EDKS-ADV-ZH-0008 · Lesson No.008 · 简体中文辅助 · Advanced English discussion practice
Listen. Find the question. Help the group think.
先听。找准问题。帮助大家把问题想清楚。
You do not have to speak the most. You do not have to know everything. You need to notice what the conversation needs next: a clearer question, a better example, a checked fact, a fair challenge, or a useful conclusion.
你不必说得最多，也不必什么都知道。你需要看出讨论下一步缺什么：更清楚的问题、更合适的例子、核实过的事实、公平的质疑，或者有用的结论。
This lesson is for Mandarin-speaking learners who can already explain an opinion in English but find complex seminars and professional discussions harder to manage. We will move beyond preparing your own answer. You will practise helping several people understand one another, examine evidence and leave with an accurate account of what has—and has not—been agreed.
本课面向已经能够用英语解释观点，却在复杂研讨会或专业讨论中感到吃力的普通话母语学习者。我们不再只练“准备自己的答案”，而是练习让几个人真正听懂彼此、共同检查证据，并准确说明最后哪些问题达成了共识，哪些还没有。
Learning target: contribute, clarify, challenge, synthesise and close a discussion. C1–C2 describes the level of language we are working towards; completing this lesson is not a CEFR assessment or certificate.
学习目标：参与、澄清、质疑、综合和收束讨论。C1–C2 表示本课面向的语言能力方向；完成本课不等于获得 CEFR 等级认证。
Start with the story · 从情境开始 · Academic seminar · 学术研讨 · Professional meeting · 专业会议 · Practice and answers · 练习与解析 · Books and references · 书籍与参考
1. The sentence that changes the room | 一句话，让讨论重新开始
Alicia has prepared an excellent paragraph. She has an opening, a reason, an example and a carefully worded conclusion. When her turn comes, she delivers it without a major grammatical error. Beatrice then gives a different answer. Ciara responds to a detail from the source pack. A few minutes later, all three are speaking confidently, but nobody is sure whether they are answering the same question.
Alicia 准备了一个很好的段落：开头、理由、例子、结论都有。轮到她时，她几乎没有语法错误。接着 Beatrice 给出另一种答案，Ciara 又回应资料中的一个细节。几分钟后，三个人都说得很有信心，但没有人确定大家是不是在回答同一个问题。
The tutor does not ask Alicia for a more impressive adjective. Instead, she asks what happened between the turns. Alicia looks at the notes. One person was discussing whether the idea could work. Another was discussing whether it was affordable. The third was asking whether the evidence was strong enough to justify a decision. These are related questions. They are not the same question.
老师没有要求 Alicia 换一个更“高级”的形容词，而是让她检查几次发言之间发生了什么。她发现：一个人在谈方法能否奏效，一个人在谈是否负担得起，另一个人在谈证据够不够支持决定。这些问题有关联，却不是同一个问题。
On the next attempt, Alicia begins differently: “Before we compare the options, could we separate effectiveness from affordability? We seem to be using ‘better’ in two different ways.” Her sentence is not a miniature speech. It is a useful intervention. It gives the other participants a clearer place from which to continue.
第二次尝试时，Alicia 换了开头：“在比较选项之前，我们能不能先把有效性和负担能力分开？我们似乎在用两种不同的意思理解 better。” 这不是一篇缩短的演讲，而是一次有用的介入。别人终于知道接下来应该讨论什么。
That is the central distinction in this lesson. Speaking well means making your own meaning available. Leading a discussion also means helping the group work with the meaning already in the room. Sometimes the most advanced contribution is a carefully limited question.
这就是本课的核心区别：说得好，是让别人接收到你的意思；带好讨论，还要帮助大家处理已经出现的意思。有时，最有水平的贡献是一句范围准确的问题。
About the examples: Alicia, Beatrice and Ciara appear here as fictional learners in invented practice scenes. All study results, project details, deadlines and meeting records in the exercises are fictional. They are not testimonials, reports of actual research or instructions for a real organisation.
例子说明：以下人物和情境用于虚构教学练习。练习中的研究结果、项目资料、期限和会议记录均为编写的材料，不是真实学生见证、实际研究报告或某个机构的工作指令。
2. Leading does not mean taking control of everyone | 带领讨论，不等于控制所有人
There are at least three meanings of “lead” in a discussion. A chair manages the process. A subject specialist contributes knowledge. A participant may lead one moment by making a distinction that everyone needs. You can perform the third role without holding the first two. You can ask a good question without being the senior person. You can improve a summary without pretending to be the expert.
讨论中的 lead 至少有三种意思：主持人管理流程；领域专家提供知识；普通参与者也可以在某一刻提出大家需要的关键区分。你不必是主持人或专家，也能完成第三种贡献。
Keep authority separate from fluency. A polished English sentence does not give you permission to approve a budget, commit a colleague, settle a technical dispute or describe a proposal as an agreed decision. Practise saying, “We can recommend this option, but final approval is outside this group.” The language remains confident while the authority boundary remains honest.
要把表达能力和决定权限分开。英语流利，不代表你有权批准预算、替同事作承诺、裁定技术问题，或把提案说成已通过的决定。可以练习：“我们可以推荐这个方案，但最终批准不属于这个小组的权限。”
Also separate discussion from debate. A debate may ask you to defend an assigned position. A seminar may ask you to understand why interpretations differ. A professional meeting may need a workable next action despite remaining disagreement. Before you prepare persuasive language, identify which activity you are in. The best contribution depends on the actual purpose, not on how forcefully you can speak.
还要区分讨论和辩论。辩论可能要求你维护指定立场；研讨会可能需要理解不同解释为何出现；工作会议可能需要在仍有分歧时确定下一步。准备“说服别人”的语言之前，先判断当前活动的目的。
3. What makes this advanced English? | 高级，体现在哪里？
The Council of Europe’s spoken-language descriptors treat interaction and coherence as part of proficiency, alongside range, accuracy and fluency. At C1, they describe connecting a contribution skilfully to other speakers and using suitable language to enter or retain a turn. At C2, they describe flexible reformulation and close integration with the shared conversation. These are broader targets than delivering an isolated speech. [1]
欧洲委员会的口语描述把互动和连贯性，与语言范围、准确度和流利度放在一起。C1 强调恰当地接上别人的发言并管理话轮；C2 更强调灵活改述，以及自然地融入共同对话。这些目标都超出了“独立说一段话”。
Cambridge C1 Advanced provides another useful reference: its collaborative speaking task involves exchanging and evaluating ideas and negotiating a decision. Use that as one example of an advanced communicative demand, not as a template for every real meeting. Actual seminars may end with a better question rather than a decision. [2]
剑桥 C1 Advanced 的合作口语任务包含交流、评价和协商决定，可以作为高级交际要求的一个参考。但真实研讨会不一定要得出决定；有时，形成一个更好的问题就是成果。
Galaczi’s research on paired speaking tests also highlights listener support and turn-taking management, not only topic development. The study concerns an assessment setting; it does not establish that any particular exercise below guarantees improvement for all Mandarin speakers. Its useful reminder is that speaking ability is partly visible in what you make possible for another speaker. [3]
Galaczi 关于双人口语测试的研究，也强调对听者和说话伙伴的支持，以及话轮管理。研究发生在测试情境中，并不证明下面每个练习对所有普通话学习者都必然有效。值得带走的是：你的口语能力，也体现在你让对方能够完成什么。
4. Prepare a question, not a script | 准备问题，不只准备稿子
A script answers the question you expected. A useful preparation sheet helps you respond when the question changes. Divide a sheet into four small areas: the issue, the evidence available, the uncertainty and a possible contribution. Keep each area short enough to scan while listening. Long notes can become a second conversation competing with the one actually happening.
稿子回答的是你预想的问题；有用的准备表，则能帮助你在问题变化后继续回应。把纸分成四小块：当前问题、已有证据、不确定之处、可能的贡献。每块都要短，避免笔记变成另一场抢走注意力的“对话”。
For a seminar, write a question such as, “Does this evidence support the proposed explanation, or only the observed association?” For a project meeting, write, “What must be decided today, and what information could change that decision?” These questions do different work. The first tests interpretation. The second organises a decision under uncertainty.
研讨会可以准备：“这个证据支持所提出的解释，还是只支持观察到的相关关系？” 项目会议可以准备：“今天必须决定什么？哪些信息可能改变决定？” 前者检查解释，后者安排不确定条件下的决策。
Next, prepare one contribution and one question. Your contribution might be a limitation in the source. Your question might ask how that limitation affects the conclusion. Do not prepare six arguments that must all be delivered. A discussion is not successful because you emptied your notes into it. Choose the contribution that remains useful after hearing everyone else.
