeong and ancestor of the line that eventually produces Kammu’s mother.
So during the reign of this strange outsider emperor whose ancestry back to Ōjin is disputed, Muryeong's own son is apparently resident closely enough within the Yamato world that his death is recorded in the Japanese imperial chronicle.
And Muryeong himself has an absurdly intimate Japan story:
according to the Nihon Shoki, he was born on an island while his mother was traveling toward Wa, hence his childhood name Sama/Shima, “island.” Modern scholarship continues to take the basic Baekje-Wa context of this tradition seriously even while arguing about his exact parentage.
Then comes Kinmei.
Keitai's succession gets even stranger.
His sons Ankan and Senka reign briefly, and then Kinmei takes over. Scholars have long debated whether some of these reigns overlapped, whether rival courts existed, and whether there was effectively a succession struggle between different branches.
Kinmei is important because his mother, Princess Tashiraka, belonged to the older imperial line.
So if Keitai really was a dynastic outsider, the marriage works beautifully as political glue:
new Keitai paternal line + old Yamato maternal line = Kinmei.
And the present Japanese imperial line descends through Kinmei.
That means if the radical Keitai theory were ever demonstrated, it wouldn't be some dead branch of the family tree. It would sit directly upstream of the later imperial house.
Again, nothing establishes Keitai as Baekje by blood. But this is exactly the kind of succession event where one should be suspicious of later perfectly smooth genealogies.
Then Kanmu, where the speculation stops.
Jump forward about 250 years.
Emperor Kammu is not an outsider. His father was Emperor Kōnin. Nothing weird about the succession comparable to Keitai.
But his mother, Takano no Niigasa, belonged to the Yamato no Fuhito lineage.
And Shoku Nihongi says that lineage descended from:
King Muryeong of Baekje
→ Prince Junda
→ ...
→ Takano no Niigasa
→ Emperor Kammu
This part isn't modern Korean nationalist theory. Scholarship on Kammu routinely describes him as a maternal descendant of the Baekje royal house, and his reign is notable for unusually favorable treatment of families claiming Baekje ancestry.
So by 781 you have a Japanese emperor whose official court genealogy itself acknowledges a Baekje royal maternal ancestor.
So,
China is uncomfortable.
Japan is pissed.
What about Korea?
Korea should be uncomfortable. Not because there are are counter-arguments using similar records claiming parts of Korea as either Chinese, or Japanese, nationalist fucks will always do that, but because think about what it means.
For thousands of years, people have been arguing, hating, killing people because of conflicting definitions of 'Us'. When the real uncomfortable truth is, we are more 'Us' than 'Not-Us'.
Now, does this map make you uncomforable?
5. They're on the 'wrong' side of Korea
I'm going to talk about an uncomfortable piece of Korean politics and history.
A wound that refuses to close.
Like most places in the world, history is written by those in power. And for most of Korea's history and to this day, power lies in the southeast Gyeongsang province, the homeland of ancient Shilla.
- 1,000 years of Shilla out of Gyeongju.
- Goryeo that came out of a collapsing Shilla, but claimed Goguryeo as their legacy (which really makes no sense if you think about it).
- Joseon dominated by Gyeongsang province aristocrats.
- Development of the Gyeongsang province under Japanese occupation.
- 9 out of 14 South Korean presidents come from the Gyeongsang province, including the current president, Lee Jae-Myung.
A whopping 69% of modern South Korea has been ruled by presidents who come from the Gyeongsang province, a region representing only 24.3% of the South Korean population.
- One, and only one, Jeolla president, Kim Dae-jung.
- One from Hwanghae in present-day North Korea, Syngman Rhee, who spent much of his adult life in the US (the dude also happened to be a genocidal, anti-communist authoritarian...but that has nothing to do with the story).
- The rest are a spattering of temporary/interim leaders that were allowed to be in their positions because it suited the military dictators, most lasting just a few months.
No one had a reason to emphasize Baekje, and its connected Jeolla history. I guess it was also not very good for your career.
6. Their discoveries came too late
Perhaps due to the lack of political interest, perhaps due to sheer bad luck, Baekje's most important archaeological discoveries came remarkably late.
