Search is not available for this dataset
image imagewidth (px) 2.7k 3.6k |
|---|
YAML Metadata Warning:empty or missing yaml metadata in repo card
Check out the documentation for more information.
What Doesn't Graft: Twenty Cases of Chinese Capital and City Destruction, 278 BCEโ1937 CE
The Failure Comparators the Entropy-Resistant Institution (ERI) Index Has Been Missing
Author: Gia Bao Huynh
Independent Researcher, Ho Chi Minh City, Vietnam
Email: huynhbao@asu.edu ยท ORCID: 0009-0008-2372-5852
Research Stance: Challenging the unchecked power, epistemic asymmetries, and monopolized governance of non-state actors across frontier technologies.
Collaborator: Claude (Anthropic)
Repositories & Data:
- ๐ GitHub: https://github.com/giabaohuynhasu/what-doesnt-graft
- ๐ค Hugging Face Hub: https://huggingface.co/datasets/giabaohuynhasu/what-doesnt-graft
- ๐ Manuscript:
What_Doesnt_Graft.pdf(18-page formal monograph, July 2026)
๐ Executive Summary & Theoretical Contribution
Endurance or Density (Huynh 2026) identified its own single highest-value vulnerability: the base 3-case Entropy-Resistant Institution (ERI) design (the Catholic Church, House of Savoy, and the Jiangnan literati lineage network) selected entirely on the dependent variable. No claim that the six structural attributes discriminate institutional survival from mere survivorship bias can be sustained without negative comparators โ polities and institutional seats that faced comparable physical and political disruption but failed to survive it.
This repository supplies twenty historical cases of Chinese capital and major city destruction spanning 2,215 years (278 BCE to 1937 CE), coded against physical destruction severity, demographic disruption, and long-run recovery trajectories.
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
โ TOTAL HISTORICAL CATALOG (N = 20) โ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโฌโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
โ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโดโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
โผ โผ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
โ CONTROL GROUP (N = 10) โ โ HIGH-INFORMATION ERI CASES (N=10) โ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโค โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโค
โ Ordinary hereditary / territorial โ โ Disentangles Replicator vs. โ
โ dynasties lacking decoupled symbolic โ โ Interactor Survival: โ
โ legitimacy. Physical destruction โ โ โข Ye (580) vs. Jiankang (589) โ
โ terminates the interactor by default.โ โ โข Daliang/Kaifeng Dual Flooding โ
โ (Ying, Handan, Ji, Shou, Linzi, etc.)โ โ โข Yongjia (311) Recolonization Seed โ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
Four Foundational Empirical Discoveries
- Interactor Death vs. Replicator Death:
- Interactor Death: A ruling house, court, or political regime is extinguished, but the underlying population, infrastructure, and administrative substrate survive intact (e.g., the Jingkang Incident 1127 CE, Case 15).
- Replicator Death: The transmissible institutional template itself is permanently obliterated at that site (e.g., Ye Cheng 580 CE, Case 10).
- Symmetrical Razing & Post-Destruction Divergence (Ye 580 vs. Jiankang 589):
- Historical sourcing overturns the prior assumption that Jiankang was merely "downgraded" while Ye was razed. Both cities suffered identical Severity-3 razing under the authority of Yang Jian (Emperor Wen of Sui) within nine years; Jiankang was explicitly ordered "flattened and turned into plowed farmland" (ๅนณ่ฉ่ๅขพ).
- The divergence occurred afterward: Ye was permanently erased (R0), its population relocated to Anyang. Jiankang's substrate (Yangtze-Huai hydrology, rice-paddy fiscal base, elite absorption) was geographically embedded rather than tied to stone walls; it reconstituted as an imperial capital in 937 CE (Southern Tang) and 1368 CE (Ming) (R2).
