第十一篇：东汉——3.0版的第一个完整周期
Essay 11: Eastern Han — Version 3.0's First Complete Cycle
刘秀重启了3.0版操作系统。框架不变，参数微调，系统重新上线。
东汉（公元25—220年）因此成为一个极其珍贵的观察样本：它是3.0版操作系统在没有构型替换的情况下完整运行一个周期的第一个案例。从建立到崩溃，将近两百年。在这两百年里，我们可以清晰地看到余项是怎样在一个不变的框架内逐步积累，从萌芽到制度化到失控到崩溃的全过程。
西汉的故事线是构型的发明和确立。东汉的故事线是构型的运行和老化。两者的教训不同：西汉教你怎么造一台机器，东汉教你一台机器怎么老死。
一、光武帝的参数调整
刘秀不改框架，但他改了几个关键参数。这些调整看起来不大，但深刻地影响了东汉两百年的走向。
第一，削弱功臣集团的政治权力。
刘秀和刘邦一样靠功臣打天下。但刘邦对功臣的处理是暴力清洗（韩信、彭越、英布先后被杀），刘秀的处理方式温和得多。他给功臣们优厚的经济待遇——封邑、田产、钱财——但不给他们实际的政治权力和军事指挥权。"退功臣而进文吏"，用文官系统替代功臣集团来运行日常行政。
这就是后人所说的"云台二十八将"的命运：荣华富贵，但远离权力中心。和西汉初年功臣们掌握兵权、参与政治决策的状态形成鲜明对比。
刘秀的逻辑很清楚：功臣集团是打天下的工具，不是坐天下的工具。打天下需要军事能力，坐天下需要行政能力。两种能力不在同一批人身上。让功臣继续掌权，等于让一群军事家来做行政管理——这不是他们擅长的事情，而且他们的军事力量本身就是对中央集权的潜在威胁。
从余项的角度看，刘秀是在预防一个他从西汉历史中看到的余项：功臣坐大。刘邦用暴力解决了这个余项（杀功臣），代价是政治信用的损失（后来的功臣不敢为汉朝卖命了）。刘秀用经济赎买解决了这个余项（给钱不给权），代价小得多。
但这个处理方式有一个附带效果：它创造了一批拥有大量土地和财富但没有政治权力的地方豪族。这些人在短期内对中央没有威胁，但长期来看，经济实力终究会转化为政治影响力。这批人就是东汉中后期困扰帝国的"豪族"的起源之一。
第二，强化尚书台。
西汉的行政中枢是丞相府。丞相权力极大，是皇帝之下的第二号人物。武帝时期已经开始用"内朝"（尚书、侍中等皇帝身边的近臣）来绕过"外朝"（以丞相为首的正式官僚系统），但没有制度化。
刘秀把这个趋势制度化了。他大幅强化了尚书台的权力，把实际的行政决策权从丞相（后来改称司徒）转移到尚书台。尚书台直接对皇帝负责，尚书令成为事实上的首席行政官。丞相/司徒变成了荣誉职位，有名无实。
这个调整的结构性意义是：皇帝的个人控制力被增强了。尚书台是皇帝的私人秘书处，它的权力来自皇帝的信任而不是制度性的授权。这意味着皇帝可以通过更换尚书来随时调整政策方向，不受制度化的丞相系统的制约。
但这个增强有一个前提：皇帝本人有足够的能力和精力来驾驭尚书台。如果皇帝英明有为（光武帝、明帝），系统运转良好。如果皇帝幼小无能呢？尚书台的权力就会被皇帝身边的其他人——太后、外戚、宦官——接管。
刘秀增强皇权的设计，恰恰为后来皇权旁落于外戚和宦官铺好了路。工具是中性的。一把好用的刀，英雄能用来切菜，小偷也能用来撬锁。
二、豪族——余项的制度化
东汉最重要的社会现象是豪族的崛起。
豪族不是东汉发明的。西汉就有大地主、大商人、大家族。但在西汉，豪族是被打压的对象——武帝时期的酷吏专门对付他们，迁徙豪强到关中（强制搬迁以削弱地方势力）是常规政策。
东汉不同。刘秀本人就出身于南阳的地方豪族（虽然他是皇族后裔，但到他这一代已经是地方大户的生活方式了）。他的核心支持者也大量来自地方豪族——南阳系、河北系的豪族是他打天下的基本盘。打完天下之后，你不太可能回过头来打压自己的基本盘。
结果是：东汉朝廷对豪族的态度从西汉的打压变成了默认容忍甚至合作。
豪族在东汉两百年间完成了一个从经济实体到政治实体的蜕变过程：
