I was in Wanhua not long ago.
Standing in front of Longshan Temple, staring at the old streets, a question hit me: during the Qing dynasty, Taiwan’s administrative center was in Tainan. So why did it end up in Taipei?
Most people would say it’s obvious — the north developed, the population grew, it just made sense to upgrade.
But after reading the Revised Taiwan Provincial Chronicles: Administrative Evolution, I realized this was anything but “natural.”
It was a three-century-long, data-driven system iteration. Every county established, every indigenous community brought under state control, every boundary adjusted — behind each one was a set of precise numbers.
Today, I’m going to walk you through this process, using what I found in these chronicles.
I. 1661: Koxinga Drew the First Line in Tainan
The starting point of Taiwan’s administrative structure is 1661.
After Koxinga expelled the Dutch, he did something most people overlook: he didn’t build walls first. He drew a map first.
He established Chengtian Prefecture in present-day Tainan, with Tianxing County and Wannian County beneath it. Then he implemented a system that would shape Taiwan for centuries: yu bing yu nong — “feeding the army through farming.”
What does that mean?
Split the troops in two. One group guards the city. The other gets farmland and grows their own food. Land was assigned by regiment; farming was done by regiment. Your regiment determined which field you worked.
The brilliance of this system: it used military organization to directly cover agricultural development. Wherever the army went, administrative control followed. Within just a few years, it had pushed from Tainan all the way to Xingang, Madou, and even as far as Banxian (present-day Changhua).
“Feed the army with grain, define the border with soldiers” — that was the first line of code in Taiwan’s administrative map.
II. 1686: Kangxi Took Over an Island with Three Numbers
After the Qing dynasty pacified Taiwan, one key figure stepped forward.
Shi Lang.
He submitted the famous memorial Respectfully Stating the Case Against Abandoning Taiwan. His argument: based on “geographic advantage” and “strategic protection,” Taiwan must have formal administrative institutions.
Kangxi listened.
In 1686, the “One Prefecture, Three Counties” structure was formally established:
- Taiwan County: 4 wards, 15 inner districts, 1 village.
- Fengshan County: 7 inner districts, 1 ward, 2 villages, 65 tribal communities.
- Zhuluo County: 4 districts, 9 wards, 9 communities.
Look at those numbers. Not “approximately.” Not “several.” Precise to the digit.
Total land and poll tax revenue for the entire island? 186,000 taels of silver. The governance reach corresponding to that number was entirely concentrated along the western coastal plains. The east? The mountains? Sorry — not in the database.
When I saw these numbers, I didn’t think of history. I thought of a product manager drawing the first version of a user map. Only the coastal plain data had been onboarded. The mountains were still in beta.
III. 1731: A Problem Called “The Whip Is Too Short”
By the Yongzheng era, things had changed.
Northern immigration was surging. “Wild lands opened daily” — an ancient phrase that means every morning, more forest had been turned into farmland. But here’s the problem: the administrative center was still in Tainan. If something happened up north, it took hundreds of li to respond.
The official documents of the time used a phrase: chang bian mo ji — “the whip is too short to reach.”
So in 1731, a crucial adjustment was made: all judicial and financial matters north of the Dajia River were transferred to the Tamsui Department, with its seat moved to Zhuqian (present-day Hsinchu). That single stroke covered 345 li of territory.
Then, in 1810, another milestone: Wu Sha led settlers into Yilan in large numbers, and the government formally established the Kavalan Department, bringing 130 li of eastern land — from Yuanwangkeng all the way to Su’ao — under administrative monitoring.
Notice the logic here: it wasn’t the government going first and settlers following. It was settlers going first, building enough scale, and then the government catching up to stamp an official seal on it.
By the late 19th century, the population of the Taipei and Yilan regions had surpassed 420,000 people, and their exports — indigo, coal, tea, camphor — had become Taiwan’s most valuable commodities.
When the data changed, the administrative boundaries had to follow.
IV. 1874: Japanese Gunboats Kicked Open Taiwan’s Back Door
In 1874, an event changed Taiwan’s fate: the Mudan Incident.
Japan used the killing of Ryukyuan fishermen as a pretext to land troops in southern Taiwan. The Qing court dispatched Shen Baozhen to handle the aftermath. When he arrived, he did one thing: he looked at Taiwan’s map again.
And he was shocked.
Vast swaths of eastern Taiwan were marked on the map as huawai zhi di — “lands beyond civilization.” Nobody was governing them. But the Japanese had entered precisely through those areas. In other words: the land you refuse to govern, someone else will govern for you.
What did Shen Baozhen and his successor Liu Mingchuan do?
- Established Taipei Prefecture, headquartered in Mengjia (Wanhua). This wasn’t random — it was precisely calculated. Mengjia sat between Keelung and Guishan mountains, directly across the strait from Fujian’s Wuhumen, making it “the key to Taiwan’s northern gate.”
- Built Hengchun County at Taiwan’s southernmost tip and moved the southern administrative post to Beinan (present-day Taitung).
- Liu’s land tax reform boosted provincial revenue from 1.1 million taels to 670,000 taels (land tax alone).
Let that number sink in: 670,000 taels — several times the previous figure. Why?
Because Liu wasn’t just raising taxes. He was registering every plot of previously untracked farmland into the database. This is what “governance coverage rate” looks like.
V. The Tiny Place Names Are Taiwan’s True Foundation
What stunned me most in the Revised Taiwan Provincial Chronicles wasn’t the macro-level adjustments. It was the exhaustive records of grassroots settlements.
Examples:
- Zhilan Ward in the Tamsui Department (present-day Beitou and Neihu): 32 villages.
- Dajiara Ward (present-day central Taipei): 16 villages.
Or take the place name “Xinwu” — New House. Do you know why it’s called that? Because five Fan-Jiang brothers built a large mansion there. Over time, the surrounding area became known by that name. One family’s house became an administrative unit.
These micro-names encode Taiwan’s true evolutionary logic: first, settlers arrived and formed clusters. When clusters multiplied, the government established bao (wards) to manage them. Above wards came li (districts), and above districts, counties. By the late Qing, the baojia system wove these scattered villages into a tight grid — tracking populations, physical records, even hygiene data.
From a place name to a complete grid-based governance system. That’s three hundred years.
Closing
Administrative red lines shift with regimes. But the settlement data in the Revised Taiwan Provincial Chronicles — those are the land’s true fingerprints.
After finishing this book, my biggest takeaway is this: not a single inch of Taiwan’s administrative structure was drawn in an office. It was tested — step by step — by pioneers operating under resource constraints and survival pressure.
From Koxinga’s “feed the army through farming” to Shen Baozhen’s “open the mountains and pacify the tribes” to Liu Mingchuan’s “register the land and raise the revenue” — each adjustment was an iteration of a governance algorithm.
History, at its core, is just a people debugging their own operating system.
And Taiwan’s operating system was debugged for three hundred years.