v1.0.0 to v1.0.0
34 added, 34 removed. Audit A to A.
---
name: abductive-hypothesis-generation
- description: 'Strategy: 面对异常的最佳解释推理'
+ description: 'Strategy: Inference to the best explanation in the face of anomalies'
version: 1.0.0
category: hypothesis-formation
type: strategy
campaign: hypothesis-formulation
tactics:
- anomaly-driven-abduction
sops:
- anomaly-characterization
- explanation-generation
- plausibility-ranking
- falsifiability-check
dependencies:
tactics:
- anomaly-driven-abduction
sops:
- falsifiability-check
---
# Abductive Hypothesis Generation
- 面对异常的最佳解释推理：当观察到现有理论无法解释的异常现象时，系统生成候选解释并选出最合理者作为假设。
+ Inference to the best explanation in the face of anomalies: when an anomalous phenomenon that existing theory cannot explain is observed, systematically generate candidate explanations and select the most plausible one as the hypothesis.
- ## 适用场景
+ ## When to Use
- - 观察到明确的异常现象（与现有理论预测不符的结果）
- - 现有理论无法充分解释某个已知现象
- - 需要在多个竞争解释中选出最值得检验的一个
- - 研究起点是"这个结果很奇怪，为什么？"
+ - A clear anomalous phenomenon is observed (a result inconsistent with existing theoretical predictions)
+ - Existing theory cannot adequately explain a known phenomenon
+ - One of several competing explanations must be selected as the most worth testing
+ - The research starting point is "this result is strange, why?"
- 不适用：没有明确异常、只是想探索一个新领域 → 改用 inductive-hypothesis-generation。
+ Not applicable: no clear anomaly, just wanting to explore a new field → use inductive-hypothesis-generation instead.
- ## 思维框架
+ ## Thinking Framework
**Anomaly → Generate candidate explanations → Rank by plausibility → Best explanation = hypothesis**
- 溯因推理的核心逻辑：
+ The core logic of abductive reasoning:
- 1. **Anomaly**：精确描述异常——什么现象、与什么预期不符、偏差有多大
- 2. **Generate candidate explanations**：系统生成所有能解释该异常的候选解释（不过早筛选）
- 3. **Rank by plausibility**：按可信度排序——哪个解释最简洁、最与已知事实一致、最可检验
- 4. **Best explanation = hypothesis**：选出最合理的解释作为工作假设，其余作为竞争假设保留
+ 1. **Anomaly**: precisely describe the anomaly — what phenomenon, inconsistent with what expectation, how large the deviation
+ 2. **Generate candidate explanations**: systematically generate all candidate explanations that can account for the anomaly (no premature filtering)
+ 3. **Rank by plausibility**: rank by plausibility — which explanation is most parsimonious, most consistent with known facts, most testable
+ 4. **Best explanation = hypothesis**: select the most plausible explanation as the working hypothesis, retaining the rest as competing hypotheses
- **溯因的核心原则**：
- - **奥卡姆剃刀**：在解释力相当时，优先选择假设更少的解释
- - **一致性**：最佳解释不应与其他已知事实矛盾
- - **可检验性**：最佳解释必须能产生可观测的预测（否则无法验证）
- - **生成完整性**：在排序前必须穷举候选解释，避免过早收敛
+ **Core principles of abduction**:
+ - **Occam's razor**: when explanatory power is comparable, prefer the explanation with fewer assumptions
+ - **Consistency**: the best explanation should not contradict other known facts
+ - **Testability**: the best explanation must be able to produce observable predictions (otherwise it cannot be verified)
+ - **Generation completeness**: candidate explanations must be exhausted before ranking, to avoid premature convergence
## Budget Gate
- | Tier | 异常描述 | 候选解释 | 假设产出 | 竞争假设 |
+ | Tier | Anomaly description | Candidate explanations | Hypothesis output | Competing hypotheses |
|------|---------|---------|---------|---------|
- | S | 1 个精确描述的异常 | ≥2 个候选解释 | 1 个最佳解释假设 | ≥1 个竞争假设保留 |
- | M | 1–2 个异常 | ≥3 个候选解释 | ≥2 个结构化假设 | 完整可信度排序 |
- | L | ≥2 个相关异常 | ≥5 个候选解释 | ≥3 个结构化假设 | 完整排序 + 区分性预测设计 |
+ | S | 1 precisely described anomaly | ≥2 candidate explanations | 1 best-explanation hypothesis | ≥1 competing hypothesis retained |
+ | M | 1–2 anomalies | ≥3 candidate explanations | ≥2 structured hypotheses | complete plausibility ranking |
+ | L | ≥2 related anomalies | ≥5 candidate explanations | ≥3 structured hypotheses | complete ranking + discriminating prediction design |
- ## 默认参考流
+ ## Default Reference Flow
- 1. 调用 `anomaly-characterization` SOP：精确描述异常（现象、预期、偏差、已排除的平凡解释）
- 2. 调用 `explanation-generation` SOP（via `anomaly-driven-abduction` tactic）：系统生成候选解释（不过早筛选）
- 3. 调用 `plausibility-ranking` SOP：按简洁性、一致性、可检验性对候选解释排序
- 4. 调用 `falsifiability-check` SOP：为最佳解释生成 falsification scenario，确认其可检验性
+ 1. Call the `anomaly-characterization` SOP: precisely describe the anomaly (phenomenon, expectation, deviation, excluded trivial explanations)
+ 2. Call the `explanation-generation` SOP (via the `anomaly-driven-abduction` tactic): systematically generate candidate explanations (no premature filtering)
+ 3. Call the `plausibility-ranking` SOP: rank candidate explanations by parsimony, consistency, and testability
+ 4. Call the `falsifiability-check` SOP: generate a falsification scenario for the best explanation, confirming its testability
## context-checkpoint
- 每轮结束后记录：
- - 异常描述（精确版本，含偏差量化）
- - 候选解释清单（含已排除的平凡解释及排除理由）
- - 可信度排序结果（含排序依据）
- - 最佳解释假设 + 竞争假设列表
- - 区分性预测（什么实验能区分最佳解释与竞争解释）
+ Record after each round:
+ - Anomaly description (precise version, with deviation quantification)
+ - Candidate explanation list (including excluded trivial explanations and exclusion reasons)
+ - Plausibility ranking result (including ranking basis)
+ - Best-explanation hypothesis + competing hypothesis list
+ - Discriminating predictions (what experiment can distinguish the best explanation from competing explanations)
<!-- BEGIN available-tables (generated) -->
## Available Tactics
可选,无固定顺序;最终叶子终为 sop。
| Tactic | 何时用 |
| --- | --- |
| anomaly-driven-abduction | Tactic: 归纳/溯因路径——描述异常现象，生成候选解释，按可信度排序 |
## Available SOPs
可选,无固定顺序;最终叶子终为 sop。
| SOP | 何时用 |
| --- | --- |
| falsifiability-check | SOP: 检验假设是否满足可证伪性标准 |
<!-- END available-tables (generated) -->