接着准备一个贡献和一个问题。贡献可以是资料的限制，问题则询问这个限制怎样影响结论。不要准备六个“非讲不可”的论点。好的讨论，不是把自己的笔记全部倒进去，而是听完别人以后仍能选出有用的那一点。
5. Keep source, claim and confidence together | 把来源、主张和把握程度放在一起
A useful source note answers three questions: where does the information come from, what exactly does it establish, and what does it leave unresolved? This is especially helpful when you are speaking in a second language, because a remembered headline can otherwise become stronger as you struggle to explain it.
一条有用的来源笔记要回答：信息来自哪里？它具体能支持什么？还有什么没有解决？用第二语言讨论时，这尤其重要，因为你在努力组织英语的过程中，很容易把记住的标题说得比原本更强。
Compare two contributions. “Research proves that the programme works.” This gives the room very little to inspect. “The report in our pack describes improvement in one pilot group, but there was no comparison group, so I would be cautious about attributing the whole change to the programme.” This second sentence identifies the source, observation and limitation. It creates several useful routes for a reply.
比较：“研究证明这个项目有效。” 大家几乎没有东西可以检查。另一种说法是：“资料包中的报告描述了一个试点组的进步，但没有对照组，因此我会谨慎看待把全部变化归因于该项目的说法。” 第二句同时交代来源、观察和限制，让别人有办法继续讨论。
When you have not checked a detail, keep that status visible. Say, “I remember the direction of the finding, but not the exact figure. Let’s not use the number until we check the source.” You have not weakened the whole discussion by admitting a memory limit. You have prevented an uncertain detail from silently becoming a shared fact.
没有核实的细节，要保留“未核实”状态：“我记得结果的大方向，但不记得准确数字。核对来源之前，我们先不要使用这个数字。” 这不是削弱讨论，而是防止不确定的记忆悄悄变成共同事实。
6. Stabilise the question before comparing answers | 比较答案之前，先固定问题
“Which option is better?” looks clear but may hide several questions. Better for whom? Better over what period? Better at reducing cost, improving access, increasing quality or making implementation easier? A group can disagree for twenty minutes while its members evaluate different outcomes. The first repair is not a stronger argument. It is a more precise question.
“哪个方案更好？”看起来清楚，却可能藏着很多问题：对谁更好？看多长时间？是成本、便利性、质量，还是实施难度？大家可能争论了二十分钟，其实一直在评价不同结果。
Use a clarification that offers a distinction rather than merely asking for repetition: “When we say ‘effective’, are we referring to immediate participation or learning that remains available later?” The listener can now choose between meanings, reject the distinction or propose another one. You are making the possible ambiguity inspectable.
澄清时，可以提供一个区分：“这里说 effective，是指即时参与度，还是指之后仍然保留的学习成果？” 对方可以选择、修正，或提出别的定义。你是在把歧义摆到大家面前，而不是只说“再讲一次”。
For professional discussion, put the decision into one sentence: “We are deciding whether to run a limited trial next month, not whether to adopt the system permanently.” This boundary can reduce unnecessary disagreement. Someone who rejects permanent adoption may still support gathering evidence through a restricted trial.
专业会议可以把任务说成一句话：“我们现在决定的是下个月要不要做一个有限试点，而不是要不要永久采用这个系统。” 拒绝永久采用的人，仍然可能支持通过小规模试点获得信息。
7. Enter the discussion by responding to something real | 进入讨论，要接住真实内容
A useful entry has an attachment point. It connects to a claim, an example, a question or a missing distinction already present. Compare “I would like to express my opinion” with “Your point about access changes how I would compare these two options.” The second begins a relationship before it begins an argument.
有用的发言入口要有一个连接点：接上已经出现的主张、例子、问题或缺失的区分。比较“我想表达一下我的观点”和“你关于使用机会的观点，改变了我比较这两个方案的方式”。第二句先建立关系，再展开论点。
Try the pattern reference → contribution → hand back. For example: “You mentioned that the evening sessions were fuller. Could that reflect timetable convenience rather than a preference for the teaching method? I’d like to hear how you would distinguish those explanations.” This is a genuine question because the answer could change your view.
可以练习“接上前文 → 提出贡献 → 交还话轮”：“你提到晚间场次人更多。这会不会反映时间更方便，而不是大家更喜欢教学方法？我想听听你会怎样区分这两个解释。” 因为对方的回答可能改变你的判断，所以这是真问题。
Avoid collecting elaborate entry phrases that you cannot complete naturally. “With due respect to the aforementioned argument” may create more distance than a simple “Can we test that assumption?” Formality should fit the setting. In a small working group, a clear question often serves the discussion better than a ceremonial introduction.
不要收集很多自己接不下去的华丽开场白。过分正式的铺垫可能制造距离。小组工作中，一句“我们能不能检查这个假设？”往往比仪式化的长开头更实用。
8. Build one useful turn | 把一次发言建好
A turn does not need to contain your entire position. It needs to do enough work for the next person to respond. A manageable contribution might acknowledge a point, name a distinction, offer a reason and ask a focused question. When the turn becomes too long, listeners may answer the final sentence while forgetting the important qualification in the middle.
一次发言不需要装下你的全部立场，只需要完成足够的工作，让下一个人能够回应。可以先承认一点，再提出区分、给出理由，最后问一个聚焦的问题。太长时，听者可能只回答最后一句，却忘了中间的重要限制。
Consider: “I agree that the feedback is quicker. My concern is whether it helps the learner revise independently. A faster response could still be difficult to use. Could we look at the revised work rather than response time alone?” Notice the movement. Agreement is limited to speed. The concern is about use. The proposed check concerns revised work. No part requires you to pretend that speed is irrelevant.
例如：“我同意反馈更快了。我担心的是，它是否帮助学习者独立修改。回复更快，也可能仍然难以使用。我们能不能看修改后的作品，而不只看回复时间？” 同意只限于速度；疑问转向使用；建议则是检查修改结果。
For practice, stop after one distinction and one question. Let another speaker respond. You can expand later if the room needs it. The constraint is not a rule for every conversation; it is a way to discover whether you can make a substantial point without hiding its purpose inside a long introduction.
练习时，可以限定自己只提出一个区分和一个问题，然后让对方回应。这不是每场对话的死规则，而是帮助你检查：能不能不靠长铺垫，也把重要观点说清楚。
9. Diagnose the disagreement | 先诊断分歧属于哪一种
Two people can use the same sentence while disagreeing at different levels. “The trial was successful” might describe a measured outcome, an interpretation of that outcome, a judgement against a criterion or a recommendation to continue. Before challenging the whole statement, locate the level that is actually contested.
“试点成功了”可能是在描述结果、解释结果、按某个标准评价，或建议继续。反对整句话之前，先找出真正有争议的层级。
Fact: “We have different attendance figures. Which record are we using?” Interpretation: “We agree attendance rose, but we have different explanations.” Criterion: “You are evaluating demand; I am evaluating learning.” Value: “We agree about the cost, but differ on who should bear it.” Action: “We agree that more evidence is needed, but not on whether to pause the trial.”
事实分歧：“我们的出席数字不同，用哪份记录？” 解释分歧：“我们同意人数上升，但解释不同。” 标准分歧：“你看需求，我看学习。” 价值分歧：“成本相同，但承担方式不同。” 行动分歧：“都认为需要更多证据，但对是否暂停意见不同。”
The repair follows the diagnosis. A factual discrepancy needs a source check. An undefined criterion needs agreement about the question. A value disagreement may remain even after every number is correct. Do not promise that one more statistic will remove a disagreement whose real basis is a different priority.
不同分歧需要不同修复。数字冲突需要核对来源；标准不明需要澄清任务；价值分歧则可能在所有数字核实后仍然存在。不要把“再找一个数字”当作解决所有分歧的方法。
10. Challenge without exaggerating the other person’s claim | 质疑时，不要把对方的话说大
The strongest challenge begins with accurate representation. If Beatrice says a small trial is worth considering, do not respond as though she recommended immediate permanent adoption. If Ciara questions the evidence, do not describe her as opposing the entire idea. Your response should preserve the other speaker’s scope before testing it.
有效质疑从准确复述开始。对方只说“小试点值得考虑”，不要当成“立即永久采用”；对方质疑证据，也不等于反对整个想法。先保留原本范围，再检查。
A useful pattern is: “You are proposing X rather than Y. My concern is Z. Would your proposal change if W?” For example: “You are proposing a trial rather than full adoption. My concern is that the trial excludes the learners most likely to struggle. Would you include an additional group before drawing a wider conclusion?”
可以练习：“你提议的是 X，而不是 Y。我担心 Z。如果 W，你会调整方案吗？” 例如：“你提出的是试点而不是全面采用。我担心试点没有包括最可能遇到困难的学习者。在得出更广泛结论前，你会加入另一组吗？”
Do not use politeness as a disguise for an accusation. “Would you agree that your conclusion is obviously unsupported?” remains confrontational despite the question form. A more workable question identifies the missing step: “Which evidence connects the improvement to the intervention rather than to extra practice?” The difference is not softness. It is precision.