第一阶段（光武帝到章帝，约25—88年）：豪族主要是经济力量。大片土地、众多佃客（依附于豪族的农民）、私人武装（护院和家兵）。他们在经济上独立于国家，但在政治上还没有形成系统性的影响力。
第二阶段（和帝到桓帝，约88—167年）：豪族开始系统性地渗透官僚体系。途径是察举制——察举制要求地方官推荐"孝廉"（道德优秀的人才），而推荐权掌握在地方官手里。地方官本身很多就来自豪族，自然优先推荐自己家族的人。几代人下来，某些大族世代出高官——"四世三公""累世公卿"成为这些家族的标签。
汝南袁氏（四世五公）、弘农杨氏（四世太尉）是最典型的例子。这些家族不是靠个人才华一代一代单独考上去的，而是靠家族网络系统性地垄断了察举通道。
第三阶段（桓帝到汉末，约167—220年）：豪族已经不只是渗透官僚系统，而是事实上控制了地方社会。他们的庄园是自给自足的经济体，有自己的农业、手工业、甚至私学（培养家族子弟和门客）。他们的佃客和家兵构成了事实上的私人军队。他们在地方上的影响力远超中央派来的官员。
到了汉末，当中央权威崩溃的时候，这些豪族就成了地方上唯一有组织能力的力量。曹操、袁绍、刘备、孙权——三国的主要玩家要么本身就是豪族出身（袁绍的汝南袁氏），要么必须依赖豪族的支持才能建立自己的政权。
从余项守恒的角度看，豪族的演化过程是一个教科书级的案例：
3.0版操作系统（郡县制+编户齐民+土地私有）的设计目标是国家直接面对个体家庭，中间不留缓冲层。秦的编户齐民打散了宗族，让国家和个人之间没有任何中间组织。
但这个设计目标不可能永久维持。因为人天然会结成群体——血缘的、地缘的、经济的、文化的。你打散了旧的中间层（宗族），新的中间层就会在其他基础上重新生长。豪族就是在土地兼并和察举制的双重作用下重新生长出来的中间层。
国家每一次清除中间层，中间层就在新的土壤上重新长出来。清除的方式可能不同（秦用编户齐民，武帝用酷吏，刘秀用退功臣），但结果相同：中间层换了一种形态重新出现。宗族被打散了，豪族长出来了。豪族被打压了（曹操在许下屯田，部分目的就是打击豪族的经济基础），门阀士族又长出来了（魏晋南北朝的九品中正制把豪族的政治垄断制度化了）。
中间层是构的永恒余项。你消灭不了它，因为它的根源不在于某种特定的制度安排，而在于人的社会性本身。人会结群。结群会产生层级。层级会产生中间层。你可以改变中间层的形态（从宗族到豪族到门阀到乡绅到……），但你取消不了中间层这个现象。
三、外戚与宦官——构内的两个余项互相消耗
东汉中后期的政治史可以概括为一句话：外戚和宦官交替掌权。
这个交替的机制是这样的：
步骤一：皇帝年幼即位（东汉从和帝开始，几乎每一个皇帝都是幼年即位——和帝十岁，殇帝百日，安帝十三岁，顺帝十一岁，冲帝两岁，质帝八岁，桓帝十五岁，灵帝十二岁）。
步骤二：太后临朝，太后的娘家人（外戚）掌握实权。
步骤三：皇帝长大后想要夺回权力。但朝廷上下已经被外戚的人把持，皇帝在正式的官僚系统中找不到自己的人。唯一和皇帝朝夕相处、完全依赖皇帝的人是谁？是宦官。
步骤四：皇帝联合宦官发动政变，消灭外戚。宦官因功得势，开始把持朝政。
步骤五：新的皇帝年幼即位（因为上一个皇帝往往短命），新的太后临朝，新的外戚掌权。宦官被清洗。
步骤六：回到步骤一。
这个循环在东汉中后期反复上演，至少经历了四五个完整的周期。每一次外戚被消灭，宦官就崛起。每一次宦官被清洗，外戚就回来。两个群体像跷跷板的两端，一上一下，永远交替。
从构的角度看，外戚和宦官都是3.0版操作系统的结构性产物。
外戚的根源是：3.0版没有设计一个独立于皇帝个人的权力交接机制。皇帝死了或幼小了，权力归谁？没有明确的制度安排。太后摄政是一个非正式的惯例，不是正式的制度。外戚掌权是太后摄政的自然延伸——太后信任谁？信任自己家的人。
宦官的根源是：3.0版把权力高度集中于皇帝个人（尚书台的设计强化了这一点），当皇帝需要对抗外朝（被外戚控制的官僚系统）时，他唯一可用的政治资源是内廷——而内廷的核心人员是宦官。
两个余项互相消耗的过程极其有害。每一次权力交替都伴随着政治清洗——杀人、抄家、流放。这些清洗消灭的不只是外戚或宦官个人，还有他们在官僚系统中的大量同盟者和依附者。几轮清洗下来，官僚系统中有能力、有经验的人被反复筛除，剩下的要么是无害的庸才，要么是善于站队的投机者。