Arguably the most important archaeological discovery in modern Korean history, King Muryeong's tomb, was only discovered in 1971, completely by accident.
Workers were digging a drainage trench beside an already-known tomb at Songsan-ri in Gongju when a shovel struck a wall of carefully laid bricks.
Behind it was something Korean archaeologists almost never get:
an untouched royal tomb.
Nobody had robbed it. Nobody had opened it. It had remained sealed for nearly 1,450 years.
And unlike almost every other royal tomb from the Three Kingdoms period, there was no need to guess who belonged inside.
Two stone epitaphs told them.
One actually identifies the dead king as King Sama (斯麻), King Muryeong of Baekje, records his death in 523 CE, gives his age, and records his burial in 525 CE. The queen had her own epitaph as well. It remains the only known royal tomb of the Three Kingdoms period whose occupant can be identified directly from an inscription found inside the tomb itself.
And then they opened it.
Gold crown ornaments.
Gold earrings and necklaces.
Gilt-bronze shoes.
Bronze mirrors.
Chinese ceramics.
A stone guardian beast.
Wooden coffins made from Japanese umbrella pine.
Thousands upon thousands of objects. The original excavation commonly reported roughly 4,600 pieces; later recataloguing brought the number to 5,232 artifacts from the tomb.
In other words, buried under a hill in Gongju was almost a perfect little version of the Baekje story:
Baekje craftsmanship.
Chinese connections.
Japanese connections.
All sitting together inside one Korean king's tomb.
Then there are the books.
Baekje wrote its own history.
During the reign of King Geunchogo, the scholar Go Heung is said to have compiled the Seogi (書記), traditionally regarded as Baekje's first written history. It is gone.
Then there are three other Baekje-derived historical works whose names survive because the Nihon Shoki, compiled in Japan in 720, actually quotes them:
Baekje-gi (百濟記)
Baekje Bon'gi (百濟本記)
Baekje Sinchan (百濟新撰)
All three are lost.
We know they existed because somebody else preserved pieces of them.
Which is almost too perfect.
Baekje's own books disappear.
Fragments survive in Japan.
Chinese dynastic histories preserve other pieces.
And the principal surviving Korean narrative, the Samguk Sagi (三國史記), was compiled in 1145 CE, nearly five centuries after Baekje disappeared, by Kim Bu-sik and his team using earlier Korean and Chinese sources that were themselves already fragmentary.
So if you go looking for Baekje today, a remarkable amount of its Stuff and Books leads you back outside Baekje.
To China.
To Japan.
Or underground.
And what survived in Korea was filtered five centuries later through a Sino-centric, Shilla-loving dude named Kim Bu-sik..
A short history of Baekje
Tradition puts Baekje's founding at around 18 BCE, as a small polity named Baekje-guk (伯濟國) inside Mahan, one of the three Samhan confederacies. In that sense, its beginning was remarkably similar to Saro-guk, the little polity inside Jinhan that eventually became Shilla.
Chinese records describe Mahan as a patchwork of 54 statelets, stretching from the Han River basin through Chungcheong and Jeolla. Baekje-guk was originally just one of them. The dominant power appears to have been Mokji-guk (目支國), associated in the Chinese records with the title King of Jin (辰王), although exactly how much authority that king exercised over the rest of Mahan is unclear.
Baekje had one enormous advantage:
location.
It sat in the fertile Han River basin, one of the richest agricultural regions on the peninsula. The river provided food, transportation and communication deep into the interior. Downstream was the Yellow Sea, and beyond it China.
As Baekje grew, it also developed increasingly sophisticated iron production. Early Baekje ironworking sites are known throughout the Han River and central peninsula, and later Baekje controlled major iron-producing centers such as Jincheon and Chungju.
Food.
Iron.
A navigable river.
And an ocean.
Not a bad place to build a kingdom.
