秦汉 Han Qin | 2026

前三篇分别在听觉（音乐）、跨通道（戏曲/歌剧）、动觉（舞蹈）中检验了"生定展固"。每一篇都增加了一个新机制：第一篇建立基本循环，第二篇引入跨通道凿，第三篇证明凿构循环不依赖特定感官通道。

现在进入最后一个检验场：话剧和电影。

话剧和电影的特殊性在于两件事。第一，叙事成为主通道——不是声音，不是身体动作，而是"故事"本身承载凿构循环。第二，电影是所有时间性艺术中通道最多的形态——叙事、对白、视觉构图、剪辑节奏、音乐、音效同时运行。

这意味着电影里可以同时运行多条凿构循环，每条在不同层级上独立运行，又互相影响。这是前三篇都没有触及的复杂度。

同时，通道越多，一个新的风险也越大：伪凿。

先把叙事单独拎出来，不加其他通道。话剧是最纯粹的形式——基本上只有对白和表演，没有剪辑，没有配乐（或极少），视觉通道被简化到舞台布景。

叙事的凿构循环跟音乐的完全同构，只是材料不同：

生：故事的开端建立了一个"世界"和一组人物关系。你的大脑开始预测——这些人物会怎样互动？冲突会往哪个方向走？你形成了一个叙事预测模型。

定：情节发展确认了你的模型。人物的行为符合你对他们性格的预期，冲突按照你预想的逻辑推进。你觉得"我理解这个故事在做什么了"。

展：某个瞬间，你的叙事预测模型被打破了。一个你没预期到的转折，一个人物做了一件你以为他不会做的事，一个你以为已经解决的冲突以新的形态重新出现。但跟音乐的"展"一样——这不是随机的，它在一个你事后能理解但事前不能预测的逻辑里。

固：故事的结局。但经历了"展"之后，同样的人物在你心理上的重量不同了。结尾不是回到开头，是带着凿的痕迹闭合。

四幕话剧，凿构循环极其完整。

生：周朴园的大宅，表面的秩序和体面建立了一个"世界"。你形成了一个预测模型——这是一个关于家庭秘密和阶级冲突的故事。

定：前两幕不断确认这个模型。周朴园的威严，繁漪的压抑，四凤的天真。你越来越确定你理解了这些人物。

展：身份关系的暴露。你以为你理解的人物关系全部被重新定义——但曹禺的厉害在于，暴露不是一次性的。每一层暴露都打破你刚刚重建的模型，余项层层叠加。你在第三幕开始觉得"我不知道这个故事会怎样结束"——但你又觉得它必须按照某种你说不出的逻辑结束。

固：最后的毁灭。所有人物都被摧毁了，但你感到的不是"故事结束了"，而是"这个结束比开始重得多"。闭合携带了前面所有凿的痕迹。

跟《雷雨》做对照。完全不同的文化和戏剧传统，但凿构循环同构。

生：威利·洛曼的日常生活建立了一个关于"美国梦"的预测模型。

定：他的自我欺骗和家庭关系不断确认这个模型——你越来越清楚这个人在做什么。

展：米勒的凿不是靠情节反转，而是靠时间结构。过去和现在的场景交替出现，你对"现实"的预测模型被时间的错乱打破了。你不确定你看到的是记忆还是幻觉还是正在发生的事。

固：威利的死。但这个闭合的重量来自前面所有时间错乱积累的余项——你在结尾感受到的不只是"一个人死了"，而是"一整个关于成功和失败的叙事模型崩塌了"。

两部话剧的编码系统完全不同（中国家庭伦理悲剧 vs 美国社会现实主义），但凿构循环的操作同构。

话剧基本上是单通道叙事加上表演。电影把通道数量推到了极致。

一部电影同时运行着：叙事线（情节的凿构循环）、视觉构图（画面的凿构循环）、剪辑节奏（时间的凿构循环）、音乐/音效（听觉的凿构循环）、表演（身体-情绪的凿构循环）。

每一条线都可以独立运行自己的"生定展固"。但更重要的是，它们可以跨层级凿——一条线在"定"的同时另一条线在"展"，就像第二篇里京剧的唱腔和身段。只不过电影有更多的通道可以做这个操作。

但电影的真正特殊性不只是"通道更多"。如果只是通道数量的叠加，那电影跟歌剧之间就只有量的差异，没有质的差异。

质的差异在于：电影（和话剧）引入了一种前三篇没有处理过的通道类型——叙事通道。叙事通道承载的不是感官层面的预测模型（"下一个音是什么""下一个动作是什么"），而是概念/逻辑层面的预测模型（"这个人物会做什么""这个故事会怎样结束""这个世界的规则是什么"）。

当概念模型与感官模型发生严重偏离时——叙事告诉你一件事，你的眼睛和耳朵告诉你另一件事——这才是电影独有的凿的形态。第二篇的跨通道凿主要发生在感官通道之间（听觉vs视觉）。第四篇的多层级凿发生在概念通道与感官通道之间。这是一个层级的跃升，不只是数量的叠加。

