IB TOK 五大AOK 知识框架 史学 科学 艺术
选 IB 的学生第一次翻开 TOK 指南，常常不是被哲学概念卡住，而是被五个领域之间的关系卡住。AOK 不是把物理、历史、艺术的知识点再背一遍，考的是这些学科凭什么说自己的结论算知识。这篇文章把历史、人文科学、自然科学、数学、艺术放在同一套四元素框架下逐条对照，也说明展评与论文里怎么用它们。
Students often read the list of areas of knowledge as a second syllabus to memorise. It is not: each area asks how that community decides what counts as knowledge, and what it cannot reach. This article compares the five areas on the same four questions, then shows how that comparison is used in the exhibition and the essay.
2022 年课程结构：一个核心主题、两个选修主题与五个必修领域
IBO 的 TOK 指南在 2022 年首次评估的版本里把课程固定成三层。核心主题是 Knowledge and the knower，所有学生都要学。选修主题从五个里挑两个，分别是知识与技术、知识与语言、知识与政治、知识与宗教、知识与原住民社会。领域知识从旧版的八个减到五个，全部必修，就是历史、人文科学、自然科学、数学、艺术。指南给出的最低总课时是 100 小时，约 50 小时给五个领域，32 小时给核心与选修主题，18 小时留给两项评估的准备。其中展评通常安排在第一年，论文题目在提交前约六个月公布，写作放在第二年。五个领域不是让学生各挑一个，而是每个都要在课堂上过一遍。
The 2022 guide fixes the course in three layers: one core theme taught to everyone, two optional themes chosen from five, and five compulsory areas. Teaching time totals a minimum of 100 hours, split roughly 50 to the areas, 32 to the themes and 18 to assessment preparation. The exhibition usually happens in the first year, while essay titles appear about six months before the deadline and the writing falls in the second year. Schools sequence the course differently, yet this structure is what every candidate is assessed on.
知识框架四元素：范围、观点、方法与工具、伦理
五个领域都用同一套四元素来拆。范围问这个领域想认识什么、边界在哪里，比如历史碰不到没有留下记录的那部分过去。观点问谁在做知识、立场怎么影响取舍。方法与工具问知识怎么产生、怎么被检验，实验、档案、证明、排练各是一套。伦理问哪些问题即使能回答也不该做。旧版指南把这四项写成建议，2022 年版改成必修，每个主题与每个领域都要覆盖。这四项可以当成四段的分析模板，写展评时一个元素一段，写论文时一个元素一层，结构就不容易散。
Every area is unpacked through the same four elements: scope, perspectives, methods and tools, and ethics. The framework is compulsory in the current guide, so an answer that only lists subject content tends to stall in the lower bands. It also works as a paragraph plan, one element per paragraph, which keeps a 950 or 1,600 word response from drifting. A useful habit is to answer each element with one concrete practice rather than with a definition.
历史：只能靠留下来的痕迹，无法重放一次
历史的范围是人类过去，而过去不能被重新观察，所以历史知识建立在留存下来的材料上。日记、档案、口述、考古遗存、影像都是痕迹，不是事件本身。方法上要判断来源的作者、目的、时间与语境，再交叉比对不同材料，一手材料与后人整理的材料分量不同。很多学生以为史料越多结论就越确定，实际经常相反，新材料会推翻旧叙述，档案馆的解密规则一改，同一段历史的研究结论就得重写。没被记录下来的普通人生活，永远是历史知识的空白区。
History studies a past nobody can re-observe, so its knowledge is built from surviving traces rather than from the events themselves. Historians date, attribute and cross-check sources, weigh primary material against later synthesis, and read each document against the context that produced it. A common mistake is to treat a bigger pile of documents as certainty, when fresh material often overturns an older narrative. What was never recorded, such as the daily life of people with no written voice, stays permanently outside what the area can know.
人文科学：被测量的人会作出反应
人文科学包括心理学、经济学、地理、社会学一类，研究对象是人和社会。它的难点在于被测者会对测量本身作出反应，问卷里的自评会随措辞改变，实验里的学生知道自己被观察之后行为就不一样。方法上有抽样、量表、统计检验、田野观察与建模，控制变量的难度比自然科学高得多。相关不等于因果，这一点在人文科学里尤其要紧，两个变量同向变化常常来自第三个没被测量的因素。伦理在这里也特别紧，欺骗、伤害、隐私都直接限制能被研究的问题。样本换一批，结论有时就翻过来，所以复现同样重要。
The human sciences study people, which means the object of study can react to being studied. Self-report shifts with wording, behaviour changes once participants know they are observed, and control is far weaker than in a laboratory. Surveys, scales, statistical tests, fieldwork and modelling are the standard tools, and a correlation is not a cause, since a third unmeasured factor often drives both variables. Consent, harm and deception set hard limits on what can be asked at all.