Over roughly three centuries, Baekje gradually expanded outward from the Han River basin, replacing Mokji-guk as the dominant power within Mahan and absorbing surrounding statelets. By the reign of King Geunchogo (r. 346–375 CE), Baekje had become the dominant power across most of southwestern Korea and had entered what is generally considered its first great golden age. Korean historical scholarship traditionally places the final political consolidation of Mahan under Baekje in this period, although, as we'll see, political conquest did not mean that Mahan simply disappeared.
Geunchogo was aggressive.
In 371 CE, Baekje attacked Goguryeo, reached Pyongyang, and killed Goguryeo's King Gogugwon in battle.
But what makes his reign particularly interesting to is what happened in every other direction.
Baekje established formal diplomatic relations with the Eastern Jin in southern China, sending an embassy in 372. Relations with Wa also intensified during this period. The Seven-Branched Sword, whatever exactly we decide its inscription means, belongs to this same Baekje-Wa diplomatic world.
By this point Baekje was no longer simply a kingdom sitting on the Korean Peninsula.
It was plugged into an East Asian network.
And a few things make Baekje stand out.
1. Their founding myth was grounded in Buyeo and Goguryeo
Baekje did not tell a story of having sprung naturally from the soil of Mahan.
Its royal tradition looked north.
The Samguk Sagi connects Baekje's founders, Onjo and Biryu, to the world of Jumong and Goguryeo, ultimately back to Buyeo. Baekje's royal house later explicitly used Buyeo (扶餘) as its clan name.
So you get this wonderful contradiction almost immediately:
a northern Buyeo/Goguryeo founding tradition ruling a southern Mahan population from the Han River basin.
That mixture is Baekje from almost the beginning.
2. They were the Phoenicians of Korea
If Goguryeo's great advantage was land and armies, Baekje's was the sea.
To the west were the Chinese dynasties.
To the southeast were Gaya and Wa.
Running through the middle was Baekje.
When the Chinese commanderies of Lelang and Daifang disappeared in the early fourth century, the old Chinese-controlled trade system around the peninsula broke apart. Baekje was perfectly positioned to insert itself into the gap. Its ports linked the Han River to the Yellow Sea, and later, after the capital moved south, the Geum River performed much the same function.
From China came ceramics, mirrors, texts, Buddhism, artistic styles and court culture.
Toward Japan went iron, weapons, textiles, Buddhist texts and images, scholars, monks, craftsmen and technologies.
And movement went the other way as well. Japanese goods appear at Baekje sites, Japanese timber ended up inside King Muryeong's coffin, and people moved constantly in both directions.
Baekje wasn't simply trading stuff.
It was moving culture.
Writing.
Religion.
Architecture.
Technology.
People.
Ideas.
That is what I mean when I call them the Phoenicians of Korea.
And that culture didn't disappear with Baekje.
After Shilla and Tang destroyed Baekje, that maritime culture did not disappear.
Two centuries later it reemerged on an extraordinary scale under Jang Bogo (장보고; 張保皐). Returning from Tang China in 828 CE, Jang established Cheonghaejin (청해진; 淸海鎭) on Wando and built something far larger than a naval base. He stitched together the ports of Shilla, Tang China and Heian Japan into a single commercial network, backed by his own ships, sailors and soldiers. Along the Chinese coast, expatriate Shilla communities, Shilla-bang (신라방; 新羅坊), temples, merchants and trading stations gave his people places to live, worship, warehouse goods and conduct business thousands of kilometers from home.
His fleets moved Chinese silk, ceramics and luxury goods eastward, Japanese products westward, and Korean goods in both directions, while also carrying monks, diplomats, books, letters and people across the Yellow Sea. Cheonghaejin sat in the middle of it all, policing the sea lanes and making Jang Bogo not merely a merchant, but the man who could make international trade happen.
For nearly two decades, a Korean admiral operating from a small island off the southwest coast commanded one of the most powerful private maritime networks in East Asia.
Baekje's ships were gone.
The sea was not.
3. Mahan didn't disappear when Baekje conquered it
This may be the most interesting part.
Political maps encourage us to imagine conquest like somebody changing the color of a region in Photoshop.
Mahan: brown.
Baekje conquers it.
Click.
Baekje: blue.
Reality wasn't remotely that clean.