这部电影的叙事线极度简约——人类发现黑石碑，去木星调查，遭遇AI失控，主角穿越某种超越性体验。如果只看情节，预测模型很快就稳定了。

但库布里克在视觉和音乐层级上做了极端的凿。

"星门"序列：叙事几乎停滞（你不知道发生了什么，但也不觉得叙事在推进），视觉通道全面接管——色彩、光线、抽象的图形以一种你完全无法预测的节奏涌来。音乐从利盖蒂的微复调到施特劳斯的《蓝色多瑙河》的重新出现，每一次音乐的选择都在打破你对"太空电影应该听起来像什么"的预测模型。

叙事在"定"（甚至在"停"），视觉在"展"，音乐在做另一种"展"。三条凿构循环在不同的相位上同时运行，它们之间的不同步制造了巨大的余项——你看完之后说不清这部电影"关于什么"，但你知道那里有东西。这个余项在五十多年后依然不可穷尽。

跟库布里克完全相反的策略。

小津的画面几乎不动。镜头固定，机位低，构图对称。剪辑节奏缓慢，场景之间用"枕镜头"（空镜头——一个花瓶，一条走廊，一片天空）过渡。音乐极少。叙事也极度日常——一对老夫妇去东京看望子女，子女们太忙没时间陪他们，他们回家了，妻子去世了。

从表面看，所有通道都在"定"。没有视觉奇观，没有情节反转，没有音乐高潮。

但这部电影看十遍还有东西。

为什么？因为小津的凿不是在任何单一通道内发生的，而是在通道和你的期待之间发生的。你期待一个关于家庭疏离的故事会有情感爆发——没有。你期待老妻子的死会被悲剧性地呈现——小津用一个枕镜头带过。你期待结尾会有某种和解或者感悟——儿媳妇说了一句极其平淡的话，然后画面切到空旷的风景。

每一次你的叙事期待被"不发生"打破，就是一次凿。小津的chisel不是"做了什么"，而是"没做什么"。他在你的预测模型中挖了一个洞——那个洞就是余项。你没办法用"看懂了"来填满它，因为那个洞的形状就是"这里本该有东西但没有"。

这种通过缺席来制造余项的手法，跟皮娜·鲍什通过过度重复来制造余项，在结构上是等价的——都是用一种极端（极端的无 vs 极端的重复）来把"定"推到相变的临界点。

现在可以正式定义一个在前三篇中已经多次暗示的概念了。

它让你惊一下，却不让"闭合后的重量"发生变化。你的预测模型被短暂扰动，但扰动过后模型恢复原样，没有留下痕迹。因此复看时余项迅速穷尽。

更精确地说：伪凿是一种可被认知系统即时吸收的低维扰动。真凿发生时，旧模型失效，你必须构建一个包含断裂痕迹的新模型——这个重建过程产生了当前认知算力无法立刻处理的剩余信息，即余项。伪凿发生时，你的认知系统在闭合的同时就把那个"意外"作为已知参数添加回了原模型。不需要重建，只需要更新一个变量。第二遍体验时，该变量已被吸收，信息熵严格归零。

伪凿在电影中特别常见，因为电影有太多通道可以制造"意外"：一个突然的镜头切换，一个音效的jump scare，一个情节的twist，一个视觉特效的奇观。这些都能让你"惊一下"。但惊不等于凿。

真凿打破你的模型之后，你需要重建模型——而重建后的模型比之前更厚，因为它包含了被打破的痕迹。这就是"固"。

伪凿打破你的注意力之后，你的模型原封不动地恢复了。没有重建，没有新的厚度。"固"的内容跟"定"一样。等于什么都没发生。

这就是为什么很多大制作电影"第一次看很爽，第二次就没意思了"——它们充满了伪凿。第一次看的时候你不知道twist是什么，所以每个twist都让你惊了一下。但第二次看，你已经知道了，而twist之外的部分（叙事的construct、人物的深度、视觉的层次）没有提供可重复发现的余项。信息量在第二遍就归零了。

第一节已经展示了这两部话剧的凿构循环。现在正式做同构对照。

两部作品的编码系统完全不同。《雷雨》用中国家庭伦理的框架，人物关系的复杂性靠血缘和阶级的交织来构建，"展"主要靠身份暴露——你以为你知道这些人是谁，你错了。《推销员之死》用美国梦的框架，"展"主要靠时间结构的错乱——你以为你知道什么是现实，你错了。

一个凿的是"人是谁"，另一个凿的是"什么是真"。凿的对象不同，但凿的操作同构：打破你的叙事预测模型，产生不可一次性消化的余项，然后在闭合时让余项的重量沉入结局。

两部话剧都能穿越周期。《雷雨》1934年首演至今仍在上演。《推销员之死》1949年首演至今是美国话剧的标准曲目。原因在认知层面是同一个：它们的余项不可穷尽——你每次看都能在"展"的部分发现新的层次。

