人生进阶指南《Glossary of Terms and Methods》术语与方法库导读:一套把"定义、证据、下一步"串起来的可检索知识索引
本文是《Life Level-up Guide / 人生进阶指南》(仓库 en/up 项目,正文位于 docs/en/threads)中核心术语与方法索引的解读。glossary 的价值不是把生活变成一场词汇考试,而是给全书反复出现的概念一个"稳定的落脚点":当某个词含义不清、或不知道该打开哪一页时,回到这里即可。读完本文,你将掌握这套"定义 → 证据 → 下一步"的检索方法,看懂 6 组术语表的分类逻辑,并能在 41 个核心概念与对应章节、模板之间快速跳转。
该文档(docs/en/reference/glossary.md)的 front-matter 明确给出了它的定位:"Define the guide's core vocabulary, evidence standards, tool entry points, and reading paths so ideas can transfer between chapters"(定义全书的核心词汇、证据标准、工具入口与阅读路径,使概念可以在章节之间迁移),文档版本更新于 2026-09-01。
为什么需要一本"术语汇",而不是一份背诵清单
Glossary 的开篇就划清了一条底线:这不是要把生活变成词汇测验,而是给那些反复出现在各个章节的词一个稳定的落点。当你遇到一个模糊的词、或不知道下一步该打开哪个页面时,回来查这里即可。它遵循的最小规则是:如果一个术语不能帮你完成任务、保存样例、或做出更好的决策,那就不需要为了记住而记住它。
这套术语体系覆盖的不只是"英语学习方法",而是把英语学习、AI 协作、失败复盘、关系经营与日常节律放在同一条时间线上的一套元方法。因此它把术语分成六个语义区,正好对应全书的结构脉络:
| 分区 | 回答的问题 | 对应正文区域 |
|---|---|---|
| Problems and Goals | "我为什么做、做到什么算数" | 学习原理与 CEFR 章节 |
| Evidence and Judgment | "什么东西能证明、什么只是猜测" | Evidence 章与叙事章 |
| Learning Actions | "用什么动作让能力留下来" | 听说读写四大技能章 |
| Tools and Delivery | "用什么模板落地、交付给谁" | 模板库与 AI 协作章 |
| Life and Boundaries | "如何护住身体、关系与成本" | 节律、恢复、决策章 |
| One Sentence to Keep | "只留一句核心流程" | 全书闭环 |
Problems and Goals:先定义问题,再谈目标
这组术语解决的是"努力的方向有没有被钉死"。它们指向全书的一个反复主张:模糊的"我想进步"不是任务,真正的任务一定包含受影响的人和一个要完成的动作。用词表的话说,Real problem 是"一个有受影响人群、并且有待完成行动的情境",而不是"I want to improve"这种愿望。
这组词的完整定义与去向如下:
| Term | Meaning in this guide | Continue here |
|---|---|---|
| Real problem | A situation with affected people and an action to complete, not "I want to improve" | Learning Principles |
| Context | When, where, and for whom a skill will be used | CEFR Goals and Self-check |
| Acceptance standard | An outcome another person can observe, check, or restate | AI Task Brief |
| Main question | One answerable question tracked for a week | 90-Day Action Plan |
把这一组与正文互证,能看出三个设计意图:
- Real problem → Learning Principles:Learning Principles 章(docs/en/threads/part-1/1-understanding.md)强调"决定学什么之前,先命名 context",并给出填空模板
My primary context is ______. In 12 weeks I will complete ______ under ______ conditions.。glossary 中的 Acceptance standard 与 AI Task Brief 模板(docs/en/templates/ai-task-brief.md)里的 "Observable definition of done" 字段直接呼应——模板还特意给出示范:验收标准要写成动作("a user can import a file in ten minutes and see an explainable error report"),而不是"The experience is good"。 - Main question 以"周"为单位:它对应 90 天计划里"一个主问题跟踪一周"的节拍,而不是一年十个大目标。
- Context 与 CEFR 挂钩:CEFR Goals and Self-check 章节强调"一个平均标签描述不了 B2 阅读 + A2 口语",所以 Context 要精确到"在什么时间、对什么人、用什么技能"。
Evidence and Judgment:把猜测与事实分开的八件武器
这是整套术语中最硬核的一组,也是全书的证据纪律所在。它的结构特点是 "Meaning + Common trap"(含义 + 常见陷阱)——每个词不是孤立的定义,而都配了一个最容易犯的错误,用来提醒读者"这个词恰恰是最容易被误用成自欺工具的"。核心词 Evidence 的主张是:能证明能力的是"能回到工作、录音、测试、出处或真实反馈的记录",而不是"小时数、打卡数、聊天记录"。
| Term | Meaning in this guide | Common trap |
|---|---|---|
