# 英语高级 · 第 12 周 Day 4|计时演练与删减两成 > 系列:语种学习 / 英语 / 高级 · 每日学习|AI生成|2026-10-27|codebuddy **今天学什么** 第一次完整演练几乎一定超时,而且超时的原因通常不是内容多,而是**句子里有大量不承载信息的成分**:关于演讲的演讲(as I said, what I want to emphasise)、双重模糊(I think it might possibly be)、同义重复(big and significant and substantial)。今天只练一件事:按句删掉 20% 的内容,且数字、机制、转折一个不动。语言上对应的是压缩句法——把三层意思压进一句,把一句废话整句拿掉。产出是两轮计时记录、一份删减后的提纲、一条 3 分钟开场录音。删减不是让步,是把注意力预算还给出证据。 ## 一、今日任务 | 环节 | 时长 | 具体动作 | 产出 | | --- | --- | --- | --- | | 1 全量第一遍 | 20 min | 看提纲完整讲一遍,计时不分段,中途不停;讲完立刻记总时长与四段各自用时 | 计时记录 1 | | 2 标记可删 | 15 min | 回听并用三种符号标提纲:M(meta 自我评论)、H(双重模糊)、D(同义重复) | 标注稿 1 份 | | 3 删 20% | 15 min | 以句为单位删,按字数核算(目标删掉原稿 20%);删完重算词数 ÷ 语速 | 二稿提纲 + 词数 | | 4 第二遍计时 | 20 min | 用二稿再讲一遍,同样计时;与第一遍逐段对比 | 计时记录 2 | | 5 录 3 分钟开场 | 15 min | 只录开场,重录不超过 3 次,取最好一条;时长 2'50"—3'10" | 录音 1 条 | **语速基准**:正式演讲 130—150 词/分钟。10 分钟约 1300—1500 词。第一遍若讲成 12 分钟,说明你的实际语速只有 110 词/分钟,或者稿子超过 1600 词——先量词数,再决定删哪里。 **为什么按句删而不是按词删**:删词最容易把句子删瘸(主谓不搭、指代失主),而删句只要保证段落逻辑链还在。一句话里删掉三个形容词,省 4 秒;整句拿掉一句 meta 评论,省 9 秒且更清楚。 ## 二、核心词汇与搭配 | 词/短语 | 音标 | 词性 | 中文释义 | 高频搭配 | 语域 | | --- | --- | --- | --- | --- | --- | | trim | trɪm | v. | 削减(篇幅、预算) | trim the script by a fifth; trim the fat | 商务、媒体 | | tighten | ˈtaɪtn | v. | 收紧(结构、措辞) | tighten the argument; tighten the opening | 商务、学术 | | redundancy | rɪˈdʌndənsi | n. | 冗余、重复多余 | cut redundancy; built-in redundancy | 学术、商务 | | verbatim | vɜːˈbeɪtɪm | adj./adv. | 逐字的 | a verbatim script; recite verbatim | 学术、媒体 | | off-the-cuff | ˌɒf ðə ˈkʌf | adj. | 即兴的、未准备的 | an off-the-cuff remark | 通用、媒体 | | stumble | ˈstʌmbl | v. | 卡壳、说绊 | stumble over a phrase; stumble at the pivot | 通用 | | filler | ˈfɪlə | n. | 填充词(um, you know) | cut the fillers; filler count | 通用、演讲 | | pace | peɪs | n./v. | 语速;控制节奏 | a measured pace; pace the opening | 演讲、通用 | | overrun | ˌəʊvəˈrʌn | v./n. | 超时;超出 | overrun by two minutes; a cost overrun | 商务 | | buffer | ˈbʌfə | n. | 缓冲(时间、余量) | a two-minute buffer; build in a buffer | 商务、通用 | **例句与用法** 1. **trim** — I trimmed the script by a fifth and the argument got sharper. 我把稿子删了五分之一,论证反而更利。(trim by + 比例,比 cut some 可核查;后半句是你需要观察到的效果。) 2. **tighten** — Tighten the opening: the first thirty seconds carry the whole talk. 收紧开场:前三秒决定整场。(tighten 的对象是结构或措辞,不是时长。) 3. **redundancy** — Most overruns come from redundancy, not from too many ideas. 多数超时来自冗余,而不是观点太多。(这句本身就是判断句,可直接用在复盘里。) 4. **verbatim** — Do not read verbatim; a memorised script breaks at the first interruption. 不要逐字念;背下来的稿子一被打断就全崩。(verbatim 常与 read、recite 连用。) 5. **off-the-cuff** — The best line in the talk was off-the-cuff and we kept it. 全场最好的一句是即兴的,我们保留了它。(off-the-cuff 含"未经准备但有效"的褒义可能,注意语境。) 6. **stumble** — I stumbled at the pivot from evidence to proposal. 