Wednesday, November 19, 2025

Thirty Years of Korean-Style Neoliberalism



IMF, the Republic of Irregular Work, Platform Labor, and the Future of Young People

From the 1997 “Gold-Collecting Campaign” to the N-Po Generation of the 2020s:
Rewinding the structural changes we actually lived through, with data and lived examples.




1. After the 1997 IMF Crisis: How Korean-Style Neoliberalism Was Installed

The 1997 Asian Financial Crisis wasn’t just an economic shock.
It was a regime change in the rules of the game.

The IMF bailout came with a familiar set of conditions:

  • Accelerated liberalization of finance and capital markets

  • Privatization and restructuring of state-owned enterprises and financial institutions

  • Labor market flexibilization – in plain language, making it easier to hire and fire

In 1998, amendments to the Labor Standards Act and the Dispatch (temp agency) Act brought in:

  • “Managerial” mass layoffs – the legal basis for firing people “for business reasons”

  • Dispatch and subcontracting as fully legalized, mainstream employment forms

Of course, layoffs and temporary jobs existed before then.
But the rulebook changed:

From “once you get in, you more or less stay” + a peripheral layer of temps
to “permanent restructuring + permanent precarity.”

Through the 2000s, the Korean economy:

  • Recovered fast as an export-led, chaebol-centric economy, and

  • Enjoyed growth built on finance and real-estate bubbles

But something snapped:

  • A company’s crisis was treated as everyone’s crisis,

  • Yet a company’s recovery no longer translated into a recovery of ordinary people’s lives.

That asymmetry is basically the starting point of Korean-style neoliberalism.


2. The “Republic of Irregular Work”: Season 1 in Numbers

2.1 Irregular Work Becomes the Norm, Not the Exception

According to Statistics Korea’s supplementary survey on non-regular workers (2023):

  • Non-regular workers: about 8.12 million people

  • Share of all wage-workers: around 37%

Right after the crisis, the share was in the low-20% range.

By the mid-2000s it had jumped to around 35%,
and since then it has hovered in the one-third to close-to-40% band.

So instead of:

“Regular workers + a small fringe of non-regulars”

what we actually got is:

Roughly “6 regular : 4 non-regular”, a dual structure that has been locked in for over 20 years.

2.2 Wages and Benefits: Same Work, Different Planet

If you pool data from government and research institutes (numbers vary slightly by year), you get something like:

  • Hourly wage of non-regulars: about 65–70% of regular workers

  • Severance pay, bonuses, welfare benefits: often less than half of regular levels

  • Unionization rate: less than one-third that of regular employees

Same work, same workplace – but if your contract type is different, then:

  • Your pay,

  • Your benefits,

  • Your promotion path

are basically on a completely different track.

For firms, this is “cost savings + flexible adjustment of manpower.”
For individuals, it is a life in which long-term planning is structurally impossible.

2.3 Income and Wealth Inequality: Both Deepen Together

The income quintile ratio (top 20% vs bottom 20%) sits around 5–6 to 1 in recent years – stubbornly high.

On the asset side, household finance and welfare surveys show:

  • The top 10% of households own over half (around 60%) of all net assets.

In one line:

30 years after IMF:
“Big corporations and asset-owners became structurally stronger,
while the lower and precarious tiers became structurally poorer.”


3. Platform Labor: The Face of Season 2

If irregular work was Season 1 of Korean neoliberalism in the 2000s,
then in the late 2010s and 2020s, platform labor is clearly Season 2.

3.1 The Numbers

Surveys by the Korea Employment Information Service and the Korea Labor Institute suggest roughly:

  • 2021: about 2.2 million platform workers

  • 2022: around 2.92 million – a **32.9% jump in just one year

Depending on how you define it, 8–10% of all workers now earn income through platforms.

Examples:

  • Delivery riders, quick-service drivers, designated drivers, app-based couriers

  • Warehouse pickers/packers for Coupang, Market Kurly, etc.

  • Online freelancers on platforms like Kmong or Soomgo (design, coding, translation, tutoring…)

  • Content creators and influencers on YouTube, Instagram, TikTok, etc.

3.2 Called “Boss,” Treated Like a Worker – With Fewer Rights

The contract logic goes roughly like this:

  • No 4 major insurances, no severance pay, no guaranteed minimum wage

  • Officially: “self-employed” or “individual contractor”

  • In reality: your work volume, pay rate, and rating are controlled by an algorithm

In other words:

You’re managed like an employee,
but your legal status is downgraded to that of a fragile micro-business.

Platform labor doesn’t just extend the old regular vs non-regular divide.
It complicates it:

  • Risk and costs are offloaded onto individuals

  • Profits and data are concentrated in the platform

That is the core architecture of Neoliberalism 2.0.


4. A Generation’s Story: From IMF Kids to Platform Kids

4.1 Official Youth Unemployment Doesn’t Tell the Whole Story

The headline youth unemployment rate (ages 15–29) in 2023 was around 6–7%.
On paper, that looks “manageable”.

But if you look at the broader “employment supplementary indicator 3” – often called youth “felt” unemployment

  • The rate is above 20% (roughly 21–22%).

This bucket includes:

  • Involuntary part-time and under-employed workers

  • People preparing for jobs but not counted as “unemployed”

  • NEETs who have effectively dropped out of job searching

So in practical terms:

More than 1 in 5 young people are outside anything resembling a decent job.

4.2 “Good Jobs Are Narrow, My Own Home Is Distant”

What 30 years of neoliberalism feels like to Korean youth can be summed up pretty simply:

  • Regular, stable jobs: fewer and harder to get

  • Even if you get in: overtime, low pay, and performance pressure as default settings

  • Seoul/metro housing prices climbing far faster than wages

  • Starting adulthood with student loans, jeonse loans, overdrafts

Hence the keywords we all know:

  • Hell Joseon,

  • the N-Po generation (the “give-up-N-things” generation),

  • yeongkkeul (all-in mortgage leverage),

  • byeorak-geoji (sudden “poverty” after missing the asset boom),

  • the “villa-king” and jeonse fraud scandals…

And holding it all up in the background is:

  • One of the highest household debt levels in the world,

  • A society built on leverage into real estate and stocks.

4.3 A Generation of Both Cynicism and Protest

There’s a paradox here.

On the one hand, this generation is soaked in political cynicism and disgust.
On the other, it has been front and center in nearly every major movement:

  • The candlelight protests,

  • Riders’ and platform-workers’ unions,

  • Feminist movements, climate activism, and new political experiments.

So this is a generation that has:

“Hyper-individualized survival competition” and “collective resistance memories”
burned into its DNA at the same time.

That mix may turn out to be the critical variable in reshaping Korean politics and society.


5. Thirty Years of Korean Neoliberalism: The Report Card

5.1 What It Gave Us

Export, digital, and platform power

  • IT, semiconductors, smartphones, games, K-pop, dramas –
    Korea undeniably climbed into the top tier of global competitiveness in many sectors.

A dramatic rise in per-capita income and material living standards

  • From around $10,000 per capita in 1997

  • To the mid-$30,000 range today

That is a real, tangible achievement.
Saying “everything just collapsed” simply isn’t accurate.

5.2 What It Took Away

But the price paid has been heavy.

The stability of work and life

  • Irregular, platform, freelance work has become a default, not an exception.

Deeply entrenched income and wealth inequality

  • Top deciles have secured a long-term lock on income and assets.

A growth model running on household debt

  • A “leverage society” where almost everything depends on real estate and financial speculation.

Structural despair among the young

  • Specs race + unstable jobs + housing crisis
    the collapse of any believable future narrative.

In short:

Korean-style neoliberalism succeeded in “growth,”
but failed spectacularly in “stability of life” and “hope for the future.”


6. Minimum Conditions for Imagining a Post-Neoliberal Korea

If we want to change this, the answer cannot just be
“try harder” or “be more efficient.”

We have to rewrite the rules themselves.
For a blog-friendly checklist, at least four pillars are hard to dodge.

6.1 Resetting Labor Rules – Ending “Boss-Cosplay Work”

Expand protections for platform, “special-type,” and freelance workers

  • Gradually extend basic safety nets (industrial accident insurance, employment insurance, minimum wage)
    regardless of contract form.

Tighten the rules on using non-regulars and strengthen conversion to open-ended contracts

  • “Non-regular as a cheap disposable category” should shrink,

  • And temporary contracts should be clearly limited to situations where they are genuinely needed.

6.2 Redistributing Income and Wealth – Because Wages Alone Won’t Cut It

Normalize taxation on assets and unearned income

  • Slightly stronger, fairer taxation on non-labor income
    from real estate, stocks, inheritance, and gifts.

Scale up public and long-term rental housing for youth and the housing-poor

  • Shift from “no home unless you inherit assets”

  • To at least “everyone can secure a place to live without life-ruining debt.”

6.3 Welfare State 2.0 – Not Just Handouts, but “Basic Infrastructure”

Cash transfers alone won’t change the structure.
We need a “basic infrastructure state”:

  • Healthcare, education, care work, housing, and transport
    closer to public goods than luxury items.

Only when “it is actually livable” can markets and individuals function without burning out.