如果说《雷雨》和《推销员之死》是"不同编码做同一个操作"，那库布里克和小津是另一个层面的证明："完全相反的凿方式，产生同一种余项效果"。

库布里克靠的是极端的感官过载——"星门"序列用视觉和声音的洪流冲垮你的预测模型。小津靠的是极端的感官克制——用"不发生"在你的预测模型中挖洞。

一个用过剩制造余项，一个用缺席制造余项。方法相反，效果等价：两部电影都看十遍还有东西，因为它们的余项都不可被一次性穷尽。

这跟第三篇的"能剧 vs 皮娜·鲍什"在结构上完全对称——能剧靠微观偏离，鲍什靠过度重复；库布里克靠感官洪流，小津靠感官真空。四种完全不同的方法，同一个余项效果。

异构同效不是巧合，是凿构循环通用性的证据：只要你的操作真实地打破了预测模型并产生了不可穷尽的余项，用什么方法打破的并不重要。

叙事的凿构循环有一个必须对话的前辈：亚里士多德的《诗学》。

两千三百年前，亚里士多德分析了希腊悲剧的结构，提出了几个核心概念：突转（peripeteia）——情节的突然逆转；发现（anagnorisis）——关键信息的揭示；净化（katharsis）——观众在悲剧结束时经历的情感释放。

这些概念跟"生定展固"的对应关系几乎是直接的：突转和发现就是"展"——它们打破了观众对故事走向的预测模型。净化就是"固"——经历了打破之后的情感闭合。

但亚里士多德的框架有两个局限。

第一，他只关注叙事通道。《诗学》讨论的是情节结构，不涉及音乐、视觉、身体动作等其他通道。所以他能解释话剧为什么有力，但解释不了电影的多层级凿构——为什么《2001太空漫游》的"星门"序列在叙事几乎停滞的情况下依然产生巨大的冲击力。在亚里士多德的框架里，没有情节的突转就没有力量。在"生定展固"的框架里，叙事只是通道之一，凿可以在任何通道上发生。

第二，他没有"余项"的概念。净化（katharsis）是一次性的——你看完悲剧，情感被释放，完成了。但这解释不了为什么有些作品能反复观看而净化效果不减——如果净化是一次性释放，第二遍应该没有效果才对。"余项"解释了这个：好的作品每次观看都暴露出新的未被消化的部分，所以"固"的内容每次都不同。净化不是一次性的释放，是每次闭合时对新暴露余项的整合。

亚里士多德的位置：他在叙事通道内精确地描述了"展"和"固"的操作（突转/发现/净化），但止步于单通道和一次性体验。本文的位置：把亚里士多德的洞察扩展到所有通道（凿不限于叙事），并引入余项概念解释为什么好的作品可以反复体验。

跟第一篇的Hanslick/Meyer、第二篇的Wagner、第三篇的Laban一样：前辈抓住了大象的一条腿，本文试图装配整头大象。

同一个导演，水准相当的制作能力，但余项深度完全不同。

《星际穿越》：叙事的凿构循环完整——父女关系建立了强大的情感construct（生定），黑洞和时间膨胀打破了你对"救援"的预测模型（展），最后的"书架场景"让所有被打破的东西在一个你没预期到但事后觉得必然的位置闭合（固）。而且音乐（汉斯·季默的管风琴）跟叙事之间有精确的跨通道凿——音乐的情感强度经常超出叙事当下的逻辑，制造通道之间的相位差。

看十遍还有东西。因为父女关系的情感余项和时间/空间的认知余项每次观看都能暴露出新的层面。

《信条》：技术上同样精湛，时间逆转的概念同样震撼，但你看第二遍就觉得少了什么。为什么？

因为《信条》的主要"展"是概念性的——时间逆转这个设定本身就是最大的打破。第一次看的时候你的预测模型被彻底冲垮，你花全部精力在理解"发生了什么"。但第二次看，你已经理解了机制，这个概念性的打破就不再有效了。而在概念之下，叙事的情感construct（主角和同伴的关系、主角的动机）相对薄弱——它被概念的复杂性挤占了空间。

用本文的语言说：《信条》的"展"是一次性的——概念性打破在第一遍就被完全吸收了。《星际穿越》的"展"是多层级的——概念层面的打破（时间膨胀）和情感层面的打破（父女关系的代价）在不同通道上运行，每一层都有自己的余项，而且层与层之间的相位差产生了额外的余项。

这就是伪凿和真凿的区别在具体作品中的表现。《信条》不是不好，它只是它的凿集中在一个层级上，那个层级在第一遍就穷尽了。《星际穿越》的凿分布在多个层级上，每个层级都有不可一次性穷尽的余项。

公平地说，早期漫威有几部凿构循环相对完整的作品——第一部《钢铁侠》里Tony Stark从军火商到英雄的人物弧线是真实的construct到chisel到带痕闭合，《美国队长2》的政治惊悚结构在叙事层面做了真正的展。这些作品能反复看，因为人物的变化和叙事的打破产生了真实的余项。