我在证据转方案的衔接处卡住了。(stumble at / over + 位置,指具体的卡点,不说 I was nervous。) 7. **filler** — Count your fillers per minute; above five and the room stops trusting you. 数一下每分钟填充词;超过五个,听众就不再信你。(filler 可数,filler count 是复盘常用指标。) 8. **pace** — Pace the opening slower than feels natural. 开场比你感觉自然的语速再慢一点。(pace 作动词时后面接宾语:pace the opening / pace yourself。) 9. **overrun** — We overran by ninety seconds, all of it in the evidence section. 我们超了九十秒,全部在证据段。(overrun by + 时长;后半句指明位置,这才是可改的信息。) 10. **buffer** — Build a two-minute buffer before Q&A. 在问答前留两分钟缓冲。(buffer 可数,build in a buffer 是固定搭配。) ## 三、语言点:删掉 20% 而不删证据 删减的判断标准只有一条:**这句话拿掉,听众会不会少知道一个事实、一个机制或一个转折**。会,就留;不会,就删。按这个标准,最容易删的三类几乎占满 20%。 **规则** 1. **删"关于演讲的演讲"**。As I said earlier、What I want to emphasise is、Let me just say、I think it is fair to say——这些句子只描述你正在说话,不提供信息。一句省 6—9 秒,十句就是一分半。 2. **双重模糊合并为一档**。I think it might possibly be the case that → It may be。两个 hedge 叠用不会更谦虚,只会显得你没把握;留一个情态动词足够。 3. **同义重复只留最具体的那一个**。big and substantial and significant → substantial,或直接给数字。三个形容词并列时,听众只记住最后一个,前两个是噪音。 4. **数字、机制、转折一律不动**。删掉 from eleven days to four 省 5 秒,但整段的可信度一起没了;删掉 but the constraint was ownership 省 6 秒,但论证的支点没了。 5. **以句为单位删,删完重算时长**。删完统计词数,除以你的实测语速(第一遍词数 ÷ 分钟数),得到新时长;不要凭感觉判断"应该够了"。 **什么时候用**:一切有时间上限的口头输出——演讲、答辩、pitch、面试回答、会议发言。写作修订时同样适用,只是标准换成"读者会不会少知道一个事实"。 **什么时候不能用**:需要建立关系、缓和冲突的场合(开场寒暄、谈判破冰、给下属反馈),那些"废话"承担的是关系功能,删了会显得冷。学术写作的限定语(to some extent, under these conditions)也不是冗余,删了会变成过度断言。 **用了会显得怎样**:删干净后显得密度高、有准备、尊重听众时间;删过头(把限定语也删了)显得绝对化、不可信;一句不删则显得在念稿,且一定超时。 **强弱对比** | 冗长版 | 压缩版 | 省下的与代价 | | --- | --- | --- | | What I would like to do now is to sort of move on and talk a little bit about the evidence. | Here is the evidence. | 省 8 秒;meta 与双重模糊全删,信息零损失 | | I think it might possibly be the case that the delay is caused by the approval process, or at least that is one of the factors. | The delay comes from the approval process. | 省 9 秒;把"可能的原因之一"改成断言,代价是你必须能扛追问——所以只在证据够硬时这样删 | | The results were really very significant and quite substantial in terms of their overall impact. | Approval time fell from eleven days to four. | 省 7 秒,且换来一个可验证的数字;这是唯一"删了反而更强"的类型 | **常见误用** 1. **删词不删句**。把 The implementation of the new approval workflow was completed ahead of schedule 删成 The implementation of new approval workflow completed ahead 就语法破碎。正确做法是整句换成 We finished the new approval workflow early。 2. **删掉转折句**。转折(but、yet、the constraint was)是论证的关节,删了以后每段都正确,但整篇没有立场。宁可删例子,不删转折。 3. **为凑时长而加速**。语速从 140 提到 180,听众的接收率会掉一半。删内容,不改语速;语速是最后才动的变量。 ## 四、句型与范例 **10 个压缩句式(左为冗长版,右为可直接用的版本)** 1. **The point is X.** 取代 What I am trying to say here is that X.(最通用的压缩句,后面直接跟名词或短从句。) 