Even as Baekje expanded politically across the southwest, local Mahan communities retained distinctive burial customs, elite networks and material culture for generations. The Yeongsan River basin, in particular, continued producing enormous local tombs and distinctive funerary traditions long after Baekje supposedly "conquered Mahan."
Some of these communities remained culturally distinctive into the fifth and even early sixth centuries, centuries after Baekje's traditional conquest chronology.
Which gives us a much better way of imagining Baekje.
It wasn't a homogeneous "Baekje people" spreading south and replacing everybody.
It was a state expanding over Mahan people, incorporating local elites, absorbing their land, trading with them, marrying them, taxing them, fighting them, and slowly turning a collection of different communities into something we later call Baekje.
That probably explains a lot about the kingdom Baekje eventually became.
Cosmopolitan.
Multi-cultural.
Comfortable with outsiders.
Connected to everywhere.
Geunchogo's kingdom would not remain at its peak forever. Goguryeo eventually came south, captured Hanseong in 475, and killed King Gaero. Baekje retreated to Ungjin, modern Gongju, and later moved again to Sabi, modern Buyeo.
But it survived.
And remarkably, even after losing the Han River basin that had created Baekje in the first place, it rebuilt itself around the same thing that had always made it different:
connections.
China across the sea.
Wa across another sea.
Gaya and Shilla beside it.
And people moving between all of them.
Until, finally, in 660 CE, Shilla and Tang brought the kingdom itself to an end.
Japonic traces in Baekje
And then there are the words.
We already found linguistic debris in Gaya.
Baekje has more.
Much more.
And, wonderfully, some of the best evidence is filed in the Samguk Sagi under Goguryeo.
That sounds strange until you remember what happened in 475 CE.
Goguryeo came south, captured Hanseong, killed the Baekje king, and took over much of Baekje's old Han River territory. So when the Samguk Sagi later records the old place names of these areas, places like Suwon, Paju, Incheon and Ganghwa can appear as "former Goguryeo" places.
But before Goguryeo conquered them?
This was the heartland of Hanseong Baekje.
Modern linguistic mapping has explicitly noticed this problem. Japonic-looking place names are concentrated through central Korea, including territory that had belonged to Hanseong Baekje before the Goguryeo conquest. The obvious question is therefore whether these are actually evidence for the Goguryeo language, or linguistic fossils left by populations Goguryeo conquered.
This matters because of the strange way the Samguk Sagi preserved them.
Sometimes it gives us two versions of the same place name.
One records approximately how the native name sounded.
The other tells us what it meant.
Which means we can occasionally reverse-engineer pieces of a language nobody ever bothered to write down.
Take Suwon (水城).
The Samguk Sagi preserves:
買忽 一云 水城
Maehul, also called "Water Fortress."
水 means water.
城 means fortress.
So if 買忽 is approximately the sound of the old indigenous name, while 水城 gives its meaning, then:
買 ≈ "water"
忽 ≈ "fortress/city"
And 買 in these place names has been reconstructed approximately as *mi, appearing repeatedly where later names mean river, water, well or pond.
Old Japanese?
mi (み) = water.
Not once.
There are 34 instances of 買 in the relevant Samguk Sagi geography books, with a substantial subset corresponding specifically to water-related meanings. Modern studies therefore treat this as one of the principal pieces of evidence for a Japonic layer in central Korean place names.
Now move north to Paju (難隱), around modern Jeokseong.
The Samguk Sagi gives:
七重縣 一云 難隱別
("Seven-fold County, also called Naneun-byeol.")
Again, one is meaningful Chinese.
七 = seven.
重 = layer/fold.
The other appears to record sounds.
So:
難隱 ≈ "seven"
And what is "seven" in Japanese?
nana (なな).
Then there is Ganghwa.
Its ancient name was:
甲比古次 (갑비고차)
And it is paired with:
穴口
literally something like "opening/mouth of a passage", hence a strait or channel.
The old form has been reconstructed roughly as:
*kapi-kuti
And there it is again.
Old Japanese:
kuti (くち) = mouth, opening.
Modern Japanese still has kuchi.
The same word participates in compounds for entrances, mouths and openings.
