Concise Summary简洁概述
Yudkowsky opens with the classic puzzle: does a falling tree make a sound if no one hears it? His diagnosis is that the two arguers anticipate exactly the same experiences — neither predicts any different observation — so they are arguing over labels, not facts. He generalizes: beliefs earn their keep by constraining anticipated experience. Beliefs that predict no particular experience, and forbid no particular experience, are "floating" — connected only to each other, not to reality. The rationalist virtue of empiricism demands you always ask: what do I expect to see because of this belief? Better still: what must not happen if this belief is true? If no experience would falsify a belief, it floats — and should be evicted.
Yudkowsky 以一个经典谜题开篇：一棵树在森林里倒下，没有人听见，它发出声音了吗？他的诊断是：争论双方预期的经验完全相同——谁都没有预测任何不同的观测结果——所以他们争的是标签，不是事实。他由此推广：信念靠限定预期经验来证明自身的价值。不预测任何特定经验、也不禁止任何特定经验的信念是「漂浮的」——只与其他信念相互关联，而不与现实挂钩。经验主义的理性美德要求你始终追问：我因为这个信念，预期会看到什么？更好的问法是：如果这个信念为真，什么经验不可以发生？若没有任何经验能推翻一个信念，它就在漂浮——应当被驱逐出门。
Infographic信息图
Beliefs must pay rent
信念必须缴租
Every belief should cash out in anticipated experiences; one that never constrains what you expect is a deadbeat tenant to be evicted.
每个信念都应兑现为预期经验；从不限制你预期的信念是一个拖欠租金的房客，应当被驱逐。
The tree-sound argument is hollow
树倒发声的争论是空洞的
Both sides predict the same recorder output, the same EEG trace, the same visual evidence — so they disagree only about a word, not the world.
两方预测的录音结果、脑电图轨迹、视觉证据完全相同——他们的分歧只在于一个词，而非世界本身。
Propositional beliefs earn their keep via inference
命题性信念通过推论证明价值
Believing "gravity = 9.8 m/s²" is not itself a sensory anticipation, but it implies precise ones — so it legitimately earns its place in the map.
相信「重力加速度=9.8米/秒²」本身不是感官预期，但它蕴含精确的感官预期——因此它在地图中的位置是正当的。
Floating belief networks
漂浮的信念网络
Phlogiston and "retropositional author" are beliefs hooked only to each other, never to any observable difference — whole webs of meaning with no anchor in experience.
「燃素」和「逆位作者」是只与彼此挂钩、永远与任何可观测差异无关的信念——整张意义之网没有任何经验锚点。
Ask what the belief forbids
追问信念禁止什么
The sharpest empiricist question is not "what do I expect?" but "what must NOT happen?" — a belief that forbids nothing constrains nothing.
最锐利的经验主义问题不是「我预期什么」，而是「什么不应该发生」——一个什么都不禁止的信念什么都不约束。
Detailed Summary详细概述
The Tree-Sound Puzzle as Diagnostic Tool
Yudkowsky opens with the ancient riddle: if a tree falls and no one hears it, does it make a sound? One person says yes — vibrations in the air. The other says no — no auditory processing in any brain. Yet if you walk the forest afterward, both expect to see the tree fallen in the same place. If you set up a sound recorder, both expect to hear the same playback. Attach an EEG to any brain — both expect the same trace. They disagree on the label, but anticipate no different experience. This is the tell: they have no real difference in world-models; they are arguing about words.
Beliefs as Maps With Inferential Consequences
It would be too simple to say that only direct sensory anticipations count as beliefs. Abstract beliefs are real and necessary. You cannot see Earth's gravity, but believing it is 9.8 m/s² allows you to predict precisely on which clock-tick a bowling ball dropped from a 120-meter building will hit the ground. That abstract belief has an inferential chain that terminates in a specific sensory anticipation. It earns its place in the map by paying rent downstream.
Floating Beliefs: Phlogiston and Literary Jargon
Trouble arises when a belief's link to experience is broken. The alchemists' phlogiston looked like it explained fire — you could draw an arrow from the "Phlogiston" node to the experience of a crackling campfire. But the link was always constructed after the experience, never constraining it in advance; the belief yielded no advance predictions. It floated.