2. **Two reasons: A, and B.** 取代 There are a couple of reasons why this happened, and the first one is A, and the second one is B.(冒号 + 编号,省一半词。) 3. **X, not Y.** 取代 It is not really Y so much as it is X.(对照替代解释,语速更快,重音更清楚。) 4. **That costs us Z.** 取代 What that means in terms of the overall cost to us is Z.(直接给代价,删掉介词链。) 5. **Say more about that.** 取代 I wonder if you could possibly elaborate a little further on that point.(追问句压到 4 词,且不失礼。) 6. **So what?** 取代 The question one might reasonably ask at this point is why this matters.(自问自答的极简版,适合段间过渡。) 7. **Which brings me to X.** 取代 This therefore leads me on to the next part of my presentation, which is X.(过渡句压到 5 词。) 8. **In one line: X.** 取代 If I had to summarise this in a single sentence, I would say X.(要求对方压缩时的指令句。) 9. **Hold that thought.** 取代 That is a really interesting point and I would love to come back to it later.(拉回而不打断,5 词完成。) 10. **Same number, different cause.** 取代 Although the figure has remained unchanged, the underlying reasons have changed considerably.(回扣句,四词,适合收束。) **3 分钟开场:压缩前后对照** | 段 | 原始(约 300 词) | 压缩后(约 205 词) | 处理 | | --- | --- | --- | --- | | 钩子 | 完整背景 + 两句自我介绍 + 数字 | 只用数字 + 一句损失 | 删自我介绍与铺垫 | | 路线 | 四项,含"最后我会总结" | 三项,删"总结" | 总结段本就冗余 | | 规则 | 三条,含"希望大家畅所欲言" | 两条,全是我方动作 | 删掉不执行的规则 | | 交接 | 两句客套 + 指名 | 一句指名 | 客套在熟人会议里是噪音 | **压缩后开场(约 205 词,带批注)** > Eleven days. That is how long it takes us to get a new hire from signed offer to first commit, and one in four leaves before day eleven. I have ten minutes on that number, because it is not a hiring problem. It is an ownership problem. > > Three things. Where the eleven days go. What the fix costs. What I need from you at the end. Two rules: anything that needs data we do not have gets parked, and if we are still on the first item at twenty past, I cut it short. > > Here is the first number. Four of the eleven days are one approval that four people sign. Three are a laptop nobody ordered. The work takes two days. The rest is queue. So the fix is not more people. It is fewer hands on one decision: collapse two approvals, audit one order in ten, put one name on the queue. > > By the end of March — one owner, one queue, four days. 批注:① 开场从 11 词压到 1 词(Eleven days),这是全篇最强的一处删减:数字本身是钩子,铺垫都是负担;② I have ten minutes on that number 一句同时给时长与主题,删掉了原来的两句过渡;③ Three things 后三项全部是名词短语,形式严格平行,比原来的四个从句快 15 秒;④ 规则从三条减到两条,且都写自己的动作(gets parked / I cut it short),删掉"希望大家配合"这类不执行的句子;⑤ 结尾 four words 收口,比原来的两句总结更有力。全段实测 2'48",词数 205。 ## 五、练习 **A 词汇与搭配(7 题)** 1. 填空:We ______ by ninety seconds, all of it in the evidence section. 2. 搭配:用 trim 写一句,含删除比例与效果。 3. 改错:I stumbled in the transition between the two sections. 4. 