- Independent Multi-Century Dormancy Outside Jiangnan (Daliang / Kaifeng):
- Daliang was drowned by canal diversion in 225 BCE (Case 4), dormant for 1,185 years, and revived as capital under Northern Song. In 1642 CE (Case 17), it was drowned by the identical Yellow River diversion tactic on the same ground. This provides an independent North China comparator answering F-ERI4, demonstrating that institutional dormancy is not an idiosyncratic artifact of Jiangnan riverine geography.
- Systemic Destruction Spikes vs. BZM Demand ($\lambda_{bz} \approx 50/\text{century}$):
- While no single institution's throughput approaches $\lambda_{bz} \approx 50/\text{century}$ (Jiankang $\delta = 1.36$, Savoy $\delta = 0.38$, Papacy $\delta = 0.44$), the wider civilizational system briefly generated local destruction rates meeting or exceeding $\lambda_{bz}$ during the Qin Unification (50/century), Yongjia Collapse (40/century), and the Ming-Qing Cataclysm (62.5/century).
๐ Repository Structure
what-doesnt-graft/
โโโ README.md # Master documentation
โโโ CITATION.cff # Formal citation metadata
โโโ LICENSE # MIT License
โโโ What_Doesnt_Graft.pdf # Complete 18-page monograph (Huynh 2026)
โโโ data/
โ โโโ chinese_capital_destructions_20_cases.csv # 20-case structured dataset
โ โโโ chinese_capital_destructions_20_cases.json # JSON-formatted dataset with rich annotations
โ โโโ local_density_spikes.csv # Quantitative density spike records (Section VIII)
โโโ figures/
โ โโโ destruction_timeline.png # Disruption severity vs. recovery timeline
โ โโโ ye_vs_jiankang_divergence.png # Symmetrical destruction vs. asymmetrical survival
โ โโโ density_spikes_vs_bzm.png # Systemic crisis density vs. BZM threshold
โโโ scripts/
โโโ make_figures.py # Generates high-res analytical charts
โโโ replicate_analysis.py # Complete replication of all paper statistics
๐ The 20-Case Catalog (Table 1)
| ID | City (EN/ZH) | Year | Dynasty / Regime | Perpetrator | Severity | Recovery | Coded Classification & Significance |
|---|---|---|---|---|---|---|---|
| 1 | Ying (้ข) | 278 BCE | Chu (Warring States) | Qin (Bai Qi) | 3.0 | R1 | Interactor death; royal tombs burned; never capital again |
| 2 | Handan (้ฏ้ฒ) | 228 BCE | Zhao (Warring States) | Qin (Wang Jian) | 2.5 | R1 | Court collapse; survived as ordinary commandery |
| 3 | Ji (่) | 226 BCE | Yan (Warring States) | Qin (Wang Ben) | 2.0 | R1 | Yan court eliminated; site underlies modern Beijing with zero institutional lineage |
| 4 | Daliang (ๅคงๆข) | 225 BCE | Wei (Warring States) | Qin (Wang Ben) | 3.0 | R2 | Canal flood diversion; dormant 1,185 yrs; capital under Northern Song (960) |
| 5 | Shouchun (ๅฃฝๆฅ) | 223 BCE | Chu (Warring States) | Qin (Wang Jian) | 2.0 | R1 | 600k Qin assault; final extinction of Chu polity |
| 6 | Linzi (่จๆท) | 221 BCE | Qi (Warring States) | Qin forces | 1.0 | R1 | Capitulation without battle; baseline control case |
| 7 | Luoyang (ๆด้ฝ) | 190 CE | Eastern Han | Dong Zhuo | 3.0 | R2 | Systematic razing & 200k forced evacuation; restored by Cao Wei (220) |
| 8 | Luoyang (ๆด้ฝ) | 311 CE | Western Jin | Han-Zhao (Liu Yao / Shi Le) | 3.0 | R2 | Yongjia Disaster; crown prince & 30k slain; seed of Eastern Jin at Jiankang |
