The Psychology of Empire - Case Study - The Han and Tang Dynasties

The Chinese Case: Two Dynasties on One Civilisational Clock

The Chinese case differs from every earlier entry in the series because the framework page already classifies China as its clearest Setting 003 example, a civilisation whose identity survived each dynastic collapse. That classification is useful at civilisational scale, but it creates a problem for case work: if the outer envelope is Continuous Genesis, every dynastic fall can be described as a deferred flashpoint, and the model risks becoming unfalsifiable at exactly the point where it ought to be tested. This page therefore treats the Han (202 BC to AD 220) and the Tang (618 to 907) as two separate imperial arcs nested inside one civilisational clock, and asks whether the stage sequence can be read independently in each.

A note on method. Within each stage the documentary record is stated first, and sentences beginning "Inference" mark the point where the mapping to the framework begins. Open questions are collected at the end. Dates are approximate, a generation is taken as roughly 25 years, and every mapping is a reading of the sources rather than a measurement.

Genesis Setting: A Nested Classification

At civilisational level the classification is Setting 003, carried by Native Memory Depth: the written script, the Classics, ritual practice and the Mandate of Heaven as a legitimising frame. The difficulty is that this tells us little about why a particular dynasty ends when it does. The proposal here is to classify each dynasty separately on Baseline Inheritance and Memory Depth, and to treat the civilisational 003 as an outer envelope that governs what survives Stage 6 rather than how fast the stages run inside the dynasty.

The Han: Setting 002. The record shows that the Han kept the Qin commandery system, the legal code, the standardised script, weights and measures, and the Qin road and canal network, while changing the public vocabulary of rule. The Qin infrastructure is Environmental Memory Depth in the sense used on the framework page. Four Constitutional DNA tensions are visible early. The first is the hybrid of commanderies and enfeoffed kings, patched rather than resolved (the Seven States revolt of 154 BC). The second is a legalist fiscal machine operating under a Confucian public vocabulary, which the Salt and Iron Debate of 81 BC later exposed. The third is free purchase of land with no landed aristocracy to restrain it, which seeded the later estate concentration. The fourth is regency by the empress dowager and her family (Lü Zhi, 195 to 180 BC), which anticipates the consort-clan problem. Inference: the pattern of patching an inherited tension rather than resolving it parallels the Roman Setting 002 case. The Imperial Academy of 124 BC looks like Acquired Memory Depth; whether it amounts to an Institutional Genesis Reset is left open.

The Tang: Setting 002 with Conquest-Driven Reset features. The record shows a lineage running from Northern Wei through Northern Zhou and Sui, in which a Xianbei-Han military and administrative elite produced institutions that neither tradition had alone: the equal-field land system, the militia system and the three-departments structure of government. Chen Yinke's argument about the Guanlong group is the best-known statement of this reading. Inference: the diagnostic the framework page gives for a Conquest-Driven Genesis Reset, institutional features that cannot be derived from either party alone, is plausibly met. The counter-reading is that this was Native continuity after Sinicisation, and the page does not yet offer a test to separate the two. Three Tang DNA tensions are visible: the Guanlong elite against the eastern clans, the frontier military against the civil bureaucracy, and the absence of a stable succession rule (the Xuanwu Gate coup of 626 and the succession crises that followed). Buddhism supplies a large Acquired Memory Depth component, carried by an organised clergy.

The Han Arc: Two Clocks in One Dynasty

Stage 1 (Pioneer): approximately 209 BC to 141 BC — approximately 3 generations. The catalyst is the collapse of Qin and the Chu-Han war of 206 to 202 BC, a 1A1 Primal Shock on a very large scale. The record shows that Liu Bang and most of his early circle were commoners or minor functionaries (the sources name a village headman, a clerk, a prison official, a butcher and a carriage driver among them), while the old aristocracies of the six states had already been broken, first by Qin and then by the war itself. Inference: that is a close match for 1A3 Cleansing of Ego, and probably the cleanest in the series. The early reigns of Wen and Jing show very light taxation, restraint in spending and a Huang-Lao preference for non-interference; the land tax is reported to have settled at one thirtieth of the harvest by the mid second century BC (figure to be verified). Sima Qian's account of the treasuries early in Wudi's reign, with strings of coin rotting and granaries overflowing, describes a surplus accumulating faster than the state knew how to spend it. Inference: in model terms this is the surplus that Stage 2 then spends. The Lü regency and the Seven States revolt are the first appearances of the DNA tensions described above, and both were survived.

