IB TOK Exhibition 选题 三件物品 950词评论
TOK exhibition 看起来像一次小型策展: 三件真实世界的物品，配一篇不超过 950 词的评论，回答一个从 35 个题目里挑出来的知识问题。真正花时间的不是写，而是选。题目选偏了，或者物品选得太泛，后面文字再讲究也补不回来。
The TOK exhibition looks like a small act of curation: three real-world objects, one commentary of no more than 950 words, and a knowledge question chosen from a fixed list of 35. The writing is the cheap part. Most of the work sits in the choosing, and a weak prompt or a vague object cannot be rescued by polished prose later on.
展览的硬指标: 三个对象 950词 10分
按 IBO 2020 年 2 月版《Theory of knowledge guide》(2022 年首次评估) 的规定，exhibition 属于内部评估，由本校 TOK 教师按 10 分制打分，IB 再抽样审核，它大约占 TOK 总成绩的三分之一。学生从 35 个 IA 题目中选一个，不许多选，也不许自己写题，三件物品必须挂同一个题目。要交的是一个文件: 标题注明所选题目的原文、三件物品的图片、逐件评论，再加引用。评论总长不超过 950 词，超出部分不读。它也不是口头展示，评分只看这份文件。
Under the TOK guide the IB published in February 2020, with first assessment in 2022, the exhibition is internally assessed. Your own teacher marks it out of 10 and the IB moderates a sample, and it carries roughly one third of the overall TOK grade. You take exactly one of the 35 IA prompts, never reword it, and tie all three objects to that single prompt. What you submit is one file: a title quoting the prompt, images of the three objects, a commentary on each, and references. The commentary is capped at 950 words, and anything past the cap is not read. Nothing is performed aloud; the file is the whole thing.
35 个题目为什么值得先通读一遍
IA 题目在整本 guide 有效期内固定不变，所以它们年年出现，网上的范例也堆得很厚。这些题目问的都是知识本身的事: 什么算知识、证据到什么程度算够、谁拥有知识、价值如何影响知识的产生。通读的用处不是挑一个"最好写"的，而是先把那些你手上没有真实物品的题目排除掉。有些题目读起来很聪明，落到三件物品上却只能靠转述别人的例子。
The IA prompts stay fixed for the life of the guide, which is why they reappear every year and why sample exhibitions pile up online. All of them ask about knowledge itself: what counts as knowledge, when evidence is enough, who owns knowledge, how values shape what gets produced. Reading the whole list is not about finding the easiest prompt. It is about crossing out the ones you cannot populate with real objects. Some prompts sound sharp until you try to put three things behind them, and then you notice you would be borrowing someone else's examples.
先定题目还是先定物品
常见的错误是先从抽屉里翻出一件有趣的东西，再去找一个能套上的题目。这样做的结果通常是评论在描述物品，最后两句才勉强拉回题目，逻辑像补丁。反过来省力得多: 先读题目，圈出三到五个有感觉的，再为每个题目试着列出三件物品，列得出来的留下，列不出来的划掉。题目决定论证方向，物品只提供证据。
The usual mistake runs the other way round: you find an interesting object first and then hunt for a prompt it might fit. What that produces is description with a prompt bolted on at the end. Working from the prompt is cheaper. Read the list, circle three to five that interest you, then try to name three objects for each. Keep the prompts you can populate and drop the rest. The prompt sets the argument; the objects only supply evidence.
十分钟能不能落地: 一个土办法
判断一个题目能不能写，用最笨的办法: 给自己十分钟，为它写出三件物品的名称、年份、出现的地点，以及当初是谁在用。写不全，说明你对这个题目的兴趣还停在抽象层面。举个例子，题目 5 "What counts as good evidence for a claim?" 可以配 2021 年某医院出具的核酸检测报告、一份法院判决书里被采纳的专家证词、一篇预印本论文下面署名的同行评审意见。三件都能指认到具体时间和场景，题目才算落地。手指碰不到的东西，写进评论里也是空的。
A crude test settles whether a prompt is workable: give yourself ten minutes and write down, for three objects, the name, the year, where it appeared and who used it. If the boxes stay empty, your interest in the prompt is still abstract. Prompt 5, "What counts as good evidence for a claim?", can carry a PCR test report issued by a hospital in 2021, an expert witness statement accepted in a court judgment, and a signed peer review attached to a preprint. Once all three can be pinned to a time and a setting, the prompt has somewhere to stand.