自然科学：可重复的实验与带不确定度的测量
自然科学的范围是物质世界与生命世界，靠可重复的观察与实验建立结论。测量一定带不确定度，同一个量在不同仪器上读数不同，所以结论通常写成区间而不是一个确定值。方法是假设、控制变量、重复、同行评审，仪器既延伸感官也设定上限。很多学生以为实验成功一次就证明了理论，实际要做的是让别的实验室也能复现同样的结果。同行评审能挡掉一部分错误，但只挡得住评审人看得懂的那一类，发表压力还会让结果偏向新奇。理论被推翻通常不是一次实验，而是一批结果慢慢对不上。
The natural sciences build knowledge from repeatable observation and experiment on the physical and living world. Every measurement carries uncertainty, so results are reported as ranges, error bars and limits rather than exact values. Peer review filters some error, though only the kind reviewers are trained to see, and publication pressure favours novel results. Replication by an independent group, not one successful run, is what moves a claim towards accepted knowledge.
数学：公理内部的确定性与系统外的解释
数学的范围是从选定的公理出发能推出什么，它不需要物理世界才能成立。方法是定义、公理、演绎、证明与反例，一个反例就能推翻一个全称命题，这一点在其他领域很少见。数学的确定性只在系统内部成立，公理换了结论跟着换，非欧几何就是这样出现的。常见的错误是把数学的严格当成对现实的描述，模型算得再准也要看假设是否成立。同一个方程可以描述人口增长，也可以描述放射性衰减，数学给的是形式，解释要由使用的人补上。证明写错一步，整条链就断了，这也是它和实验科学最大的不同。
Mathematics derives conclusions from axioms a community agrees to accept, so its certainty is internal to the system rather than a claim about the world. Proof and counterexample do the work that experiment does elsewhere, and a single counterexample can end a general claim. The same equation may describe population growth and radioactive decay, since mathematics supplies form while the user supplies interpretation. When a model is applied to reality, the axioms and assumptions become the part worth questioning.
艺术：制作与解释共同构成知识
艺术的范围是人的经验怎么被表达出来，形式本身就是内容的一部分。方法上有技法训练、材料试验、传统与流派、批评与观众反应，标准存在但不是测量。作品的意义随时代与观众变化，同一部戏在 1955 年与今天读出的东西并不相同，这种多义不是缺陷。伦理问题集中在谁有资格讲谁的故事、展览该呈现什么、作者生平该不该改变对作品的评价。技法上的对错是可讨论的，一个和弦解决得对不对，同行能给出理由。解释没有唯一答案，但也不是随便说，要能被作品细节与批评传统支撑。
The arts produce knowledge through making and interpreting, so the meaning of a work is part of what is known rather than noise around it. Technique, material, tradition and critical response are the methods, and judgements about craft are arguable with reasons even though they are not measurements. Interpretation changes with audience and period, which is why a work can support two honest readings at once. What separates a reading from a guess is whether the details of the work and the critical tradition can carry it.
五个领域放在一起看：什么算证据
同一句知识问题在五个领域里的答案不一样，原因在证据的门槛。自然科学要可重复的测量，人文科学接受统计上的概率结论，历史接受交叉核对过的材料，数学只接受证明，艺术接受作品与批评传统支撑起来的解释。把证据门槛混用是展评里最常见的失分点，比如用实验数据去回答一个关于历史解释的问题。反过来，用一句审美判断去否定一个测量结果，同样不成立。比门槛更细的一层是证伪方式，实验靠反例数据，数学靠反例命题，历史靠新档案，艺术靠新的解读。
Put the five areas side by side and the difference appears in what counts as evidence. Measurement, statistical inference, cross-checked sources, proof and critical interpretation are not interchangeable currencies, so mixing them is a reliable way to lose marks. The finer question is how each area is corrected: data overturns a hypothesis, a counterexample kills a theorem, a new archive rewrites a narrative, a fresh reading reframes a work. Answering that question is usually what separates a mid-band response from a high one.