| Baseline | A raw sample saved before tool support or concentrated training | Replacing the first version with feelings or a polished result |
| Evidence | A record that returns to work, recording, test, source, or real feedback | Treating hours, saves, or chat logs as proof of ability |
| Fact | Something with a source, date, sample, or direct observation | Writing an explanation as a proven cause |
| Interpretation | A reading of the facts that may still need revision | Stopping verification because it sounds plausible |
| Hypothesis | A judgment waiting for the next action to test it | Writing a wish or model answer as a conclusion |
| Inference boundary | Separating what a source says, what evidence supports, and what you add | Treating a headline, emotion, or one anecdote as the whole picture |
| Narrative compression | Turning a complex experience into a shareable sentence or causal chain | Mistaking a shortened story for the whole event after removing chance, others' work, and unknowns |
| Hindsight bias | Explaining a past choice with outcomes that were known only later | Sending later information backwards into the past |
把这一组展开,能还原出 Evidence: How Change Becomes Visible 一章的核心方法论:
- Baseline 的位置:Evidence 章指出四条时间线——Baseline(集中训练/工具介入前的原始样本)、Immediate performance(练习后立刻观察)、Delayed retention(撤掉最近提示后的延迟再现)、Transfer(换话题/听众/媒介的迁移)。glossary 中 Baseline 的陷阱是"用感觉或打磨过的结果替换第一版"——这正是 Evidence 章警告的"只留最终成品,就忘了你如何从不会走到会"。
- Fact / Interpretation / Hypothesis 的链条:Decision-Making 章与 Recovery 章都用了"事实 / 我的解读 / 我承担的责任"三栏写法;Recovery 章明确要求把羞耻感拆回事实。Narrative compression(叙事压缩)在 Narrative and Evidence 章中被进一步展开:人把复杂经历压成一句话因果链,删掉了偶然性、他人贡献与未知,就会把"被缩短的故事"误当成"完整事件"。
- Hindsight bias(后见之明偏差):决策章主张留下决策记录(decision record),"让未来的自己可以复查,而不是只用结果审判过去"——这正是防止把事后才知道的信息倒灌进过去的手段。
Learning Actions:让"重复"变成"回声"的八个动作
这一组解决的核心问题是:哪些学习动作真的能留下能力? 它承接 Learning Principles 章"重复暴露制造熟悉感,而提取、讲解、应用与反馈才暴露知识是否可用"的原则,并把方法落实到 8 个可执行动作上。特别注意其中三个词的严谨边界:
- Retrieval practice / Distributed practice 是 Learning Principles 章引证过的两个广受支持策略(Dunlosky et al. 2013、Roediger & Karpicke 2006 等文献对应其 Sources 清单),但术语表与正文都强调:没有任何方法承诺"普遍百分比或普适日历"。
- Transfer 与 Delayed retention 严格分开:Rhythm 章把变化分三层——同日表现、留存、长期迁移,"别让一次成功给整条路背书"。
- Variation(变换):保持核心动作不变,改变话题、听众、时间或约束,用于检验学习是否真实——这是 Rhythm 章四大"回声"之一。
| Term | Meaning in this guide | Related chapter |
|---|---|---|
| Retrieval practice | Close the source and answer, retell, write, or perform from memory | Learning Principles |
| Distributed practice | Return at multiple points instead of cramming once | Vocabulary |
| Intensive listening/reading | Use short material to check detail, structure, evidence, and language form | Listening · Reading |
| Parallel material | A different source with similar difficulty and task for testing transfer | Reading Evidence Card |
| Transfer | Complete a related action after changing topic, listener, time, or task | 90-Day Cycle Map |