| 9 | Chang'an (้ทๅฎ) | 316 CE | Western Jin | Han-Zhao (Liu Yao) | 3.0 | R2 | Prolonged starvation; Western Jin ended; rebuilt across 6 northern dynasties |
| 10 | Ye (้ด) | 580 CE | Northern Zhou / Qi | Yang Jian (Sui) | 3.0 | R0 | Total erasure; population moved to Anyang; permanent archaeological ruin |
| 11 | Jiankang (ๅปบๅบท) | 589 CE | Chen (Southern Dynasties) | Sui (Yang Su / Gao Jiong) | 3.0 | R2 | Razed into farmland (ๅนณ่ฉ่ๅขพ); substrate revived capital in 937 & 1368 |
| 12 | Chang'an (้ทๅฎ) | 756 CE | Tang | An Lushan rebels | 2.0 | R2 | Court evacuated to Shu; Tang restored within 2 years; decentralization began |
| 13 | Chang'an (้ทๅฎ) | 881 CE | Tang | Huang Chao rebels | 2.5 | R2 | Sack, massacre of aristocracy; recovered hollowed out |
| 14 | Chang'an (้ทๅฎ) | 904 CE | Tang | Zhu Wen | 3.0 | R1 | Palaces dismantled, timber floated downriver; permanent capital de-escalation |
| 15 | Kaifeng (้ๅฐ) | 1127 CE | Northern Song | Jurchen Jin | 1.5 | R2 | Jingkang Incident; emperors captured; pure interactor death without razing |
| 16 | Xiangyang (่ฅ้ฝ) | 1273 CE | Southern Song | Mongols (Aju) | 1.5 | R3 | 5-year siege; trebuchet breach of Fancheng; intact strategic surrender |
| 17 | Kaifeng (้ๅฐ) | 1642 CE | Ming | Ming Gov. (vs. Li Zicheng) | 3.0 | R2 | Yellow River breached; ~300k drowned; buried in silt; rebuilt 1662 |
| 18 | Yangzhou (ๆๅท) | 1645 CE | Southern Ming | Qing (Prince Dodo) | 2.5 | R3 | 10-day civilian massacre; rain stopped arson; rapid commercial recovery |
| 19 | Jiading (ๅๅฎ) | 1645 CE | Southern Ming | Qing (Li Chengdong) | 2.0 | R3 | 3 successive massacres; resistance crushed; built fabric standing |
| 20 | Nanjing (ๅไบฌ) | 1937 CE | Republic of China | Imperial Japanese Army | 2.0 | R3 | 6-week atrocities (200kโ300k killed); modern structures survived; rapid reconstitution |
๐ Key Visualizations
1. Timeline of 20 Destructions: Disruption Severity vs. Recovery Trajectory
2. Symmetrical Destruction vs. Asymmetrical Post-Disruption Divergence (Ye vs. Jiankang)
3. Systemic Crisis Density Spikes vs. BZM Demand Rate ($\lambda_{bz}$)
๐ฌ How to Reproduce
Prerequisites
pip install pandas numpy matplotlib
Execute Replication
python scripts/replicate_analysis.py
Expected output validates:
- Total N = 20 cases across 2,215 years.
- Cross-tabulation matching Table 1.
- Proof of symmetrical razing between Case 10 (Ye) and Case 11 (Jiankang) with opposite recovery tiers ($R0$ vs $R2$).
- Verification of 1,867-year interval in Kaifeng flood repetition ($225\text{ BCE} \rightarrow 1642\text{ CE}$).
- Quantitative verification of systemic crisis spikes meeting $\lambda_{bz} \approx 50/\text{century}$.
๐ Citation
@article{huynh2026whatdoesntgraft,
author = {Gia Bao Huynh},
title = {What Doesn't Graft: Twenty Cases of Chinese Capital and City Destruction, 278 BCE--1937 CE, and the Failure Comparators the Entropy-Resistant Institution Index Has Been Missing},
journal = {Case Studies in Institutional Endurance and QCA},
year = {2026},
month = {July},
url = {https://github.com/giabaohuynhasu/what-doesnt-graft}
}
โ๏ธ License
Distributed under the MIT License.
- Downloads last month
- 103