这是一种系统的自我降解。每一次内部权力斗争都降低了官僚系统的整体质量。系统像一件反复被洗的衣服，每洗一次就薄一层。洗到最后，布料已经不能穿了。
更深一层看，外戚宦官交替揭示了3.0版操作系统的一个根本性缺陷：它的最高权力节点是单点的。整个系统的运转依赖于皇帝一个人的能力和意志。皇帝英明，系统运转良好。皇帝幼弱，系统立刻被争夺这个节点的各方势力撕裂。
单点系统没有冗余。一个节点失灵，整个网络失灵。现代的分布式系统设计要求关键节点有冗余备份，中国的帝制系统没有这个概念。皇帝是唯一的，不可复制的，不可替代的。当这个唯一的节点被一个两岁的婴儿占据的时候，系统就只能靠非正式的权力安排（太后、外戚、宦官）来维持运转。而非正式的权力安排没有规则约束，完全取决于参与者的实力博弈。
博弈的结果就是外戚宦官的跷跷板。
四、党锢之祸——当维护者变成挑战者
公元166年和169年，东汉发生了两次"党锢之祸"。
事件的经过是：一批以太学生和地方士人为核心的儒生群体（史称"清议"或"党人"），公开批评宦官专权，品评人物，臧否时政。宦官集团反击，指控这些人"结党营私"，说服桓帝和灵帝下令逮捕、禁锢党人——不只是当事人本人被禁止做官，他们的亲属、学生、朋友也被牵连。两次党锢，受牵连者数以百计，波及面数以千计。
党锢之祸的政治后果——大量优秀人才被排斥出官僚系统——是显而易见的。但从凿构的角度看，更值得分析的是它的结构性意义：儒生阶层从构的维护者变成了构的挑战者。
回顾一下3.0版操作系统的设计。独尊儒术的目的是用儒家意识形态来编码余项——让社会自愿接受秩序。儒生群体在这个设计中的角色是什么？他们是意识形态的生产者和传播者——他们学习经典、解释经典、用经典的标准来评价政治、教育下一代。他们是3.0版的"软件维护团队"。
在系统正常运行的时候，这个角色是稳定性的来源。儒生批评皇帝的某些行为（天人感应框架下的灾异批评），但他们不质疑系统本身。他们的批评是建设性的——指出系统偏离了正确参数，要求回调。这种批评实际上帮助系统自我修复。
但当系统的偏离不是参数问题而是结构性问题的时候——当宦官系统性地腐蚀了整个官僚体系，当皇帝系统性地站在宦官一边——儒生的批评就从建设性的变成了对抗性的。他们不再只是说"参数调错了"，他们开始说"操作者有问题"。品评人物、臧否时政，这些行为的实质是：我们——知识分子——要来判断谁有资格掌权，谁没有。这是对皇帝用人权的直接挑战。
宦官集团看到了这个威胁。他们给党人贴的标签——"结党"——虽然是政治攻击，但不完全没有道理。党人确实在形成一个有组织的、有共同价值观的、试图影响政治走向的群体。这个群体如果成功，宦官就完了。宦官要活下去，就必须消灭党人。
从余项的角度看，党锢之祸暴露了独尊儒术的一个深层矛盾：你培养了一个以道德批评为天职的知识分子阶层，然后你不允许他们批评。
独尊儒术的设计意图是让儒生成为系统的软件维护团队。但维护软件的人必然会发现软件的bug。发现了bug，他们会报告。如果报告被接受（皇帝纳谏），系统修复，皆大欢喜。如果报告被拒绝（皇帝和宦官结盟抵制），维护团队就会越来越沮丧，越来越激进，最终从维护者变成反对者。
你不能既要一支独立思考的知识分子队伍（用来生产合法性和提供政策批评），又要他们在你不想听的时候闭嘴。独立思考不是一个可以按需开关的水龙头。你打开了它，就要接受它流出你不想看到的水。
党锢之祸是3.0版操作系统的一次自残行为。系统切除了自己的软件维护团队。短期效果是宦官集团安全了。长期效果是系统失去了自我修复能力。没有人再批评了——不是因为没有问题，是因为批评者被禁言了。问题继续积累，但不再有人指出。沉默不等于健康。沉默等于放弃治疗。
党锢之祸后二十年，黄巾起义爆发。系统的软件维护团队被切除之后，硬件的崩溃只是时间问题。
五、黄巾起义——底层余项的宗教化释放
