🕵️ 逻辑谬误：表面有理的陷阱🕵️ Logical Fallacies: Traps That Only Look Reasonable
上一片里，我们学会了正确的拼法：形式对了，结论必然为真。可身边很多话，听着像在推理，其实把积木硬塞进了卡扣。这些坑就叫逻辑谬误（Logical Fallacy）。今天请机器人侦探出马，认认四个最常见的。Last leaf we learned the valid snaps: right form, guaranteed conclusion. But plenty of everyday talk only sounds like reasoning — the blocks are jammed into the wrong buckles. These traps are called logical fallacies. Today the robot detective takes the case and walks us past four of the most common ones.
第 1 步：稻草人——拳头打在假人上Step 1: Straw man — a punch that lands on a dummy
第一个坑最直观：对方说的，和你听到的，根本不是同一句话。The first trap is the most visual: what they said and what you heard are not the same sentence.
右边的人对着一个稻草假人挥拳，假人身上的牌子写着「他要砸掉所有手机」；可左边的女孩原话只是「少玩点游戏」。把对方的话换成夸张版本再打倒，打倒的只是假人，原来的观点还站着——这叫稻草人（Straw man）。它的邻居是人身攻击（Ad hominem）：不理你的理由，只攻击说话的人，比如“你自己都没考好，凭什么说这道题该先算括号”。The angry figure on the right throws a punch at a straw dummy whose sign reads “He wants to smash all phones!” — yet the girl on the left only ever said “fewer games”. Swap someone's claim for an exaggerated copy, knock that down, and the real claim is still standing: that's a straw man. Its neighbour is ad hominem — never answering the reason, only attacking the person, as in “You didn't even ace the test yourself — why should we do brackets first?”.
第二个坑更隐蔽：每个“所以”都像真的，串在一起就悬了。The second trap is sneakier: every “therefore” feels plausible, but strung together they dangle.
第 2 步：滑坡——一步滑到底Step 2: Slippery slope — one step to the bottom
小算盘机器人在坡顶举牌喊停，可小球已经从「今天迟到」滚向「明天旷课」，一路滑到坡底「最后进监狱」的铁窗前。滑坡（Slippery slope）借了「如果……就……」的外壳：链条上每一环都得有证据，一环撑不住，整条链就断——迟到和进监狱之间，隔着无数个“不一定”。它的邻居是以偏概全（Hasty generalization）：只拿一两个例子，就给整个群体下结论，比如“我碰到两个迟到的外卖员，所以外卖员全都不守时”。The little abacus robot holds up a stop sign at the top, but the ball is already rolling — “Late today”, “Skip class tomorrow”, straight down to the barred window of “Jail in the end”. A slippery slope borrows the “if…then…” shell: every link in the chain needs evidence, and one weak link breaks the whole chain — between being late and going to jail sit countless “not necessarilies”. Its neighbour is hasty generalization: one or two examples, then a verdict on an entire group — “Two delivery riders were late on me, so delivery riders are never on time.”
认得出坑，还得会拆——放大镜里藏着三句追问。Spotting a trap is half the job; the detective's glass holds three follow-up questions.
第 3 步：放大镜——防坑三问Step 3: Magnifying glass — three questions that unearth traps
机器人侦探把放大镜对准一段对话，裂缝里藏着一句「理由够吗？」。防坑其实只要三问：①他在回答我的理由，还是在攻击我这个人？（人身攻击）②他复述的是我的原话，还是一个夸张的假人？（稻草人）③这串“所以”的每一环，都有证据撑住吗？（滑坡、以偏概全）问完三句，大多数坑都会现形——你甚至不用赢下争论，只要不上当。The robot detective aims its magnifying glass at a conversation, and inside the crack sits one question: “Does the reason hold?” Three questions are enough to stay out of traps: ① Is she answering my reason, or attacking me? (ad hominem) ② Is she repeating my actual words, or an exaggerated dummy? (straw man) ③ Does every link in that chain of “therefore”s have evidence under it? (slippery slope, hasty generalization) After those three, most traps show themselves — you don't even have to win the argument, just avoid the fall.
🎮 你来当侦探：谬误找茬（4 段对话）🎮 Your turn, detective: spot the fallacy (4 chats)
下面四段对话，每段都踩了一个坑。读懂对话，点出谬误的名字，看看你的判断对不对。Four short conversations, each stepping into one trap. Read the chat, tap the fallacy's name, and see whether your call holds up.
一句话记住它：逻辑谬误 = 伪装成推理的坑；它借了推理的外壳，却经不起放大镜。Remember it in one line: a logical fallacy is a trap dressed as reasoning — it borrows the shell of logic but cannot survive a magnifying glass.
人身攻击：不回答理由，只攻击说话的人Ad hominem: never answer the reason, just attack the speaker稻草人：把对方的话夸张成假人再打倒；滑坡：小事一路滑到最坏结局Straw man: exaggerate the claim into a dummy, then knock it down; slippery slope: slide from a small step to the worst ending以偏概全：样本太小就下结论；防坑三问：回答了吗？是原话吗？每一环都有证据吗？Hasty generalization: verdicts from tiny samples; the three checks: answered the reason? actual words? every link backed by evidence?
内容参考 OpenStax 等公开教材，多来源核对 · AI 生成、人工审核 · 发现错误欢迎指正，帮这片叶子长得更好。Based on OpenStax and other open textbooks, cross-checked across sources · AI-generated, human-reviewed · Spotted a mistake? Tell us — help this leaf grow.
Back to tree