观点与范式：研究者、经费与档案都在其中
观点这一元素容易被写成个人感受，它问的其实是知识生产的位置。谁拿到经费、谁进得了档案馆、哪本期刊愿意发表，都会改变一个领域里被研究的问题。自然科学里范式一换，旧数据会被重新解释，历史里被记录的人与被忽略的人，本身就是权力结构的结果。写这一段时给一个具体机制，比写一句每个人都有立场更有用。课堂上做的实验往往是被简化过的版本，学生看到的步骤与真正研究里的争论并不一样。能说清这个差距，观点这一元素才算写到位。
Perspectives are not private opinions; they are the positions from which knowledge is produced. Funding, archive access and publication decisions shape which questions get asked at all, and a paradigm shift can re-read data that was already there. Naming one mechanism, such as who sits on a funding panel or whose records were never kept, does more than repeating that everyone has a viewpoint. School practicals are simplified versions of real research, and saying how they differ is often the strongest move available.
伦理：从独立领域变成贯穿四个元素的线索
旧版 TOK 把伦理当成八个领域之一，2022 年版取消了这个独立领域，把伦理写进四元素。变化带来的实际后果是每个领域都要回答一次伦理问题。自然科学要谈人体试验与动物实验的审批，人文科学要谈欺骗与隐私，历史要谈档案封存与用今天标准评判过去，数学要谈算法被拿去做什么。只在一个领域里谈伦理，等于放弃了另外四个领域的分数。写的时候要具体到一个决定，谁批准、谁承担后果、有没有替代方案。抽象地说科学要负责任，阅卷时拿不到分。
Ethics used to be one of the eight areas; in the current guide it is a thread running through the framework instead. Each area therefore has its own version of the question, from human trial approval in the sciences to sealed archives in history and the use of published mathematics in pricing. A strong answer names a decision, the body that authorises it and who carries the consequences, rather than stating that science should be responsible. Leaving ethics to a single area gives away marks in the other four.
展评与论文：35 个提示词、3 件物品与 1600 词
展评从指南给出的 35 个提示词里挑一个，用三件真实物品写 950 词评论，校内评分、IBO 外部调节，占 TOK 总分三分之一。论文从每年公布的六个题目里选一个，写 1600 词，外部评分，占三分之二，讨论要落在两个领域上。两项任务都不要求覆盖五个领域，要求的是写清一个领域里知识怎么被生产出来、谁在检验它。展评的三件物品要围绕同一个提示词，各管一个侧面，不能三件说同一件事。论文里两个领域要有真正的对照，不是各写一半互不相干。
The exhibition takes one prompt from the 35 in the guide and links three real objects to it in 950 words, marked in school and moderated by the IB. The essay answers one of six prescribed titles in 1,600 words and is marked externally, with the argument built across two areas. The three objects should each carry a different angle on the same prompt, not repeat one point three times. In the essay the two areas need genuine comparison, since two separate halves that never meet score poorly.
常见的错误：把领域当科目、把描述当分析
阅卷时最常见的两类问题都出在层次上。一类是把 AOK 当成科目复习，写光合作用的过程却不说生物学家凭什么相信这个结论。另一类是把描述当分析，花三句话讲一件物品长什么样，只用一句说它和提示词的关系。改法很具体，每写一个事实，接一句这个事实是怎么被确立的、谁可能不同意。还有一个隐蔽的坑是例子太熟，图灵测试、哥白尼、薛定谔的猫反复出现，写的人多但分析常常停在介绍。换一个自己真正用过的例子，分析反而容易写深。
Two problems dominate the weaker scripts, and both are about level rather than amount of knowledge. One describes the subject instead of asking how the subject knows; the other describes an object at length and links it to the prompt in a single clause. The fix is mechanical, since every factual claim can be followed by how it was established and who might dispute it. Overused examples are the quieter problem, as familiar cases get introduced at length while the analysis stays thin.
牛津高阶英汉双解词典 第10版 商务印书馆¥168📱 去京东购买京东精选
需要一对一辅导？Need One-on-One Tutoring?
IB TOK 展评与论文、DP 各科 IA 与 EE 辅导
16621398022 同微信
Visit aleveler.com
或截图后用「扫一扫 → 相册」选择本图