不要用礼貌句型包装指责。“你是否同意你的结论显然没有依据？”仍然很有攻击性。更准确的是：“哪个证据把进步与干预联系起来，而不是与额外练习联系起来？” 区别不在于软弱或强硬，而在于是否具体。
11. Keep confidence proportionate to what is known | 让把握程度与证据匹配
Academic discussion needs ways to distinguish an observation, an interpretation and a recommendation. The University of Manchester’s Academic Phrasebank describes cautious language as a way to avoid unwarranted certainty and overgeneralisation. This is useful for speaking too, provided that you adapt written examples to natural conversation rather than reciting them unchanged. [4]
曼彻斯特大学 Academic Phrasebank 说明，谨慎表达可避免不合理的确定语气和过度概括。口语也需要这种区分，但要把书面表达改成自然说法，而不是原封不动背出来。
Try: “The attendance record shows an increase.” That reports an observation. “One possible explanation is that the new time was more convenient.” That proposes an interpretation. “I would test that explanation before expanding the programme.” That recommends an action. Each sentence makes its job visible.
“出席记录显示人数增加。” 是观察。“一种可能解释是新的时间更方便。” 是解释。“扩大项目前，我会先检验这个解释。” 是建议。每句话都让自己的任务清楚可见。
Do not turn modal verbs into fixed probability percentages. “May” and “might” do not provide a universal numerical confidence scale. Nor does adding three hedges automatically make a sentence responsible. Replace “It may perhaps possibly be somewhat useful” with a clear limit: “It may help learners who already understand the basic concepts; we do not yet know how it works for beginners.”
不要给 may、might 安排固定的概率百分比；也不要认为叠三个缓和词就更负责。可以直接说：“它可能帮助已经理解基础概念的学习者，但对初学者怎样起作用，我们还不知道。” 具体边界比含糊的弱化更清楚。
12. Worked seminar: did a discussion activity improve learning? | 完整研讨情境：讨论活动真的改善了学习吗？
Fictional source pack. Twenty-four volunteers joined four weekly discussion sessions. Their average score on the same short quiz rose from 6/10 before the sessions to 8/10 afterwards. There was no comparison group. Eighteen participants completed an optional feedback form; sixteen of those respondents described the sessions as useful. The participants were volunteers from one course, and nobody checked performance a month later.
虚构资料包：24 名志愿参与者参加了四次每周讨论活动。同一份小测的平均分由活动前的 6/10 升至活动后的 8/10。没有对照组。18 人填写了自愿反馈表，其中 16 人认为活动有用。参与者来自同一课程，一个月后的表现没有测量。
The seminar question is deliberately narrow: “What does this information justify us saying about the activity?” It is not “Should every course adopt it?” The narrower question allows the learners to practise source accuracy before moving to a recommendation.
研讨问题是：“这些信息允许我们对该活动作出什么判断？” 不是“所有课程都应该采用吗？”先缩小问题，才能练习证据范围。
Alicia: “The scores increased by two points, and most respondents found the sessions useful. That seems encouraging, but I would not yet call it proof that the discussion activity caused the improvement.”
中文：“分数上升了两分，多数反馈者也认为有用。这令人鼓舞，但我还不会说讨论活动被证明导致了进步。”
Beatrice: “What other explanation do you have in mind? The change happened after the sessions.”
中文：“你想到什么其他解释？变化是在活动之后发生的。”
Alicia: “The order is consistent with that explanation, but the participants took the same quiz twice. Familiarity with the questions could have contributed. We also do not know what else they studied during those weeks.”
中文：“时间顺序与这个解释相符，但他们做了同一份测验两次。题目熟悉度可能有影响，而且我们不知道期间还学了什么。”
Ciara: “So are you saying the activity did not work?”
中文：“所以你是说活动没有效果？”
Alicia: “No. I am separating an encouraging observation from a causal conclusion. The data do not establish that the activity failed either.”
中文：“不是。我在区分令人鼓舞的观察和因果结论。数据也没有证明活动失败。”
This short repair matters. Alicia prevents the conversation from collapsing into a false choice between “proved effective” and “proved ineffective”. She does not answer uncertainty with another unsupported certainty. She identifies the narrower position that the evidence permits.
这次修复避免了“已经证明有效”和“已经证明无效”的假二选一。Alicia 没有用另一种没有依据的确定说法回答不确定性，而是保留证据允许的较窄判断。
Beatrice: “That is fair. Then perhaps the feedback tells us that participants valued the experience, even if the quiz cannot tell us how much learning it caused.”
中文：“有道理。那么反馈或许告诉我们参与者重视这段体验，即使测验无法说明它带来了多少学习。”
Ciara: “I would narrow that slightly: sixteen of the eighteen respondents said it was useful. We do not know the views of the six people who did not return the form.”
中文：“我会再缩小一点：18 名反馈者中有 16 人认为有用。剩下 6 名没有填表的人怎样想，我们不知道。”
Beatrice: “Agreed. So we have positive feedback among respondents, not a complete measure of every participant’s view.”
中文：“同意。我们有的是反馈者中的积极评价，而不是每个参与者意见的完整记录。”
Notice the exact language in the repair: “among respondents” protects the denominator. Ciara is not accusing Beatrice of dishonesty. She is making the population claim match the source. Beatrice accepts the correction without surrendering the useful information that remains.
among respondents 保护了分母范围。Ciara 不是指控 Beatrice 不诚实，而是让“关于谁的结论”与资料一致。Beatrice 接受修正，也没有把仍然有用的信息全部丢掉。
Alicia: “Could we now separate what we observed, what we suspect and what we would test next?”
中文：“现在能不能把观察到的、推测的，以及下一步要检验的分开？”
Ciara: “We observed a higher average quiz score and positive feedback from most respondents. We suspect the sessions may have contributed, but the design does not isolate that contribution.”
中文：“观察到的是平均测验分数提高，以及多数反馈者的积极评价。我们推测活动可能有贡献，但设计没有把这种贡献单独分离出来。”
Beatrice: “For the next trial, I would want a suitable comparison and a later check using a different but comparable task. That would help us ask whether any improvement lasts and transfers beyond the original questions.”
中文：“下一轮我会希望有合适比较，并用不同但可比较的任务做延后检查。这样才能问：进步是否保持，是否超出原题。”
Alicia: “Then our conclusion is not that discussion activities always improve learning. It is that this small pilot gives us a reason to investigate further, with important uncertainty about cause and transfer.”
中文：“所以结论不是讨论活动总能改善学习，而是这个小试点提供了进一步研究的理由，但因果和迁移仍有重要不确定性。”
Model closing statement: “In this fictional pilot, average performance on the repeated quiz improved and most feedback respondents valued the sessions. Those findings are encouraging, but they do not isolate the activity’s causal effect or establish longer-term transfer. A further study would need a stronger comparison and a delayed, different task.”
示范结语：“在这个虚构试点中，重复测验的平均表现提高，多数反馈者也认可活动。这些结果令人鼓舞，但不能分离活动的因果效果，也不能证明长期迁移。进一步研究需要更有力的比较和延后的不同任务。”
13. What changed in that seminar? | 刚才的讨论，具体进步在哪里？
The participants did not discover a new statistic. They changed the structure of the claim. The initial sentence risked blending improvement, satisfaction and causation. The final statement separated them. This is a valuable speaking target because it can be heard in the actual turns: a challenged assumption, a corrected denominator, a distinction between absence of proof and proof of absence, and a more defensible conclusion.
参与者没有找到新数字，而是改变了主张的结构。原来的说法容易混合成绩进步、满意度和因果；最后则把它们分开。你能从发言中直接听见学习目标：检查假设、修正分母、区分“尚未证明”和“证明没有”，形成更能站得住的结论。
Several language choices deserve attention. “Could have contributed” leaves room for other causes. “I would narrow that slightly” introduces a precise correction without rejecting the whole contribution. “The design does not isolate…” questions what the information supports, not the competence of the speaker. “A reason to investigate further” states a legitimate outcome short of certainty.
几个表达值得注意：could have contributed 保留其他原因；I would narrow that slightly 引入局部修正；the design does not isolate 检查证据能力而不是攻击人；a reason to investigate further 则承认“值得继续研究”本身就是合理结果。
For practice, repeat the seminar with different roles. Let the person who originally questioned the evidence make the optimistic interpretation. Let another person correct the source boundary. The goal is not to assign one learner the permanent identity of “the critical one”. Every learner should practise building, checking, accepting correction and summarising.