6.4 Bringing the Young to the Center of Politics

Lower the barriers to youth participation in parties and policymaking

  • Party membership, candidate registration, access to policy budgets, etc.

Public support for organizing youth, platform workers, and freelancers

  • Not just traditional unions,

  • but also new forms of co-ops, guilds, and professional associations.


7. Conclusion – “Can We End This Regime?” Is the Wrong Question

If we grade 30 years of Korean neoliberalism, the report roughly looks like this:

  • Economic growth and global competitiveness: B+ to A-

  • Stability of life, equality, and the future of the young: D

The real issue is not:

“Can we overthrow this system overnight?”

but rather:

“Do we have the imagination and political will
to upgrade this system into something fairer?”

The good news is that the raw materials for that upgrade are already here:

  • The memory of the candlelight movement,

  • The small but persistent strikes and organizing efforts of platform workers,

  • The political experiments of young people, women, and the climate generation.

They are, in a sense, the trailer for whatever comes after Korean-style neoliberalism.

What the full movie looks like
will depend on what choices we make now.



<

Is Neoliberalism Really Over?





China, Africa, Automation – Reading the Next Round of the Global Economy


If you had to pick one word that haunted the global economy for the last 40 years, a lot of economists would still say:

“Neoliberalism.”

Deregulation, privatization, free trade, small government.
On paper it sounds like “efficiency and growth.”
On the ground, it often felt more like this:

  • Restructuring and mass layoffs,

  • Precarious contracts and gig work,

  • The constant fear of being fired “at any time,”

  • Soaring housing and education costs,

  • While your paycheck hardly moves.

Bundle all of that together and you get what many now call:

“The dark side of the neoliberal order.”

So how long does this system last?
Is it really about to collapse and be replaced by something completely different?

Rather than shouting “it’s all doomed” or “everything’s fine”,
this piece takes a slower walk through what is actually changing right now –
and what those changes might mean for a “post-neoliberal” world.


1. Neoliberalism: What Was the Actual Problem?

Very roughly, neoliberalism pushed four big ideas as policy:

  1. Free movement of capital
    – Looser controls on cross-border investment and finance.

  2. Shrinking the public sector
    – Privatization, selling state-owned firms, marketizing public services.

  3. Labor market “flexibility”
    – Easier layoffs, more temp and contract work, wage restraint.

  4. Fiscal austerity
    – Smaller welfare states, tighter budgets, cutting social spending.

The result, in many countries:

  • The top 1% saw their share of income and wealth grow,

  • The middle saw wages stagnate or slide,

  • Insecure work became a normal life condition, not an exception,

  • And even with all that, growth was often weaker than promised.

In other words:
the pain was widely socialized, the gains were highly concentrated.


2. China, India and the Limits of the “Cheap Labor” Model

The first playground for global capital under neoliberal rules was simple:

“Go where labor is cheap and plentiful.”

China, after “reform and opening,” rode that wave for decades with near double-digit growth.
India plugged itself into global value chains through IT and services.

But that model is now hitting its limits:

  • Wages in China have steadily risen,

  • Environmental rules are tighter,

  • Labor rights claims and social tensions are growing.

In other words:

“Just being the world’s cheap factory floor”
is no longer a stable long-term strategy.

Factories have already started shifting to:

  • India,

  • Southeast Asia,

  • Bangladesh,

  • Vietnam…

But we only have one planet.
You can’t keep outrunning rising wages forever by moving to “the next cheap place.”
At some point, there is no “next.”


3. Is Africa the “Last Labor Reservoir”?

The original post that inspired this essay imagined a grim scenario:
global capital “invading Africa” as the final reservoir of cheap labor.

Reality is messier.

  • Africa’s population is already about 1.4 billion,

  • And is projected to reach 2.5 billion by 2050.

So the issue isn’t “too few people.”
The real problem is:

Jobs, education and infrastructure are not keeping up with that population growth.

At the same time, Africa today is not just a passive victim of “foreign exploitation”:

  • China is funding railways, ports and industrial zones,

  • Western governments and firms are pushing renewables and digital infrastructure,

  • Local fintech, agri-tech and other startups are popping up and drawing capital.

Are there new forms of inequality and conflict in that process? Absolutely.
Do many projects fail? Of course.

But the simplistic vision –

“One day the US sends in troops and runs the whole continent like a giant colony”

– really doesn’t match what’s happening on the ground.


4. Automation and AI: The End of Work, or Just a Re-Shuffle?

The original piece we’re riffing on argued:

“Science will reduce the need for labor so much that permanent mass unemployment becomes inevitable.”

That’s overstated, but the direction of concern is worth taking seriously.

Robots, factory automation and AI are already:

  • Replacing repetitive manual tasks,

  • Eating into middle-skill office and technical jobs,

  • And reshaping whole sectors at high speed.

That does not mean “all jobs disappear.”
New technologies always create new types of work too.

The deeper problem is this:

The new jobs often demand more education and higher skills,
and those who can’t clear that bar fall into even more precarious work.

The classic neoliberal answer was:

“Reskill yourself. Compete harder. Survive as an individual.”

But in many countries, we’ve reached the political and social limits of that mantra.

People increasingly feel:

  • The risks are individualized,

  • While the gains are privatized.

That’s a recipe for backlash.


5. Is Neoliberalism Really Heading for an End?

Since the 2008 financial crisis, then COVID-19, the climate crisis, and rising US–China tensions, the global landscape has shifted a lot.

We now see:

  • The US and EU returning to industrial policy and subsidies,

  • Supply chains being treated as national security assets,

  • Welfare, emergency income support and public healthcare being politically rehabilitated.

What’s striking is that these moves aren’t just coming from left-leaning governments.
Even many conservatives now accept that:

The era of shouting “markets solve everything” is over.

We’re in a post-neoliberal mood, where:

  • The state is expected to step back in strategically.

But this doesn’t mean:

  • Capitalism collapses tomorrow,

  • And Marx’s final stage communism rolls in automatically.

A more realistic picture:

  • Some regions drift toward “repaired capitalism” – more welfare, more regulation.

  • Others evolve into “neo-capitalist blocs” – heavy state intervention plus digital surveillance and authoritarian controls.

  • Still others experiment with hybrid models that don’t fit any neat textbook.

So instead of:

“One Great Depression → Worldwide Revolution → One New System”

we’re more likely entering a long, messy multipolar contest among several kinds of regimes.


6. So What Should We Actually Be Thinking About?

The original article ends with a big claim:

“We need a globally unified planned economy under the banner of humanism.”

As an ideal, that sounds noble.
As a realistic political program, it feels… very far away.

Still, there are concrete questions we can extract and actually work with:

1. In an age of “growth without jobs,” how do we share income?

  • Universal basic income?

  • Social dividends from natural resources or data?

  • Shorter working hours?

  • Stronger regulation of platform work?

2. How do we rebuild welfare states without repeating past inefficiencies?

  • How do we tax large corporations and ultra-wealthy households effectively, not symbolically?

  • How do we guarantee quality in public services while keeping them democratically accountable?

3. Can we redefine “growth” under climate and resource constraints?

  • Is there a model that prioritizes quality of life, safety and ecological sustainability
    over just “a bigger number each year”?

4. How do we stop AI and automation from becoming a pure monopoly game?

  • Rules for data as a public resource,

  • Transparency and fairness in algorithms,

  • Support for open-source ecosystems and public interest tech.

Depending on how societies answer these, the “end of neoliberalism” could mean:

  • A transition to a more just and stable order,

  • Or a slide into something more unequal, more repressive, and more brittle.

The direction is not predetermined. It’s a political choice.


7. In the End, It’s Not “Will It Collapse?” but “How Do We Change It?”

One healthy instinct in the original piece is this:

It refuses to believe that the current system will last forever.

Where it trips up a bit is here:

  • A hyper-dramatic Africa invasion scenario,

  • A binary choice between “human extinction or world revolution,”

  • Old numbers and very absolute, prophetic language.

For today’s readers, that can start to feel like:

“Retro left-wing sci-fi” rather than serious analysis.

If you’re writing this for an ad-driven blog, you’re usually better off if you:

  • Update facts and figures with the latest data,

  • Offer scenarios and options, not “I alone know the future,”

  • End with concrete questions that readers can relate back to their own lives
    – housing, job insecurity, pensions, student debt, and so on.

Use the reframed structure above as your backbone,
then wrap it with:

  • A few of your own experiences,

  • Some vivid examples from Korean society (real estate, temp work, pension debates…).

That alone is enough to build a piece that grabs clicks and reading time,
without sacrificing depth or honesty.



From Seungja Chongtong to Cheonbochong How Joseon Learned the Firepower Lessons of the Imjin War



1) “Arquebus Shock”: When One Weapon Tilted the Battlefield


Any time Koreans talk about the Imjin War, one phrase pops up without fail:

“조총 쇼크 – the arquebus shock.”

The matchlock arquebus (조총, tanegashima) that Japanese forces brought to Korea was, from Joseon’s point of view, a completely different beast. The Annals of the Joseon Dynasty even record claims that the arquebus could:

“Hit birds flying in the forest,”
and that “eight or nine out of ten shots would find their mark.”

