SAE 经济学系列

SAE Economics Series

经济学漏掉了什么 What Economics Left Out

理性不止一层,效用函数装不下底线 Rationality has four layers; the utility function cannot hold a bottom line

DOI: 10.5281/zenodo.19358010  ·  学术原文 ↗Full Paper ↗

经济学有一个最有力的假设,也有一个最深的盲区。

最有力的假设是:人是理性的。给定选项,给定偏好,人会选效用最高的那个。这个假设撑起了价格理论、博弈论、宏观模型——绝大部分现代经济学的骨架。它不是错的。在大量的真实市场里,它非常准确。

盲区在于:这个"理性"只是理性的一个层级。

四层,不是一层

维度序列理论(SAE框架)区分了四种结构性不同的理性,对应四个认知层级:

层级操作经济学对应
12DD利益最大化计算标准理性人
13DD不确定性监控,"我可能是错的"行为经济学修正
14DD"我不得不",有不可交换之物sacred values, identity
15DD"我们各自不得不,所以"——在双方底线间产出新方案目前无对应概念

经济学的理性人是12DD的。Kahneman的行为经济学修正的是13DD的执行偏差。但14DD和15DD几乎在主流经济学的词汇表里消失了——不是因为它们不存在,而是因为语言装不下它们。

翻译损耗

当一个人拒绝出售某样东西"无论出价多高",经济学只有一种翻译:他对它的效用评估极高。但这个翻译损耗了关键的东西——那个人说的不是"价格不够高",而是"这个操作不存在"。这两句话描述的是完全不同的现实。

当两个谈判者产出了一个双方事先都没有的新方案,经济学翻译为:他们在更大的方案空间中找到了帕累托改进。但那个新方案不是被"找到"的——它是在双方各守底线的前提下被创造出来的。C不是妥协,不是折中,不是在已有选项里选个大家都能接受的。C是一个进入谈判前不存在的东西。

翻译不算错,但损耗是真实的。

诊断

经济学把大量14DD/15DD行为判定为"非理性"——不卖是非理性,no deal是谈判失败,不愿妥协是固执。但从14DD的视角看,这些恰恰是理性的:守底线是因为那不是可交换的东西,no deal是因为C没有出现,不妥协是因为妥协意味着交出不可交出的。

经济学所谓的大量"非理性行为",是更高层级的理性被低层级框架系统性误读的产物。

Kant的目的王国(Kingdom of Ends)说的是:每一个理性存在者都同时是目的,不只是手段。经济学的核心语法描述的是手段王国(Kingdom of Means):一切价值可以被表达为偏好,一切偏好可以被交换,一切主体互为工具。

手段王国不是贬义词——它准确描述了大量真实的经济活动。问题在于,它不是全部。剩下的那部分,就是本系列要展开的。

Economics has its most powerful assumption — and its deepest blind spot.

The assumption: people are rational. Given options and preferences, people choose the highest utility. This assumption supports price theory, game theory, macroeconomic models — the skeleton of most of modern economics. It isn't wrong. In vast stretches of real markets, it's very accurate.

The blind spot: this "rationality" is only one layer of rationality.

Four Layers, Not One

Dimensional Sequence Theory (the SAE framework) distinguishes four structurally different kinds of rationality, corresponding to four cognitive layers:

LayerOperationEconomics analog
12DDutility maximizationstandard rational agent
13DDuncertainty monitoring, "I might be wrong"behavioral economics corrections
14DD"I cannot not" — having non-tradeable commitmentssacred values, identity economics
15DD"we each cannot not, therefore" — producing new solutions between two sets of bottom linesno current concept

Economics describes a 12DD rational agent. Kahneman's behavioral economics corrects 13DD execution biases. But 14DD and 15DD have largely vanished from mainstream economics vocabulary — not because they don't exist, but because the language cannot hold them.

Translation Loss

When someone refuses to sell something "no matter how high the price," economics has only one translation: they place extremely high utility on it. But this translation loses the crucial thing — that person isn't saying "the price isn't high enough," they're saying "this transaction does not exist." These two sentences describe entirely different realities.

When two negotiators produce a solution neither had before entering the room, economics translates: they found a Pareto improvement in a larger solution space. But that new solution wasn't "found" — it was created from the premise that both parties held their ground. C is not compromise, not splitting the difference, not selecting from existing options. C is something that didn't exist before the negotiation began.

The translation isn't wrong. But the loss is real.

Diagnosis

Economics labels large amounts of 14DD/15DD behavior as "irrational" — refusing to sell is irrational, no-deal is negotiation failure, unwillingness to compromise is stubbornness. But from the 14DD perspective, these are precisely rational: holding the bottom line because it is not tradeable, no-deal because C hasn't appeared, not compromising because compromise means surrendering the unsurrenderable.

Much of what economics calls "irrational behavior" is higher-layer rationality being systematically misread by a lower-layer framework.

Kant's Kingdom of Ends says: every rational being is simultaneously an end, not merely a means. Economics' core grammar describes a Kingdom of Means: all value can be expressed as preference, all preferences can be traded, all agents are tools for each other.

Kingdom of Means isn't a slur — it accurately describes enormous amounts of real economic activity. The problem is that it isn't all of it. The remainder is what this series unfolds.