Yudkowsky's other example is equally sharp: suppose a professor says the writer Wulky Wilkinsen is a "retropositional author" characterized by "alienated resublimation." "Resublimation" is defined by its connection to "retropositionality," and vice versa. What do you expect to see in Wilkinsen's books? Nothing specific. The labels connect to each other — forming a coherent-seeming web — but the web has no anchor in experience. You can memorize it and pass a quiz, but the beliefs float.
The Empiricist Test — and Its Sharpest Form
The rationalist virtue of empiricism, as Yudkowsky frames it, means constantly demanding that every belief cash out in anticipated experiences. Ask: What do I expect to happen, because of this belief? But the sharpest form of the test is negative: What must NOT happen? A belief that is compatible with any possible experience constrains nothing. It permits everything. It floats.
If you believe in élan vital as the explanation of life, what experience would definitely falsify it? A null answer reveals a null belief. If you cannot say what would count against phlogiston, the belief is barnacle, not structure.
Label Arguments and the Prescription
Yudkowsky closes with a practical rule for disputes: always identify the difference of anticipation you are arguing about. If no such difference exists, you are arguing labels — and should notice that and stop, or reframe. The final injunction is vivid: do not ask what to believe — ask what to anticipate. Every belief should begin and continue by paying rent in specific predictions. If a belief turns deadbeat, evict it.
树倒发声之谜作为诊断工具
Yudkowsky 以那个古老谜题开篇：树在没人听见的情况下倒下，它发出声音了吗？一个人说有——空气里的振动。另一个人说没有——没有任何大脑在做听觉处理。然而，如果事后走进森林，两人预期看到树倒在同一个地方。如果架一台录音机，两人预期听到同样的回放。给任何一个大脑接上脑电图——两人预期看到同样的轨迹。他们在标签上意见相左，却对任何经验没有不同预期。这正是症状所在：他们对世界模型没有真正分歧；他们争论的是词语。
信念作为具有推论后果的地图
说只有直接感官预期才算信念，会过于简单化。抽象信念是真实的，也是必要的。你看不见地球的重力，但相信它是9.8米/秒²，就能精确预测从120米高楼丢下一个保龄球，会在时钟秒针指到哪一格时听到撞击声。这个抽象信念有一条推论链，终止于一个具体的感官预期。它通过在下游缴租，赢得了在地图中的位置。
漂浮的信念：燃素与文学行话
当信念与经验之间的链接断裂，麻烦就来了。炼金术士的「燃素」看起来解释了火——你可以从「燃素」节点画一条箭头指向篝火噼啪作响的经验。但这条链接总是在经验之后建立，从不事先约束经验；这个信念不产生任何提前预测。它在漂浮。
Yudkowsky 的另一个例子同样犀利：假设一位教授说作家沃基·威尔金森是一位「逆位作者」，以「异化再升华」为特征。「再升华」由其与「逆位性」的关联来定义，反之亦然。你对威尔金森的书预期看到什么？没有任何具体的东西。这些标签彼此关联——形成一张看似连贯的网——但这张网没有任何经验锚点。你可以背诵它并通过测验，但这些信念在漂浮。
经验主义检验——及其最锐利的形式
Yudkowsky 所框定的经验主义理性美德，意味着不断要求每个信念兑现为预期经验。问：因为这个信念，我预期会发生什么？ 但这项检验最锐利的形式是否定式：什么不应该发生？ 一个与任何可能经验都兼容的信念什么都不约束。它允许一切发生。它在漂浮。
如果你相信「生命力」解释了生命，什么经验能确凿地推翻它？无法回答，揭示的是一个空洞的信念。若你说不出什么算是反对燃素的证据，这个信念是藤壶，不是支柱。
标签之争与处方
Yudkowsky 以一条实践规则收尾：在争论中，始终要确定你在争论的预期差异是什么。若不存在这种差异，你争的是标签——应当认出这一点，停下来，或重新框架。最后的告诫生动有力：不要问该相信什么——要问该预期什么。每个信念都应从一开始就持续地以具体预测缴租。若信念开始赖账，驱逐它。
FAQ常见问答
Is the tree-sound argument genuinely meaningless, or does it have a real answer?树倒发声的争论是真的无意义，还是其实有一个正确答案？
Yudkowsky's point is not that the question has no answer, but that the two arguers are not actually disagreeing about anything experiential. Once you specify what you mean — air vibrations vs. auditory processing — both propositions can be true or false independently, and both sides already agree on every observation. The word "sound" is doing all the work, and it can be dropped once you see that.