公元184年，张角领导的太平道在全国范围内同时起义。信众以黄巾裹头，故称黄巾起义。口号是"苍天已死，黄天当立，岁在甲子，天下大吉"。
这是中国历史上第一次以宗教形式组织的大规模农民起义。
为什么是宗教？
此前的农民起义——陈胜吴广、绿林赤眉——都是世俗的。他们的组织基础是地缘关系（同乡）或军事关系（同一支部队），动员口号是政治性的（"王侯将相宁有种乎""复汉"）。
黄巾起义不同。太平道是一个完整的宗教组织，有教义（太平经），有仪式（符水治病），有组织体系（三十六方，每方数千到数万人），有领袖崇拜（张角被称为"大贤良师"）。信众遍布全国八州，横跨今天的河北、河南、山东、湖北、安徽、江苏。
为什么底层的反抗采取了宗教形式？
因为3.0版操作系统堵住了所有世俗的反抗通道。
儒家意识形态把政治秩序道德化了——君臣父子的关系是天经地义的，反抗就是不忠不孝。这个编码在正常时期有效地压制了世俗层面的反抗意识。一个受过基本儒学教化的农民，很难用世俗的政治语言来为自己的反抗行为辩护——因为儒学告诉他，反抗本身就是错的。
但宗教提供了一个绕过这个编码的通道。"苍天已死"——我不是在反抗君主，我是在服从一个更高的权威（天/黄天）。天命已经转移了。旧的天（苍天=汉朝）已经死了，新的天（黄天=太平道）要取而代之。这不是叛乱，这是天命所归。
宗教话语把反抗行为重新编码了：从"不忠不孝的叛乱"变成了"顺应天命的革命"。编码的层级更高——天命高于君臣伦理，所以天命的变动可以合法化对君臣关系的颠覆。
从余项的角度看，黄巾起义是底层余项（农民的贫困、愤怒、绝望）在世俗释放通道被堵死之后寻找到的替代释放通道。儒家意识形态堵住了世俗通道（你不能说"我反抗是因为朝廷对我不公"），宗教打开了超世俗通道（你可以说"我反抗是因为天命变了"）。
这个模式此后反复出现。东汉有太平道和五斗米道。元末有白莲教和弥勒教（红巾军）。明末有闻香教。清代有白莲教、太平天国。每一次底层反抗的宗教化，都说明同一件事：官方意识形态在堵住世俗反抗通道的同时，把反抗推向了更难控制的宗教维度。你堵住了一楼的门，人就从二楼的窗户跳出去。
黄巾起义本身很快被平定了（不到一年主力就被击溃）。但它的后果是致命的：为了平定起义，朝廷不得不允许地方豪族和州郡长官自行招募军队。这等于把军事权力从中央下放到了地方。军事权力下放之后再也收不回来了。各地的军阀——董卓、曹操、袁绍、刘表、孙坚——就在这个权力真空中崛起。
东汉不是被黄巾起义推翻的。东汉是被它自己为了对抗黄巾起义而释放出来的军事力量瓦解的。治病的药变成了致命的毒。
六、东汉的死因——一份完整的验尸报告
东汉的灭亡不是某一个因素造成的。它是多个余项同时到达临界点的结果。可以写一份完整的验尸报告：
直接死因：军阀割据导致的中央权威崩溃。
董卓入京（189年）之后，中央政府名存实亡。此后三十年（到220年曹丕代汉）是一个漫长的死亡过程——皇帝还在，但已经是傀儡。权力分散到了各地军阀手中。
间接死因一：外戚宦官交替消耗了官僚系统的质量。
反复的政治清洗把有能力的人淘汰出局，让官僚系统充满了庸才和投机者。到了需要系统应对危机的时候（黄巾起义），系统的行政能力已经不足以有效应对。
间接死因二：党锢之祸切除了系统的自我修复能力。
儒生群体被禁锢之后，系统失去了发现和报告问题的能力。问题继续积累但不被处理，直到积累到无法挽回的程度。
间接死因三：豪族的崛起掏空了中央对地方的实际控制力。
豪族的庄园经济、私人武装和垄断性的人才网络，使得地方社会在事实上独立于中央。当中央权威崩溃的时候，地方没有向心力把它拉回来，反而有离心力把它进一步撕裂。
根本死因：3.0版操作系统的单点结构。
所有上述问题的共同根源是：整个系统的运转依赖于皇帝一个人这个单点。皇帝年幼→外戚掌权→宦官反击→官僚系统降解→自我修复能力丧失→底层反抗→军事权力下放→军阀割据→中央崩溃。这条因果链的起点是"皇帝年幼"——一个系统无法控制的随机事件。
一个会因为一个随机事件而启动不可逆的崩溃链条的系统，不是一个好系统。但3.0版在这个关键节点上没有设计冗余。它没有摄政制度（有惯例但没有制度），没有权力分立（所有权力集中于皇帝），没有独立的选拔最高权力者的机制（血缘继承，嫡长子优先，没有能力筛选）。