练习时交换角色。不要让某个人永远负责“挑错”。每个学习者都应练习提出观点、检查证据、接受修正和总结。
14. Synthesise without inventing agreement | 综合观点，但不要制造共识
A summary reports contributions. A synthesis explains how they fit together. “Alicia discussed cost, Beatrice discussed access and Ciara discussed staffing” records three topics. “We agree that access is limited, but the disagreement is about whether a longer opening schedule is workable with the available staff” identifies a shared problem and a contested explanation.
总结可以报告各自说了什么；综合则解释这些内容如何连接。“Alicia 谈成本，Beatrice 谈使用机会，Ciara 谈人员”只是列题目。更有用的是：“大家同意使用机会有限，分歧在于现有人员能否支持更长开放时间。”
Be careful with “we agree”. It is an important social claim. Do not infer agreement from silence, from a nod whose meaning is uncertain, or from the fact that nobody interrupted you. Offer your synthesis for checking: “That is my understanding of our current position. Is that accurate, or have I missed a reservation?”
we agree 是重要的社会性主张。不要把沉默、含义不明的点头或没人打断，当成同意。可以说：“这是我对当前立场的理解。准确吗？还是我漏掉了谁的保留意见？”
A strong synthesis can preserve several outcomes at once: agreement on the problem, disagreement on the preferred solution, uncertainty about a fact and agreement on how to investigate it. There is no need to smooth these into a false single position. Precision often means keeping the unresolved parts visible.
好的综合可以同时保留：对问题的共识、对方案的分歧、对事实的不确定，以及对查证方法的同意。不必把它们抹平成一个假的统一立场。
15. Translate the communicative job, not the Chinese surface | 翻交际功能，不照搬中文表面
Mandarin can be an excellent resource for understanding what you intend. The next step is to identify the conversational job. A Chinese expression that signals modesty, a soft challenge or a temporary reservation may need a different English structure depending on the audience. There is no single correct English replacement for every Chinese phrase.
中文可以帮助你理解自己的意图，但接着要判断交际任务：谦逊、轻度质疑、暂时保留，还是不同意？不同对象可能需要不同英文结构，不存在每个中文短语都对应一个永久固定英文说法。
我补充一点。 When you are extending a point, try “There is one further implication of what you said.” When you are adding a practical detail, try “Could I add a practical consideration?” The first continues reasoning; the second introduces a new constraint. Neither requires the stiff phrase “I supplement one point”.
我补充一点：延伸推理时，可以说“你刚才那一点还有一个进一步含义”；加入实际限制时，可以说“我能不能补充一个实际考虑？”两者的功能不完全相同。
这个问题不能一概而论。 Instead of stopping at “It depends”, identify the variable: “The answer changes with the learner’s starting level.” Or: “We should distinguish a temporary trial from permanent adoption.” You have turned a broad warning into a distinction the group can use.
不能一概而论：不要只停在 it depends。要指出依赖什么：“答案随学习者起点而变化”，或者“要区分临时试点与永久采用”。
我不是反对，只是有顾虑。 Protect both parts: “I support the aim, but I have a reservation about the proposed schedule.” This is more informative than “I have some concerns”, because the room can see exactly where agreement ends and the unresolved issue begins.
我不是反对，只是有顾虑：可以说“我支持目标，但对目前安排的时间有保留”。这样大家知道你同意什么，又把什么保留下来。
原则上可以。 Do not translate this as unconditional approval. Say “I can support the proposal in principle, subject to confirming the staffing arrangement.” Then make clear whether you are offering a personal view, a recommendation or an authorised decision.
原则上可以：不能自动说成无条件批准。可以说“原则上我可以支持，但要先确认人员安排”，并说明这是个人意见、建议还是有权限的决定。
16. Facilitate participation without putting people on the spot | 让别人参与，但不要当众施压
Making space is not the same as demanding an immediate performance. A quiet participant may be thinking, checking the source, waiting for a suitable gap or choosing not to contribute yet. Do not diagnose the person from one moment of silence. Offer a clear route in: “We have heard the cost argument. Does anyone want to examine the access side before we move on?”
给空间，不等于要求立刻表演。安静的人可能在思考、核对来源、等待话轮，或暂时选择不发言。不要从一次沉默诊断一个人。可以提供具体入口：“成本方面已经谈过了，继续之前，有没有人想检查使用机会这一面？”
When you do invite a named person, explain why their perspective is relevant and allow an alternative: “Ciara, you reviewed the feedback forms. Is there a limitation we should keep in mind? You can also add it to the shared notes.” This asks for a bounded contribution rather than “Tell us everything you think”.
点名邀请时，说明相关性并给另一种参与方式：“Ciara，你看过反馈表。有没有我们应该保留的限制？也可以先加在共同笔记里。”任务范围清楚，也不要求即刻说出全部想法。
British Council guidance on group discussion recommends teaching component skills and giving focused feedback rather than merely announcing a topic and telling learners to talk. Its questioning guidance also emphasises allowing thinking time. Our practical application is to make participation routes clear, then examine whether they actually help the particular group. [5] [6]
英国文化教育协会建议把讨论拆成可教的技能，并给予聚焦反馈，而不只是给个题目就让学生说；提问指导也强调思考时间。本课把这些原则应用为清楚的参与入口，再观察它是否适合当前小组。
17. Handle interruption and return to the question | 处理打断，再回到问题
An interruption can mean several things: a clarification, an urgent correction, enthusiasm, disagreement or a struggle for the floor. Respond to its function. If someone corrects an important number, check it before continuing. If they introduce a different issue, acknowledge it and decide when to return. Treating every interruption as disrespect can make useful repair harder.
打断可能是澄清、重要纠正、兴奋、不同意或争取话轮。先看功能。别人纠正关键数字，就先核实；引入新问题，就承认它并安排何时回来。把所有打断都当成不尊重，会妨碍修复。
Use a short holding move: “Let me finish the distinction, then I will come to the cost question.” Use a repair when you entered too early: “Sorry, I interrupted before you gave the condition. Please finish that part.” Use a return when the room has drifted: “How does this change the decision we need to make today?”
保持话轮：“让我先说完这个区分，再回应成本问题。” 修复自己的打断：“抱歉，你的条件还没讲完，我就插进来了。请先说完。” 拉回主线：“这一点怎样影响我们今天需要作出的决定？”
If one speaker dominates, discuss process rather than personality: “We have several views still to hear. Could we take one concise point from each person before returning to the open discussion?” This creates a temporary structure. It does not label someone as selfish or assume that equal seconds will produce equal value.
有人占用过多时间时，调整流程，不攻击人格：“还有几个观点没听到。能不能先让每个人简要说一点，再恢复开放讨论？”这是暂时结构，不是在给某人贴标签。
18. Worked meeting: recommend a limited pilot without overpromising | 完整专业情境：提出有限试点，而不作过度承诺
Fictional brief. A training centre is considering a new room-booking process. The group can recommend a trial but cannot authorise a permanent change. Option A keeps the current spreadsheet. Option B tests a shared booking form with twelve volunteer users for two weeks. Option C switches every user immediately. The proposed form has not yet been tested by the volunteer group. One member will be unavailable during part of the second week. The group must identify a recommendation and the checks required before it starts.
虚构任务：一个培训机构考虑更换教室预约流程。小组只能推荐试点，无权批准永久变更。A 保留现有表格；B 让 12 名志愿用户试用共享预约表两周；C 立即切换全部用户。志愿组尚未测试新表格，第二周有一名成员部分时间不在。小组需要提出建议，并说明启动前必须完成什么检查。
The task is not to deliver a persuasive speech in favour of Option B. It is to decide what can responsibly be recommended from the information available. You may choose another option if you justify it and preserve the facts. The dialogue below is one possible response, not the only acceptable answer.
任务不是替 B 写一篇说服稿，而是根据现有信息提出负责任的建议。只要理由和事实成立，也可以选择别的方案。下面是一个可能的讨论版本，不是唯一正确答案。
Alicia, chair for the exercise: “By the end of this discussion, we need a recommendation, not a final authorisation. Shall we compare the three options against disruption, support needs and the information each would give us?”
中文：“讨论结束时，我们需要的是建议，不是最终批准。我们按干扰程度、支持需求和能获得的信息来比较三个方案，好吗？”
Beatrice: “I favour the limited trial. It gives us experience without asking everyone to change at once.”
中文：“我倾向有限试点。它让我们获得经验，又不用所有人同时改变。”
Ciara: “I agree with the scale, but I would not describe the form as ready yet. The volunteer group has not tested it.”
中文：“规模上我同意，但我还不会说表格已经准备好了。志愿组还没测试。”
Alicia: “Then let’s distinguish support for the idea from readiness to start. What would we need to check before inviting the volunteers?”
中文：“那我们把支持想法和已经可以启动分开。邀请志愿者之前，需要检查什么？”
Beatrice: “At minimum, that a user can book, correct and cancel a reservation, and that the administrator can see the final booking state. I would also want a clear route back to the current process if the trial is stopped.”
