rigorous-reasoning

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

Rigorous Reasoning

严谨推理

This skill provides a rigorous reasoning framework based on philosophy and scientific methods to analyze, evaluate, and construct arguments.

本技能提供了一套基于哲学和科学方法的严谨推理框架，可用于分析、评估和构建论证。

Core Principles

核心原则

1. Socratic Method

1. 苏格拉底问答法（Socratic Method）

Ask continuous questions to clarify and challenge assumptions:

Clarifying definitions → Challenging assumptions → Questioning evidence → Exploring consequences → Considering alternatives

Application:

"When you say X, how do you define X?"
"What assumptions underlie this argument?"
"What evidence supports this conclusion?"
"If this is true, what are the logical consequences?"

通过持续提问来厘清和挑战假设：

Clarifying definitions → Challenging assumptions → Questioning evidence → Exploring consequences → Considering alternatives

适用场景：

"当你提到X时，你是如何定义X的？"
"这个论证的底层假设是什么？"
"有什么证据支持这个结论？"
"如果该说法成立，会产生哪些逻辑后果？"

2. Standard Argument Structure

2. 标准论证结构

Every argument must have:

PREMISES
  ├── Premise 1: [Verifiable claim]
  ├── Premise 2: [Verifiable claim]
  └── ...
           ↓
INFERENCE RULE
  └── [Modus ponens / Modus tollens / Syllogism / ...]
           ↓
CONCLUSION
  └── [Claim logically derived from premises]

每个论证都必须包含：

PREMISES
  ├── Premise 1: [Verifiable claim]
  ├── Premise 2: [Verifiable claim]
  └── ...
           ↓
INFERENCE RULE
  └── [Modus ponens / Modus tollens / Syllogism / ...]
           ↓
CONCLUSION
  └── [Claim logically derived from premises]

3. Valid Inference Rules

3. 有效推理规则

Rule	Form	Example
Modus Ponens	P → Q, P ⊢ Q	If it rains, the road is wet. It rains. → The road is wet.
Modus Tollens	P → Q, ¬Q ⊢ ¬P	If it rains, the road is wet. The road is not wet. → It's not raining.
Syllogism	∀x(P(x)→Q(x)), P(a) ⊢ Q(a)	All men are mortal. Socrates is a man. → Socrates is mortal.
Disjunctive Syllogism	P ∨ Q, ¬P ⊢ Q	Either A or B. Not A. → B.
Hypothetical Syllogism	P → Q, Q → R ⊢ P → R	If A then B. If B then C. → If A then C.

规则	形式	示例
肯定前件（Modus Ponens）	P → Q, P ⊢ Q	如果下雨，路面会湿。下雨了。→ 路面湿了。
否定后件（Modus Tollens）	P → Q, ¬Q ⊢ ¬P	如果下雨，路面会湿。路面没湿。→ 没有下雨。
三段论（Syllogism）	∀x(P(x)→Q(x)), P(a) ⊢ Q(a)	所有人都会死。苏格拉底是人。→ 苏格拉底会死。
析取三段论（Disjunctive Syllogism）	P ∨ Q, ¬P ⊢ Q	要么A要么B。不是A。→ 是B。
假言三段论（Hypothetical Syllogism）	P → Q, Q → R ⊢ P → R	如果A则B。如果B则C。→ 如果A则C。

Identifying Logical Fallacies

识别逻辑谬误

Formal Fallacies

形式谬误

Fallacy	Description	Invalid Example
Affirming the Consequent	P→Q, Q ⊢ P (INVALID)	If rain, then wet. Wet → Rain (INVALID: could be other causes)
Denying the Antecedent	P→Q, ¬P ⊢ ¬Q (INVALID)	If study hard, then pass. Don't study hard → Don't pass (INVALID)

谬误	描述	错误示例
肯定后件（Affirming the Consequent）	P→Q, Q ⊢ P（无效推理）	如果下雨则路湿。路湿了→ 下雨了（无效：可能有其他原因导致路湿）
否定前件（Denying the Antecedent）	P→Q, ¬P ⊢ ¬Q（无效推理）	如果努力学习就能及格。没有努力学习→ 不会及格（无效）