但随着系列扩展，一个退化模式越来越明显。后期大部分漫威电影的结构变成了：建立英雄的初始状态（生），确认他的能力和使命（定），大反派出现制造危机（展？），英雄战胜反派（固）。

看起来四步完整。但为什么后期漫威电影看三遍就够了？

因为"展"的位置上放的越来越多是伪凿。大反派的出现确实制造了"意外"，但这个意外不改变你对整个叙事框架的模型——你知道英雄最终会赢，你知道牺牲的人可能在下一部复活，你知道宇宙的规则不会真的被打破。你的预测模型被短暂扰动，但扰动过后恢复原样。

视觉层面更是如此。每一场战斗都是视觉奇观，但奇观之间没有凿构循环——一个爆炸接一个爆炸，你的视觉预测模型始终在"展"的状态却从不进入"固"。这就是第三篇描述的"展-展-展-展"——纯炫技结构。

这条退化曲线本身就是论证：同一个系列，早期有真凿，后期伪凿占比越来越高，观众的"反复看"意愿相应下降。不是制作水准下降了，是凿构循环的完整性在流水线化的过程中被牺牲了。construct越来越精致（视觉特效越来越好），但chisel越来越少（你的预测模型越来越不会被真正打破）。

在结束之前，补一个被前面几篇跳过的形态：口述传统。

荷马史诗的吟诵、中国评书、日本落语、西非griot的口头叙事——这些是"叙事作为主通道"的最古老形态。它们是话剧和电影的前身。

口述传统的特殊性在于：它只有声音和叙事两个通道（没有视觉布景、没有剪辑、没有配乐），但它的凿构循环一样完整。

中国评书的"扣子"（悬念设置）就是"生"和"定"——建立你对故事走向的预测。"翻包袱"就是"展"——打破预测，暴露余项。最后的收束就是"固"。

日本落语更极端——一个人坐在那里，没有道具，只靠声音的变化（不同角色的声线切换、节奏的快慢交替）来运行凿构循环。它证明了一件事：即使通道被压缩到最少（只有声音+叙事），凿构循环依然可以完整运行。

这跟第三篇中"舞蹈在接近沉默中依然成立"的论证是对称的：舞蹈证明了去掉声音凿构循环还在，口述传统证明了去掉画面凿构循环还在。通道可以增加也可以减少，循环不变。

四篇写完了。回到开头的问题。

为什么有些旋律听一百遍还有东西？为什么有些戏看十遍还能看到新的东西？为什么有些舞蹈让你觉得"结束之后比开始更重"？为什么有些电影反复看不厌？

答案是同一个：

这个判断标准跨越了：

感官通道——适用于听觉（音乐）、动觉（舞蹈）、视觉+叙事（电影） 文化传统——适用于贝多芬也适用于K-pop，适用于京剧也适用于街舞 历史时期——适用于巴赫也适用于Kendrick Lamar，适用于能剧也适用于库布里克 媒介复杂度——适用于单通道（纯器乐）也适用于全媒介（电影）

因为它不是任何特定编码系统的产物。它是人类认知系统的基本操作：建模→确认→模型被打破→带痕重建。

乐理是这个操作在声音上的编码。舞蹈词汇是这个操作在身体上的编码。叙事结构是这个操作在故事上的编码。每一套编码系统都精妙而独特，但它们编码的是同一个东西。

"生定展固"直接取自Self-as-an-End框架的生命周期表。音乐让你感动，舞蹈让你屏息，电影让你在结束后久久不能起身——都是因为它们在各自的通道上复现了生命本身的节奏：选择、构建、被凿开、重新闭合。

本系列不讲乐理，讲乐理的道理。

不讲舞蹈技法，讲技法的道理。

不讲叙事理论，讲叙事理论的道理。

道理只有一个。

最后说一个从这个框架里自然推导出来的预测：时间性艺术没有上限，没有尽头。

原因不是"人类创造力无穷"这种空话。原因是凿构循环的结构本身保证了它不可能终止。

任何形式，一旦被观众熟悉，就变成construct。construct一旦稳定，就可以被凿。凿暴露余项。余项推动新的construct形成。新的construct再被凿。

奏鸣曲式是construct，勋伯格凿了它。十二音技法变成新的construct，凯奇凿了它。4'33"变成construct，下一个人会凿它。京剧的板式是construct，梅兰芳在里面做了微观凿。梅兰芳的风格本身变成了construct，下一代演员会凿它。古典芭蕾是construct，皮娜·鲍什凿了它。鲍什的重复手法变成了construct，下一代编舞会凿它。

没有终点。因为余项——Self-as-an-End框架贯穿始终的核心概念——永远不可能被完全收编。每一次闭合都携带着不可穷尽的剩余，这个剩余就是下一次凿的起点。

这不是乐观，是结构。

Han Qin | 2026

The first three essays tested Arise-Settle-Unfold-Fix in audition (music), cross-channel (opera / Chinese opera), and kinesthetic (dance) domains. Each essay added a new mechanism: Essay I established the basic cycle, Essay II introduced cross-channel chisel, Essay III proved the cycle does not depend on any particular sensory channel.