这个根本缺陷不是东汉特有的。它存在于3.0版的每一次安装中——西汉有同样的问题（外戚干政在西汉末年也是关键因素），只是西汉的皇帝们平均在位时间更长、平均即位年龄更大，问题没有像东汉这么集中爆发。东汉的不幸在于连续出了一串幼帝，把这个缺陷推到了极限。
但连续出幼帝也不完全是偶然。东汉后期外戚和宦官都倾向于选择年幼的皇位继承人——因为幼帝更容易被控制。操纵继承人选择的能力本身就是外戚和宦官权力的一部分。系统的缺陷（单点结构→幼帝风险）被利用缺陷的人（外戚、宦官）主动放大了。这是正反馈——缺陷制造了利用缺陷的人，利用缺陷的人又放大了缺陷。
七、东汉对周期律的贡献
东汉两百年是3.0版操作系统的第一个完整运行周期。它提供了几条永久性的经验。
第一，余项会制度化。
豪族的演化过程证明：余项不是一个静态的东西。它会生长，会从小变大，从松散变紧密，从非正式变正式。你在系统建立初期看到的那个小小的余项（几个退休功臣的庄园），在一百年后可能变成一个吞噬系统的怪物（控制了整个地方社会的豪族网络）。
余项的制度化是构老化的核心机制。不是构本身的设计变坏了——3.0版的框架两百年间没有根本变化——而是余项在框架内部不断积累和固化，最终把框架掏空。
第二，构的自我修复机制可以被自我切除。
党锢之祸证明：一个构如果拥有自我修复能力（儒生群体的批评功能），这个能力本身也可以被构内部的权力斗争所消灭。系统会杀死自己的医生。
这是一个令人沮丧的发现。它意味着即使你在系统中设计了自我修复机制，你也不能保证这个机制不会在某一天被系统自己关掉。系统的免疫系统可以被系统内部的"自身免疫病"摧毁。
第三，堵住一个释放通道，余项会寻找另一个通道。
黄巾起义证明：你用意识形态堵住了世俗反抗的通道，反抗就会穿上宗教的外衣从另一个维度冒出来。余项是水。你堵住一条路，它就找另一条路。你堵住所有的路，它就从地下渗出来。
第四，治病的药可能变成致命的毒。
为了对抗黄巾起义而下放的军事权力，最终摧毁了中央权威本身。这个教训后来反复出现：安史之乱后唐朝为了平叛而扩大藩镇权力，结果藩镇成了新的威胁。明朝为了对抗蒙古而设立的边军系统，最终成了李自成起义军的兵源。
每一次用力量解决问题，力量本身都有可能变成新的问题。这不是说不应该使用力量——有时候你别无选择——而是说使用力量的人应该意识到力量释放之后的不可控性。药都有副作用。用药之前要评估副作用是否比疾病更致命。
第五，构的崩溃不是一个事件，是一个过程。
东汉的崩溃不是发生在某一天的。从和帝即位（88年）到曹丕代汉（220年），是一百三十多年的缓慢崩解过程。在这个过程中，有过看起来像恢复的时刻（邓绥太后的政治清明期、顺帝初年的短暂稳定），但这些恢复只是崩解过程中的暂停，不是逆转。
崩溃的过程性意味着：在崩溃的每一个阶段，都有人认为系统还能修复。他们不是在自欺——系统在那个阶段确实还没有完全坏掉。但他们低估了已经积累的余项的总量。表面上看系统还在运转，底层的余项已经积累到了不可逆的程度。
什么时候从"还能修复"变成"不可挽回"？没有人在当时能准确判断。这个临界点只有在事后才清晰。这是构的老化过程中最残酷的特征：你永远不知道自己是在临界点之前还是之后，直到一切已经太晚。
下一篇：三国——天然的对比篇。东汉崩溃留下的核心余项是豪族。四个主要玩家对这个余项给出了四种截然不同的处理方式：曹操压制它，刘备绕过它，孙权和它谈判，司马家拥抱它。四种方式，四种代价。最后胜出的是拥抱方案——司马家把豪族从余项变成执政基础，以阶级共治换取天下。看起来最聪明，但代价最大：你和余项融为一体，就再也没有制衡的空间了。这个代价在西晋八王之乱和此后北方诸族入主中原的大分裂中全额清算。
Liu Xiu rebooted version 3.0. Framework unchanged, parameters slightly adjusted, system back online.