If we blamed Joseon’s early defeats at Busan, Dongnae, Sangju, and at the Tangeumdae battlefield only on Japanese guns, that would be an oversimplification. Command, logistics, and mobilization all collapsed together.

But it’s equally hard to deny this:

Against an army built around bows and spears, the Japanese matchlock squads put a new level of concentrated firepower on the field.

The story doesn’t end there, though.

As the war dragged on:

  • Joseon captured Japanese guns and reverse-engineered them,

  • Studied Chinese and Japanese firearms side-by-side,

  • And began producing Korean-pattern matchlocks in quantity.

By the later Imjin phase, through the Jeongyu War (the later invasion), and into the Manchu and northern campaigns, gunners (포수, 銃手) had become the core of Joseon’s field firepower.

In other words, the Imjin War was both:

  • the war where Joseon was knocked down by arquebuses,

  • and the war where Joseon learned to shoot back.


2) Korean Matchlocks and the Cheonbochong: The Age of Range Wars

After any major war, generals everywhere ask the same question:

“Can we shoot farther and harder than the other side next time?”

Joseon was no exception.
After the Imjin War, Korean gunsmiths took a hard look at Japanese long guns (장총) and started developing their own long-range matchlocks.

In the records we see names like:

  • Daejochong (大鳥銃) – “great bird gun,”

  • Cheonbochong (千步銃) – literally “thousand-pace gun.”

Just from the name, the image writes itself:

  • A gun with a longer barrel than standard arquebuses,

  • Packing more powder,

  • Designed to push the effective range out to several hundred meters
    a deliberate counter to Japanese long guns.

Exact numbers differ from source to source, but the very phrase “cheonbo – a thousand paces” tells us this was a serious long-range weapon by the standards of the time.

As for patriotic claims like:

“Joseon completely outclassed Japanese gun tech after the war!”

that’s where we need to tap the brakes a bit.

From what modern research can reconstruct, it’s safer to say:

  • In range, accuracy, and production quality,

  • Joseon and Japan were constantly leapfrogging each other,

  • Each tuning their guns to their own terrain and tactics, rather than one side clearly “winning the arms race.”

Still, the very fact that Joseon responded to arquebus shock by developing its own long guns is a crucial part of the story.


3) Seungja Chongtong: Gun or Cannon?

One of the most heated internet arguments is about the Seungja Chongtong (勝字銃筒).

Some insist:

“There’s no way that thing was fired by hand. It’s a small cannon, period.”

Measured against modern scholarship, that’s… half true and half exaggerated.

So what was the Seungja Chongtong?

Roughly:

  • Length: about 75–80 cm,

  • Bore: around 2 cm,

  • Function: a small artillery piece that could fire

    • multiple lead balls at once, or

    • special arrow-like bolts.

It had:

  • A wooden shaft (stock) fitted into the rear, so a single soldier could carry and aim it,

  • And could also be mounted on tripods, battlements, or ship railings.

In modern terms, it sits somewhere between:

“oversized personal firearm” and “miniature cannon.”

You could aim it like a gun,
but in terms of structure, ammo, and employment, it was closer to artillery.

So was it fired from the shoulder, or braced?

From surviving examples and period notes:

  • The long wooden stock suggests it was at least partly shoulder- or hip-fired,

  • But due to recoil, it was often braced against a wall, parapet, ship’s side, or carriage.

So:

  • Calling it a “light, shoulder-fired arquebus” is misleading.

  • But saying, “It’s a full-blown cannon; no one could ever fire it by hand” is also too absolute.

The most honest, blog-friendly description is something like:

“Seungja Chongtong: a portable small cannon — a Korean hand-cannon.”


4) Cheonja, Jija, Hyeonja, Hwangja, and the Folangji: How Big Guns Really Worked

As the forum post that inspired this piece notes, Joseon already had a famous “four-tier” artillery system by the Imjin War:

  • Cheonja (天字)

  • Jija (地字)

  • Hyeonja (玄字)

  • Hwangja (黃字) Chongtong

These were the pinnacle of the late Goryeo / early Joseon gunpowder artillery tradition:

  • Impressive range and hitting power,

  • Very effective in fortresses and on warships,

  • But enormously heavy—not something you drag around for mobile field battles.

On land, what really shone were lighter, more mobile guns. The star of that show:

The Folangji gun (佛郎機砲) – often just called bulranggi.

The Folangji was a Western-derived, breech-loading small cannon:

  • Short, relatively light barrel,

  • A breech chamber you could swap in and out for speedy reloading,

  • Perfect for mounting on castle walls, ship railings, carts, or makeshift tripods.

Yes, breech-loading at the time meant:

  • Leaky gas,

  • Less power than a solid-breech heavy cannon.

But as a mobile, quick-firing gun that could keep up with troops, it was loved by Joseon commanders.

If you think in game terms:

  • Cheonja / Jija / Hyeonja / Hwangja were the “on paper” S-tier units—massive stats, but expensive and sluggish.

  • The Folangji was the meta-defining unit for actual field fights: not the biggest numbers, but the most practical.


5) Janggunjeon, Singijeon, Land Mines: Flashy Weapons vs. Cold Numbers

Some weapons are guaranteed to start a fight in any history forum:

  • Janggunjeon / Daejanggunjeon – massive arrow or rocket-assisted bolts,

  • Singijeon – the legendary “rocket arrows” fired from hwacha (rocket carts),

  • Jirwipo / flying thunder bombs – early bomb and “land mine”-style weapons.

In paintings and museum displays, they look spectacular.
You can’t help but think:

“Surely one volley of that must have wiped out entire Japanese formations.”

But when we look at production numbers and actual deployment,
a more sober picture emerges. These were more “special effects” than “main damage dealers.”

Low production volume

  • Janggunjeon-type heavy bolts consumed a lot of high-quality materials and labor.

  • Some studies even suggest that, when you compare pre-war and post-war inventories,
    the number of these “prestige arrows” actually declined during the war,
    because they were so hard to replenish.

Hard to aim and control

  • Rocket arrows look amazing in flight, but:

    • Their trajectory is at the mercy of burn rate, wind, and manufacturing inconsistencies.

  • Gunners didn’t have infinite time or powder to train only with these exotic projectiles.

Their real value: “psychological weapons” and niche tools

  • At night or during fortress defense,
    the noise, flame, and smoke from rocket arrows and bombs could seriously rattle the enemy.

  • They were great for:

    • Panic,

    • Confusion,

    • Setting fires,

    • Breaking up tightly packed formations.

But the day-to-day killing power on the battlefield still came from:

  • Matchlocks,

  • Seungja chongtong hand-cannons,

  • Folangji guns,

  • And the classic Chongtong artillery.

As for Jirwipo and “land mines”, most scholars agree these were closer to:

“Bombs fired from guns that exploded on or near the ground,”

rather than modern buried land mines.
Again: useful in specific situations, but not the main pillar of Joseon’s firepower.


6) Bows and Guns Together: The Imjin War’s Real Takeaway

People like to summarize the Imjin War with lines like:

“Japan was strong because of guns.
Joseon was strong because of cannons.”

That’s… half right and half wrong.

If you look at the units that actually performed well in battle, you see a pattern:

They used bows, spears, guns, and cannons together.

  • In mountain ambushes, bows and light arrows (like pyeonjeon, short armor-piercing shafts) were still brutally effective.

  • On plains and in siege warfare,

    • arquebuses,

    • Seungja chongtong,

    • Folangji,

    • And the heavier Chongtong provided the backbone of firepower.

Famous victories like Kwon Yul’s defense at Haengju combined:

  • Close-range aimed volleys from firearms,

  • Massed longbow and pyeonjeon fire,

  • And cannon fire into a layered defensive system.

Joseon adopted new weapons fast—but didn’t throw away the old ones.

King Seonjo himself, for all his flaws, recognized the power of the arquebus,
yet he and his officers never stopped thinking about how to use bows and pyeonjeon to their advantage.

Later, during King Hyojong’s northern-campaign plans,
there were even proposals to reorganize artillery units as mixed gun-and-bow formations
(called sapo chambandae, 사포참반대) rather than pure gunner units.

In the end, the key wasn’t the type of weapon, but the ability to use it properly.

  • Give the same matchlock to two armies:

    • If one has poor training, command, supply, and terrain usage,
      its “firepower advantage” evaporates.

    • If the other knows how to combine its weapons and exploit the terrain,
      it can hold out even without the very latest gear.

So perhaps we should stop telling the story as:

“A tragic tale of a kingdom crushed by one foreign gun.”

and instead say:

“The Imjin War was the moment Joseon threw every bit of its centuries-old gunpowder and archery tradition into the forge,
and hammered old and new weapons into a new kind of combined-arms warfare.”




The Bronze Age Just Got 500 Years Older?



Did the 10th-Century BCE Bronze Age Theory Really Collapse?

How Namgang, Gangneung and Sokcho Are Rewriting Korea’s Prehistory


“The Bronze Age on the Korean Peninsula begins around the 10th century BCE.”

You’ll still find that line in plenty of textbooks and introductory histories.
It’s basically the opening sentence of “official” Korean history.