Stage 2 (Builder): approximately 141 BC to 87 BC — approximately 2 generations. The record shows Wudi pursuing the Xiongnu wars, opening the western regions through Zhang Qian's missions, founding the Imperial Academy (124 BC), establishing the state monopolies on salt and iron (119 BC) and introducing the Wu Zhu coin (118 BC), which remained the nominal standard for much of the following seven centuries, with interruptions. That is a 2D1 Standard and a 2D2 Commercial Extension. Wudi's finance also included fiduciary experiments, among them the white deerskin money of 119 BC. Inference: this is closest to 2D4 Speculative Sovereignty, credibility issued against future conquest, and it places a monetary signature in Stage 2 rather than in Stage 4, which sits awkwardly with the usual ordering of the monetary indicators. On the Institutionalisation Bottleneck the Han differ from the Mongol illustration. The record shows a recommendation system for office from 134 BC, the Academy, the growth of the inner court, and the regency of Huo Guang after 87 BC holding the structure together. Inference: the charisma-infrastructure gap appears to have closed, probably because the Qin inheritance supplied institutional depth that did not have to be built from nothing. The Salt and Iron Debate of 81 BC shows the legalist-Confucian tension being argued openly rather than resolved.

Stage 3 (Satiated): approximately 87 BC to 33 BC — approximately 2 generations. The record shows the Zhao and Xuan reigns restoring fiscal order and extending Han authority over the western regions, with the Protector General of the Western Regions established in 60 BC. The destruction of the Huo family in 66 BC, two years after Huo Guang's death, shows consort-clan power reaching the point of conspiracy within a single generation. Inference: the Xuan reign (74 to 49 BC) is the best candidate for the Peak, and the Huo affair is an early 3C1 Gated Enclave signature. I can identify no clear Han counterpart for 3B1 Physical Labor Stigma at this point, and the indicator is not forced here.

Stage 4 (Anxious): approximately 33 BC to AD 23 — approximately 2 generations, compressed. The record shows the Wang consort family dominating the Yuan, Cheng, Ai and Ping reigns, with land and slave concentration widely complained of. The proposal of Shi Dan and Kong Guang in 7 BC to limit holdings failed in the face of resistance from the great households. Wang Mang took the regency in AD 1 and the throne as the Xin emperor in AD 9, prohibited the sale of land and the slave trade, and reissued the currency repeatedly in multiple denominations; the land and slave measures were rescinded in AD 12. The Yellow River changed course in AD 11, and the Red Eyebrows and Lülin risings followed from AD 17. Inference: the 7 BC proposal is a failed Intervention I, and Wang Mang's programme is a larger failed Intervention I carried out by the ruler himself, consistent with the model's expectation that economic decompression attempted late is overwhelmed. The Wang Mang takeover also complicates the Stage 5 pathway, because it runs through the court and the literati rather than through a mass populist demand, and I know of no evidence in the sources of a public tribal binary at this date.

The Guangwu Reactivation: approximately AD 25 to AD 57 — approximately 1 generation. The record shows Liu Xiu restoring the dynasty in AD 25 after the civil wars of 17 to 25, cancelling the regular commandery militia training in AD 30, and ordering a survey of land and households in AD 39 to 40 that provoked falsification by local officials and revolts, after which the regime leaned on the landed families, notably those of Nanyang, from whom many of its own commanders were drawn. Inference: this is a Stage 1 Reactivation whose outputs were not constitutionally protected, and it also resembles the Factional Hardening limitation, since the restored settlement was hardened around the landed elite whose holdings the Stage 4 reformers had tried to limit.