挂哪个主题: 核心主题与五个可选主题
guide 建议把展览放在一个 TOK 主题下，核心主题是 knowledge and the knower，五个可选主题是知识与技术、知识与语言、知识与政治、知识与宗教、知识与原住民社会。主题不进评分表，但三件物品有共同的落脚点，评论读起来才像一次展览，而不是三个互不相干的例子。选主题的实用办法是看三件物品是否共享同一个知识社群或同一套工具。
The guide recommends anchoring the exhibition in a TOK theme. The core theme is knowledge and the knower, and the five optional themes cover knowledge and technology, language, politics, religion, and indigenous societies. The theme does not appear in the marking, yet shared ground is what makes three objects read as one exhibition instead of three unrelated examples. A practical way to pick is to ask whether your objects belong to the same community of knowers or lean on the same set of tools.
真实世界语境写到什么颗粒度
评估工具要求识别物品与它的 real-world context，这一句决定评论的下限。"一张地图"不合格，"2016 年某市规划局发布的防汛风险图，用来决定哪些街区优先加固"就合格。语境不是装饰，它说明这件物品在现实中怎样参与知识的生产与流通。写作时把年份、发布者、使用场景三样写齐，一般就够了。有日期和机构名，审核者才确认这不是一张示意图。
The instrument asks you to identify each object and its real-world context, and that sentence fixes the floor of the mark. "A map" will not do. "A flood risk map issued by a city planning bureau in 2016 to decide which districts were reinforced first" will. Context is not decoration: it shows how the object took part in producing or circulating knowledge. Year, issuer and use are usually enough, and the date and the institution name are what tell a moderator this is not a diagram someone drew for the task.
三件物品之间要有张力
三件物品如果都在证明同一句话，评论会显得单薄，读起来像三个并列的例子。更耐读的排法让它们站在知识生产的不同环节: 一件显示知识如何被生成，一件显示它如何被检验，一件显示它如何被传播或走样。也可以让它们互相拆台，比如一份官方统计与一份现场记录对同一件事给出不同数字，评论立刻有话可说。很多学生把"相关"当成"同向"，这两件事不一样。真正需要的不是三票赞成，而是三种不同的证据形态。
Three objects that all prove the same sentence make a thin commentary; it reads as three parallel examples. A stronger arrangement puts them at different stages: one showing how a claim was generated, one how it was tested, one how it travelled or got distorted. They can also work against each other. An official statistic and a field record that report different numbers for the same event hand you an argument. Many students read "relevant" as "agreeing", and the two are not the same thing. What the task rewards is three different shapes of evidence, not three votes in favour.
每件物品先写贡献句
评分标准里反复出现的是 justification of the contribution each object makes。落到写法上，每节第一句先交代这件物品让题目里的哪个词变得具体，再描述物品本身。顺序反过来，读者要先读三百词描述才等到论证，而论证往往已经被挤到最后一行。常见的错误是把物品当主角，把题目当结论。改法很简单: 把那句贡献句挪到段首，剩下的描述围绕它筛。
The phrase that keeps returning in the criteria is the justification of the contribution each object makes. In practice, open each section with the word in the prompt that this object makes concrete, then describe the object. Reverse the order and the reader waits through three hundred words of description for the argument, which by then has been squeezed onto the last line. The familiar error is treating the object as the subject and the prompt as the conclusion. Moving one sentence to the front fixes most of it, and the remaining description can then be filtered against it.
950 词怎么切
950 词是总上限，不是每件物品的上限。通行的切法是每件三百词上下，留四十到五十词做开头与收束。guide 写明超过上限后审核者只读前 950 词，所以把最吃重的论证放在前面不是风格问题，是分数问题。图片上的文字、致谢和参考文献不计入词数，这条可以放心用。写得快的人常犯一个反向错误: 三件物品各写七百词，第一件的评论还没进入正题就被截断。
The 950 word cap covers the commentary as a whole, not each object. A workable split gives about three hundred words per object and leaves forty to fifty for the opening and the close. The guide states that examiners stop reading at 950 words, so putting the load-bearing claim early is a scoring decision rather than a stylistic one. Text inside the object images, acknowledgments and references sit outside the count, and that is fair to use. Fast writers tend to make the opposite error: seven hundred words on the first object, cut off before the commentary has reached its point.
评分看的是知识如何显形
官方评估工具给出的驱动问题只有一句: 这次展览是否成功展示了 TOK 如何在周围世界中显形。五个等级靠整体印象判断，取最贴合的一档，而不是逐条打勾。落到评论上，审核者看四件事: 物品与语境是否被识别清楚、贡献是否被论证、证据是否被使用、论述是否回到题目。四件事缺一件，等级就往下掉。所以通篇不点名回到题目，即使内容相关，也会被判成"与题目联系薄弱"。
The driving question in the official instrument is a single sentence: does the exhibition successfully show how TOK manifests in the world around us? Five level descriptors are applied by global impression, taking the best fit rather than ticking boxes one by one. On the page, moderators look at four things: whether the object and its context are identified, whether its contribution is justified, whether evidence is used, and whether the discussion returns to the prompt. Let one of the four slip and the level drops. A commentary that never names the prompt again can be marked as weakly linked even when the content is relevant.