So the ancient name of Ganghwa contains something sounding remarkably like kuti, while its Chinese semantic translation contains 口, "mouth/opening."
That one is difficult to look at without raising an eyebrow.
There are more.
Old central Korean place names contain forms reconstructed as *tan, meaning "valley," compared with Old Japanese tani, "valley."
Forms written 達 occur where the corresponding meaning is "mountain" or "high," and have been compared with Old Japanese take / taka, "mountain/high."
Not every one of these equations is equally convincing. Some have competing Koreanic, Tungusic or Sinitic explanations. Some depend on reconstructing how Chinese characters were being pronounced locally fifteen hundred years ago.
But that is almost beside the point.
There isn't one suspicious word.
There is a geographic cluster of them.
And the cluster sits right across central Korea, including the former Hanseong Baekje world.
Which brings us to something even stranger.
The place names aren't our only evidence.
Chinese historians actually wrote down a few Baekje words.
The Book of Liang (梁書) tells us:
號所都城曰固麻，謂邑曰檐魯
(Baekje called its great fortified capital 固麻, *kuma / koma)
and a settlement or administrative district:
檐魯, *yemlu / tamna-like reconstruction depending on reading.
Not a later Korean place name that might have belonged to somebody else.
Baekje words.
And then the Book of Zhou gives us perhaps the strangest linguistic sentence about Baekje:
王姓夫餘氏，號於羅瑕，民呼爲鞬吉支
(The king belonged to the Buyeo clan)
He was called:
於羅瑕,
Eoraha.
the people called him:
鞬吉支
Geon-gilji.
Two different words for the same king.
One associated with the royal/ruling world.
One explicitly associated with the people.
Perhaps Baekje was not merely a state containing people who spoke several languages.
Perhaps Baekje itself had a linguistic hierarchy:
Buyeo-derived ruling elite → one language
Indigenous Mahan population → another language
What the Chinese passage gives us is something different:
Baekje itself apparently distinguished between the vocabulary of the royal world and the vocabulary used by ordinary people.
Now remember how Baekje began.
According to its own tradition:
- A Buyeo/Goguryeo-linked ruling lineage comes south.
- It establishes itself inside Mahan.
- It grows outward from the Han River.
- It spends centuries incorporating Mahan communities.
And archaeological Mahan culture does not immediately disappear when the political map turns Baekje-colored.
Now add the linguistic evidence.
Suddenly this model becomes extremely attractive:
Koreanic-speaking political power expanding over an older population, at least part of which spoke Peninsular Japonic.
Over time, Koreanic wins.
The state speaks Koreanic.
The court speaks Koreanic.
Eventually everybody speaks Koreanic.
Japonic disappears.
But rivers are stubborn.
Mountains are stubborn.
Villages are stubborn.
Place names are fucking stubborn.
So the people disappear into Baekje.
Their language disappears.
And fifteen hundred years later we discover that around Suwon somebody once apparently called water mi.
That is what a linguistic substrate looks like.
One last thing about Baekjae
How did it build such a powerful maritime trading apparatus?
Not just sending the occasional diplomatic/trade mission abroad.
Baekje appears to have maintained a physical presence in major Chinese and Wa maritime centers, moving merchants, diplomats, scholars, monks, craftsmen, books, technology and luxury goods across an enormous stretch of East Asia.
Which raises a few obvious questions.
1. Why was Wa so reliant on Baekje for access to China?
Wa was perfectly capable of sailing.So why not simply cut out the middleman?
2. What the hell were Baekje ships actually like?
The waters immediately surrounding Baekje demanded a very particular kind of vessel.
The Yellow and Bohai Seas are shallow.
The west coast of Korea has enormous tidal ranges.
Mudflats stretch for kilometers.
Rivers and estuaries were essential transportation arteries.
A ship operating there benefits enormously from a shallow draft and the ability to enter rivers, cross shoals, beach itself, or settle at low tide.
But Baekje had another problem.
3. A hostile Goguryeo?
To Baekje's north sat a hostile Goguryeo with a powerful navy and a chain of fortified positions along the Liaodong coast.