Eastern Han Stages 3 to 4: approximately AD 57 to 159 — approximately 4 generations. The record shows the Ming, Zhang and He reigns as a second period of consolidation, with Ban Chao's campaigns in the west. After 105 the sources describe recurrent child emperors, consort clans (Dou, Deng, Liang) and eunuchs struggling for control of the court, the long and costly Qiang wars from 107 to 118, and the rule of Liang Ji from 141 until a eunuch coup removed him in 159. Inference: the second Stage 4 is dominated by the DNA tension of regency, and the Qiang wars are a 4A2 Resource Drain signature, though the strain on the treasury needs checking against the fiscal sources.

Stage 5 (Angry): approximately AD 159 to 189 — approximately 1 generation. The record shows eunuch dominance of the court after 159, the Partisan Prohibitions of 166 and 169 which purged and disqualified scholar-officials, and the Yellow Turban rising of 184, organised through the Way of Great Peace as a religious network. Inference: the literati and eunuch confrontation is a plausible 5B1 Tribal Binary, the Prohibitions a 5C1 Weaponization Blitz directed at the carrier class itself, and the Way of Great Peace a rival ideological carrier building a parallel social order, which is a functional match for 5A2 Parallel Reality Split rather than a literal one.

Stage 6 (Defeated): approximately AD 189 to 220 — approximately 1 generation. The record shows the massacre of the eunuchs by Yuan Shao's troops in 189, Dong Zhuo's seizure of Luoyang and the melting of bronze objects to mint small coins in 190, the sacking of the capitals, and a long period in which coin largely gave way to grain and cloth as a medium of exchange until the Wei state restored the Wu Zhu in 227. Recorded population fell by a large fraction between the second-century registers and the Three Kingdoms registers, though registration failure makes this a lower bound on the problem rather than a measurement. Inference: Dong Zhuo's coinage and the move to barter are a clear 6A1 Faith Vaporization, the warlord armies a 6B3 Warlord Normalisation, and Cao Pi's accepted abdication ceremony in 220 a textbook 6C2 Pragmatic Reset. Western Han runs about 211 years (202 BC to AD 9) and Eastern Han about 195 (AD 25 to 220), roughly 420 years in total across the Xin interval.

The Tang Arc: A Compressed Break and a Long Plateau

Stage 1 (Pioneer): approximately 617 to 649 — approximately 1 to 2 generations. The record shows the Sui reunifying the country in 589 after about three centuries of division and then collapsing within three decades under the burden of the Goguryeo campaigns and the Grand Canal labour, with widespread rebellion. Li Yuan took Chang'an in 617, was enthroned in 618, and consolidated by about 624. Li Shimin's coup at the Xuanwu Gate in 626 was followed by the Zhenguan reign, with its culture of remonstrance (Wei Zheng) and a codified legal framework. Inference: 1A1 Primal Shock and 1B1 Duty Matrix fit well. 1A3 Cleansing of Ego fits poorly, because the founding house belonged to the Guanlong aristocracy and the northwestern elite carried on through the transition. A plausible reading is that the higher Baseline Inheritance of a Setting 002 genesis produces a less pure Pioneer phase than the Han, whose circle of commoners was much closer to the model's description. Whether the Sui reunification should be counted as the true Stage 1 of a longer Sui-Tang arc from 581 is an open question.

Stage 2 (Builder): approximately 649 to 713 — approximately 2 to 3 generations. The record shows Gaozong's campaigns in Korea and against the Western Turks, the rule of Wu Zetian (as empress and then as emperor of her own Zhou dynasty, 690 to 705) widening the examination route and favouring officials from outside the old aristocracy, a flourishing Silk Road trade, and the Kaiyuan Tongbao coin introduced in 621 as a durable standard. The Japanese Wado Kaichin of 708 was modelled on it. Inference: the coinage is a 2D1 Standard with a 2D2 Commercial Extension. The Institutionalisation Bottleneck is expressed in Tang as succession instability, with the coups of 705 and 710 showing a reach that outran its institutional rules for the transfer of power.

Stage 3 (Satiated): approximately 713 to 755 — approximately 1 to 2 generations. The record shows the Kaiyuan (713 to 741) and Tianbao (742 to 756) eras as the high point of registered population and cultural display, with Li Linfu as chancellor from 734 to 752 and the court of Xuanzong and Yang Guifei. Inference: 3A3 Spectacular Distraction is a fair description of the court culture, and Li Linfu's long chancellorship a 3C1 Gated Enclave in the form of a narrowing of access to the emperor. I can identify no clear Tang counterpart for 3B1 Physical Labor Stigma, and it is not forced here.