Yudkowsky 的论点不是说这个问题没有答案，而是说两个争论者在任何经验层面上其实没有分歧。一旦你说清楚你指的是什么——空气振动还是听觉处理——两个命题可以独立地为真或为假，而双方对每项观测结果本来就已达成一致。「声音」这个词承担了所有工作；一旦看清这一点，它就可以被丢弃了。
Does this mean all valid beliefs must be directly observable?这是否意味着所有有效信念都必须是可直接观察的？
No. Yudkowsky is explicit that abstract, propositional beliefs — like "gravity is 9.8 m/s²" or "this building is 120 meters tall" — are legitimate precisely because they generate a chain of inference that terminates in a specific sensory anticipation. The test is not directness; it is whether the belief constrains experience somewhere downstream.
不是。Yudkowsky 明确指出，抽象的命题性信念——如「重力加速度为9.8米/秒²」或「这栋楼有120米高」——之所以正当，恰恰是因为它们产生一条推论链，在下游终止于一个具体的感官预期。检验标准不是直接性，而是信念是否在下游某处约束了经验。
What exactly are floating beliefs, and why are they a problem?「漂浮的」信念究竟是什么，为什么它是个问题？
Floating beliefs are connected to other beliefs but not — even indirectly — to any anticipated experience. Phlogiston "explains" fire but forbids no observation; "retropositional" defines "resublimation" and vice versa. The problem is not that they are invisible — gravity is invisible — but that they yield zero advance predictions and can be neither confirmed nor disconfirmed by any experience. They are semantic dead weight.
漂浮的信念与其他信念相连，但——即便是间接地——也不与任何预期经验相连。燃素「解释」了火，却不禁止任何观测；「逆位性」定义了「再升华」，反之亦然。问题不在于它们不可见——重力也不可见——而在于它们不产生任何提前预测，无法被任何经验证实或否定。它们是语义上的死重。
What is the "what must NOT happen" test, and why is it sharper than "what do I expect"?「什么不应该发生」的检验是什么，为何它比「我预期什么」更锐利？
Asking what you expect can be satisfied by listing any experience loosely consistent with a belief. Asking what must not happen forces the belief to actually exclude possibilities. A belief with no forbidden experiences is compatible with any world — it has zero content. Falsifiability is the sharper edge of the empiricist blade.
问你预期什么，随便列举任何与信念松散相符的经验就能敷衍过去。问什么不应该发生，则迫使信念真正排除某些可能性。一个没有任何被禁止经验的信念，与任何世界都相容——它毫无内容。可证伪性是经验主义这把刀最锋利的刃口。
How does this apply to everyday disputes that seem factual?这如何应用于日常中看似事实性的争论？
Yudkowsky recommends always asking: what difference of anticipation are we arguing about? If you and someone else disagree about a seemingly factual claim, trace the chain: what would each of you expect to see if you were right? If no difference of observation exists, you are arguing labels. Identifying that typically dissolves the argument or reveals the real question.
Yudkowsky 建议始终追问：我们争的是哪种预期差异？如果你和某人对一个看似事实的主张意见相左，请追溯链条：如果你是对的，你们各自预期看到什么？若不存在任何观测差异，你们争的是标签。认清这一点，通常能化解争论，或揭示真正的问题所在。
Is pay rent just a colorful way of saying be falsifiable (Popper)?「缴租」只是「可证伪」（波普尔）的花哨说法吗？
There is significant overlap, but Yudkowsky's framing is subtly broader. Popper's falsifiability focuses on scientific theories and their potential refutation. Yudkowsky's rent applies to every belief in everyday cognition — even a single conversational claim must anticipate some experience. He also emphasizes the positive side: beliefs can earn rent by generating advance predictions, not only by being falsifiable in principle.