Eastern Han (25–220 CE) is therefore an extraordinarily valuable observation sample: the first case of version 3.0 running a complete cycle without construct-type replacement. From establishment to collapse, nearly two hundred years. In those two hundred years, we can watch with clarity how the remainder accumulates within an unchanged framework — from germination to institutionalization to loss of control to collapse.
Western Han's story was the invention and establishment of a construct type. Eastern Han's story is the operation and aging of that construct type. The lessons differ: Western Han teaches you how to build a machine; Eastern Han teaches you how a machine dies of old age.
I. Guangwu's Parameter Adjustments
Liu Xiu did not change the framework but changed several key parameters. Two deserve attention.
Restraining the meritocracy bloc. Liu Xiu had relied on meritorious generals to win the realm, but he gave them wealth (fiefdoms, land, money) rather than real political or military power. "Retire the meritorious and advance the civil servants" — the "Twenty-Eight Generals of the Cloud Terrace" became wealthy but politically marginal. This was smarter than Liu Bang's purges (which damaged political credibility) but had a side effect: it created a class of locally powerful families with enormous land and wealth but no official power. These people were the seed of the great clans (豪族) that would dominate later Eastern Han.
Strengthening the Secretariat (尚书台). Liu Xiu institutionalized Emperor Wu's practice of bypassing the formal bureaucracy (headed by the chancellor) in favor of the imperial Secretariat, which answered directly to the emperor. This enhanced the emperor's personal control — when the emperor was competent (Guangwu, Emperor Ming), the system worked well. But it also meant that if the emperor was young and weak, whoever stood closest to the emperor would seize control of the Secretariat. Liu Xiu's design for strengthening imperial authority inadvertently paved the way for consort clans and eunuchs.
II. Great Clans: The Institutionalization of the Remainder
Eastern Han's most important social phenomenon was the rise of the great clans.
Great clans had existed in Western Han, where they were actively suppressed — Emperor Wu's harsh officials specifically targeted them; forced relocation to the capital region was standard policy. Eastern Han was different. Liu Xiu himself came from a local clan in Nanyang. His core supporters were overwhelmingly from local clan networks. You cannot turn around after winning the realm and suppress the very people who were your base.
The result: the Eastern Han court shifted from Western Han's active suppression to default tolerance.
Over Eastern Han's two hundred years, the great clans underwent a transformation from economic entities to political entities in three phases.
Phase one (Guangwu to Emperor Zhang, 25–88 CE): Clans were primarily economic forces — large landholdings, dependent tenant farmers, private armed retainers. Economically independent from the state but not yet systematically politically influential.
Phase two (Emperor He to Emperor Huan, 88–167 CE): Clans began systematically penetrating the bureaucratic system through the recommendation system (察举制). The recommendation system required local officials to recommend moral exemplars; the recommendation power was in local officials' hands; local officials were largely themselves from great clans; they naturally recommended their own family members first. After several generations, certain great families monopolized the bureaucracy across generations — "four generations, three grand councilors" became a family's badge. The Yuan clan of Runan (four generations, five grand councilors) and the Yang clan of Hongnong (four generations as Grand Marshal) were the most prominent examples.
Phase three (Emperor Huan to Han's end, 167–220 CE): Great clans no longer merely penetrated the bureaucracy — they effectively controlled local society. Their manor estates were self-sufficient economic entities with their own agriculture, handicrafts, and even private schools. Their dependent farmers and household soldiers constituted de facto private armies. By the end of Han, when central authority collapsed, these clans were the only organized forces in the localities.