But since the 2000s, radiocarbon (C-14) dates from sites in the Namgang Dam reservoir, along the Gangwon East Coast, and around Honam/Yeongnam have started to shake that orthodoxy.

Some settlements now date back to the 15th century BCE.
Dolmens containing bronze axes and bipa-shaped bronze daggers are consistently falling in the 10th century BCE—or even earlier.

So what do we do with a timeline like that?

This article is not a “who wins, North vs South academia” match report.
Instead, it’s a humanities-style field note on three questions:

  1. How did the “10th century BCE” Bronze Age orthodox view emerge in the first place?

  2. What exactly are these new C-14 dates telling us?

  3. Where is a reasonable middle ground in current scholarship?


1. How Did the Textbook “10th-Century BCE Bronze Age” Come to Be?

Let’s start with the orthodox position.

The Encyclopedia of Korean Culture (published by the Academy of Korean Studies) defines the Bronze Age in South Korea roughly like this:

“A period centered on the early 1st millennium BCE, during which plain pottery (Mumun pottery) was in full use, bronze artifacts were produced and used, and dolmens were constructed in large numbers.”

In other words, three key markers:

  • Full-scale use of Mumun (plain) pottery,

  • Bronze tools and weapons such as axes, daggers and spearheads,

  • Dolmens and other large burial monuments.

The moment these three elements appear together in a stable, systematic way across much of the peninsula
was rounded off as “around the 10th century BCE.”

Why that date?

  • Up until the 1960s–80s, there was very little in the way of absolute dating.

  • Archaeologists mostly relied on:

    • Pottery and weapon typology,

    • Stratigraphy (which layer lies above which),

    • And relative comparison with northeast China and Liaodong.

With so much uncertainty, the general mood was:

“Let’s err on the conservative side, not drag things too far back.”

That’s how the start of the Bronze Age in the central and southern peninsula ended up as the nice round number:

“roughly the 10th century BCE.”


2. The Surprise Numbers from Namgang, Gangneung and Sokcho

2-1. Namgang Dam reservoir: house floors pointing to the 16th–14th centuries BCE

In the 2000s, large-scale excavations around the Namgang Dam in Jinju, Gyeongsangnam-do began to shift the mood.

At sites like Okbang and Daepyeong, archaeologists uncovered very large dwelling sites—some with sides over 15 meters long.
Charcoal from the house floors was sent for radiocarbon dating.

The results?

  • ca. 1590–1310 BCE

  • ca. 1620–1400 BCE

Originally, the excavation team had thought these houses were maybe 5th–4th century BCE.
The radiocarbon dates pulled them back by almost a thousand years.

Other sites in the same reservoir area repeatedly yielded dates in the 14th–11th centuries BCE.

Taken together, these results strongly suggest that in the Namgang basin,
a Mumun pottery + early bronze + dolmen(-like) cultural package was already present
by around the 14th–12th centuries BCE.

2-2. Gangwon & Honam: older lights along the East Coast

Similar signals show up on the Gangwon East Coast and in Honam.

Bronze Age settlement sites at Gyodong and Bangnae-ri in Gangneung:

  • return radiocarbon dates in the 19th–15th centuries BCE,

  • and even under cautious interpretation, most scholars now place them around the 15th century BCE.

In Juknae-ri, Suncheon (Jeollanam-do), a Bronze Age dwelling has yielded C-14 dates in the 16th–15th centuries BCE in overseas lab analyses.

Put simply, along a belt running:

Gangwon East Coast → Namgang region in Gyeongnam → eastern Jeonnam,

there is now a high probability that:

  • Mumun pottery,

  • simple early bronzes,

  • and the earliest dolmen-style mortuary practices

were already present by the 15th century BCE.

2-3. A bronze axe from Sokcho: evidence of “already mastered” technology

The Joyang-dong site in Sokcho is one of the rare South Korean sites where a bronze axe has been found in a clear context.

Radiocarbon dating at the site places the relevant layer before the 9th century BCE,
i.e. roughly 3,000 years ago.

The important point here is the quality of the artifact.

  • The Joyang-dong axe is not a crude, experimental piece.

  • It shows refined casting technique—it’s closer to a “mature type” of bronze artifact.

If a bronze like that already exists by the 10th–9th centuries BCE,
then more primitive early bronzes almost certainly existed earlier
very likely before the 10th century BCE.

In the same vein:

  • Bipa-shaped bronze daggers and bronze axes from Birae-dong in Daejeon,
    found in dolmen contexts,

  • are now being dated to roughly the 10th–9th centuries BCE.

Putting this all together:

  • Settlement layers at Namgang, Gangneung, Suncheon15th century BCE,

  • Dolmens and bronze weaponry at Joyang-dong, Birae-dongaround the 10th–9th centuries BCE,

we get a picture where:

“A fairly advanced Bronze Age culture already existed considerably earlier than once thought.”


3. So Do We Now Just Declare a “15th-Century BCE Bronze Age”?

Here comes the key question:

“So should we simply rewrite the textbooks to:
‘The Korean Bronze Age begins in the 15th century BCE’?”

Most specialists would currently answer:

“Not yet.”

Roughly for three reasons.

3-1. Radiocarbon dates are ranges, not pinpoint years

C-14 dating produces results like:

  • 1620–1400 BCE (95% confidence interval)

It’s not saying:

“This house was built exactly in 1620 BCE,”

but rather:

“There is a 95% probability that it dates somewhere within this band.”

So you can’t take a handful of early dates and immediately proclaim:

“The entire Korean Peninsula’s Bronze Age starts in the 16th century BCE.”

The data simply aren’t that neat.

3-2. Different regions, different speeds

Resources like Our History Net (우리역사넷) summarize the Bronze Age of Liaodong, Liaoxi and Manchuria as:

  • largely emerging around the 13th century BCE,

  • and, after calibration, potentially traceable back to the 15th century BCE in some areas.

For the Korean Peninsula, it’s becoming likely that:

  • Northern regions / parts of the East Coast were early adopters,

  • while central inland and southwestern coastal areas picked it up somewhat later.

In other words, we may be looking at a “step-wise spread”:

  • 15th–13th centuries BCE: early Bronze + Mumun cultures in leading regions,

  • 12th–10th centuries BCE: diffusion to the south, consolidation into a “classic” Bronze Age society.

So rather than rewriting one line from:

“Bronze Age starts ca. 10th century BCE”

to:

“Bronze Age starts ca. 15th century BCE,”

it’s more accurate to say:

“Bronze Age cultures developed regionally between the 15th and 10th centuries BCE.”

3-3. The very definition of “Bronze Age” is changing

In the past, the rule of thumb was simple:

“If you find bronze, you’re in the Bronze Age.”

Today, that’s no longer enough.

Researchers now look at whole social packages:

  • Level of agriculture,

  • Signs of social stratification,

  • Dolmens / stone mounds and other monumental tombs,

  • Size of settlements and presence of defensive works,

and then ask:

“When do all these elements come together in a robust way?”

That’s when many would mark the “practical” beginning of the Bronze Age.

From that angle:

  • The 10th century BCE in older literature is still a “safe lower bound,”

  • While the Namgang–Gangneung–Suncheon C-14 dates are candidates for a plausible upper bound.


4. North vs South: Less About “Who Was Right” and More About “Who Can Adjust”

Online discussions often boil this debate down to something like:

“The North has said ‘20th century BCE’ since ages ago.
The South insisted on ‘10th century BCE’ and got wrecked by C-14.”

Reality is, as usual, more complicated.

  • North Korean archaeology has indeed argued since the 1960s
    that the Korean Bronze Age began as early as the 20th century BCE
    (this predates even the famous “Dangun tomb” claims).

  • South Korean archaeology, hampered by lack of data and lab tools,
    settled on the 10th century BCE as a cautious baseline.

If you only look at the direction of the new data, you could say:

  • The North’s “much earlier than you think” instinct was closer to the eventual trend,

  • While the South’s “10th century BCE and not much earlier” is clearly in need of adjustment.

But:

  • There is still nowhere near enough data to simply rubber-stamp “20th century BCE”.

  • Even if we gather all the radiocarbon results so far, the safest statement is something like:

“In certain leading regions, early Bronze-age cultures appear around the 15th–13th centuries BCE.”

So the real test isn’t:

“Who predicted an older date first?”

but:

“When new evidence appears, can we revise our narratives regardless of ideology?”

C-14 dating, cross-lab checks, and sending samples overseas are all part of that maturing process.


5. Three Big Questions Raised by a Moving Timeline

The exact number of centuries is interesting, but maybe not the most interesting part.
The deeper questions lie underneath.

5-1. If the Korean Bronze Age started “earlier than we thought,” what kind of society was it?

If Mumun villages, early bronze artifacts and dolmens are in place by the 15th century BCE,
then the Korean Peninsula–Manchuria zone was not just a scatter of tiny hamlets.

It implies a society that:

  • Had significantly adapted to agriculture,

  • Could mobilize large labor groups to build massive stone tombs,

  • And took part in metalworking networks that spanned multiple regions.

Regardless of what we think about “Dangun Joseon” as a historical state,
we are getting more and more grounds to imagine:

“Some form of complex early polities or confederations” in this area.