中文：“至少要检查用户能预约、更正和取消，管理员能看到最终预约状态。若试点停止，也要有回到现有流程的清楚办法。”
This turn converts “ready” into observable actions. The language is advanced because it makes the criterion specific. It does not rely on a vague assurance such as “the system should be fine”. The condition is something the group can check.
这次发言把 ready 变成可观察动作，而不是“应该没问题”的模糊保证。高级之处在于标准具体，大家知道怎样检查。
Ciara: “There is also the support issue. I am away on Thursday and Friday of the second week. I can prepare the guide, but I cannot promise to handle questions on those days.”
中文：“还有支持安排。我第二周周四和周五不在。我能准备指南，但不能承诺那两天处理问题。”
Alicia: “Thanks for separating those commitments. Preparing the guide is one task; providing live support is another. Who could cover the two days, or should the trial dates change?”
中文：“谢谢你把承诺分开。准备指南和即时支持是两件事。谁能覆盖那两天？还是要改试点日期？”
Beatrice: “I can cover Thursday, but Friday is uncertain. Rather than assume coverage, I suggest we make a confirmed support arrangement a condition of the recommendation.”
中文：“周四我可以，但周五不确定。我建议把支持安排已确认，写成推荐方案的条件，不要直接假设有人。”
Alicia: “So far, we support recommending a two-week trial with twelve volunteers, subject to a successful process check and confirmed support. We have not agreed a start date. Is that an accurate summary?”
中文：“目前我们支持建议 12 人参加两周试点，条件是流程检查通过、支持安排确认。开始日期还没定。这样总结准确吗？”
Ciara: “Yes, with one addition. We should decide what we will observe during the trial. Otherwise we may finish with anecdotes rather than enough information to compare the options.”
中文：“准确，但还要加一点：要先确定试点观察什么，否则结束时可能只有零散故事，仍不足以比较方案。”
Beatrice: “We could record whether users complete each booking task, where they need help and whether any booking needs to be corrected. Those observations would guide our recommendation; they would not prove that the process is suitable for every user.”
中文：“可以记录任务是否完成、哪里需要帮助、哪些预约需要纠正。这些观察用于建议，不证明流程适合所有用户。”
Alicia: “Then the next action is to prepare a trial proposal with those conditions and observations, and send it to the person authorised to approve it. No one should invite users before that approval.”
中文：“下一步是把这些条件和观察项目写入试点提案，交给有批准权限的人。在批准之前，不邀请用户。”
In this dialogue, the strongest English is not the most enthusiastic English. “Subject to”, “we have not agreed”, “I can prepare…but I cannot promise” and “the person authorised to approve” protect important distinctions. The participants move the discussion forward without making the outcome sound more complete than it is.
这段对话中，最好的英语不是最热情的英语。subject to、we have not agreed、I can…but I cannot promise 等表达保护了条件、未定状态和权限。讨论前进了，但没有把结果说得比实际更完成。
19. Write the outcome so another person can act | 把结果写到别人能接着做
A meeting can feel productive and still leave an unusable record. “We discussed the new form and will move forward” hides the recommendation, conditions, responsibilities and unresolved points. The reader may assume that approval already exists or that a start date has been agreed. A useful record separates those items.
“我们讨论了新表格，接下来推进”听起来积极，却隐藏了建议、条件、责任和未定事项。读者可能误以为已经批准或已经定日期。可用记录必须把这些分开。
Model record for the fictional meeting: “The group recommends a two-week booking-process trial with twelve volunteers. This recommendation is conditional on approval, completion of the booking/correction/cancellation checks, a documented return to the existing process and confirmed support coverage. The start date is not agreed. Ciara will prepare the guide; live support responsibilities remain partly unresolved. The proposal will identify task completion, requests for help and corrected bookings as trial observations. Permanent adoption is outside the group’s authority.”
虚构会议示范记录：“小组建议 12 名志愿用户进行两周预约流程试点。建议以取得批准、完成预约/更正/取消检查、有记录的旧流程恢复办法和支持人员确认作为条件。开始日期未定。Ciara 准备指南；即时支持责任仍有部分未解决。提案将记录任务完成、求助和预约更正情况。永久采用不属于小组权限。”
Compare that record with the dialogue. Each statement should have a traceable basis. The guide assignment was accepted; Friday support was not. A recording or automated transcript is not a substitute for this check, because accurate words can still be summarised into an inaccurate commitment. Ask participants to confirm consequential responsibilities rather than relying on an optimistic paraphrase.
把记录和对话逐项比较。准备指南已经接受，周五支持没有。即使逐字稿准确，摘要仍可能变成错误承诺，所以重要责任要请当事人确认，不能靠乐观转述。
20. Close a discussion without forcing an artificial ending | 收束讨论，不制造假的圆满
A discussion can end in a decision, a recommendation, a clarified disagreement or a next investigation. Name which one you have. “We have agreed the next check” is not the same as “We agree on the explanation”. “We prefer Option B” is not the same as “Option B is approved”. Each ending is useful when it accurately represents the conversation.
结束可能是决定、建议、澄清后的分歧，或下一项调查。说清是哪一种。“同意下一步查什么”不等于“同意解释”；“偏好 B”不等于“B 已批准”。准确的不同结果都可以有价值。
Use a three-part close: what is established → what remains open → what happens next. For a seminar: “We agree that the result is encouraging. We differ on how much causal weight it carries. The next useful step is to inspect the comparison and follow-up evidence.” For a meeting: “We recommend the limited trial. Support coverage remains unconfirmed. The proposal will return for approval after that gap is resolved.”
可以用三段式收束：“已确认 → 仍未解决 → 下一步”。这样既不拖延，也不假装所有问题都解决。
A final check is often worth more than another closing speech: “Does this summary preserve the reservation you raised?” It gives someone one last opportunity to correct the shared record. You are not reopening everything. You are checking whether the ending accurately represents where the group arrived.
最后一句“这个总结保留了你刚才的顾虑吗？”往往比再讲一段漂亮结语更有用。你不是重开所有问题，而是在检查结束是否准确。
21. When new information arrives late | 新信息最后才出现，怎么办？
Imagine that the group is ready to recommend Option B when another participant says the room is unavailable during the proposed trial. Do not protect the recommendation merely because everyone has invested time in it. Equally, do not throw away the entire discussion. Identify which part the new information affects.
假设小组即将推荐 B，最后却有人指出试点期间教室不可用。不要因为大家花了时间就保护原建议，也不要把整场讨论全部推翻。先看新信息影响哪一部分。
Try: “That changes the proposed dates, but not necessarily our view of the trial design. Let’s check whether an alternative room or a later period is available before revisiting the whole recommendation.” If the new fact instead concerns a missing essential capability, the affected scope may be larger. Your language should show the size of the change rather than treating every update as equally serious.
可以说：“这改变了日期，但未必改变我们对试点设计的看法。先检查其他教室或较后时段，再决定是否重审整个建议。” 若新信息涉及核心能力缺失，影响范围可能更大。要说清改变有多大。
Useful phrases include “That changes one assumption”, “This does not affect the earlier observation, but it weakens our explanation”, and “We need to reopen the decision because the condition on which it depended is no longer met.” These are not interchangeable. Practise matching the phrase to the actual dependency.
“一个假设变了”“观察没变，但解释变弱了”“决定依赖的条件不成立，需要重开决定”是不同关系，不能随意互换。高级讨论需要把依赖关系说出来。
22. Keep spoken, written and online contributions connected | 把口头、书面和线上贡献接起来
In a mixed discussion, one person may speak while another adds a question in shared notes. A remote participant may have heard only part of an exchange. Without a deliberate bridge, the room can split into parallel conversations. Use a neutral connection: “There is a question in the notes about the sample. Let’s include it before closing the evidence section.”
有人发言，有人在共同笔记里提问，远程参与者可能只听到部分交流。若没有连接，讨论会分成平行对话。可以说：“笔记里有一个关于样本的问题，结束证据部分前，我们把它一起处理。”
Do not interpret a delayed response as agreement or indifference. First check the communication channel: “Did the question come through clearly?” If someone lost the thread, offer the last stable point and the exact open issue, rather than restarting the whole meeting. “We have agreed the trial scope; we are now checking support coverage” is an efficient re-entry.
不要把延迟回应解释成同意或不在乎。先确认问题有没有传到。有人掉线时，给出最后稳定状态和当前问题，不必重讲全部内容：“试点范围已经同意，现在检查支持人员。”
For practice recordings, ask everyone’s permission and use fictional or de-identified material. Do not upload confidential workplace discussions into an external transcription service merely to practise English. A short invented role-play can provide useful language evidence without exposing another person’s private information.