The old safe route to China was obvious:
hug the Korean coast northward, follow Liaodong, move from harbor to harbor, cross through the islands, and eventually reach Shandong.
Stay close to land.
Find shelter when the weather turns.
Take on water.
Repair the ship.
Wait for better winds.
Keep going.
And then Goguryeo took control of the route.
Suddenly Baekje had a choice:
stop going to China,
or
go across the fucking sea.
And Baekje went across.
Not only toward Shandong.
Baekje maintained relations with the Southern Dynasties far down the Chinese coast, and Chinese geographical traditions later placed Yuezhou, around modern Shaoxing and Hangzhou Bay, directly across the sea from Baekje.
Whatever ships Baekje was using, they were capable of sustained open-water travel.
That means a ship capable of surviving several days without putting into port.
It needed:
- reliable sails and steering
- sufficient freeboard
- structural strength
- enough water and food
- weather knowledge
- knowledge of seasonal winds
- pilots who have done this before
- some way of maintaining a heading offshore
So what did those ships look like?
We don't know.
Historical records and archaeology cannot give us a Baekje shipwreck, pull it out of the mud, and say:
This. This is what crossed the Yellow Sea.
Historical records and archaeology cannot give us a Baekje shipwreck, pull it out of the mud, and say:
This. This is what crossed the Yellow Sea.
No such ship survives.
But that does not mean such ships did not exist.
Quite the opposite.
We know the ships, or something functionally equivalent to them, had to exist, because we know what Baekje actually did.
Baekje crossed the Yellow Sea. Repeatedly.
People crossed.
Diplomats crossed.
Books crossed.
Monks crossed.
Craftsmen crossed.
Luxury goods crossed.
Entire communities appear to have formed on the other side.
Baekje maintained relationships with ports and courts stretching from Liaoxi and Shandong down toward the lower Yangtze, while simultaneously operating another maritime network through southern Korea, Tsushima, Kyushu and Wa.
You cannot run that system by occasionally standing on a beach and waiting for a Chinese ship to pick you up.
Absence of archaeological evidence tells us that we cannot reconstruct the ship.
It does not erase the voyages.
And the voyages tell us something important:
Baekje possessed a serious blue-water maritime capability.
Whatever the hardware looked like.
But even that probably wasn't Baekje's greatest advantage.
Baekje's competitive advantage was not merely the ship.
It was the network.
- Ports.
- Pilots.
- Weather knowledge.
- Interpreters.
- Chinese-speaking scribes.
- Diplomatic recognition.
- Safe harbors.
- Repair facilities.
- Water and provisioning points.
- Merchants already living overseas.
- Relationships with Chinese courts.
- Relationships with coastal communities.
- Knowledge of when to leave.
- Knowledge of where to land.
- Knowledge of how to get from Shandong to the Yangtze, from Baekje to Tsushima, from Tsushima to Hakata, and from Hakata onward to Yamato.
That begins to answer the first question.
Why was Wa so reliant on Baekje?
Because building a boat is easy.
Building a maritime system is not.
Wa could sail to China.
Eventually, it did.
But Baekje already had the ships, pilots, ports, languages, relationships, diplomatic experience and overseas communities necessary to make the entire system work.
Baekje wasn't merely the middleman between China and Japan.
Baekje was the network connecting them.
And once you stare at the map long enough, that suddenly makes perfect sense.
---
It's time to update our window of time. We are still at ~1500 BCE and 668 CE. But now that we have covered Shilla and Baekje (Gaya doesn't affect the timeline at all), we're in a good position to move the upper band backwards.
The question is to when?
We have three potential dates:
~18 BCE (when Baekje was founded),
and
~c. 57 BCE (when Shilla was founded),
and
~c. 194 BCE (when Wiman Joseon was founded).
Let's examine each date. At this early date, the evidence is pretty thin. Myth, history, and archaeology collide. No one knows what happened. We're all looking at a few scraps of stuff, some words in a book written a thousand years after the events happened that refers to sources that are long lost. Remember we're talking about glowing eggs falling from the sky, and a box filled with gold and a glowing child drifting on the ocean currents here.