5-2. Where does the peninsula sit on the Bronze Age map of Northeast Asia?

If we line this up with:

  • Liaodong–Liaoxi–Liaoning Bronze Ages starting in the 13th century BCE,
    (again, potentially back to the 15th after calibration),

then a picture starts to emerge of:

A very early Bronze Age corridor running
Manchuria → Korean East Coast → South Sea coast.

In that scenario, the Korean Peninsula wasn’t just a passive endpoint receiving “civilization” from China.

It was more like one active node in a larger Northeast Asian network of:

  • bronze technology,

  • pottery styles,

  • and megalithic funerary culture.

5-3. How might textbooks change?

Instead of one neat, definitive line like:

“The Bronze Age on the Korean Peninsula began around the 10th century BCE.”

we’ll likely see more nuanced phrasing such as:

“Bronze Age cultures developed between the 15th and 10th centuries BCE,
with regional variation in timing.”

For students and readers, that’s a small but meaningful shift:

from “history as a set of fixed dates to memorize”
to “history as a timeline constantly revised by new data.”


6. Summary: Where the Debate Stands Right Now

To wrap up, here’s the current situation in one view.

Old orthodoxy

  • South Korean textbooks:
    Start of the Bronze Age (especially in the central–southern peninsula) =
    around the 10th century BCE.

New C-14 data

  • Multiple settlement sites in Namgang reservoir, Gangneung, Suncheon, etc.
    dating to the 16th–14th centuries BCE.

  • Dolmens and bronzes (axes, bipa-shaped daggers) from Joyang-dong (Sokcho) and Birae-dong (Daejeon)
    falling around the 10th–9th centuries BCE.

Direction of interpretation

  • We cannot yet safely declare:

    “The Korean Bronze Age starts in the 15th century BCE, full stop.”

  • But it is increasingly persuasive that:

    • In some leading regions,

    • early Bronze-Age cultures were already present by the 15th–13th centuries BCE.

What it means

  • Social complexity on the peninsula and in Manchuria emerged earlier than once assumed.

  • We gain more grounds to see Korea not as a “passive fringe,”
    but as an active axis in the Bronze Age landscape of Northeast Asia.

So the headline “The 10th-century BCE Bronze Age theory has collapsed!”
is a bit overdramatic.

A more precise summary would be:

“The 10th century BCE remains a solid lower bound,
but the upper bound has now been pushed back to around the 15th century BCE.”

History, in the end, is not a finished answer key.
It’s the ongoing work of quietly erasing and rewriting numbers
every time a new charred beam, a stray seed, or a cracked bronze blade
comes out of the ground and into the lab.

The Korean Bronze Age dating debate is simply one vivid example
that we’re standing right in the middle of that process.





Tuesday, November 18, 2025

A Northern Song Emperor’s Tomb in Hoeryeong?



A Northern Song Emperor’s Tomb in Hoeryeong?




Chasing the Mystery of Undusan Fortress and “Wuguo Fortress”

“Somewhere in the hills along the Tumen River, there lies the tomb of a Northern Song emperor.”

At first glance it sounds less like a documentary and more like the title of a light novel.
But this line actually comes from a late Joseon scholar’s own written account.
And behind that one line lies a long shadow of Northeast Asian border history:
the fall of the Northern Song, the rise of Jin, and the tangled frontier of Goguryeo–Jurchen–Joseon–Qing.

In this piece, instead of jumping straight to

“This is DEFINITELY the Northern Song imperial mausoleum!”

we’ll take a slower route:

  1. What do the actual historical sources say?

  2. How does modern scholarship interpret them?

  3. And why is the Hoeryeong theory such an intriguing little mystery?


1. The “Tomb of the Song Emperor” Puzzle in Taengniji

The late Joseon geographer Yi Jung-hwan, author of the famous geography book Taengniji (택리지), passed through Hoeryeong in Hamgyeong Province and left a curious note. Summarized:

  • If you follow the Tumen River upstream to Undusan Fortress near Hoeryeong,
    you’ll find several large burial mounds on a hill outside the fortress.

  • Local people call these mounds “Hwangjerung” – the Emperor’s tombs.”

  • In 1712, when the Qing envoy Mukedeng (穆克登) visited to set up the border stele for the Tumen and Yalu,
    his party saw these tombs and tried to open one.

  • In the process, they found a small, short stele. On it were carved four characters:
    “宋帝之墓” – “Tomb of the Emperor of Song.”

  • Mukedeng then ordered the grave to be repaired and the mound rebuilt on a grander scale before leaving.

Because of this, the rumor that

“This might be the tomb of a Northern Song emperor”

spread widely around Hoeryeong. Yi Jung-hwan himself notes that he doesn’t know exactly which emperor it might be, but the idea stuck with him.

If we then add later Joseon and early modern gazetteers, travelogues, and even Japanese colonial-era photographs, we get at least this much:

  • Inside Undusan Fortress at Hoeryeong, there was an old cluster of tomb mounds
    traditionally called “Hwangjerung” or “Tomb of the Song Emperor.”

  • Japanese investigators even photographed it, labeling the images
    “皇帝塚 (Emperor’s Mound)” or “宋皇帝塚 (Tomb of the Song Emperor)” in the captions.

So far, so good. The place existed, and the name existed.

Which naturally raises the next question:

Then which “Song emperor” is this supposed to be?

Of course, it could simply be a tomb of some local elite from the Song region.
But popular imagination jumped straight to the two tragic last emperors of the Northern Song.


2. The Fall of the Northern Song and the Two Emperors’ Bitter End

The Northern Song collapsed in 1127 when the Jurchen Jin dynasty stormed the capital in what is known as the Jingkang Incident (靖康之變).

  • Emperor Huizong (徽宗, Zhao Ji) and his son

  • Emperor Qinzong (欽宗, Zhao Huan)

were both captured and taken north by the Jin.

The Jin first kept them near the capital area, then transferred them:

  1. To Shangjing (上京),

  2. Then to Hanzhou (韓州, roughly in modern Jilin Province),

  3. And finally to a fortress called “Wuguo Fortress” (五國城).

According to the standard Chinese histories (Song History, Jin History and related texts):

  • Huizong died in Wuguo Fortress in 1135.

  • Qinzong lived longer, dying sometime in the 1160s, also within Jin territory.

Later, in the south, the newly established Southern Song under Emperor Gaozong (Zhao Gou) negotiated with Jin to reclaim Huizong’s coffin.

  • Huizong’s remains were then ceremonially reburied in Song territory,
    at Yongyou Mausoleum (永祐陵) near modern Shaoxing in Zhejiang.

Put together, the standard storyline is:

  • Both emperors spent their final years near Wuguo Fortress, under Jin control;

  • At least Huizong’s coffin was later moved south and reinterred in a newly built imperial tomb in Zhejiang.

On paper, this sounds neat and tidy:

The “final resting place” is in Zhejiang – end of story.

But the trouble starts when we ask:

“Okay, then where exactly was Wuguo Fortress?”


3. Where Was Wuguo Fortress? – Majority vs Minority Views

3-1. Majority view: Near Yilan in Heilongjiang

Most Chinese and Korean scholars today identify Wuguo Fortress with a site near Yilan (依蘭) in Heilongjiang Province.

  • Near the confluence of the Heilong (Amur) and Songhua Rivers,
    there is a Jin-period fortress site with visible walls and moats.

  • This site is officially listed and marked as “Wuguo Fortress Ruins (五国城遗址)”,
    protected by the local authorities.

When you combine:

  • The geographical hints in Song History, Jin History, and other old gazetteers, and

  • Modern mapping of the Songhua–Mudan River region,

the Yilan area matches quite well. That’s why, at present, the Yilan identification carries the most academic weight.

3-2. Why the confusion? Conflicting coordinates in old texts

Even so, it’s hard to slam the gavel and declare,

“Case closed. Wuguo Fortress = Yilan and nowhere else.”

Why?

Because over the Jin–Yuan–Ming centuries, different gazetteers gave slightly different descriptions of Wuguo Fortress:

  • Some texts place it “a few hundred li toward Liaodong,”

  • Others say “near the border with Joseon,”

and so on. The distances and directions don’t always line up cleanly,
which left plenty of room for later scholars to get confused.

Out of that fog of ambiguous coordinates came the question:

“Is it possible there was another fortress called Wuguo closer to the Joseon border—say, right across the Tumen River?”

3-3. Minority view: Undusan Fortress at Hoeryeong = Wuguo Fortress?

That’s where the “Hoeryeong Undusan Fortress = Wuguo Fortress” hypothesis steps in.

Undusan Fortress (sometimes also called Oguk Fortress / O-guk-seong (五國山城) in some older materials) is:

  • A massive stone fortress clinging to cliffs along the Tumen River,

  • With a perimeter of about 6 km.

  • Its gates, command posts, and water gates are clearly laid out,
    marking it as a key node in the northeastern defense line, likely of Goguryeo origin.

On the hill near this fortress is a group of large tomb mounds that:

  • Were already known in late Joseon times as “Hwangjerung” (Emperor’s tomb),

  • And, according to Taengniji, bore the stele inscription “Tomb of the Emperor of Song (宋帝之墓).”