Stage 4 (Anxious): approximately 737 to 755 — less than 1 generation, compressed. The record shows the militia system replaced by long-service professional armies on the frontiers from 737, ten frontier commands by the early 740s whose governors held military, fiscal and civil powers, equal-field allocations failing as peasants fled the registers, and costly defeats at Talas in 751 and against Nanzhao in 751 and 754. An Lushan by 751 held three commands. Inference: the defeats are a 4A1 Imperial Overreach Shock in double form. The unusual feature is that the structural change, the professionalised frontier armies, preceded the shock, so that Stage 4 is compressed to under a generation, and the 4A2 Resource Drain Panic is expressed as fiscal and military reorganisation rather than concealed accounting.

Stage 5 (Angry), interleaved with partial stabilisation: approximately 755 to 875 — approximately 4 to 5 generations. The record shows the An Lushan rebellion from 755 to 763, the emperor's flight and the killing of Yang Guifei at Mawei in 756, reliance on Uighur military assistance, and a brief Tibetan occupation of Chang'an in 763. The northeast passed under semi-autonomous military governors. On the fiscal side the record shows the salt monopoly developed by Liu Yan from about 759 and the Two-Tax Law of 780 by Yang Yan, which shifted assessment from registered persons to land and property and leaned heavily on the Yangtze economy. Large-denomination coins (Qianyuan zhongbao, reported at ten cash and later fifty cash) were issued from 758. Eunuch control of the Shence army is recorded from the 780s, the Niu and Li factional conflict from about 808 to 849, and the Ganlu Incident of 835, in which an attempted purge of the eunuchs failed and officials were massacred. Inference: the Two-Tax Law and the salt monopoly are partial Intervention I deployments, and they are a plausible explanation for the plateau, since the dynasty lasted roughly 150 years after the break. The Stage 5 psychology is expressed as elite faction (eunuchs against bureaucrats, Niu against Li) rather than as mass tribalism, and I know of no evidence in the sources of a populist saviour demand.

Stage 6 (Defeated): approximately 875 to 907 — approximately 1 generation, followed by the Five Dynasties and Ten Kingdoms to 960. The record shows the risings of Wang Xianzhi and Huang Chao from 874, Huang Chao's capture of Chang'an in 880, and Arab sources reporting the killing of large numbers of foreign merchants at Guangzhou in 879 (the figures are not reliable). Zhu Wen, a former Huang Chao commander who had gone over to the Tang in 882, compelled the last emperor to abdicate in 907. Inference: the fragmentation that follows is a 6B2 Tribal Retreat and 6B3 Warlord Normalisation, and Zhu Wen is a 6C2 Pragmatic Reset by the figure already commanding the capital's army, as Cao Pi was in 220.

Migration, Carriers and Memory Depth in Both Cases

The Han. The record shows the Southern Xiongnu settled within the frontier from about AD 48 to 50, Qiang populations resettled in the northwest, and Wuhuan and Xianbei auxiliaries used by the Eastern Han court and later by the warlords, with Dong Zhuo's power base in the northwest drawing on Qiang and Hu troops. Inference: this resembles the Gothic counter-case on the framework page, migration without absorption into a transmitting institution, which adds military dependence rather than Acquired Memory Depth. The Partisan Prohibitions struck the carrier class directly, yet Confucian transmission persisted through the Three Kingdoms and Jin periods, which is consistent with Native Memory Depth being distributed widely enough not to depend on a single institution.

The Tang. The record shows foreign commanders of Sogdian-Turkic, Korean, Khitan and Turkic origin in senior frontier commands (An Lushan himself was of mixed Sogdian and Turkic descent), alongside foreign merchants and a large foreign clergy in the capital. Inference: Tang absorbed foreign carriers into the military and commercial layers but, on the evidence I know, not into the examination-based administrative class, so the Acquired Memory Depth they brought was partly separate from the institutions that sustained the state. The masking problem on the framework page also fits: foreign commanders on professional frontier armies concealed the failure of the domestic militia rather than repairing it. After 755 the record shows hostility to foreign influence, and the edict of 845 against Buddhism extended to Manichaean, Zoroastrian and Nestorian foundations. Open: whether the later cultural turn, from Han Yu's Confucian revival towards Song Neo-Confucianism, is Native Memory Depth reasserting itself after the loss of Acquired carriers.