| Delayed retention | Complete a similar task after time has passed and recent prompts are removed | Evidence |
| Evidence chain | One page connecting baseline, immediate performance, delayed retention, transfer, and next step | Evidence Chain Template · Evidence |
| Variation | Keep the core action while changing the topic, listener, time, or constraint to test whether learning is real | Rhythm |
在正文中的对应实操包括:
- Evidence chain(证据链):Evidence 章给出一页式模板,把主问题、真实情境、完成标准、四个时间点样本、最重要的错误、仍属假设的结论、隐私/许可/健康边界、下次要改的一个变量、复测日期与停止条件压缩在一页;并明确要求"保留错误的版本,而不只是概括它的一句话"。
- Parallel material(平行材料):指向 Reading Evidence Card 模板,用难度相似但来源不同的材料检验迁移。
- Intensive listening/reading:是 Listening 与 Reading 两章共用的精细加工动作,强调用短材料核对细节、结构、证据与语言形式,而不是"看完即懂"。
Tools and Delivery:把学习变成"可交付的东西"的七个入口
这一组把前几组的动作装进具体工具。它贯穿一个主张(也是 Prologue 的承诺之一):学习要回到生活,行动要允许回到学习,而中间需要"造物/交付/检验"的环节。Artifact(产物) 被定义为"别人能读、能用、能质疑、能改进的输出,不一定是产品";Delivery(交付) 则要求"在真实受众、真实用户或真实约束下完成工作并接受其结果"。
| Term | Meaning in this guide | Related chapter |
|---|---|---|
| Task brief | A one-page record of context, input, action, boundaries, acceptance, and ownership | AI Task Brief |
| Case review | A record that separates a public experience into source, facts, judgment, outcome, and transferable principle | AI Case Review |
| Evidence card | A template for preserving a first take, feedback, and retest for one language skill | Four-skill path |
| Artifact | An output others can read, use, question, or improve; not necessarily a product | Artifacts |
| Delivery | Completing work under a real audience, user, or constraint and accepting the result | 90-Day Action Plan |
| Human gate | AI may assist, but a person confirms facts, permissions, privacy, cost, and final judgment | AI Project Development |
| Rollback | Pause, degrade, switch, or return to a known usable state after failure | AI Learning and Project Practice |
这一组与模板库(docs/en/templates)的对应关系最直接:
- Task brief → AI Task Brief 模板。它的字段设计本身就是在回答术语定义里的六个要素:真实情境、用户/受众、要完成的决策或动作、截止时间,再加"输入与边界"(允许的文件、数据敏感性分级 public/internal/confidential/restricted、刻意扣下的材料)与"输出与验收"(最终交付物、可观察的 done、需人工确认的项目)。
- Case review / Human gate / Rollback 共同撑起 AI 协作的"责任边界":Evidence 章把 AI 介入拆成三版记录——Unaided version(先无 AI 出一版)、Assisted version(让 AI 提问/解释/反馈,记录提示词、来源、编辑与返工)、Handover version(关掉聊天后仍能解释、维护、继续),并强调"人类闸门":AI 可以辅助,但事实、权限、隐私、成本与最终判断必须由人确认;失败后的回滚是暂停、降级、切换或回到已知可用状态。
- Evidence card 指向前言(Prologue)里的"听说读写四技能路径":每种语言技能各保一张卡片,保存第一次尝试、反馈与重测。
Life and Boundaries:护住身体、关系与成本的十二个词
这一组是全书"不把生活做成仪表盘"的部分。它承认:有些变化适合计数(交付数量、延迟重测正确率、测试是否通过),有些变化不应被压缩成一个分数(更早求助、在关系中造成更少伤害、在疲惫来临时真正停下)。因此这组术语多数描述的是"可观察的事实记录"而非 KPI。
| Term | Meaning in this guide | Related chapter |
|---|---|---|
| Boundary | What I will do to protect myself and state a consequence when something happens | Relationships |