Some Korean researchers take this a step further and argue:

  • If we re-read the fuzzy coordinates in the old texts,
    it’s possible to interpret Wuguo Fortress as not only the Songhua–Heilongjiang region,
    but also the Tumen–Hoeryeong region.

  • Therefore, the fortress where Huizong and Qinzong spent their final years could be Undusan Fortress,
    and the “Tomb of the Song Emperor” stele might refer to one of these two captured emperors.

This is very much a minority hypothesis, not a mainstream conclusion.

Still, it’s a fascinating thought experiment at the crossroads of:

  • Border fortresses,

  • Old tombs and local legends,

  • Joseon scholars’ field notes, and

  • The fall of the Northern Song.


4. So What Is Actually in Hoeryeong?

Time for the practical question:

“Okay, but what is that ‘Emperor’s tomb’ on the hill at Undusan Fortress?”

4-1. Could it really be a “true” Northern Song imperial tomb?

Based on what we know so far,
the chances that this is the final, primary tomb of a Northern Song emperor are extremely low.

  • We already have records that Huizong’s coffin was brought south in the Southern Song era
    and reburied at Yongyou Mausoleum near modern Shaoxing.

  • That tomb complex, although looted and damaged many times over the centuries,
    has remained known and locatable in historical memory.

If the Hoeryeong mound were actually the real burial site of the emperor’s remains,
we’d have to assume that the Southern Song court:

  • Faked the location of the imperial tomb, or

  • Entirely fabricated a major part of their own ritual record.

That would require a huge historical conspiracy—and we currently have no solid evidence to support anything that dramatic.

4-2. More plausible: a cenotaph, symbolic tomb, or later memorial

Still, the stele inscription “Tomb of the Emperor of Song” is not something we can shrug off.
If we force ourselves to map out a few plausible scenarios:

1. Memorial or cenotaph

  • At some point in the Jin–Yuan–Ming–Qing continuum, someone—
    perhaps officials or migrants with a cultural memory of the Song—
    might have erected a symbolic tomb or memorial without actual imperial remains.

  • Both in China and Korea, building “empty tombs” (가묘) or symbolic shrines
    for fallen dynasties and exiled rulers was not unheard of.

2. Tomb of a local elite claiming Song heritage

  • It’s also possible that some local power-holder or community in the north,
    of Song origin or loyalist sentiment,
    used the title “Emperor of Song” for themselves or for a revered ancestor.

  • But we have no corroborating texts or inscriptions supporting this scenario yet.

3. Misreading, exaggeration, or later reinterpretation

  • The original stele is now lost; we only have Yi Jung-hwan’s transcription from Taengniji.

  • It’s entirely possible that “宋帝之墓” was:

    • Misread in the field,

    • Mis-copied later, or

    • Interpreted in a way the original stone never intended.

  • In late Joseon popular speech, titles like “emperor,” “princess,” “general” were often used loosely.
    The label “Hwangjerung (Emperor’s tomb)” may itself be a product of such loose, legendary naming.

4-3. The core problem: we lack decisive archaeological data

The biggest issue is this:

There has never been a fully published, modern archaeological excavation
of the Undusan fortress tomb cluster that answers the key questions.

From:

  • Late Joseon texts,

  • Japanese colonial-era photos,

  • and some North Korean heritage designations,

we know only that:

  • There are large mounds,

  • They were locally known as the Emperor’s Tomb / Song Emperor’s Tomb, and

  • The Taengniji records a stele reading “Tomb of the Emperor of Song.”

But we still don’t have:

  • A detailed excavation report,

  • Clear plans of the internal structure,

  • Artifact lists,

  • Radiocarbon dates,

  • Or stratigraphic analysis.

So, as of today, the most academically honest conclusion is something like this:

“The ‘Emperor’s tomb’ on the hill at Undusan Fortress in Hoeryeong
is undoubtedly a historically interesting site with a rich local tradition.
However, there is currently no hard evidence that it is the actual primary tomb
of a Northern Song emperor.
Even if it does have some connection to the Northern Song,
it is safer to treat it as a possible cenotaph or later memorial
rather than a confirmed imperial mausoleum.”

And that, for now, is where the mystery quietly sits—
half in the documents, half in the earth, and very much still open to future digging,
both literal and scholarly.




Goguryeo Army vs Roman Army – Who Would Win?



Goguryeo Army vs Roman Army – Who Would Win?




An Imaginary Matchup to Read East–West Military Civilizations

If you hang around online history forums long enough, you’ll see this question pop up over and over:

“If the Goguryeo army fought the Roman army, who would win?”

Goguryeo was the military powerhouse of Northeast Asia that held out against Sui and Tang.
Rome was the giant empire wrapped around the entire Mediterranean.

It’s the kind of matchup where fandom divides instantly:

  • “Goguryeo had more cavalry, so they’d steamroll Rome,” vs

  • “Roman legionaries would just cut them to pieces.”

And the debate usually degenerates into pure emotion.

In this piece, instead of crowning a winner outright, we’ll ask:

  • Under what conditions and which time periods is it even fair to compare them?

  • How far can we reconstruct army size, troop types, tactics, and command systems from the sources?

  • And what does this comparison tell us about the actual level of ancient Eastern and Western civilizations?

Let’s walk through it calmly.


1. First, Sync the Timelines

Goguryeo and Rome aren’t really “exact contemporaries” in the strict sense.

Goguryeo’s heyday

Roughly 4th–6th century CE, around the reigns of Gwanggaeto, Jangsu, and Yeongyang.

Militarily, Goguryeo is most famous for its wars against Sui and Tang in the 6th–7th centuries. (Wikipedia)

Rome’s heyday

Late Republic to early Empire (roughly 2nd century BCE to 2nd century CE).

Especially under Trajan–Severus, the empire fielded about 28–33 legions,
and with auxiliaries, some 400–450,000 standing troops. (Wikipedia)

By the time Goguryeo was fighting Sui and Tang in the 7th century,
“Rome” in the West was already in its Eastern Roman / Byzantine phase.

So most historians don’t compare “Goguryeo vs Rome” directly, but rather:

  • Goguryeo ↔ Sui–Tang, and

  • Rome ↔ Han / Later Han,

matching them with their primary regional rivals. (scholarsarchive.byu.edu)

Here, we’ll allow ourselves one controlled fantasy:

Compare the respective peak capabilities of each side.


2. Population and Mobilization – The Reality of Numbers

Goguryeo’s mobilization capacity

If you combine the Samguk Sagi, Chinese histories, and modern scholarship,
Goguryeo at full mobilization is usually estimated at up to ~300,000 troops. (Wikipedia)

Recent research puts Goguryeo’s population roughly in the 2–3 million range. (De Gruyter / Brill)

So a realistic upper bound might be:

“A military state with around three million people
that could, in a crisis, drag close to 300,000 into the field.”

Rome’s mobilization capacity

At its peak, the Roman Empire’s population is estimated at 50 million+. (Wikipedia)

  • End of Augustus’ reign: c. 250,000 standing troops (legionaries + auxiliaries). (Wikipedia)

  • Under Septimius Severus (c. 211 CE):
    33 legions + over 400 auxiliary units = around 450,000 standing troops. (Wikipedia)

In sheer population and permanent military manpower,
Rome absolutely dwarfs Goguryeo.

So those DC-style claims like “Goguryeo 400k vs Rome 350k, similar numbers”
don’t line up with current scholarship.

That said, it is true that Goguryeo showed an unusually high mobilization rate
for its population size—classic behavior for a small frontier state
forced to face giant empires head-on.


3. Troop Types and Equipment – What Did They Actually Fight With?

3-1. Goguryeo’s forces

Direct written descriptions of Goguryeo’s army are scarce,
but murals, archaeological remains, and Chinese accounts allow us to reconstruct roughly:

(Wikipedia; art-and-archaeology.com)

Heavy cavalry (armored horsemen)

  • Horse and rider covered in lamellar armor—essentially cataphract style.

  • Armed with long lances, ring-pommel swords, and shields.

  • Elite shock troops designed to punch holes in enemy lines.

Horse archers

  • Mounted archers firing from the saddle.

  • Goguryeo / Maek composite bows were famous for high draw weight and penetration.

Infantry (heavy and light)

  • Lamellar or scale armor, armed with spears, shields, axes, etc.

  • Used for fort defense, choke points, siege work (assault and defense), and as support in field battles.

Specialized siege and defense units

  • Troops trained to carry ladders, scale walls, handle siege engines, and so on. (Wikipedia)

Strengths:

  • High proportion of heavy cavalry and horse archers.

  • Excellent use of mountain fortresses and terrain,
    especially in defensive battles and attritional warfare. (Wikipedia)

3-2. The Roman army

The Roman army changes structure over time,
but a typical early imperial (Principate) setup looks like this: (Wikipedia)

Legionaries

  • Heavy infantry. Armed with scutum (large rectangular shield), gladius (short sword), and pila (heavy javelins).

  • Each legion: roughly 4,800–5,500 men, forming the core combat block.

Auxilia (auxiliaries)

  • Non-citizen troops from the provinces.

  • Provided more than half of the empire’s total manpower.