But some things are myth based on fact.
Goguryeo existed.
Gojoseon existed.
Wiman overthrew King Jun (준왕).
Iron stuff appeared in the area where Baekje and Shilla were.
First: ~ 18 BCE
Around ~c. 18 BCE, according to the Baekje Annals (백제본기; 百濟本紀) of the Samguk Sagi, Baekje originated as a splinter migration from early Goguryeo, driven by a royal succession dispute.
When Jumong established Goguryeo in the north, he married Soseono, who bore him two sons, Biryu and Onjo. However, when Jumong's older son from his previous life in Bukbuyeo, Yuri, suddenly arrived and was anointed crown prince, Biryu and Onjo realized their future in Goguryeo was dead.
Accompanied by their mother Soseono and ten trusted retainers, the two brothers and their followers abandoned the northern mountains and migrated south across the strategic rivers to the fertile Han River basin.
Biryu chose to settle on the western coast at Michuhol (modern Incheon), but the land was swampy and the water was salty, making it difficult to sustain. Onjo settled further inland at Wiryeseong (in the modern Seoul area), establishing a fortified capital surrounded by rich agricultural plains.
Then Biryu's settlement failed, his remaining followers migrated to join Onjo at Wiryeseong. Onjo initially named the fledgling state Sipje ("Ten Ministries/Households; 济"), but as waves of commoners and refugees happily flocked to join his expanding polity, he rebranded the nation Baekje ("Hundred; 百 + Clans/Crossings; 济)").
→
This connects Baekje directly to the Koreanic-speaking Goguryeo world, through a weirdly specific founding myth:
- named father,
- named mother,
- named half-brother,
- named brothers,
- named succession dispute,
- two named retainers out of ten,
- large body of followers,
- named destination region,
- two competing settlement sites,
- specific reason one failed,
- specific aftermath: Biryu fails, his followers join Onjo
Most founding myths I've read generally involve gods appearing, the sun dying, the oceans splitting. This one reads like a daytime soap opera.
Second: ~57 BCE
In the founding myth of Shilla, early Shilla kings came from three clans, the Three Royal Clans (삼; 三姓):
Bak/Park (박; 朴): Founded according to legend by Bak Hyeokgeose, who famously hatched from a mysterious purple egg shaped like a gourd (박). They represent Silla's earliest foundational chiefdom layer in the southeastern corner of the peninsula.
The Seok Clan (석; 昔): Founded by Seok Talhae, a figure whose myth cycles describe him arriving on a boat from a distant northeastern maritime territory (often associated with coastal regions near the ocean or northern states; 용성국; 龍城國). They injected a distinct, competitive regional faction into early Silla politics.
The Kim Clan (김; 金): Founded through the legendary discovery of Kim Alji (김알지; 金閼智), who was supposedly found as an infant inside a glowing golden chest guarded by a white rooster in a sacred forest (hence Kim, meaning "gold"). Though they entered the power structure later than the Bak and Seok, the Kim clan ultimately broke the rotation cycle, monopolized the throne permanently from the mid-Silla period onward, and gave rise to the Gyeongju Kim lineage.
Of particular interest are the two foreign clans; Seok (석) and Kim (김).
The Seok (~c. 1C BCE) legend states that his mother placed him in a large box loaded with treasures, gold, and weapons, setting him adrift on the ocean. The chest floated down the coast and landed on the shores of Jinhan (southeastern Korea). An old woman discovered him, and he eventually worked his way into the Silla court through sheer intellect and military competence, marrying a princess and seizing the throne as King Talhae (4th ruler).
→
This is a Shilla connection to a Koreanic-speaking people, since logically to arrive off the coast of Shilla, they must have sailed southward along the east coast of Korea, making the origin Okjeo or Dongye, both Koreanic-speaking peoples. An examination of currents off the east coast of Korea confirms the possibility of this, if these people were sailing on a boat (if in a floating box that kid would probably end up in northern Honshu discovered by a bunch of Ainu folk wondering what the fuck to do with a shining golden boy).