What the Chinese Case Adds to the Model

First, it argues for a nested classification. Applying Setting 003 at civilisational level and 002 at dynastic level lets the same page explain both why the civilisation survives and why each dynasty ends, but it raises the question of when a dynastic change is a Political Reset without Genesis change and when it is a true Genesis Reset. On the evidence here, Han to Tang is closer to the former, with the Sui-Tang synthesis the strongest candidate for the latter.

Second, the Han arc supplies a clean example of a Stage 1 Reactivation, and Guangwu's restoration shows the limitation the framework page predicts: outputs not constitutionally protected, and a settlement hardened around the elite whose capture the earlier reformers had tried to prevent. The Tang arc supplies a different pattern, a sharp Stage 4 to 5 break followed by a long plateau, partly explained by the fiscal reforms.

Third, the Stage 5 pathway in both cases runs through court and military factions, not a mass populist demand. That is either a limit on the generality of the Stage 5 indicators as currently worded, or a sign that the stage should be generalised to elite tribalisation in states where there is no mass public politics. I lean towards the second reading, but two cases are too few to settle it.

Fourth, the monetary indicators do not behave identically. In the Han the fiduciary experiments appear in Stage 2, and the collapse of coin appears at Stage 6. In the Tang, the large-denomination issues appear after the 755 break, which suggests that the monetary track lagged the structural break rather than leading it, although the eighth-century coin-quality record should be checked before the point is relied on. If it holds, it weakens the claim that monetary indicators are a leading diagnostic.

Fifth, 6C3 operates differently under Setting 003. The framework page describes the imperial past passing into legend and becoming a story about a vanished age. In the Chinese case the dynastic names survive as living ethnonyms (Han and Tang), which is arguably the flywheel effect itself, with the golden age retained as an identity rather than mythologised as a different world.

Sixth, on duration, the Han total of about 420 years sits inside the 400 to 650 year range quoted for Setting 002 on the British case page, while each Han half, at about 200 years, and the Tang, at about 290 years (about 326 if counted from the Sui in 581), sit below it. I have not applied the Stage Duration table, and whether a Setting 003 outer envelope should lengthen or shorten these figures is left open. The model does not claim that the purity of Stage 1 predicts duration, and with two cases nothing can be concluded.

Evidence, Inference and Open Questions

Evidence. The Han inherited and kept the Qin administrative framework. Its founding circle was largely commoner. The Wu Zhu coinage and the salt and iron monopolies date from Wudi's reign. A land-limitation proposal failed in 7 BC and Wang Mang's land and currency measures failed after AD 9. The Partisan Prohibitions, the Yellow Turban rising, Dong Zhuo's coinage and the abdication of 220 are recorded. The Tang inherited Sui institutions, and the militia gave way to professional frontier armies by 737. The 755 rebellion, the 780 Two-Tax Law, the Ganlu Incident, the 845 edict and the abdication of 907 are recorded.

Inference. Han Stage 1 fits 1A3 closely and Tang Stage 1 fits it poorly. Guangwu's restoration is a Stage 1 Reactivation with a Factional Hardening limitation. The Tang plateau after 755 is partly explained by partial Intervention I. Foreign carriers in the Tang were absorbed into the military layer without being absorbed into the administrative class. Both arcs end in a 6C2 Pragmatic Reset.

Open questions. Whether Han to Tang counts as a Genesis Reset, and whether the Sui should be counted as the start of the Tang arc. Whether Stage 5 needs generalising to elite factional conflict. Whether the monetary track lags the structural break. Whether the lower Stage 1 purity of the Tang bears on its shorter arc. Whether Tang absorption of carriers can be measured independently of the outcome it is meant to explain, which is the circularity risk already noted on the framework page.

Where this site says "no evidence", it means none in the published record, not that none exists; what follows is reconstruction from documents, place-names and landscape logic, flagged as inference, not fact.