  • Included most of the cavalry, archers, and special units. (Wikipedia)

Cavalry (ala, cohors equitata)

  • Many recruited from traditional horse cultures: Gauls, Germans, Numidians, etc.

  • Mixed heavy and light cavalry using lances, javelins, swords, and sometimes bows. (Wikipedia)

Archers (sagittarii)

  • Often from Eastern regions such as Crete, Syria, and Palmyra.

  • Deployed throughout Roman forces as dedicated archer units. (Roman Army Museum)

Strengths:

  • A thick, disciplined heavy infantry line of legionaries,

  • Combined with cavalry, archers, and slingers arranged around them in a fairly standardized combined-arms system.

So the online cliché, “Romans had no cavalry and couldn’t use bows,”
is based on overextending a few early Republican examples.
By the imperial period, Rome was actively using allied cavalry and archers through the auxilia. (Wikipedia)


4. Tactics and Command – How Systematic Were They?

4-1. Goguryeo command and tactics

We lack detailed organizational charts for Goguryeo,
but Chinese sources and modern research point to a few key traits. (Wikipedia)

Five-division (五部) structure + regional forces

  • About 12,500 elite cavalry as the capital guard.

  • Multiple regional forces of 20–30,000 stationed in key areas.

Large-scale royal hunts doubled as military exercises,
training soldiers’ riding and archery skills.

Strategically, Goguryeo excelled at:

  • Using fortresses, rivers, and mountain passes for delay and chokepoint defense,

  • Employing cavalry and archers for feigned retreats, ambushes, and counter-encirclement.

Famous examples include the Battle of Salsu and
multiple campaigns where Goguryeo wore down massive Sui/Tang armies
and then smashed them. (Wikipedia)

Some scholars, however, argue that because units were organized more by weapon type or troop kind,
it may have been harder for a single formation to execute fully integrated combined-arms tactics. (Wikipedia)

4-2. Roman command and tactics

Rome is famous for its small-unit command structure and standardized training. (Wikipedia)

Centuries and cohorts

  • The basic building block was the century (about 80–160 men),
    led by a seasoned centurion.

  • Six centuries formed a cohort,
    and ten cohorts formed a legion.

Because each century trained together with its own officers over long periods,
Roman units could repeatedly execute quite complex battlefield maneuvers.

Under capable commanders (Caesar, Trajan, etc.), Roman armies could pull off:

  • Double-envelopment maneuvers like Cannae,

  • Feigned retreats, night attacks,

  • Sophisticated siege warfare and river crossings. (scholarsarchive.byu.edu)

We don’t have enough detail to run a fine-grained “century vs Goguryeo unit” matchup,
but purely in terms of documented layered command structure and standardization,
the Roman side is more fully reconstructed—and that almost certainly reflects a highly organized training system.


5. “If They Actually Fought?” – A Thought Experiment

From a historian’s standpoint, forcing two states that never met onto the same battlefield
is asking for trouble. But as a thought experiment, it’s fun.

Let’s assume a neutral plain, and equal numbers—say 50,000 on each side.

Conditions favoring Goguryeo

A wide plain with gentle rolling hills,

where heavy cavalry and horse archers can swing around the flanks and rear,
probing and harassing the Roman line.

Rome might be caught off guard at first by:

  • Eastern-style mobile warfare,

  • Constant harassment, feints, and “Parthian shot”–style backward archery.

As in the wars against Sui and Tang,
Goguryeo hitting supply lines, luring the enemy toward mountain fortresses,
and then counterattacking could be extremely dangerous for Rome. (Wikipedia)

Conditions favoring Rome

A narrower, more constrained plain

where large flanking moves are difficult,
and Roman heavy infantry and siege equipment can do their work.

If Rome has time to:

  • Adapt its allied cavalry and archers to Goguryeo-style warfare,

  • Build up defenses with fieldworks, stakes, shield walls, and layered formations,

then the shock power of Goguryeo heavy cavalry is greatly reduced. (Wikipedia)

If the fight turns into a prolonged siege campaign where Roman engineering shines,
defending Goguryeo fortresses frontally could become nightmarishly hard. (Wikipedia)

In the end, who wins depends entirely on:

  • Terrain,

  • Logistics,

  • Intelligence,

  • And above all, the quality of commanders.

No serious scholar today claims, “Side A is definitely superior to Side B in all respects.”


6. What This Comparison Actually Tells Us About East–West Civilizations

The biggest trap in “Goguryeo vs Rome” threads is how fast they slide into:

  • Inferiority complex, or

  • Blind nationalism.

“Eastern strategy is sophisticated; Western warfare was just thuggish brawling.”

vs

“Rome was on a totally different level; East Asia lagged across the board.”

Both are heavy overstatements given current research.

In modern comparative history, scholars usually focus on “Han vs Rome” and conclude roughly this: (scholarsarchive.byu.edu)

Military and administration

  • Rome: decentralized provincial governance, flexible legion system,
    strong bond between citizenship and military service.

  • Han / Goguryeo sphere: bureaucratic document culture, counties-commanderies + army integrated,
    excellent at large-scale mobilization and frontier defense.

Technology and economy

  • Rome: roads, aqueducts, concrete, glass, iron smelting—
    exceptional civil engineering and urban infrastructure. (Reddit)

  • East Asia: iron agricultural tools, irrigation and flood control, paper, advanced fortification methods—
    high agricultural productivity and a state built around administration.

War culture

  • East Asia: treatises like The Art of War, a massive military-text tradition,
    and a view of war as an extension of political technique. (Taylor & Francis Online)

  • Rome: practical experience converted into training manuals, engineering,
    and a legal/organizational culture around the legions.

So rather than one side being “universally superior,”
we’re looking at two highly developed civilizations that optimized in different directions.

Goguryeo, within this East Asian military world, was:

  • Not “a random small border kingdom of China,”

  • But a state with its own fairly sophisticated military and administrative system,

  • And a tough frontier power that held out for a long time
    against the Sui and Tang “superpowers” to its south. (Wikipedia)


7. How to Enjoy the “Goguryeo vs Rome” Debate Productively

To sum up:

On raw numbers

  • In population and standing armies, Rome is much larger.

  • Goguryeo is a high-mobilization military state relative to its population size.

On troop types and weapons

  • Goguryeo: strong in heavy cavalry, horse archers, and fortress defense in rough terrain.

  • Rome: strong in heavy infantry and standardized combined arms with auxiliaries; excellent at siege/engineering.

On tactics and command

  • Goguryeo: fewer records, but clearly very capable at terrain use, mobile warfare, and fortress warfare.

  • Rome: multi-layered command and training so systematic that we still have the details.

On “civilization levels”

Current scholarship leans much more toward:

“They developed in different, equally sophisticated ways

than toward:

“X was definitely above Y in every respect.”

So the best way to enjoy “Goguryeo vs Rome” isn’t to boost one side and trash the other.

It’s to ask:

“Where does Goguryeo really sit on the global ancient-civilization map?”

and try to see that position in 3D.

Viewed that way, Goguryeo was neither:

  • A trivial border state overshadowed by China, nor

  • A transcendent super-civilization that “outclassed Rome.”

It was:

A hyper-efficient military frontier state
that had to fight giant empires with a relatively small population,
and one of the key players in Northeast Asia’s ancient history.




Monday, November 17, 2025

The Most Henpecked Emperor in Chinese History?



The Most Henpecked Emperor in Chinese History?




The Strange Romance of Emperor Wen of Sui and Empress Dugu

When we hear the word “emperor,” we usually picture something like this:

Hundreds of maids and concubines,
court in the morning, banquets at night,
and a man who enjoys every possible privilege under the sun with his so-called “three thousand palace women.”

But in Chinese history, there is one imperial couple who openly broke this script.

The man is Emperor Wen of Sui (Yang Jian), the ruler who ended the chaos of the Northern and Southern Dynasties and reunited China.

At his side stood a woman who insisted on “one man, one woman”,
and who turned even an emperor into a textbook henpecked husband:
the formidable Empress Dugu (Dugu Qieluo / Kara Dugu / “Dugu Chela”).

Their story is more than a cute “love story.”

It’s a fascinating case study of:

  • An emperor’s private life and public power,

  • How marriage intersects with politics,

  • And how one woman’s personality and principles
    could reshape an entire dynasty.


1) The Unifier Yang Jian – and the Young Girl Standing Next to Him

Yang Jian grew up as a scion of a northern aristocratic family,
moving through the courts of Western Wei and Northern Zhou.

As a youth, he was the son of the general Yang Zhong,
and he caught the eye of Dugu Xin, a major strongman of Northern Zhou.
At sixteen, Yang Jian married Dugu Xin’s daughter, Dugu Qieluo.

The Dugu clan was one of the famed “Eight Great Noble Families of the North.”

  • The eldest daughter became empress of Northern Zhou,

  • Another daughter married into the imperial family,

  • And Dugu Qieluo herself would later become empress of Sui.

In other words, it was literally a “one family, three empresses” situation.

Yang Jian and Dugu Qieluo, who had known each other since childhood,
are said to have been not just a political match,
but emotionally very close as husband and wife.