录制练习前取得参与者同意，使用虚构或去识别材料。不要只是为了练英语，就把保密工作讨论上传到外部转录服务。短小的虚构角色扮演也能提供有用练习。
23. Recover when you miss a point | 漏听以后，怎样恢复？
Recovery begins by naming the smallest gap you can identify. “I don’t understand anything” gives the speaker no route to help. “I followed the cost comparison, but I missed why the second option needs more support” identifies a specific missing connection. It also shows that you were listening to the earlier part.
恢复时，说出能定位的最小缺口。“我全都不懂”很难帮助对方修复；“成本比较我跟上了，但我漏听了第二方案为什么需要更多支持”则明确指出缺失关系。
After the clarification, demonstrate the repaired connection: “Right—the extra support is needed during the changeover, not permanently. That makes the short-term comparison different.” This checks understanding and returns something useful to the group. It is better than saying “yes” and silently hoping you have reconstructed the point correctly.
澄清后把修复结果说出来：“明白，额外支持是转换期间需要，不是永久需要。这会改变短期比较。” 既检查理解，也回到讨论，而不是只说 yes。
When the gap concerns a word, first ask whether the word is essential. If it is, request a definition in context: “When you say ‘robust’, do you mean that the result survives different assumptions, or simply that the effect is large?” This may help fluent speakers too, because technical-sounding words can conceal differences in meaning.
若缺的是词，先判断是否关键。关键时问语境中的定义：“这里 robust 是指换了假设结果仍成立，还是只指效果大？”技术感很强的词也可能隐藏歧义。
24. A phrase clinic for advanced Mandarin-speaking learners | 普通话学习者的高级表达诊所
Do not memorise this section as a performance. Choose a function that repeatedly causes difficulty, learn two alternatives, and use them in a new situation. The examples below are original practice formulations; they are not mandatory academic formulas.
不要把这部分背成表演。挑一个反复困难的功能，学两种说法，换情境使用。以下为原创练习表达，不是必须使用的学术公式。
Clarifying scope: “Are we talking about all users, or only those in the trial?” / “Would you limit that conclusion to the group we observed?”
澄清范围：“说的是所有用户，还是试点用户？”／“这个结论要不要限于观察到的小组？” The second form is useful when the evidence is sound but the conclusion is too wide.
Separating issues: “Could we treat cost and reliability as two separate questions?” / “I think we are mixing the aim with the method.”
分开问题：“能否把成本和可靠性分开？”／“我们似乎把目标和方法混在一起了。” Explain the distinction immediately; otherwise this can sound like an abstract criticism.
Building on a contribution: “Your example suggests a condition we have not considered.” / “That point strengthens the case for a smaller first step.”
接上观点：“你的例子提示了一个尚未考虑的条件。”／“这一点让先走小一步更有理由。” Name the condition or explain why a smaller step follows.
Testing an assumption: “Would the argument still hold if the participants had no prior experience?” / “What would have to be true for that recommendation to work?”
检查假设：“参与者没有经验时，还成立吗？”／“建议要奏效，需要哪些条件？” These invite examination rather than pretending the answer is already settled.
Correcting the record: “One detail needs correcting: the report describes a recommendation, not an approval.” / “I may have stated that too strongly; the result suggests an association.”
修正记录：“一处要更正：报告写的是建议，不是批准。”／“我刚才可能说得太强；结果提示相关关系。” Correct the exact problem, then continue.
Offering a qualified position: “I would support the trial, provided that the support arrangement is confirmed.” / “I agree about the potential benefit, but not yet about the scale.”
有条件支持：“支持安排确认后，我支持试点。”／“潜在好处我同意，但规模还不能同意。” These preserve support and its boundary together.
Marking an unverified memory: “I recall a similar finding, but I would need to check the source before relying on it.” / “I do not have the exact figure in front of me.”
标明未核实：“我记得类似结果，但要核实才能依靠它。”／“准确数字不在手边。” Do not follow the disclaimer by treating the remembered figure as established anyway.
Inviting a reservation: “What objection would remain even if we resolved the timing issue?” / “Does anyone see a condition under which this would fail?”
邀请保留意见：“时间解决后，仍有什么反对理由？”／“什么条件下可能失败？” Keep the invitation open long enough for a real answer.
Synthesising: “We agree on the problem; the unresolved point is the proposed mechanism.” / “Both examples can be true if the starting conditions differ.”
综合：“问题有共识，机制仍有分歧。”／“起点不同，两个例子可能同时成立。” Check that the sources actually support your reconciliation.
Closing: “Our next step is a source check, not a final decision.” / “The recommendation is ready to submit, but approval remains outstanding.”
收束：“下一步是查证，不是最终决定。”／“建议可提交，但仍待批准。” These prevent the ending from overstating the state of the work.
25. Practice set: choose the next useful move | 练习：选择下一步有用行动
Try the tasks aloud before reading the model responses. More than one answer may work. Evaluate whether your response preserves the facts, matches the conversational purpose and gives the other person a sensible next turn. Difficult vocabulary is optional; accurate relationships are not.
先说答案，再看示范。题目可能有多个合理答案。检查是否保留事实、符合交际任务，并让对方能够继续。难词不是必须，准确关系才是。
Task A: an undefined comparison | 未定义的比较
A participant says, “The second method is clearly better.” You do not yet know the criterion. Produce one question that can clarify the comparison without rejecting it.
对方说“第二种方法明显更好”，标准尚不清楚。用一句问题澄清，而不是立即反对。
Task B: a claim made wider | 主张被说大了
Your colleague says, “I would consider a limited trial.” Another person replies, “So you want us to replace the entire process immediately.” Repair the representation before the debate continues.
同事说“可以考虑有限试点”，却被解释成“立即替换整个流程”。先修复复述。
Task C: a misleading denominator | 分母不清楚
Twenty people joined a fictional workshop. Twelve returned a form and ten respondents gave positive feedback. Someone says, “Almost everyone liked it.” Offer a precise correction without discarding the positive feedback.
虚构工作坊 20 人参加，12 人交表，10 名反馈者积极评价。有人说“几乎所有人都喜欢”。准确修正，但不要抹掉积极信息。
Task D: recommendation mistaken for approval | 建议误作批准
The group can recommend an option but has no approval authority. A participant begins assigning implementation work. What should you say?
小组只能推荐，没有批准权限，却有人开始分派实施工作。你怎样回应？
Task E: a missing connection | 关系漏听
You heard the difference in cost but missed why one option takes longer. Ask for the smallest useful clarification.
你听懂成本差别，漏听其中一个方案为什么更慢。请求最小范围的澄清。
Task F: a condition disappears | 条件消失
You said, “I can support this if the schedule changes.” The summary says, “Everyone supports the current proposal.” Repair it.
你说“时间调整后可以支持”，总结却写“大家支持目前提案”。修复。
Task G: a value disagreement | 价值分歧
Both speakers accept the same cost figures. One prioritises equal access; the other prioritises the smallest total expense. What distinction could stop them repeating the figures?
两人同意成本数字，一人优先平等机会，另一人优先最低总支出。怎样避免重复报数字？
Task H: apparent agreement in silence | 沉默被当成同意
No one replies immediately after the chair’s proposal. The chair says, “Good, we have unanimous agreement.” Offer a process check.
提案后没人立即回应，主持人说“很好，一致同意”。给出流程检查。
Task I: a late correction | 最后才出现的更正
The group recommended a trial on the assumption that the room would be available. The room is now confirmed unavailable. Preserve the useful earlier work while reopening the affected part.
试点建议依赖教室可用，但现在确认不可用。保留前面有用工作，同时重开受影响部分。
Task J: an unknown statistic | 不确定的数字
You remember a statistic that would support your view but cannot identify its source. Make a contribution that does not turn the memory into evidence.
你记得一个支持自己观点的数字，却说不出来源。怎样贡献而不把记忆说成证据？
Task K: a one-sided synthesis | 单方面综合
A summary records the preferred option but omits a participant’s concern about support coverage. Add the reservation accurately.
总结记录偏好方案，却漏掉支持人员顾虑。准确补充。
Task L: an unresolved but productive ending | 未解决，也可以有成果
A seminar has clarified two competing explanations but cannot distinguish between them with the available sources. Close it in two sentences without declaring a winner.
研讨会澄清了两个竞争解释，但来源不足以区分。用两句收束，不宣布胜者。
26. Model responses and why they work | 示范回应与解析
A: “Better in terms of which outcome—cost, reliability or ease of use?”
中文：“是哪种结果更好：成本、可靠性还是易用性？” This identifies possible criteria while allowing the original speaker to choose another one. It does not assume that the claim is false.
B: “I understood that as a proposal to test a limited change, not to replace the whole process. Could we evaluate the trial on that basis?”
中文：“我理解是测试有限变化，不是替换全部流程。能否按试点来评价？” The correction restores scope before the argument grows around a distorted position.
C: “The feedback was positive among most respondents: ten of twelve. We do not know the views of the eight people who did not return the form.”