From the remainder-conservation perspective, the evolution of the great clans is a textbook case. Version 3.0's design goal was the state facing individual households directly, with no intermediate layer. Qin's household registration system had dispersed the clans. But the design goal could not be permanently maintained — human beings naturally form groups (blood-based, geographic, economic, cultural). Destroy the old intermediate layer (the clans), and new intermediate layers grow on new foundations. Great clans grew from the dual action of land concentration and the recommendation system.
The intermediate layer is the construct's permanent remainder. You cannot eliminate it because its roots lie not in any specific institutional arrangement but in human sociality itself. You can change its form (from clans to great clans to gentry-families to local strongmen), but you cannot cancel the phenomenon.
III. Consort Clans and Eunuchs: Two Remainders Consuming Each Other
Eastern Han's mid-to-late political history can be summarized in one sentence: consort clans and eunuchs alternately held power.
The mechanism worked as follows: emperor ascends as a minor (from Emperor He onward, virtually every Eastern Han emperor ascended as a child — Emperor He at ten, Emperor Shang as an infant, Emperor An at thirteen, Emperor Shun at eleven, Emperor Chong at two, Emperor Zhi at eight, Emperor Huan at fifteen, Emperor Ling at twelve). The empress dowager governs; her natal family (consort clan) holds real power. The emperor grows up and wants power back, but the court is staffed with consort clan allies. The only people who live with the emperor day and night and depend entirely on him: eunuchs. The emperor allies with the eunuchs to stage a coup and destroy the consort clan. Eunuchs seize power. New emperor ascends as a minor. New empress dowager governs. New consort clan holds power. Eunuchs are purged. Repeat.
This cycle repeated at least four or five complete rotations. Each time the consort clan was eliminated, eunuchs rose. Each time eunuchs were purged, consort clans returned. The two groups were like the ends of a seesaw — perpetually alternating.
Both groups were structural products of version 3.0. Consort clans arose because version 3.0 had no mechanism for power transfer independent of the emperor's person — when the emperor was young, whoever was closest to him seized power, and that was typically the empress dowager's family. Eunuchs arose because version 3.0 concentrated power in the emperor's person (the Secretariat design); when the emperor needed to fight the outer court (controlled by the consort clan), his only political resource was the inner court, whose core personnel were eunuchs.
The mutual-consumption process was enormously harmful. Each power transition brought political purges — killings, confiscations, exiles. These eliminated not just individuals but their networks throughout the bureaucracy. After several rounds of purging, capable and experienced people had been repeatedly removed; what remained were either harmless mediocrities or opportunistic fence-sitters. The system was self-degrading. Each internal power struggle lowered the bureaucracy's overall quality, like a garment washed too many times — thinner and thinner, until the fabric could no longer hold.
IV. The Partisan Prohibitions: When Maintainers Become Challengers
In 166 and 169 CE, Eastern Han saw two rounds of the "Partisan Prohibitions" (党锢之祸).
A network of Confucian scholars centered on Imperial Academy students and local literati — the "pure critics" (清议) movement — had been publicly criticizing eunuch dominance and offering assessments of officials and political affairs. The eunuch faction counterattacked, accusing these scholars of "forming factions," and persuaded Emperors Huan and Ling to arrest and prohibit the "partisans" from office — not just the individuals, but their relatives, students, and associates. Hundreds were directly affected; thousands were peripherally caught.
From the Chisel-Construct perspective, the Partisan Prohibitions' most significant effect was structural: the Confucian scholar class was transformed from the construct's maintainers into the construct's challengers.
Recall version 3.0's design: "exclusive reverence for Confucianism" aimed to use Confucian ideology to encode the remainder — making society willingly accept order. The Confucian scholar class's role in this design was to produce and transmit ideology — studying classics, interpreting them, evaluating politics through them, educating the next generation. They were the system's software maintenance team.
When the system ran normally, this role was a source of stability. Scholars criticized specific imperial conduct (within the Heaven-humanity resonance framework), but did not question the system itself. Their criticism was constructive — pointing out deviation from correct parameters and requesting adjustment. This kind of criticism actually helped the system self-repair.
But when the system's deviation was not a parameter problem but a structural one — when eunuchs systematically corrupted the entire bureaucracy, when emperors systematically sided with eunuchs — the scholars' criticism shifted from constructive to oppositional. They were no longer saying "the parameters are misconfigured." They were beginning to say "the operators are the problem." Evaluating officials and assessing politics — the substance of these activities was: we, the intellectuals, claim the right to judge who deserves to hold power. This was a direct challenge to the emperor's personnel prerogative.