Now we enter the final testing ground: theatre and film.

Theatre and film are special in two respects. First, narrative becomes the primary channel — not sound, not bodily movement, but "story" itself carries the chisel-construct cycle. Second, film is the temporal art form with the greatest number of simultaneous channels — narrative, dialogue, visual composition, editing rhythm, music, sound design all running at once.

This means film can run multiple chisel-construct cycles simultaneously, each at a different level, operating independently yet mutually influencing one another. This is a degree of complexity untouched in the first three essays.

At the same time, the more channels available, the greater a new risk becomes: pseudo-chisel.

Isolate narrative first, without other channels. Theatre is the purest form — essentially only dialogue and performance; no editing, no score (or very little), the visual channel reduced to stage design.

The narrative chisel-construct cycle is fully isomorphic with that of music; only the material differs:

Arise: The story's opening establishes a "world" and a set of character relationships. Your brain begins predicting — how will these characters interact? Which direction will the conflict take? You form a narrative predictive model.

Settle: Plot development confirms your model. Characters behave consistently with your expectations of their personalities; conflict advances according to the logic you anticipated. You feel: "I understand what this story is doing."

Unfold: At some moment, your narrative predictive model is broken. An unanticipated turn; a character does something you thought they would not do; a conflict you thought resolved reappears in a new form. But as with musical Unfold — this is not random; it follows a logic comprehensible in retrospect but not predictable in advance.

Fix: The story's ending. But having passed through Unfold, the same characters now carry different weight in your psychology. The ending is not a return to the beginning; it is closure carrying the trace of chisel.

A four-act spoken drama with an extremely complete chisel-construct cycle.

Arise: The Zhou mansion; its surface order and respectability establish a "world." You form a predictive model — this is a story about family secrets and class conflict.

Settle: The first two acts continuously confirm this model. Zhou Puyuan's authority, Fanyi's repression, Sifeng's innocence. You grow increasingly certain you understand these characters.

Unfold: The revelation of identity relationships. The character relationships you thought you understood are entirely redefined — but Cao Yu's mastery lies in revelation not being a single event. Each layer of revelation breaks the model you have just rebuilt; remainder accumulates layer upon layer. By Act III you feel: "I do not know how this story will end" — yet you also feel it must end according to some logic you cannot articulate.

Fix: The final destruction. All characters are destroyed, but what you feel is not "the story is over" but "this ending weighs far more than the beginning." Closure carries the trace of every preceding chisel.

In counterpoint with Thunderstorm. Entirely different cultural and theatrical tradition, but isomorphic chisel-construct cycle.

Arise: Willy Loman's daily life establishes a predictive model of the "American Dream."

Settle: His self-deception and family relationships continuously confirm this model — you grow increasingly clear about what this man is doing.

Unfold: Miller's chisel does not rely on plot twist but on temporal structure. Past and present scenes alternate; your predictive model of "reality" is broken by the disorientation of time. You are unsure whether what you see is memory, hallucination, or present event.

Fix: Willy's death. But the weight of this closure comes from all the temporal disorientation that preceded it — what you feel at the end is not merely "a man died" but "an entire narrative model of success and failure has collapsed."

The two plays' encoding systems are entirely different (Chinese family-ethics tragedy vs. American social realism), but their chisel-construct operations are isomorphic.

Theatre is essentially single-channel narrative plus performance. Film pushes channel count to the extreme.

A film simultaneously runs: narrative line (the chisel-construct cycle of plot), visual composition (the cycle of the image), editing rhythm (the cycle of time), music/sound design (the auditory cycle), performance (the body-emotion cycle).

Each line can independently run its own Arise-Settle-Unfold-Fix. But more importantly, they can chisel across levels — one line in Settle while another is in Unfold, just as in Essay II where Peking opera's singing and body work ran at different phases. Film simply has more channels available for this operation.

But film's true specificity is not merely "more channels." If it were only quantitative stacking, the difference between film and opera would be of degree, not kind.

The qualitative difference is this: film (and theatre) introduce a channel type the first three essays did not address — the narrative channel. The narrative channel carries not a sensory-level predictive model ("what is the next sound," "what is the next movement") but a conceptual/logical-level predictive model ("what will this character do," "how will this story end," "what are the rules of this world").

When the conceptual model and the sensory model are in severe misalignment — narrative tells you one thing while your eyes and ears tell you another — this is the chisel form unique to film. Essay II's cross-channel chisel occurred primarily between sensory channels (audition vs. vision). Essay IV's multi-level chisel occurs between the conceptual channel and sensory channels. This is an escalation in level, not merely in quantity.

This film's narrative line is extremely spare — humanity discovers the monolith, goes to Jupiter to investigate, encounters AI failure, the protagonist undergoes some transcendent experience. Looking only at plot, the predictive model stabilizes quickly.

But Kubrick performs extreme chiseling at the visual and musical levels.