The Kim (~c. 65 CE) legend states that the Shilla court minister Hogong (himself an immigrant from an eastern/northern background) heard a mystical white rooster crowing in a sacred forest west of the capital. Investigating the source, he found a glowing golden chest hanging from a tree. Inside was an infant, whom King Talhae joyfully adopted. Because he came from a golden box, the child was given the surname Kim (金), meaning "gold". The forest was renamed Gyerim ("Rooster Forest"), which later became an alternative name for the entire nation of Silla.
While the myth dresses it up in divine bird-and-box lore, mainstream historical and archaeological consensus views the Kim clan's arrival as the upward-mobility migration of a powerful, heavily armed northern elite group. Some modern historical theories (such as those exploring ties to Xiongnu-influenced refugee networks or northern nomadic iron-working groups drifting through Manchuria and Lelang) suggest the Kims brought advanced metallurgy, elite equestrian warfare, and external political clout that allowed them to slowly out-maneuver the older Bak and Seok houses.
Kim Alji himself never sat on the throne; he graciously stepped aside early on. However, his seventh-generation descendant, King Michu (미추왕; 味鄒王 / 味鄒尼師今), became the first Kim king in 262 CE. From the 4th century onward, the Kim lineage locked down absolute hereditary monopoly over the Silla crown, ending the rotational power-sharing game with the Baks and Seoks for good.
→
The Kim clan founding myth does not affect the timelines since they were late, but it does suggest something important: Whether Kim Alji was of Xiongnu origin, as some theories suggest, or of Buyeo/Goguryeo origin, it still strengthens Shilla's ties to the north. The fact that Kim (김; 金) is used extensively in Jurchen legends and dynasties, Kim in Chinese is Jin, another interesting thing to note.
Third : ~194 BCE
Here the records split.
The Chinese Sanguozhi (三國志) says that after Wiman (위만; 衛滿) overthrew King Jun (준왕; 準王) of Gojoseon around 194 BCE, Jun and some of his people fled south into the Han world.
Another, preserved much later in the Samguk Sagi, says that before Silla existed:
先是, 朝鮮遺民, 分居山谷之間, 爲六村
(“The remnants of Joseon lived scattered among the mountain valleys, forming six villages.”)
Those six villages became the communities around Saro-guk.
Are these two stories talking about the same people?
I think they probably are.
Wiman takes Gojoseon.
People from Gojoseon flee south.
A later Silla tradition remembers its foundational villages as being populated by refugees from Joseon.
Could they be completely unrelated migrations?
Sure.
But I don't need a time machine to notice when two footprints point in the same direction.
→
This establishes another connection between a Koreanic-speaking people establishing itself in the southern part of the Korean Peninsula.
Three in a row:
One from Goguryeo/Buyeo.
One (possibly) from Okjeo/Dongye.
One from Gojoseon.
⇒
Three potential dates. All checking boxes:
Soil (Land): Yes
Stuff (Artifacts): Yes
Blood (DNA): Yes
Words (Language): Yes
Books (Written records): Yes
Are they Korean ('Us')?
→ Yes
As to which date to choose is up to you:
① ~18 BCE
② ~57 BCE
③ ~194 BCE
If you choose ①, then our window is: ~1,500-~18 BCE (1,492 years)
If you choose ②: ~1,500-~57 BCE (1,443 years)
If you choose ③: ~1,500-~194 BCE (1,306 years)
There is no right and wrong answer, and even if there was, there is no evidence that remains that can dispute any of this. Lack of evidence, does not mean lack of existence.
But whichever one you choose, we are making one very important decision:
The arrival of a small northern Koreanic-speaking elite group
into a sea of Japonic-speaking peoples counts as "Korean" from now on.
Personally, I'm going to choose ③: ~1,500-~194 BCE (1,306 years), not because I have some secret knowledge that noboday has, but because it makes writing this series more fun. And probably ~194 BCE isn't going to survive anyway.
---
Heavy shit. Sorry it only gets more confusing from now on.
Next is Goguryeo. The unapologetically kick-ass, take-no-prisoner Korean state that lights a fire inside the souls of all Korean little boys and girls. The ultimate WhatIf of Korean history.
Comments
Post a Comment