两者有很大重叠，但 Yudkowsky 的框架微妙地更宽泛。波普尔的可证伪性聚焦于科学理论及其潜在反驳。Yudkowsky 的「租金」适用于日常认知中的每一个信念——即便是随口说的一句话，也必须对某种经验有所预期。他还强调了积极的一面：信念可以通过产生提前预测来缴租，而不仅仅是原则上可被证伪。
In-depth Analysis · Pros & Cons深入解读 · 优缺点
This essay is part of the Mysterious Answers sequence and introduces empiricism as a practical cognitive tool rather than a philosophical doctrine. Its target is not skepticism but a specific failure mode: beliefs that have become semantically self-contained, disconnected from the world they purport to describe.
这篇文章是《神秘答案》系列的一部分，它将经验主义作为一种实践的认知工具而非哲学教条来介绍。它针对的不是怀疑主义，而是一种特定的失败模式：信念变得语义上自给自足，与它们声称描述的世界脱节。
- The tree-sound opener is a perfectly chosen diagnostic树倒发声的开场是完美的诊断工具It is familiar, non-threatening, and immediately reveals the problem — both sides predict the same world — without needing to invoke abstract philosophy. The example does real work rather than just illustrating.它熟悉、无威胁性，且立即揭示了问题所在——双方预测的是同一个世界——无需援引抽象哲学。这个例子真正在做工，而不仅仅是在图解。
- The "forbids" test is a genuine advance over "predicts"「禁止」检验是对「预测」检验的真正推进Asking what a belief prohibits — rather than what it predicts — is a sharper tool, equivalent to Popper's falsifiability but applied to any belief, not just scientific theories. This is practical and teachable.询问一个信念禁止什么——而非它预测什么——是一个更锐利的工具，等价于波普尔的可证伪性，但适用于任何信念，而不仅仅是科学理论。这是可实践、可教授的。
- The distinction between floating and inferential beliefs is precise漂浮信念与推论信念的区分是精确的By contrasting phlogiston (broken link) with gravity (inferential chain to sensation), Yudkowsky shows that abstraction is not itself the problem — it is disconnection from eventual experience that disqualifies a belief.通过对比燃素（断链）与重力（通向感觉的推论链），Yudkowsky 表明抽象本身不是问题——使信念失格的是与最终经验的断连。
- The "evict" metaphor makes the prescription memorable and actionable「驱逐」的隐喻让处方令人难忘且可付诸行动Framing beliefs as tenants that must pay rent in anticipated experiences gives readers a concrete mental action — not just a philosophical stance, but a practice they can apply to their own belief inventory.将信念框架为必须以预期经验缴租的租户，给读者提供了一个具体的心理行动——不只是一种哲学立场，而是一种可应用于自己信念库存的实践。
- The scope of experience is left underspecified「经验」的范围未被明确界定The essay never defines what counts as an experience or how indirect is too indirect. Phlogiston is condemned, but gravity is approved — yet both involve multi-step inference chains. Without a principled criterion for how long the inferential chain may be, the test cannot be applied rigorously.文章从未定义什么算作「经验」，或间接到什么程度算太间接。燃素被谴责，重力被批准——但两者都涉及多步推论链。若没有关于推论链可以多长的原则性标准，这项检验就无法严格应用。
- The literary-jargon example may prove too much文学行话的例子可能证明得过头了Yudkowsky dismisses "retropositional" as meaningless because it has no experiential cash-out. But much of mathematics, logic, and theoretical physics involves concepts equally remote from direct sensation yet clearly not mere floating labels. The essay does not explain where the line falls.Yudkowsky 以没有经验兑现为由，将「逆位性」斥为无意义。但许多数学、逻辑和理论物理的概念同样远离直接感觉，却显然不仅仅是漂浮的标签。文章没有解释界线落在哪里。
- The framing conflates semantics with epistemology框架混淆了语义学与认识论The tree-sound dispute is arguably a semantic disagreement about the word sound, not an epistemic one. Yudkowsky treats it as a failure of empiricist discipline, but arguably the real failure is equivocation — a logic issue, not a belief-pays-rent issue. The two problems are related but not identical.树倒发声的争论可以说是一个语义分歧（关于「声音」这个词），而非认识论分歧。Yudkowsky 将其视为经验主义纪律的失败，但可以说真正的失败是歧义谬误——这是一个逻辑问题，不是信念缴租问题。两个问题相关，但并不相同。
- No guidance on genuinely theoretical disagreements对真正的理论分歧没有指导Real scientific disputes — string theory, interpretations of quantum mechanics — involve beliefs that currently lack direct empirical settlement but are not obviously floating. The essay's binary (earns rent / floats) has no bucket for beliefs that are not yet empirically settled but are meaningfully constrained.真正的科学争论——弦理论、量子力学诠释——涉及目前缺乏直接经验解决但又显然不是漂浮的信念。文章的二元分类（缴租/漂浮）没有为「尚未经验性确定但有实质性约束」的信念留出一个桶。