The "Star Gate" sequence: narrative is nearly static (you do not know what is happening, but neither does narrative seem to advance); the visual channel completely takes over — color, light, abstract patterns arrive at a rhythm you cannot predict. Music shifts from Ligeti's micropolyphony to the reappearance of Strauss's "Blue Danube," each musical choice breaking your predictive model of "what a space film should sound like."

Narrative is in Settle (or even in suspension); vision is in Unfold; music performs another form of Unfold. Three chisel-construct cycles run at different phases simultaneously; their desynchronization produces enormous remainder — after viewing, you cannot articulate what this film is "about," but you know something is there. This remainder remains inexhaustible after more than fifty years.

The precise opposite strategy from Kubrick.

Ozu's images are almost motionless. Fixed camera, low angle, symmetrical composition. Editing is slow; scenes transition through "pillow shots" (empty images — a vase, a corridor, a patch of sky). Music is minimal. The narrative is extremely mundane — an elderly couple visits their children in Tokyo; the children are too busy to spend time with them; the couple returns home; the wife dies.

On the surface, all channels are in Settle. No visual spectacle, no plot twist, no musical climax.

Yet this film still yields something after ten viewings.

Why? Because Ozu's chisel does not occur within any single channel but between channels and your expectation. You expect a story about family estrangement to have an emotional eruption — there is none. You expect the wife's death to be presented tragically — Ozu handles it with a single pillow shot. You expect the ending to contain some reconciliation or insight — the daughter-in-law says something extremely bland, then the image cuts to an empty landscape.

Each time your narrative expectation is broken by "non-occurrence," that is a chisel. Ozu's chisel is not "what he did" but "what he did not do." He carved a hole in your predictive model — and that hole is remainder. You cannot fill it with "having understood," because the hole's shape is precisely "something should have been here but was not."

This technique of producing remainder through absence is structurally equivalent to Pina Bausch's technique of producing remainder through excessive repetition — both use an extreme (extreme absence vs. extreme repetition) to push Settle to the threshold of phase transition.

A concept repeatedly intimated in the first three essays can now be formally defined.

It startles you but does not alter the weight of closure. Your predictive model is briefly perturbed, but after the perturbation the model restores itself unchanged, leaving no trace. On repeated viewing, remainder is rapidly exhausted.

More precisely: pseudo-chisel is a low-dimensional perturbation that can be immediately absorbed by the cognitive system. When genuine chisel occurs, the old model fails and you must construct a new model containing the trace of rupture — this rebuilding process produces residual information exceeding your current cognitive processing capacity, i.e., remainder. When pseudo-chisel occurs, the cognitive system absorbs the "surprise" as a known parameter appended to the original model simultaneously with closure. No rebuilding is needed, only updating a variable. On the second experience, that variable has already been absorbed; information entropy is strictly zero.

Pseudo-chisel is particularly common in film because film has so many channels available for manufacturing "surprise": a sudden cut, a jump scare in the sound design, a plot twist, a visual-effects spectacle. Each can startle you. But startling is not chiseling.

Genuine chisel, after breaking your model, requires you to rebuild — and the rebuilt model is thicker than its predecessor because it incorporates the trace of breaking. This is Fix.

Pseudo-chisel, after breaking your attention, leaves your model restored identically. No rebuilding, no new thickness. The content of Fix equals the content of Settle. Nothing has happened.

This is why many large-budget films "are exciting the first time but uninteresting the second" — they are filled with pseudo-chisel. On first viewing you do not know what the twist is, so every twist startles you. On second viewing you already know, and the portions beyond the twists (narrative construct, character depth, visual layering) provide no repeatedly discoverable remainder. Information reaches zero on the second pass.

Section I already presented the chisel-construct cycles of both plays. Now the formal isomorphic comparison.

The two works' encoding systems are entirely different. Thunderstorm uses the Chinese family-ethics framework; the complexity of character relationships is built through the interweaving of blood kinship and class. Its Unfold is primarily through identity revelation — you thought you knew who these people were; you were wrong. Death of a Salesman uses the American Dream framework; its Unfold is primarily through temporal-structural disorientation — you thought you knew what was real; you were wrong.

One chisels "who people are"; the other chisels "what is real." The object of chiseling differs, but the operation is isomorphic: break the narrative predictive model, produce remainder that cannot be absorbed in a single pass, then in closure let the weight of remainder sink into the ending.

Both plays cross cycles. Thunderstorm has been continuously performed since its 1934 premiere. Death of a Salesman has been a staple of the American theatrical canon since its 1949 premiere. The reason, at the cognitive level, is the same: their remainder is inexhaustible — each viewing reveals new layers in the Unfold.

If Thunderstorm and Death of a Salesman are "different encodings performing the same operation," then Kubrick and Ozu demonstrate a further level: "entirely opposite chisel methods producing the same remainder effect."

Kubrick relies on extreme sensory overload — the "Star Gate" sequence uses a flood of vision and sound to overwhelm your predictive model. Ozu relies on extreme sensory restraint — using "non-occurrence" to carve holes in your predictive model.