| Recovery | Restore safety, body, relationships, and agency before efficiency and growth | Recovery |
| Minimum order | A safe action and visible result that remain repeatable during disorder | Life Practice Toolkit |
| Minimum viable day | A version of the day that preserves safety, one minimum output, and a cue for the next step at the capacity available | Daily System |
| Capacity budget | A rough estimate that includes time, attention, body, and responsibility rather than time alone | Daily System |
| Return action | The smallest output that can be completed within 24 hours and bring the system back after interruption | Daily System |
| Rhythm | One pattern of action that remains recognisable and executable at different capacities, connecting repetition, feedback, and recovery over time | Rhythm |
| Compounding | Today's choice reduces tomorrow's hesitation, rework, or recovery cost, allowing ability, trust, judgment, and recovery to support one another | Rhythm |
| Minimum contract | An agreement for bad days that states when to downgrade, what action to keep, and when to ask for help or escalate risk | Rhythm |
| Rhythm ledger | One page each week recording the main question, protected action, useful feedback, repeated resistance, and next condition to change | Rhythm Ledger Template · Rhythm |
| Cost | Time, money, health, relationships, privacy, and responsibility that must be named | Decision-Making |
| Reversibility | Whether a choice can be paused, undone, or changed back at manageable cost | Decision-Making |
这组术语与正文的呼应非常紧密,足以看出它是"从日课到长跑"的桥梁词汇:
- Minimum viable day / Capacity budget / Return action 三个词同源。Daily System 章要求给一天写三个容量档位(Full day 一次输入一次输出一次记录 / Compressed day 一个 10–25 分钟的保底输出 / Recovery day 保安全、饮食、休息与一个回来的提示),并按时间、容量、注意力、关系、身体五列做预算——预算要先保住不可让渡的底线,再把剩余空间留给野心。中断后用四行记录"事实发生了什么 / 哪个条件变了 / 24 小时内最小的输出 / 我明确不会补做的部分",这就是 Return action。
- Rhythm / Compounding / Minimum contract / Rhythm ledger 指向 Rhythm: Let Small Things Travel Through Time,它是连接日课与 90 天计划的桥梁章。能力、信任、判断、恢复四种东西互相复利(compounding);坏日子签"最低契约"(minimum contract)回答三个问题——何时降级而非硬撑、降级后还保留哪个最小动作、何时该求助或交给专业人士;每周一页 rhythm ledger 记录主问题、被保护的最小动作、带来新信息的反馈、反复出现的阻力、下周要改的一个条件、以及"停止跟踪的一个数字"——最后一行是刻意的:每条记录都消耗注意力,删掉一个无法指导决策的数字本身就是节律设计的一部分。
- Cost / Reversibility 指向 Decision-Making:成本必须点名时间、金钱、健康、关系、隐私与责任;可逆性决定你该多快下注、需要多厚的证据——高度可逆的选择尽早变成小实验,难以逆转的选择(大额负债、卖掉重要资产、公开有损指控)要加冷却期、第二意见与退出计划。
- Boundary / Recovery 表明这套术语并不回避生活暗面:Recovery 章要求在任何"效率与未来"的讨论之前先确认安全(content note 明确说明该章不是医疗/心理/法律/财务建议);Relationships 章则在 daily 执行层面补充"边界要说出口"的句子模板("19:00–19:30 我在赶工不会看消息,非紧急 20:00 回复,必须今天处理请直接告诉我截止时间")。
一句话保留:整套方法论的压缩版本
Glossary 用一段"可带走的一句话"收束全篇。它实际上是六组词表串起来后唯一的推荐序列:
If you keep one sequence: name the problem, save a baseline, use the right tool for one small task, let a real person test it, write the feedback into the next action, and recover before adding pressure.