A lean, precise essay that delivers a tool every reader can immediately use: the "what does this belief forbid?" test. Its central insight — that floating beliefs are a uniquely human pathology arising from our capacity for abstract thought — is genuine and important. The weaknesses are real but largely downstream problems; for an introductory essay establishing a core empiricist habit, the compression is appropriate.
一篇精简、精确的文章，提供了一个每位读者都能立即使用的工具：「这个信念禁止什么？」检验。它的核心洞见——漂浮的信念是人类特有的病理，源于我们抽象思维的能力——是真实且重要的。文章的弱点是真实存在的，但大多是下游问题；作为一篇建立核心经验主义习惯的入门文章，这种压缩是恰当的。
Original Text原文
Thus begins the ancient parable:
If a tree falls in a forest and no one hears it, does it make a sound? One says, “Yes it does, for it makes vibrations in the air.” Another says, “No it does not, for there is no auditory processing in any brain.”
If there’s a foundational skill in the martial art of rationality, a mental stance on which all other technique rests, it might be this one: the ability to spot, inside your own head, psychological signs that you have a mental map of something, and signs that you don’t.
Suppose that, after a tree falls, the two arguers walk into the forest together. Will one expect to see the tree fallen to the right, and the other expect to see the tree fallen to the left? Suppose that before the tree falls, the two leave a sound recorder next to the tree. Would one, playing back the recorder, expect to hear something different from the other? Suppose they attach an electroencephalograph to any brain in the world; would one expect to see a different trace than the other?
Though the two argue, one saying “No,” and the other saying “Yes,” they do not anticipate any different experiences. The two think they have different models of the world, but they have no difference with respect to what they expect will happen to them; their maps of the world do not diverge in any sensory detail.
It’s tempting to try to eliminate this mistake class by insisting that the only legitimate kind of belief is an anticipation of sensory experience. But the world does, in fact, contain much that is not sensed directly. We don’t see the atoms underlying the brick, but the atoms are in fact there. There is a floor beneath your feet, but you don’t experience the floor directly; you see the light reflected from the floor, or rather, you see what your retina and visual cortex have processed of that light. To infer the floor from seeing the floor is to step back into the unseen causes of experience. It may seem like a very short and direct step, but it is still a step.
You stand on top of a tall building, next to a grandfather clock with an hour, minute, and ticking second hand. In your hand is a bowling ball, and you drop it off the roof. On which tick of the clock will you hear the crash of the bowling ball hitting the ground?
To answer precisely, you must use beliefs like Earth’s gravity is 9.8 meters per second per second, and This building is around 120 meters tall. These beliefs are not wordless anticipations of a sensory experience; they are verbal-ish, propositional. It probably does not exaggerate much to describe these two beliefs as sentences made out of words. But these two beliefs have an inferential consequence that is a direct sensory anticipation—if the clock’s second hand is on the 12 numeral when you drop the ball, you anticipate seeing it on the 1 numeral when you hear the crash five seconds later. To anticipate sensory experiences as precisely as possible, we must process beliefs that are not anticipations of sensory experience.
It is a great strength of Homo sapiens that we can, better than any other species in the world, learn to model the unseen. It is also one of our great weak points. Humans often believe in things that are not only unseen but unreal.