中文：“多数反馈者是积极评价：12 人中 10 人。未交表的 8 人怎样想，我们不知道。” The answer keeps the favourable result but does not silently treat non-response as approval.
D: “Before we assign implementation tasks, we need to distinguish our recommendation from authorisation. Could we prepare the proposal for approval first?”
中文：“分派实施任务前，要分清建议和批准。先准备提案报批，好吗？” The response protects authority without rejecting the substantive idea.
E: “I followed the cost comparison, but missed the link to timing. Why would the second option take longer?”
中文：“成本比较我跟上了，但漏了与时间的关系。第二方案为什么更久？” Naming what you understood makes the repair small and useful.
F: “Could we keep my condition in the summary? I support the proposal only if the schedule is revised; I have not agreed to the current timetable.”
中文：“总结能否保留我的条件？时间表修改后我才支持，现时间表尚未同意。” This is a correction of the shared record, not a new objection.
G: “It seems we agree on the figures but are weighting access and total cost differently. Shall we make that trade-off explicit?”
中文：“数字有共识，但对机会和总成本的权重不同。能否明确这个取舍？” More repetition of the same figures would not resolve the underlying priority difference.
H: “Could we check for reservations before recording agreement? Some participants may still be considering the proposal.”
中文：“记录同意前，能否检查有无保留意见？可能有人还在考虑。” This does not claim that anyone disagrees; it asks for confirmation of the claimed consensus.
I: “That removes one condition for the proposed dates. We can retain the trial design while checking another room or a later period, then confirm whether the recommendation still works.”
中文：“原日期的一个条件不成立。先保留试点设计，查其他教室或较后时段，再确认建议是否仍可行。” The response identifies the affected dependency without protecting an invalid schedule.
J: “I recall a figure on this, but I cannot verify it now, so I would not rely on the number. The question it raises is whether the effect differs across groups; could we check that in the sources?”
中文：“我记得一个数字，但现在无法核实，先不依靠它。它提示的问题是不同群体效果是否不同，我们能否查来源？” A memory becomes a question, not evidence.
K: “The group prefers the limited trial, with a reservation about support coverage that remains unresolved. The recommendation should retain that condition.”
中文：“小组偏好有限试点，但支持人员仍有未解决顾虑，建议应保留该条件。” Preference and readiness remain separate.
L: “We have clarified two plausible explanations, but the present sources do not allow us to choose between them. The next useful step is to identify evidence that would distinguish their predictions.”
中文：“两个解释已澄清，但现资料不足以选择。下一步是找能区分两者预测的证据。” The ending identifies progress and a testable gap without manufacturing certainty.
27. Extend the practice: one conversation, three outcomes | 延伸练习：同一对话，三种成果
Use the fictional room-booking brief again, but change the required output. In the first round, the group must write a recommendation. In the second, it must identify why the evidence is insufficient. In the third, it must prepare three questions for the person who holds the missing information. Keep the source facts identical.
继续使用预约流程材料，但改变成果要求：第一轮写建议，第二轮说明证据为何不足，第三轮准备三个问题，询问掌握缺失信息的人。资料不变。
This reveals whether you can adapt language to purpose. “We recommend…” is appropriate in the first task. “The current information does not establish…” may be central in the second. “Could you confirm whether…” becomes useful in the third. A learner who repeats the same prepared speech in all three rounds is not yet using the task to guide the language.
这能检查你是否根据目的选择英语。第一轮需要推荐，第二轮需要说明边界，第三轮需要确认问题。如果三轮都讲同一篇稿，就还没有让任务指导语言。
For an additional challenge, introduce one new constraint after the first few turns. Do not reward the participant who ignores it and finishes the original answer smoothly. Ask the group to state exactly what changed, which earlier claims remain intact and which part must be reconsidered. Flexible control matters more here than uninterrupted delivery.
进一步练习：开始几次发言后加入新限制。不要奖励忽略新信息却顺利背完的人。要求小组说明什么变了、什么没变、哪里需要重审。
Finish by comparing the outputs rather than judging only the speaking style. Does the recommendation preserve the conditions? Does the evidence-gap statement distinguish missing information from negative evidence? Do the questions ask for information that could actually change the decision? These checks turn a speaking exercise into a meaning-preservation exercise.
最后比较成果，而不只评价说话风格：建议有没有保留条件？证据缺口有没有区分“缺信息”和“负面证据”？问题的答案能否改变决定？这样，口语练习也成为保护意义的练习。
28. A practice cycle that keeps the task manageable | 一套可执行的练习循环
The following timings are suggested practice designs, not research-established prescriptions or official examination timings. Adjust them to the difficulty of the source pack and the learner’s needs. A shorter, well-examined exchange can be more useful for your current target than a long discussion nobody reviews.
以下时间是建议练习安排，不是研究确定的处方或官方考试时间。按材料和学习者需要调整。当前目标可能更适合一段被认真检查的短交流，而不是没人复盘的长讨论。
Twenty-minute practice: spend four minutes reading a short fictional brief, two minutes writing a question and a provisional view, six minutes discussing, four minutes checking the outcome against the source and four minutes repeating the weakest exchange. Choose only one main target, such as preserving conditions or checking an unsupported claim.
20 分钟：读材料 4 分钟，写问题和暂定观点 2 分钟，讨论 6 分钟，按来源检查成果 4 分钟，重练最弱片段 4 分钟。一次只设一个主要目标。
Three-person practice: rotate contributor, challenger and synthesiser. These are temporary tasks, not personalities. The challenger must first state the original claim accurately. The synthesiser must invite correction. The contributor must respond to at least one challenge rather than returning to a memorised paragraph. Rotate after each short round.
三人练习：轮换提出观点、检查观点和综合三个任务。检查者先准确复述，综合者邀请纠正，提出者至少回应一次质疑。每轮换人，不把任务变成人格标签。
Independent practice: record a response to a source card, then write a plausible challenge from another perspective. Answer the challenge aloud and produce a revised summary. This cannot reproduce all the unpredictability of a real partner, so add actual interaction when possible. It still lets you practise changing your wording when a condition or objection matters.
独立练习：先回应资料卡，再从另一个视角写一个合理质疑，口头回应并重写总结。这不能完全替代真实伙伴的不可预测性，但可以练习条件变化后的表达。
29. Progress from support to independent discussion | 从有支持，到独立讨论
In the first stage, use a visible question and a small phrase bank. The aim is to understand the function of a move. Can you distinguish a clarification from a challenge? Can you give a conditional recommendation without dropping the condition? Do not make speed the main test while the function is still unclear.
第一阶段给清楚问题和小型表达表。先理解发言功能：澄清与质疑有何不同？有条件建议能否保留条件？功能还不清楚时，不以速度为主要考核。
In the second stage, keep the source pack but remove the phrase bank. Ask for the same functions in the learner’s own words. Then change the topic while retaining the reasoning structure. A learner who can qualify a workshop result should also attempt to qualify an unrelated fictional service trial, without copying the original nouns.
第二阶段保留资料，撤掉表达表，用自己的话完成相同功能。然后换话题，保留推理结构。能限制工作坊结论，也要尝试限制另一服务试点的结论。
In the third stage, add interactional uncertainty: a partner asks an unexpected follow-up, corrects a source detail or introduces a missing stakeholder. The learner must adapt rather than replay. Finally, ask for a written outcome that preserves the discussion. Independent control is not proved by one successful rehearsal; look for the same useful behaviours across more than one task.
第三阶段加入意外追问、来源更正或新的相关方，再写出保留讨论结果的记录。一次排练成功不证明独立掌握；要在多个任务中观察同样行为。
30. Assess what the conversation achieved | 评价讨论真正完成了什么
Use a simple observation sheet instead of calling every fluent answer “C1”. For each area, mark needs support, works in this task or holds across changed tasks. These labels organise teaching evidence; they do not convert into a CEFR score.
不要把每个流利答案都叫 C1。每项可标“需要支持”“当前任务可完成”“换任务仍可完成”。这些是教学观察，不转换成 CEFR 分数。
Question control: did the group keep the actual task visible? Uptake: did a response accurately use something another person said? Evidence: were sources, populations and uncertainty preserved? Challenge: did disagreement test the claim rather than distort it? Synthesis: did the closing reflect both agreement and reservation? Language: were relationships, conditions and responsibilities expressed clearly?
问题：任务是否清楚？ 接续：是否准确使用前一个人的话？ 证据：来源、对象和不确定性有没有保留？ 质疑：是否检查而非歪曲？ 综合：是否同时保留共识和保留意见？ 语言：关系、条件、责任是否清楚？
Record one short example for each judgement. “Good interaction” is less useful than “After Ciara corrected the denominator, Beatrice changed ‘participants’ to ‘respondents’ in her summary.” The second note makes the observed improvement specific. It also gives the learner a behaviour to repeat in the next task.