用 filler 写一句复盘结论,含数字。 5. pace 作动词与作名词各造一句。 6. buffer 与 margin 在时间语境下有何差别?各造一句。 7. verbatim 与 off-the-cuff 是反义关系吗?用一句话说明并各造一句。 **B 句法与压缩(6 题)** 1. 压缩到 8 词以内:What I would like to do now is to move on and talk about the evidence. 2. 去掉双重模糊:I think it might possibly be that the delay is caused by approvals. 3. 合并同义重复并给数字:The improvement was very significant and quite substantial. 4. 整句改写(不删词,改结构):The implementation of the new approval workflow was completed ahead of schedule. 5. 把转折句保留、把例子删掉:原段含一个完整案例(5 句)与一句 But the real constraint is ownership. 6. 用实测语速算时长:稿子 1720 词,第一遍讲了 12 分 10 秒,删到 1380 词后预计多久? **C 输出(2 题)** 1. 交删减前后对照:原稿词数、删后词数、删除比例(须 ≥20%)、被删句按 M/H/D 分类计数。 2. 录 3 分钟开场(2'50"—3'10"),词数控制在 390—450;附 3 条回听记录:填充词次数、最长停顿位置、哪句是 meta。 ## 六、参考答案 **A** 1. overran(overrun by + 时长;过去式 overran。) 2. 示例:I trimmed the script by 22 per cent and kept every number.(trim by + 比例;后半句说明删减边界。) 3. 错。stumble at / over 更常用:I stumbled at the transition between the two sections.(at 指具体位置。) 4. 示例:Fillers fell from nine to three per minute after I slowed the opening.(filler 用 per minute 计量。) 5. 动词:Pace the opening slower than feels natural.(控制开场)/名词:The pace dropped in the evidence section.(语速本身) 6. buffer 指预留的可用余量(可主动设置);margin 指离失败线的距离(多为被动拥有)。例:Build in a two-minute buffer.(预留)/We finished with a margin of forty seconds.(剩下) 7. 不是严格反义:verbatim 指逐字照稿,off-the-cuff 指未经准备。例:He read verbatim from the note.(照念)/Her best line was off-the-cuff.(即兴) **B** 1. Here is the evidence.(4 词) 2. The delay comes from approvals.(或:Approvals cause the delay.) 3. 示例:Approval time fell from eleven days to four.(删掉两个评价形容词,换成一个可验证的变化。) 4. 示例:We finished the new approval workflow early.(名词化还原为动词,主语恢复为人。) 5. 示例:Keep one clause of the case, then: But the real constraint is ownership.(案例只留能支撑转折的那一句,其余删;转折句一字不动。) 6. 实测语速 = 1720 ÷ 12.17 ≈ 141 词/分钟;1380 ÷ 141 ≈ 9.8 分钟,即 9 分 48 秒。结论:删 20% 刚好回到 10 分钟内,无需加速。 **C 评分维度** 删减对照看三项:① 删除比例 ≥20%(按词数算,不按段落数);② 被删句中 M 类占比最高,数字句与转折句零删除;③ 删后每段仍有结论句。开场录音看三项:① 2'50"—3'10",词数 390—450(语速约 140);② 填充词 ≤4 次/分钟;③ 前 20 秒内出现一个数字,且该数字后不接解释性从句。 ## 七、今日输出与自查 产出文件:第12周_Day04_计时记录.md(两轮)、第12周_Day04_删减对照.md、第12周_Day04_开场录音(1 条,3 分钟)。 | 自查项 | 合格线 | 怎么查 | | --- | --- | --- | | 删减比例 | ≥20%(按词数) | 原词数 − 新词数 ÷ 原词数,写进文档 | | 证据保全 | 数字、机制、转折零删除 | 对照两版逐段点名,缺一项即回补 | | 时长 | 第二遍 9'30"—10'30";开场 2'50"—3'10" | 秒表,不用估算 | | 语速 | 130—150 词/分钟,全程波动 ≤20 词 | 分段词数 ÷ 分段秒数 | | 填充词 | ≤4 次/分钟 | 回听数十次,记位置不记感受 | **关于删减的心理障碍**:你会觉得删掉的那句"讲得很好"。判断标准不是好不好,而是拿掉之后听众是否少知道一个事实。若答案是"不少",那句好只是对你自己好。今天写的每一句 M 类句子,都值得单独看一眼——它们通常是你最满意、也最该删的句子。 ## 八、明日预告 Day 5 做 30 分钟模拟面试:自问自答并录音,每题至少追问一层,目标是 6 题里 4 题扛得住。今天练的"删 meta、留机制"会直接决定追问的质量——追问打的多半是你没说清的机制,而不是你说错的数字。明天还要用上 Day 1 的六张卡,先把卡找出来。