拆开看,这句话正好是六个分区的浓缩:name the problem(Problems and Goals)→ save a baseline(Evidence and Judgment)→ use the right tool for one small task(Learning Actions)→ let a real person test it(Tools and Delivery 里的 delivery)→ write the feedback into the next action(Evidence chain)→ recover before adding pressure(Life and Boundaries)。
全文的两个入口与一套检索流程
文档结尾给出两个入口:阅读入口 Prologue: Do Not Rush to Change Your Life,行动入口 90-Day Cycle Map。前者回答"为什么要这么慢、这么小步地改自己",后者把术语翻译成以 90 天为周期、可勾选的行动地图。
在实际检索时,建议按 glossary 自己在 "How to Use It" 里定义的流程操作:
- 定位:先找到你此刻面对的"任务",而不是先翻方法论;
- 查词:在六个分区表中找到对应术语,看清它的定义与陷阱列(或"去哪里继续"列);
- 跳转:点开指向的正文章节获得方法,或打开指向的模板直接复制使用;
- 落地检验:如果一个术语不能帮你完成任务、保存样例或改善决策,就允许自己忘掉它——这不是知识漏洞,而是这套索引的设计意图。
小结:一本"不要求你背下来"的词典
作为全书"术语的固定降落点",这份 Glossary 的独特之处在于:它不提供新的方法论,而是把方法论词汇化、表格化、可跳转化。正文中的每个抽象主张(证据链、人类闸门、可逆性、最低契约、四种复利……)都能在这个文件里 30 秒内定位到标准定义、常见误区与下一步文档。对读者而言,它既是一份复习索引,也是一张"迷路时的地图";对研究者或二次创作者而言,它则是从文档层面快速索引全书结构的最短路径。当你忘记某个词、或不知道该打开哪一页时,回到 glossary.md 即可——它保证你只记住那些值得记住、并能导向下一个动作的词。
atomcodeClaude Code 的开源替代方案。连接任意大模型,编辑代码,运行命令,自动验证 — 全自动执行。用 Rust 构建,极致性能。 | An open-source alternative to Claude Code. Connect any LLM, edit code, run commands, and verify changes — autonomously. Built in Rust for speed. Get StartedRust0629
MiniCPM5-2BMiniCPM5-2B 是一款面向端侧、本地部署和资源受限场景的 2B 稠密 Transformer,能够达到同尺寸开源模型 SOTA 水平。Markdown00
GLM-5.3GLM-5.3 与 GLM-5.2 使用相同的基座模型——所有提升均来自后训练。与 GLM-5.2 相比,它在复杂编程和长程任务上的表现显著提升。Jinja00
HivisionIDPhotos⚡️HivisionIDPhotos: a lightweight and efficient AI ID photos tools. 一个轻量级的AI证件照制作算法。Python07
DragonOSDragonOS is an operating system developed from scratch using Rust, with Linux compatibility. It is designed for **Serverless** scenarios. 使用Rust从0自研内核,具有Linux兼容性的操作系统,面向云计算Serverless场景而设计。Rust00
Spark-X2.5-1.7BSpark-X2.5-1.7B 旨在让强大的 AI 更加实用、高效且易于获取。这些模型在广泛的日常任务中表现出色,涵盖对话、写作、翻译、推理、编程、工具调用和智能体工作流,并在同等规模的开源模型中取得领先结果。Spark-X2.5 将面向效率的架构与最高 1M tokens 的原生上下文窗口相结合,并支持 200 多种语言。Python00