The same brain that builds a network of inferred causes behind sensory experience can also build a network of causes that is not connected to sensory experience, or poorly connected. Alchemists believed that phlogiston caused fire—we could simplistically model their minds by drawing a little node labeled “Phlogiston,” and an arrow from this node to their sensory experience of a crackling campfire—but this belief yielded no advance predictions; the link from phlogiston to experience was always configured after the experience, rather than constraining the experience in advance.
Or suppose your English professor teaches you that the famous writer Wulky Wilkinsen is actually a “retropositional author,” which you can tell because his books exhibit “alienated resublimation.” And perhaps your professor knows all this because their professor told them; but all they're able to say about resublimation is that it's characteristic of retropositional thought, and of retropositionality that it's marked by alienated resublimation. What does this mean you should expect from Wulky Wilkinsen’s books?
Nothing. The belief, if you can call it that, doesn’t connect to sensory experience at all. But you had better remember the propositional assertions that “Wulky Wilkinsen” has the “retropositionality” attribute and also the “alienated resublimation” attribute, so you can regurgitate them on the upcoming quiz. The two beliefs are connected to each other, though still not connected to any anticipated experience.
We can build up whole networks of beliefs that are connected only to each other—call these “floating” beliefs. It is a uniquely human flaw among animal species, a perversion of Homo sapiens’s ability to build more general and flexible belief networks.
The rationalist virtue of empiricism consists of constantly asking which experiences our beliefs predict—or better yet, prohibit. Do you believe that phlogiston is the cause of fire? Then what do you expect to see happen, because of that? Do you believe that Wulky Wilkinsen is a retropositional author? Then what do you expect to see because of that? No, not “alienated resublimation”; what experience will happen to you? Do you believe that if a tree falls in the forest, and no one hears it, it still makes a sound? Then what experience must therefore befall you?
It is even better to ask: what experience must not happen to you? Do you believe that Élan vital explains the mysterious aliveness of living beings? Then what does this belief not allow to happen—what would definitely falsify this belief? A null answer means that your belief does not constrain experience; it permits anything to happen to you. It floats.