There’s a famous vow that Yang Jian is said to have made to her in his youth:

“I will not let any woman but you bear my children.”

This wasn’t just romantic drama someone made up centuries later;
versions of this vow appear repeatedly in more formal historical traditions as well.

And in practice, the couple stuck to it:

Dugu Qieluo personally gave birth to ten children
five sons and five daughters, all of them Yang Jian’s.

In a world where emperors were expected to scatter offspring across dozens of concubines,
this level of de facto monogamy makes Emperor Wen almost unique in Chinese imperial history.


2) “These Two Rule the Empire Together” – The Couple Called the “Two Sages”

In 581, Yang Jian deposed the young emperor of Northern Zhou,
founded the Sui dynasty, and became Emperor Wen.

He then went on to end centuries of north–south division,
destroying the Chen dynasty in 589
and reunifying China.

Throughout this process, Empress Dugu was not content to be
the classic “virtuous wife quietly supporting from behind the curtain.”

She:

  • Read state documents together with the emperor,

  • Offered opinions on rewards, punishments, and appointments,

  • Accompanied him right up to the door of important meetings,

  • Had eunuchs report to her on what was happening inside,

  • And when the emperor made a bad decision, she corrected him with blunt remonstrance.

Because of this, courtiers sometimes referred to Emperor Wen and Empress Dugu as:

“The Two Sages in the Palace (二聖, er sheng).”

In that phrase you can hear two feelings mixed together:

  • On the one hand, recognition of a genuine political partnership,

  • On the other, a hint of unease—
    that the empress’s power might be too great for everyone’s comfort.


3) “If You Mess Around With Another Woman, I’ll Kill Her”

Empress Dugu was not only a sharp political partner—
she was also an intensely jealous woman of iron principles.

The official histories preserve one particularly vivid scene that sums up her jealousy.

● The Emperor’s Slip – and Dugu’s Bloody Response

One day, in the palace, a woman named Lady Yuchi was working as a lowly maid.
She was the granddaughter of Yuwen Shu/Yuwen Rong’s clan—
a general whose family had risen in rebellion, lost,
and been reduced to slavery.

Emperor Wen noticed her beauty
and slept with her.

When Empress Dugu found out, she did not leave the matter alone.

She had Lady Yuchi hunted down and killed—
cutting off her husband’s “moment of weakness” in blood.

When the emperor learned of this after the fact,
he was furious. He mounted his horse and rode out of the palace, declaring in effect:

“I’d rather quit being emperor than live like this, under such interference.”

● The Emperor Storms Out – and the Ministers Step In

For a time, Emperor Wen did not return.
Senior officials like Gao Jiong and Yang Su went after him and tried to reason with him:

“Your Majesty, how can you abandon the empire over a single woman?”

By nightfall, the emperor finally returned to the palace.

According to the story, he found Empress Dugu waiting for him at the gate,
having stayed there all night,
tearfully apologizing and begging for reconciliation.

From this episode alone, you can see why people say:

“It wasn’t just the emperor—
even the ministers were constantly watching Empress Dugu’s mood.”

The Korean internet meme of Sui Wen as
“the ultimate henpecked emperor”
did not come out of thin air.


4) Was He Really Henpecked, or a Man Who “Feared and Respected” His Wife?

Here’s where it gets interesting:
Emperor Wen is not just a tragic husband who “lived in fear of his wife.”

① More Than Henpecked – An Emperor Who Respected His Political Partner

Emperor Wen didn’t only fear Empress Dugu;
he also deeply trusted her as a political ally.

  • When members of her own clan abused their influence,
    Dugu was the one who demanded their punishment.

  • When the emperor handed down excessively harsh decisions,
    she would go on hunger strike and submit petitions until lives were spared.

So she was:

A wife who kept her husband under tight control,
and at the same time
a cool-headed politician who wouldn’t bend to kinship or factional pressure.

From Emperor Wen’s perspective,

  • Politically, she was his most reliable advisor,

  • Privately, she was the one person he truly feared.

Those two roles overlapped in a single human being.

② The Fall of Gao Jiong – The Price of Saying “A Woman Is Just a Woman”

The story about Gao Jiong’s line
“Will you abandon the empire over a woman?”
actually has a fascinating sequel.

On the surface, his words helped bring the emperor back to the palace.

But Empress Dugu quietly remembered that phrase—
the way he had referred to her as “just a woman.”

Later, she began to dig into Gao Jiong’s private life:

  • After vowing not to remarry when his wife died,
    he quickly took a concubine and had a son with her.

Empress Dugu repeatedly reported this to the emperor as hypocrisy:

“His words and deeds do not match.”

In the end, when the Sui campaign against Goguryeo faltered,
Gao Jiong was saddled with responsibility for the failure
and pushed out of high office.

In modern terms, you could sum it up like this:

“He butted into the boss’s marital drama,
dropped one poorly chosen line,
and his career trajectory went from ‘glory’ to ‘crash and burn.’”


5) After Dugu’s Death – Did the Emperor Really Become a “Freed Man”?

In 602, Empress Dugu died, around the age of fifty.

The official record says that Emperor Wen:

  • Was devastated by her death,

  • And sank into a deep period of grief and loss.

Soon after, though, another change appears in the sources:

“After that, the emperor took two palace women as consorts—
Lady Chen and Lady Cai—and began to lavish favor upon them.”

In other words:

  • While Dugu was alive, he was almost a model husband
    who kept his vow in practice.

  • Once she was gone, he seems to have tried—finally and belatedly—
    to enjoy a bit of what imperial status traditionally allowed.

But by then Emperor Wen was already in his early sixties,
and he died just two years later, in 604.

Online jokes that he “overindulged in women after being freed and shortened his life”
are obviously colored by modern imagination.

Still, the basic pattern holds:

  • It was only after Dugu’s death that he seriously took concubines,

  • And he died not long after.

That much is broadly consistent with the historical record.


6) How Was He Different from Other “Henpecked Emperors”?

Chinese history has a few emperors often labeled as “henpecked”:

  • Tang Gaozong, overshadowed by Empress Wu Zetian,

  • The Xianfeng Emperor of Qing, overshadowed by Empress Dowager Cixi, and others.

But the relationship between Emperor Wen and Empress Dugu
is of a different kind.

Gaozong and Xianfeng were emperors who gradually lost power
due to political weakness, health, or circumstances.

By contrast, Emperor Wen and Empress Dugu were:

A relatively normal married couple
who simultaneously functioned as near co-rulers of the state.

That’s why some scholars suggest that, for Emperor Wen:

  • Instead of calling him a simple “henpecked husband,”

  • It’s more fitting to see him as a “respect-and-fear husband”
    a man who deeply respected and simultaneously feared his wife.


7) Looking at Emperor Wen’s “Henpecked Legend” from Today’s Perspective

Seen with modern eyes, this imperial couple can be read on several levels.

● On the level of personal love

An emperor who spends effectively his whole life with one woman,
and swears that no other woman will bear his children—
that alone is an extraordinary exception in the long history of East Asian monarchies.

● On the level of power and jealousy

From a today’s perspective, Empress Dugu’s jealousy and punishments
can look extreme and violent.

But behind that is a very clear understanding of palace politics:

“If a concubine bears a son,
she instantly becomes a political rival.”

Her brutality is inseparable from the brutal logic
of succession and faction struggle inside the court.

● On the level of political partnership

Empress Dugu was not a background supporter.
She was a genuine co-governor who shaped policy and personnel decisions.

Emperor Wen, for his part, acknowledged this
and shared power with her.

That joint rule is a big reason why
the early Sui dynasty achieved such stable and efficient government.

So the “henpecked emperor” legend of Sui Wen

is less about a “pitiful man crushed by his wife”
and more about a rare case where love, jealousy, power, and trust
were tightly braided together in one marriage at the very top of the state.


8) Conclusion – Why Talk About Emperor Wen and Empress Dugu Now?

At first glance, the story of Emperor Wen and Empress Dugu
looks like a small side episode in the vast sweep of Chinese history.

But it’s a case that makes us think simultaneously about:

  • Relationships between men and women,

  • Power and responsibility,

  • Monogamy and the harem system,

  • And where the line lies between family and state.

By contemporary standards,
Empress Dugu’s actions were clearly problematic and violent in many ways.

Yet at the same time, she was:

  • A woman who discussed state affairs on equal footing with her husband,

  • A figure whose influence was so great that even the emperor bent before it,

  • And someone who played a deep role in the rise and fall of a dynasty.

In that sense, she stands out as one of the most intense empresses in Chinese history.

And beside her,
moving constantly between love and fear,
respect and exhaustion,
stood Emperor Wen of Sui.

His life reminds us that:

“Even an emperor is, in the end, someone’s husband.”

In a long history full of emperors who collected concubines like stamps,
there was exactly one who swore:

“I will have children with one woman only.”

That man is the one Korean netizens love to call:

“The most legendary henpecked emperor in Chinese history – Emperor Wen of Sui.”



Not Rich, But Beautiful: What Kim Gu Really Meant by a “Cultural Nation”

Was Kim Gu naïve when he said he wanted Korea to be “the most beautiful nation,” not the richest? A closer reading shows a hard-edged bluep...