One uses excess to produce remainder; the other uses absence. Methods are opposite; effects are equivalent: both films still yield something after ten viewings, because their remainder is inexhaustible.

This parallels Essay III's "Noh theater vs. Pina Bausch" in structure — Noh relies on micro-deviation; Bausch on excessive repetition; Kubrick on sensory flood; Ozu on sensory vacuum. Four entirely different methods; the same remainder effect.

Heteromorphic equivalence is not coincidence; it is evidence for the universality of the chisel-construct cycle: as long as the operation genuinely breaks the predictive model and produces inexhaustible remainder, the method of breaking does not matter.

The narrative chisel-construct cycle has one predecessor with whom dialogue is necessary: Aristotle's Poetics.

Twenty-three hundred years ago, Aristotle analyzed the structure of Greek tragedy and proposed several core concepts: peripeteia — a sudden reversal of fortune; anagnorisis — a critical recognition or discovery; katharsis — the emotional purgation the audience undergoes at the tragedy's conclusion.

The correspondence between these concepts and Arise-Settle-Unfold-Fix is nearly direct: peripeteia and anagnorisis are Unfold — they break the audience's predictive model of the story's direction. Katharsis is Fix — the emotional closure following the breaking.

But Aristotle's framework has two limitations.

First, he addresses only the narrative channel. The Poetics discusses plot structure; it does not address music, visuals, bodily movement, or other channels. He can therefore explain why theatre is powerful, but cannot explain the multi-level chisel of film — why the "Star Gate" sequence of 2001: A Space Odyssey produces enormous impact even as narrative has nearly halted. In Aristotle's framework, without peripeteia in the plot there is no power. In Arise-Settle-Unfold-Fix, narrative is only one channel; chisel can occur in any channel.

Second, he has no concept of remainder. Katharsis is a one-time event — you watch the tragedy, emotion is released, done. But this cannot explain why some works can be viewed repeatedly without diminished katharsis — if katharsis is a one-time release, the second viewing should have no effect. "Remainder" explains this: good works expose, with each viewing, new unabsorbed material, so the content of Fix differs each time. Katharsis is not a one-time release but, with each closure, an integration of newly exposed remainder.

Aristotle's position: within the narrative channel he precisely described the operations of Unfold and Fix (peripeteia / anagnorisis / katharsis), but stopped at a single channel and a single experience. The present essay's position: extending Aristotle's insight to all channels (chisel is not limited to narrative) and introducing remainder to explain why good works sustain repeated experience.

As with Hanslick/Meyer in Essay I, Wagner in Essay II, and Laban in Essay III: a predecessor grasped one leg of the elephant. The present essay attempts to assemble the whole animal.

Same director, comparable production capability, but entirely different depth of remainder.

Interstellar: the narrative chisel-construct cycle is complete — the father-daughter relationship builds a powerful emotional construct (Arise-Settle); the black hole and time dilation break your predictive model of "rescue" (Unfold); the final "bookshelf scene" closes everything at a location you did not expect but in retrospect feel was necessary (Fix). Moreover, the music (Hans Zimmer's organ) performs precise cross-channel chisel with the narrative — musical emotional intensity often exceeds the narrative's current logic, producing phase difference between channels.

Still yields something after ten viewings. Because the emotional remainder of the father-daughter relationship and the cognitive remainder of time/space each reveal new facets with each viewing.

Tenet: technically equally accomplished, the concept of time inversion equally stunning, yet on second viewing you feel something is missing. Why?

Because Tenet's primary Unfold is conceptual — the time-inversion premise is itself the greatest breaking. On first viewing your predictive model is completely overwhelmed; you spend your entire cognitive budget on understanding "what is happening." But on second viewing, you already understand the mechanism; this conceptual breaking is no longer effective. And beneath the concept, the narrative emotional construct (the protagonist's relationship with his partner, the protagonist's motivation) is relatively thin — squeezed out by conceptual complexity.

In the language of this essay: Tenet's Unfold is single-pass — conceptual breaking is fully absorbed on the first viewing. Interstellar's Unfold is multi-level — conceptual breaking (time dilation) and emotional breaking (the price of the father-daughter relationship) operate on different channels, each with its own remainder, and the phase difference between levels produces additional remainder.

This is how pseudo-chisel and genuine chisel manifest in specific works. Tenet is not bad; its chisel is concentrated on a single level, and that level is exhausted after one pass. Interstellar's chisel is distributed across multiple levels, each with remainder that cannot be exhausted in a single pass.

To be fair, early Marvel had several works with relatively complete chisel-construct cycles — the first Iron Man features Tony Stark's character arc from arms dealer to hero as a genuine construct-to-chisel-to-traced-closure, and Captain America: The Winter Soldier's political-thriller structure performs genuine Unfold at the narrative level. These works bear repeated viewing because character change and narrative breaking produce real remainder.

But as the series expanded, a degradation pattern became increasingly pronounced. Later Marvel films' structure tends toward: establish the hero's initial state (Arise), confirm abilities and mission (Settle), the villain appears and creates crisis (Unfold?), the hero defeats the villain (Fix).