每项评价保留一个例子。“互动好”不如“Ciara 更正分母后，Beatrice 在总结中把 participants 改成 respondents”。后者让进步可见，也给下次练习明确目标。
Do not confuse a quiet style with weak contribution. A short question that repairs the shared definition may matter more than a long fluent answer that repeats an established point. Equally, a useful question is not evidence of mastery of all speaking skills. Keep the conclusion as specific as the observation.
不要把安静风格等同于贡献弱；也不要因为一句好问题，就宣布所有口语能力都已掌握。观察多具体，结论就多具体。
31. A tutor’s guide: intervene at the first broken connection | 给教师：在第一个断点介入
Suppose a learner gives an irrelevant answer. Before correcting vocabulary, check whether the learner understood the previous turn. If the source was misunderstood, a polished paraphrase will preserve the wrong meaning. If the source was understood but the response failed to connect, model the attachment point. If the connection was good but the sentence became unclear, work on language.
学生答非所问时，改词之前先检查是否听懂前文。来源误解了，漂亮改写只会保留错误；听懂却接不上，就示范连接点；关系对了但句子不清，再修语言。
A useful teacher intervention is small: “Which part of Beatrice’s statement are you responding to?” If that question is enough, do not immediately provide a complete answer. If more support is needed, model one turn and explain its purpose: “I am keeping her condition before I ask whether the evidence supports the wider claim.” Then let the learner attempt a parallel turn.
可以先问：“你回应的是 Beatrice 哪一部分？”够用就不要立刻给完整答案。需要更多支持时，示范一次并说明：“我先保留她的条件，再问证据能否支持更广结论。”随后让学生尝试。
Goh and Burns’ Teaching Speaking: A Holistic Approach is a relevant teacher reference because it connects speaking theory, task design and classroom practice. Goh and Vandergrift’s listening volume complements it by focusing on how learners manage listening processes. The exercises in this lesson are original applications, not reproduced chapters or independently validated versions of those books. [7] [8]
Goh 与 Burns 的口语教学书连接理论、任务设计和课堂实践；Goh 与 Vandergrift 的听力书补充学习者怎样管理听力过程。本课练习是原创应用，不是复制章节，也不声称已经独立验证。
32. The first weak link: choose the repair that matches the failure | 第一薄弱环节：按原因选择修复
You prepare well but miss the discussion: reduce the notes and practise referring to the previous speaker before adding your own point. You enter quickly but challenge the wrong claim: practise accurate restatement before disagreement. You sound cautious but say little: replace vague hedging with one specific condition or missing fact.
准备很多却跟不上：减笔记，先接前文。 很快反对却反错对象：先准确复述。 很谨慎却没信息：用具体条件或缺失事实替代模糊弱化。
You can explain alone but lose precision with partners: keep the language target small while increasing interaction. You summarise smoothly but omit reservations: check the ending against each participant’s conditions. You use exact vocabulary but cannot close: practise distinguishing a decision, recommendation, evidence gap and next check.
独自说好，与人说乱：增加互动时缩小语言目标。 总结顺却漏顾虑：逐个检查条件。 词准却收不住：区分决定、建议、缺口和下一检查。
These are diagnostic possibilities, not labels to attach permanently to a learner. One behaviour can have several causes. Observe a second task, ask the learner what they were trying to do, and change the support when the evidence points elsewhere. The purpose of diagnosis is to choose useful work, not to turn a moment of difficulty into an identity.
这些只是诊断可能，不是永久标签。同一行为可能有多个原因。换任务观察，也问学习者当时想完成什么。诊断是为了选择有用训练，不是把一次困难变成身份。
33. Connect this lesson to the rest of your English | 把讨论接回其他英语能力
A complex discussion draws on reading, listening, vocabulary, grammar, pronunciation and writing. Work on the component that is actually limiting the task. If you cannot hear the distinction your partner made, return to listening. If you hear it but cannot express a condition, practise the sentence. If you can say it but the summary loses it, work on the written outcome.
复杂讨论需要阅读、听力、词汇、语法、发音和写作。修真正限制任务的部分：听不出区别，就回听力；听出却说不出条件，就修句子；说得出却写漏，就修结果记录。
For the current series, Lesson 005 supports listening for structure and stance; Lesson 006 supports real-time formulation; and Lesson 007 supports intelligible delivery. Use them as targeted returns, not as a requirement to restart everything.
第 005 课帮助听结构和立场，第 006 课帮助实时组织，第 007 课帮助说清楚。按问题返回，不必每次全部重学。
School learners can also use the Secondary 3 Academic Discussion and Seminar guide for stage-specific practice. The Advanced English Vocabulary Chinese Edition gives a separate route when the main difficulty is the depth and use of words.
学校学习者可以进入中三学术讨论指南；若主要困难在词语理解和使用深度，则使用高级英语词汇中文版。
34. References and books: use each for a clear purpose | 参考网站与书籍：每个工具都有具体用途
[1] Council of Europe, qualitative aspects of spoken language use. The official descriptors provide a reference for range, accuracy, fluency, interaction and coherence. Use them to understand the target, not to award yourself a level after a single exercise.
[2] Cambridge English, C1 Advanced format. The official format page explains the test’s speaking tasks. It is useful for learners preparing for that qualification; this lesson’s fictional meetings are broader language practice, not official sample test material.
[3] Evelina D. Galaczi, “Interactional Competence across Proficiency Levels: How do Learners Manage Interaction in Paired Speaking Tests?” Published in Applied Linguistics, 35(5), 2014, with earlier online publication. The publisher abstract highlights topic development, listener support and turn-taking. The discussion here uses that reported scope, not a claim to have replicated the study.
[4] University of Manchester, Academic Phrasebank. Being cautious and Being critical help distinguish careful qualification from unsupported certainty. Adapt the functions to speech rather than importing long written phrases unchanged.
[5–6] British Council, TeachingEnglish. Group discussion skills and Question techniques and strategies provide teaching ideas for participation, preparation and feedback. The original tasks in this lesson should be adjusted to the actual learners.
[7] Christine Goh and Anne Burns, Teaching Speaking: A Holistic Approach. Cambridge University Press, 2012. This is principally a teacher-development reference, not a collection of ready-made answers for a speaking test. Its publisher description outlines the integration of theory, pedagogy and speaking tasks.
[8] Christine C. M. Goh and Larry Vandergrift, Teaching and Learning Second Language Listening: Metacognition in Action, second edition. The Routledge book page describes a learner-oriented approach to managing listening processes. Use it when the discussion difficulty begins with listening rather than speaking.
[9] Council of Europe, mediation. The official explanation describes helping meaning and communication across and within languages. This supports a useful view of bilingual ability: Chinese is a resource for understanding, while English can be practised as the language through which the group builds shared meaning.
阅读这些资源的方法：先确定自己的问题，再选资源。研究、官方描述和教学建议不是同一种证据；本课的虚构资料、示范答案、练习时间和观察表是教学设计，不是官方评分标准，也不保证在固定时间内达到某一等级。
35. Final assignment: leave an accurate shared result | 最终作业：留下准确的共同结果
Choose either fictional case in this lesson. Work with two partners, or begin with the independent version and later add partners. Each person prepares a provisional view, one source limitation and one question. Do not write a full speech. During the discussion, each participant must accurately use a previous contribution, make or answer a fair challenge, and help check the final summary.
选本课一个虚构案例，三人合作；也可以先独练，再加入伙伴。每人准备暂定观点、一个来源限制和一个问题，不写完整演讲。讨论中准确接续前文，提出或回答公平质疑，并参与核对总结。
After the discussion, write three short paragraphs: what the group established, what remains unresolved, and what should happen next. Compare the writing with the source pack and the participants’ actual positions. Repair any lost condition, invented agreement or unsupported certainty. Then repeat one difficult exchange with a different person taking the lead.
讨论后写三小段：已建立什么，未解决什么，下一步是什么。对照资料和实际立场，修复丢失条件、虚构共识或没有根据的确定说法，再换人带领重练一个困难片段。
Your final question is not simply, “Did I speak fluently?” Ask: “Did the group leave with a clearer question, a more accurate view of the evidence and a truthful account of what it can do next?” When that happens, your English is doing more than displaying knowledge. It is helping people think together.
最后不只问“我说得流利吗？”还要问：“大家离开时，问题是否更清楚，证据是否理解得更准确，对下一步的说法是否真实？” 做到这些，英语就不只是在展示知识，而是在帮助人共同思考。
Batch 02 learning sequence: listening in Lesson 005 → real-time speaking in Lesson 006 → intelligible delivery in Lesson 007 → collaborative discussion in Lesson 008.
第二组学习路线：第 005 课听懂，第 006 课及时说，第 007 课说清，第 008 课共同讨论。
Return to the Advanced English Chinese Edition hub · 返回高级英语中文版学习主页