When you argue a seemingly factual question, always keep in mind which difference of anticipation you are arguing about. If you can’t find the difference of anticipation, you’re probably arguing about labels in your belief network—or even worse, floating beliefs, barnacles on your network. If you don’t know what experiences are implied by Wulky Wilkinsens writing being retropositional, you can go on arguing forever.
Above all, don’t ask what to believe—ask what to anticipate. Every question of belief should flow from a question of anticipation, and that question of anticipation should be the center of the inquiry. Every guess of belief should begin by flowing to a specific guess of anticipation, and should continue to pay rent in future anticipations. If a belief turns deadbeat, evict it.
古老的寓言从这里开始：
若一棵树在森林里倒下，没有人听见，它发出声音了吗？一人说：「有，因为它在空气中产生了振动。」另一人说：「没有，因为没有任何大脑在进行听觉处理。」
如果说理性这门武术有什么根基性技能，一种所有其他技法都依赖的心理姿态，或许正是这个：察觉自己头脑里的心理信号——你是否真的在心里有某件事的地图，还是没有。
假设树倒之后，两个争论者一起走进森林。一个会预期看到树向右倒，另一个会预期看到树向左倒吗？假设在树倒之前，两人在树旁放了一台录音机。回放录音时，一个预期听到的声音会与另一个不同吗？假设他们给世界上任何一个大脑接上脑电图，一个预期看到的脑电轨迹会与另一个不同吗？
尽管两人各执一词，一个说「没有」，另一个说「有」，他们却不预期任何不同的经验。两人以为自己对世界有不同的模型，但他们对于将要发生在自己身上的事没有任何分歧；他们的世界地图在任何感官细节上都没有差异。
人们很容易想用这样的方法来消除这类错误：坚持认为唯一正当的信念就是对感官经验的预期。但世界上确实存在大量无法被直接感知的东西。我们看不见砖块背后的原子，但原子确实在那里。你脚下有一块地板，但你并不直接体验地板；你看见的是从地板反射的光，更准确地说，你看见的是视网膜和视觉皮层对那束光的处理结果。从「看见地板」推断「地板存在」，是退后一步，进入经验背后那些不可见的原因。这看起来也许是非常短暂、非常直接的一步，但它仍然是一步。
你站在一栋高楼的顶部，旁边有一座有时针、分针和跳动秒针的祖父钟。你手里拿着一个保龄球，从楼顶丢下去。保龄球撞击地面时，时钟的秒针会指到哪一格？
要精确地回答这个问题，你必须用到诸如地球重力加速度约为9.8米每秒的平方，以及这栋楼大约120米高这样的信念。这些信念不是对感官经验的无言预期；它们是语言性的、命题式的。把这两个信念描述为由文字构成的句子，大概不算太夸张。但这两个信念有一个推论性的后果，那是一个直接的感官预期——如果你丢下保龄球时秒针指向12，你预期五秒后听到撞击声时秒针会指向1。为了尽可能精确地预期感官经验，我们必须处理那些本身并非感官经验之预期的信念。
智人的一大优势，正是我们能比世界上任何其他物种都更好地学习对不可见事物建模。这也是我们的一个重大弱点。人类常常相信那些不仅不可见、而且根本不存在的东西。
同一个能在感官经验背后构建推论原因网络的大脑，也能构建出与感官经验毫无关联、或关联薄弱的原因网络。炼金术士相信燃素导致了火——我们可以用一个简单的模型来描述他们的心智：画一个标着「燃素」的小节点，再从这个节点画一条箭头，指向他们感官上感受到的篝火噼啪声——但这个信念没有产生任何提前预测；从燃素到经验的连接，总是在经验发生之后才被建立，而从未提前约束经验。
或者假设你的英语教授告诉你，著名作家沃基·威尔金森实际上是一位「逆位作者」，你可以从他的书表现出「异化再升华」中看出这一点。也许你的教授知道这些，是因为他们的教授告诉了他们；但关于「再升华」，他们能说的只有：它是逆位思维的特征；而关于「逆位性」，它以异化再升华为标志。这意味着你对沃基·威尔金森的书应该有什么预期？
什么都没有。这个信念，如果能叫做信念的话，根本不与感官经验相连。但你最好记住「沃基·威尔金森」具有「逆位性」属性，同时也具有「异化再升华」属性，这样你才能在即将到来的测验中将它们背出来。这两个信念彼此相连，但仍然与任何预期经验都不相连。
我们可以建立整张只与彼此相连的信念网络——姑且称之为「漂浮的」信念。这是动物物种中一种独属于人类的缺陷，是智人建立更通用、更灵活信念网络这一能力的变态。
经验主义的理性美德，在于不断追问我们的信念预测了哪些经验——或者更好，禁止了哪些经验。你相信燃素是火的原因吗？那么你因此预期会看到什么发生？你相信沃基·威尔金森是一位逆位作者吗？那么你因此预期会看到什么？不，不是「异化再升华」；什么经验将要降临到你身上？ 你相信若树在森林里倒下、无人听见，它仍然发出了声音吗？那么因此，什么经验必然将降临到你身上？
更好的问法是：什么经验不可以降临到你身上？你相信生命力解释了生命体神秘的生命性吗？那么这个信念不允许发生什么——什么将确凿地推翻这个信念？若回答为空，意味着你的信念并不约束经验；它允许任何事降临到你身上。它在漂浮。
当你争论一个看似事实性的问题时，始终要牢记你争论的是哪种预期差异。如果你找不到预期差异，你很可能在争论信念网络中的标签——甚至更糟，是在争论漂浮的信念，你网络上的藤壶。如果你不知道沃基·威尔金森的作品具有「逆位性」意味着什么样的经验，你可以永无止境地争论下去。
最重要的是，不要问该相信什么——要问该预期什么。每一个关于信念的问题，都应当从一个关于预期的问题中流淌而出，而那个关于预期的问题，应当是探究的核心。每一个对信念的猜测，都应当从流向一个具体的预期猜测开始，并应当持续地以未来的预期缴纳租金。若信念开始赖账，驱逐它。