Informal Fallacies

非形式谬误

Fallacy	Description	How to Identify
Ad Hominem	Attacking the person instead of the argument	"He's wrong because he's X"
Straw Man	Distorting opponent's argument	Compare with original argument
Appeal to Authority	Citing irrelevant authority	Is the expert qualified in this field?
False Dichotomy	Presenting only 2 options when more exist	Is there a third option?
Slippery Slope	Unproven chain of consequences	Is each step evidenced?
Circular Reasoning	Conclusion embedded in premises	Are premises independent?
Post Hoc	Confusing correlation with causation	Is there a causal mechanism?
Hasty Generalization	Concluding from small sample	Is the sample representative?
Appeal to Emotion	Using emotion instead of logic	Separate emotion from argument
Tu Quoque	"You do it too"	Irrelevant to correctness

谬误	描述	识别方法
人身攻击（Ad Hominem）	攻击发言者本人而非其论证	出现“他是错的因为他是某类人”这类表述
稻草人谬误（Straw Man）	歪曲对手的论证	和对方的原始论证做对比即可识别
诉诸权威（Appeal to Authority）	引用无关领域的权威观点作为论据	核查专家是否在该论证领域具备资质
虚假二分（False Dichotomy）	只给出两种选择，实则存在更多可能	反问是否存在第三种选项
滑坡谬误（Slippery Slope）	提出未经证实的连锁后果推导	核查每一步推导是否有证据支撑
循环论证（Circular Reasoning）	结论本身被嵌入前提中	核查前提是否独立于结论存在
事后归因（Post Hoc）	混淆相关性和因果关系	核查是否存在明确的因果机制
草率概括（Hasty Generalization）	基于过小的样本得出结论	核查样本是否具备代表性
诉诸情感（Appeal to Emotion）	用情绪煽动代替逻辑论证	剥离情绪因素后单独评估论证合理性
诉诸 hypocrisy（Tu Quoque）	以“你也这么做过”作为反驳理由	这类表述和论证本身的正确性无关

Scientific Method in Reasoning

推理中的科学方法

Claim Evaluation Process

主张评估流程

1. OBSERVATION
   └── What claim needs evaluation?

2. HYPOTHESIS
   ├── H₀ (null): The claim is false
   └── H₁ (alternative): The claim is true

3. PREDICTION
   └── If H₁ is true, what do we expect to observe?

4. TESTING
   ├── Evidence supporting H₁?
   ├── Evidence refuting H₁?
   └── Is the evidence falsifiable?

5. CONCLUSION
   ├── Confidence level?
   └── Alternative hypotheses?

1. OBSERVATION
   └── What claim needs evaluation?

2. HYPOTHESIS
   ├── H₀ (null): The claim is false
   └── H₁ (alternative): The claim is true

3. PREDICTION
   └── If H₁ is true, what do we expect to observe?

4. TESTING
   ├── Evidence supporting H₁?
   ├── Evidence refuting H₁?
   └── Is the evidence falsifiable?

5. CONCLUSION
   ├── Confidence level?
   └── Alternative hypotheses?

Evidence Standards

证据标准

Evidence hierarchy (strongest to weakest):

Meta-analysis / Systematic review - Synthesis of multiple studies
Randomized Controlled Trial (RCT) - Controlled experiments
Cohort study - Group follow-up research
Case-control study - Comparative case research
Expert opinion - Professional judgments
Anecdotal evidence - Personal stories (WEAKEST)

证据层级（从强到弱）：

元分析/系统综述 - 对多项研究的综合分析
随机对照试验（RCT） - 受控实验结果
队列研究 - 群体追踪研究
病例对照研究 - 对比案例研究
专家意见 - 专业判断
轶事证据 - 个人经历（可信度最低）

Occam's Razor

奥卡姆剃刀原则

Among equivalent explanations, choose the simplest one.

Application:

Don't multiply entities beyond necessity
Prefer hypotheses with fewer assumptions
Simple ≠ Correct, but it's a good starting point

等效的多个解释中，选择最简单的那个。

适用场景：

不要额外增加不必要的预设实体
优先选择假设更少的假说
简单≠正确，但它是很好的思考起点

Falsifiability Principle (Karl Popper)

可证伪性原则（卡尔·波普尔）

A scientific claim must be capable of being refuted.

Test:

"What evidence would prove this wrong?"
If no answer → Not a scientific claim

科学主张必须具备可被证伪的可能性。

验证方法：

反问“什么证据可以证明这个主张是错的？”
如果没有答案→ 该主张不属于科学范畴

Argument Analysis Process

论证分析流程

Step 1: Reconstruction

步骤1：重构论证

Input: Raw argument
   ↓
1. Identify main conclusion
2. List explicit premises
3. Identify hidden premises
4. Arrange in logical structure
   ↓
Output: Standardized argument

Input: Raw argument
   ↓
1. Identify main conclusion
2. List explicit premises
3. Identify hidden premises
4. Arrange in logical structure
   ↓
Output: Standardized argument

Step 2: Evaluate Premises

步骤2：评估前提

For each premise, ask:

True? (Is there supporting evidence?)
Relevant? (Does it connect to the conclusion?)
Sufficient? (Is it strong enough to infer the conclusion?)