常见的失分点
改写题目是最常见的一种。官方口径是按原题评，但相关性上的损失会落在分数里，如果完全看不出用的是哪道题，直接零分。物品太泛排第二，"社交媒体"、"一本教科书"、"互联网"都不算具体物品。第三是只描述不论证，把评论写成说明书。第四是引用缺失，尤其是图片和统计数据。还有一类失分很隐蔽: 把 exhibition 写成 essay，通篇讨论知识理论，物品只在开头露一次面。
Rewording the prompt is the most common loss. The official line is that the work is marked against the original prompt, with any loss of relevance showing up in the mark, and if the prompt cannot be identified at all the score is zero. Vague objects come second: "social media", "a textbook" and "the internet" are not specific objects. Third is description without justification, a commentary that reads like an information sheet. Fourth is missing references, especially for images and figures. A quieter failure is writing an essay instead of an exhibition, theory throughout with the objects mentioned once at the start.
一个反例与一个可用版本
拿题目 12 "Is bias inevitable in the production of knowledge?" 对比两次选择。反例是"一部手机、一本书、一台电脑"，三件都太泛，评论只能写"手机会带来信息偏差"这类放之四海皆准的话，没有语境可识别，也没有证据可用。可用版本换三件: 2019 年某新闻机构公开撤回的一篇报道截图、同一关键词在两个账号下得到不同结果的推荐页截图、某平台 2020 年公布的内容审核准则。三件分别落在生产、分发、规范三个环节，评论自然有推进。
Take prompt 12, "Is bias inevitable in the production of knowledge?", and compare two selections. The weak one is "a phone, a book, a laptop". All three are too vague, so the commentary can only say that phones carry biased information, which is true of everything and therefore of nothing, with no context to identify and no evidence to use. The workable version swaps in three others: a screenshot of a report a news organisation publicly retracted in 2019, a screenshot of a recommendation page returning different results for the same keyword on two accounts, and a platform's published content moderation rules from 2020. Production, distribution and regulation each get one object, and the commentary has somewhere to go.
图片 引用 与文件形式
提交件里必须有三件物品的图片，可以是自己拍的，也可以是来源清楚的公开图片。引用和参考文献不计入 950 词，缺了它们，"有效使用证据"这一项就没支撑。各校的实物展出、拍摄和存档要求不同，动手前先问清本校 TOK 协调员的安排，别照抄网上的模板。截图类物品要保留可见的时间、平台标识与上下文，裁得太干净反而失去语境。
The file has to include images of the three objects, either your own photographs or clearly sourced public images. Citations and the bibliography fall outside the word count, but without them the strand on effective use of evidence has nothing behind it. Schools differ on physical display, photography and archiving, so check the local arrangement with your TOK coordinator before building anything rather than copying a template from the internet. Screenshots should keep the visible date, platform marks and surrounding context; cropping them too cleanly removes the very thing that makes them evidence.
草稿只给一次反馈机会
官方允许教师对一份草稿提供反馈，机会只有一次。用在已经成型的版本上，别用在大纲上。倒推时间表比较实用: 确认题目、写清三件物品的语境、写好三句贡献句、扩到 950 词、对照四个方面自评一遍，最后交草稿。很多学生把这次机会花在三千字的初稿上，拿回来的批注只能解决结构问题。
Teachers may give feedback on one draft, and only one. Spend it on a version that is already finished rather than on an outline. Working backwards helps: fix the prompt, pin the context of three objects, write the three contribution sentences, expand to 950 words, self-check against the four strands, then hand in the draft. Many students use the single chance on a draft three times too long and get back comments that can only fix the structure.
一个可以直接套用的骨架
标题用题目原文，下面一句说明三件物品各自回答题目的哪一部分。三节各三百词，顺序固定: 物品是什么并注明来源，它在现实中如何参与知识生产，它对题目的哪一处给出答案。收尾两句说明三件合起来补上了单独一件补不上的东西。写完再拿 12 个核心概念过一遍用词，evidence、certainty、truth、interpretation、power、justification、explanation、objectivity、perspective、culture、values、responsibility，哪些真用上了、哪些只是提了一句，一眼能看出来。
Quote the prompt as the title, then one sentence saying which part of it each object answers. Three sections of about three hundred words, always in the same order: what the object is and where it comes from, how it took part in producing knowledge, and which part of the prompt it answers. Close with two sentences on what the three together show that none of them shows alone. Then run the twelve concepts across your vocabulary: evidence, certainty, truth, interpretation, power, justification, explanation, objectivity, perspective, culture, values, responsibility. Which ones you actually used, and which you merely mentioned, becomes obvious.
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?
A-Level / GCSE / IB / KS2-KS3 全科辅导
16621398022 同微信
Visit aleveler.com
或截图后用「扫一扫 → 相册」选择本图