The eunuch faction saw the threat. The label they attached to the partisans — "forming factions" — was a political attack, but not entirely without basis. The partisans were indeed forming an organized group with shared values attempting to influence political direction. If this group succeeded, the eunuchs were finished.
The Partisan Prohibitions exposed a deep contradiction in the exclusive reverence for Confucianism design: you cultivated an intellectual class whose vocation was moral criticism, then you forbade them to criticize. You cannot have an independently thinking intellectual corps (to produce legitimacy and provide policy critique) and also have them fall silent on demand. Independent thought is not a tap you can turn on and off as needed. Turn it on, and you must accept the water it produces even when you don't want it.
The Partisan Prohibitions were an act of self-mutilation. The system cut off its own software maintenance team. Short-term effect: the eunuch faction was safe. Long-term effect: the system lost its self-repair capacity. Problems continued accumulating but no longer had anyone pointing them out. Silence is not health. Silence is abandonment of treatment.
Twenty years after the Partisan Prohibitions, the Yellow Turban Uprising broke out.
V. Yellow Turbans, Great Clan Militarization, and Collapse
In 184 CE, Zhang Jue's Way of Supreme Peace launched a simultaneous nationwide uprising. Believers wrapped their heads in yellow cloth — the Yellow Turban Uprising. Slogan: "The blue sky is dead; the yellow sky shall stand. The year is jiazi; great fortune for the realm."
This was China's first large-scale peasant uprising organized in religious form. Why religion?
Because version 3.0's operating system had blocked all secular channels of resistance. Confucian ideology had moralized the political order — ruler-minister and father-son relationships were the way of heaven and earth; resistance was disloyalty and filial impiety. A farmer with even basic Confucian education could not justify resistance in secular political language — Confucianism told him resistance itself was wrong.
But religion provided a bypass. "The blue sky is dead" — I am not resisting the ruler; I am obeying a higher authority (Heaven). Heaven's mandate has transferred. The old heaven (Han) has died; the new heaven (the Way of Supreme Peace) is to take its place. This is not rebellion — this is following Heaven's mandate. Religious discourse re-encoded the act of resistance: from "disloyal, unfilial rebellion" to "revolutionary compliance with Heaven's mandate."
The Yellow Turban Uprising was itself quickly suppressed (the main force was broken in under a year). But its consequences were fatal: to suppress it, the court had to allow local great clans and regional officials to recruit armies independently. This devolved military power from the center to the localities. Military power once devolved could never be recovered. The regional warlords — Dong Zhuo, Cao Cao, Yuan Shao, Liu Biao, Sun Jian — rose in this power vacuum.
Eastern Han was not toppled by the Yellow Turban Uprising. Eastern Han was dissolved by the military forces it had released to combat the Yellow Turban Uprising. The medicine became the fatal poison.
VI. The Death Certificate: A Complete Autopsy
Eastern Han's death was not caused by any single factor. It was the result of multiple remainders simultaneously reaching their critical point.
Direct cause: collapse of central authority through regional militarization.
Indirect cause one: the consort clan-eunuch alternation degraded bureaucratic quality through repeated purges, leaving the system unable to effectively respond to crisis.
Indirect cause two: the Partisan Prohibitions eliminated the system's self-repair mechanism — problems accumulated unchecked.
Indirect cause three: the great clans' rise hollowed out central control over localities; when central authority collapsed, there was no centripetal force to pull it back, only centrifugal forces to pull it further apart.
Root cause: version 3.0's single-point structure. The entire system's operation depended on one person — the emperor — functioning as its central node. Minor emperor → consort clan dominance → eunuch counterattack → bureaucratic degradation → loss of self-repair capacity → popular uprising → military devolution → regional warlordism → central collapse. The starting point of this causal chain was "minor emperor" — a random event the system could not control.
A system that, when triggered by a random event, initiates an irreversible collapse chain, is not a good system. But version 3.0 had no redundancy at this critical juncture. No institutionalized regency (custom existed, not institution). No separation of powers. No capacity-based mechanism for selecting the supreme power holder (succession by blood, eldest legitimate son first, no ability screening). These defects were not unique to Eastern Han — they existed in every installation of version 3.0. Eastern Han's misfortune was that it produced a string of consecutive child-emperors, pushing these defects to their absolute limit.
Next: Essay 12 — Three Kingdoms: four approaches to the great clan remainder, and why the one that looked cleverest proved most costly.