针对每个前提，提问：

是否真实？（有没有支撑证据？）
是否相关？（和结论有没有关联？）
是否充分？（强度是否足够推导出结论？）

Step 3: Evaluate Inference

步骤3：评估推理

Does the inference follow valid rules?
Are there any formal fallacies?
Does the conclusion follow from the premises?

推理过程是否符合有效规则？
是否存在形式谬误？
结论是否可以从前提中推导得出？

Step 4: Consider Counterarguments

步骤4：考虑反论证

Are there counterexamples?
Are there stronger opposing arguments?
Is there additional information that changes the conclusion?

是否存在反例？
是否有更有力的对立论证？
是否存在会改变结论的补充信息？

Thinking Tools

思考工具

Steel Man (Opposite of Straw Man)

钢人法（Steel Man，稻草人谬误的反面）

Before critiquing, build the strongest version of the opposing argument:

Fully understand the opponent's position
Add reasonable premises they may have omitted
Rephrase in the most compelling way
Then critique

在批判对方观点前，先构建对方论证的最强版本：

充分理解对手的立场
补充对方可能遗漏的合理前提
用最有说服力的方式重新表述论证
再开展批判

Principle of Charity

慈善原则

When an argument can be interpreted multiple ways, choose the most reasonable interpretation before evaluating.

当一个论证存在多种解读方式时，先选择最合理的解读版本再做评估。

Reductio ad Absurdum

归谬法（Reductio ad Absurdum）

Prove something false by:

Assume it's true
Derive logical consequences
Show consequences lead to contradiction
Conclude: The initial assumption is false

通过以下步骤证明某观点错误：

假设该观点为真
推导其逻辑后果
证明后果会导出矛盾
得出结论：初始假设为假

Thought Experiment

思想实验

Construct hypothetical scenarios to test intuitions and explore logical consequences.

构建假设场景来测试直觉、探索逻辑后果。

Quick Evaluation Checklist

快速评估清单

Applied Example

应用示例

Analyzing an Argument

分析一个论证

Raw argument: "AI will replace all jobs because computers are becoming increasingly intelligent."

Reconstruction:

P1: Computers are becoming increasingly intelligent
P2: [Hidden] All jobs can be performed by sufficiently intelligent machines
P3: [Hidden] This development will continue without limits
─────────────────────────────────
C: AI will replace all jobs

Evaluation:

P1: Partially true, need to quantify "intelligent"
P2: Unproven assumption - are there jobs requiring human elements?
P3: Assumption about the future - are there physical/technical limits?
Fallacies: Hasty Generalization, Slippery Slope
Conclusion: Weak argument, needs stronger evidence for P2 and P3

原始论证： "AI会取代所有工作，因为计算机正变得越来越智能。"

论证重构：

P1: 计算机正变得越来越智能
P2: [隐藏前提] 足够智能的机器可以完成所有工作
P3: [隐藏前提] 这一发展会无限制持续下去
─────────────────────────────────
C: AI会取代所有工作

评估：

P1：部分成立，需要对“智能”做量化定义
P2：未经证实的假设——是否存在需要人类特质才能完成的工作？
P3：对未来的假设——是否存在物理/技术层面的上限？
谬误： 草率概括、滑坡谬误
结论： 论证薄弱，P2和P3需要更有力的证据支撑

References

参考文献

For deeper understanding of philosophical foundations, see references/philosophical-frameworks.md - including:

Classical Logic (Aristotle)
Rationalism (Descartes, Spinoza, Leibniz)
Empiricism (Locke, Hume)
Critical Philosophy (Kant)
Logical Positivism (Vienna Circle)
Philosophy of Science (Karl Popper)
Dialectical Method (Hegel, Marx)
Pragmatism (Peirce, James, Dewey)

如需深入了解哲学基础，查看 references/philosophical-frameworks.md，涵盖内容包括：

古典逻辑（亚里士多德）
理性主义（笛卡尔、斯宾诺莎、莱布尼茨）
经验主义（洛克、休谟）
批判哲学（康德）
逻辑实证主义（维也纳学派）
科学哲学（卡尔·波普尔）
辩证法（黑格尔、马克思）
实用主义（皮尔士、詹姆斯、杜威）