This appears to be a complete four steps. But why are most later Marvel films sufficient after three viewings?

Because increasingly, what occupies the Unfold position is pseudo-chisel. The villain's appearance does create "surprise," but this surprise does not change your model of the entire narrative framework — you know the hero will ultimately win, you know those who are sacrificed may return in the next installment, you know the universe's rules will not truly be broken. Your predictive model is briefly perturbed, then restores itself unchanged.

The visual level is even more so. Each battle is a visual spectacle, but between spectacles there is no chisel-construct cycle — one explosion follows another, your visual predictive model is constantly in Unfold without ever entering Fix. This is the "Unfold-Unfold-Unfold-Unfold" pure-virtuosity structure described in Essay III.

This degradation curve is itself an argument: within the same series, early works had genuine chisel, later works were increasingly dominated by pseudo-chisel, and audience willingness to re-watch declined correspondingly. Production quality did not decline — the chisel-construct cycle's completeness was sacrificed in the process of industrialization. Construct grew increasingly refined (visual effects increasingly impressive), but chisel grew increasingly scarce (your predictive model was increasingly unlikely to be genuinely broken).

Before concluding, a supplementary form skipped in previous essays: the oral tradition.

The recitation of Homeric epic, Chinese pingshu storytelling, Japanese rakugo, West African griot oral narrative — these are the oldest forms of "narrative as primary channel." They are the ancestors of theatre and film.

The oral tradition's specificity lies in this: it has only two channels — voice and narrative (no stage design, no editing, no score) — yet its chisel-construct cycle is equally complete.

The Chinese pingshu storyteller's "hook" (扣子, suspense setup) is Arise and Settle — building your prediction of the story's direction. "Flipping the bundle" (翻包袱, surprise reveal) is Unfold — breaking prediction, exposing remainder. The final gathering (收束) is Fix.

Japanese rakugo is even more extreme — a single person sitting, without props, relying only on vocal variation (shifting character voices, alternating fast and slow rhythms) to run the chisel-construct cycle. It proves one thing: even with channels compressed to a minimum (only voice + narrative), the cycle can still run completely.

This is symmetrical with Essay III's argument that "dance holds even in near-silence": dance proved that removing sound leaves the cycle intact; the oral tradition proves that removing visuals leaves the cycle intact. Channels can increase or decrease; the cycle does not change.

Four essays are complete. Return to the opening question.

Why do some melodies still have something after a hundred listens? Why do some plays still reveal new things after ten viewings? Why does some dance make you feel "the ending weighs more than the beginning"? Why can some films be watched repeatedly without exhaustion?

The answer is the same:

This criterion crosses:

Sensory channels — applies to audition (music), kinesthesia (dance), vision + narrative (film) Cultural traditions — applies to Beethoven and to K-pop, to Peking opera and to street dance Historical periods — applies to Bach and to Kendrick Lamar, to Noh theater and to Kubrick Media complexity — applies to single-channel (pure instrumental music) and to full-media (film)

Because it is not the product of any specific encoding system. It is the basic operation of the human cognitive system: model → confirm → model broken → rebuild carrying trace.

Music theory is this operation's encoding in sound. Dance vocabulary is this operation's encoding in the body. Narrative structure is this operation's encoding in story. Each encoding system is sophisticated and unique, but what they encode is the same thing.

Arise-Settle-Unfold-Fix is taken directly from the Life Cycle Table of the Self-as-an-End framework. Music moves you, dance holds your breath, film keeps you motionless long after the credits — all because they replicate, in their respective channels, the rhythm of life itself: selection, construction, being chiseled open, re-closure.

This series does not discuss music theory. It discusses the rationale behind music theory.

It does not discuss dance technique. It discusses the rationale behind technique.

It does not discuss narrative theory. It discusses the rationale behind narrative theory.

The rationale is one.

One final point: a prediction that follows naturally from the framework. Temporal arts have no ceiling, no terminus.

The reason is not the platitude "human creativity is infinite." The reason is that the structure of the chisel-construct cycle itself guarantees it cannot terminate.

Any form, once familiar to an audience, becomes construct. Once construct stabilizes, it can be chiseled. Chiseling exposes remainder. Remainder drives the formation of new construct. New construct is chiseled again.

Sonata form was construct; Schoenberg chiseled it. Twelve-tone technique became new construct; Cage chiseled it. 4'33" became construct; the next person will chisel it. Peking opera's modal framework was construct; Mei Lanfang performed micro-chisel within it. Mei Lanfang's style itself became construct; the next generation will chisel it. Classical ballet was construct; Pina Bausch chiseled it. Bausch's repetition technique became construct; the next generation of choreographers will chisel it.

There is no terminus. Because remainder — the core concept running throughout the Self-as-an-End framework — can never be fully absorbed. Every closure carries inexhaustible residue, and that residue is the starting point of the next chisel.

This is not optimism. It is structure.