常见逻辑谬误
说明：
本文主要介绍自己或者他人写作中可能发生的常见逻辑谬误，每一谬误均给出了相应定义、示例，以及如何防止这些谬误的建议。
关于论证
学术写作大都要求进行论证，即：对要提出或者阐释的见解，应给出理据。也许你得到了这样的评语：你的论证逻辑不够严密，也不够有力，而你又担心自己在逻辑性方面较弱，或者不清楚论证要有力到底什么意思。虽说学会出色的论证是一个发展的过程，然而却并非不可能做到：“讲求逻辑”任何人都能做到，多加实践即可！所有论证都包括一些“论据”（即给出理由或证据的语句），按照恰当的顺序组织到一起，从而能够支撑“结论”（即所论证的核心观点或者阐释）。要确保论证有力，应该做到：
1． 论据质量要高（确有把握认定论据的理据充分，因此论据不但真实而且与论题相关）；
2． 确保论据能够支撑结论（不能是其他结论，甚或完全不能得出任何结论）；
3． 检查是否已经讨论了问题的最重要或者最相关的那些方面（亦即论据与结论的焦点均着落在所论证的问题的真正重要部分）；
4． 论点不要过大过泛，以免超出自己的论证能力。
此外，还应确保要点的陈述要连贯有序，以便读者跟上思路。这方面请参考我们关于论证及其谋篇布局的相关建议，提高自己的论证水平。
本文介绍了一些论证中往往造成背离上述几条的问题形式；这些问题即是“逻辑谬误”。如果觉得自己的论证有些问题，不妨看看其中是否存在逻辑谬误！
若对论题抱有强烈情感，则最易陷入逻辑谬误——如果觉得结论显而易见，就更容易认定其为真，从而对相关证据不够仔细小心。为了便于了解人们如何出现此类常见错误，作为示例本文选取了许多政治论题——例如堕胎、枪支管制、死刑、同性恋婚姻、安乐死、以及色情等等问题。不过，本文的目的并非要论证任何这些问题的任何一种观点立场，而是要籍此阐明什么叫做推理有欠缜密，而这一问题可能发生在几乎任何论证之中！请注意：这些示例中的观点，均是为了说明问题而编拟出来的，并未经过科学论证，因此不应在自己的文章中采用它们作为证据。
逻辑谬误的概念
逻辑谬误是削弱论证的那些缺陷。学会在自己以及他人的文章中找出逻辑谬误，可以提高对自己的、读到的、听到的论证的评判能力。关于逻辑谬误，重要的是应该认清两点：首先，存在逻辑谬误的论证实在极为常见，且可能显得非常令人信服，至少对于那些诱因读者或者听众是如此；在报纸、广告及其他信息源中，都可以找到众多存在逻辑谬误的实例。其次，有时很难判断某论证是否存在逻辑谬误；某一论证实际上可能非常无力、有些无力、有些有力，或者非常有力；包含若干步骤或者若干部分的论证，可能其中某些章节很有力，而另外一些则很无力。因此，本文的目标不是教人如何给论证贴上逻辑谬误或者逻辑严密的标签，而是教人如何审查自己的论证，然后使其从“无力”变为“有力”。
逻辑谬误的判别
以下介绍的每一逻辑谬误，都分别给出了它的定义或解说、示例，以及论证中可以如何避免的建议。
以偏概全
定义：依据不充分的例证（通常不具代表性或者过于琐碎）得出普遍的结论。诸如“大学男生联谊会的人都是醉鬼”以及“大学生都很书呆子气”等等成见，就是“以偏概全”的很好例子。
示例：“我的舍友说她的哲学课很难，我的哲学课也很难，因此，所有的哲学课必定都很难！”本例的依据仅为两个人的感受，是不足以得出这样的结论的。
建议：检查所选用的例证：是否只有不多几个人的观点或者经历？如是，则应考虑是否需要增加证据，抑或缩小结论涵盖范围。（注意：本例结论若改为不那么绝对的“某些哲学课对某些学生来说是很难的”，则不属于“以偏概全”。）
推不出
定义：给出的论据的确能够支撑某种结论，然而却非欲论证的结论。
示例：“惩罚的严厉程度应当与违法的严重程度相吻合。现如今，对酒后驾车的惩罚也许不过只是罚款。然而酒后驾车乃是非常严重的违法，可以导致无辜行人的丧命。因此，酒后驾车应当适用死刑。”这里的论据实际上可以支撑若干结论，例如“对酒后驾车应当严惩”等等，但是就本例而言，却不能支持适用死刑的正当性。
建议：分清论据与结论。检查论据，看它们可以客观地给出什么结论。检查结论，看它需要什么论据作为支撑，然后检查实际是否给出了这样的论据。“推不出”的逻辑谬误往往发生在结论过于宽泛或者过于绝对之时，因此，如果所提主张较大，则应特别小心仔细。
事后归因（亦称“假性因果”）
这一谬误的名称来自拉丁语“post hoc, ergo propter hoc”，意为“后此，故因此”。
定义：认为事件甲之后发生了事件乙，因此事件甲导致了事件乙。当然，有时的确是前一事件导致了后一事件，例如，我选修了某门课，之后我的名字便出现在该课学生名单之中，因此确实是前者导致了后者。但是，有时时间上似乎相互关联的两件事，实质上并不存在因果关系。换言之，相互关联并不等于存在因果关系。
示例：“琼斯总统提高了税收，之后暴力犯罪上升。因此，琼斯总统对犯罪上升负有责任。”税收增加可能是也可能不是犯罪上升的一个因素，但本例并未揭示二者存在因果关系。
建议：要避免“事后归因”，本例的论证就需要给出一定的解释，阐明税收增加如何最终就会引起犯罪上升。因此，避免这一谬误的办法就是：如果要说甲导致乙，就还应说明甲如何导致乙，而不能仅仅说是因为甲先发生乙则后发生！
滑坡谬误
定义：声称某事之后将会发生一连串通常是可怕的后果，但却并无充分证据支撑该推论。这样的推论断定，如果再往前一步踏上“滑坡”，就必定会一路滑跌到沟底，亦即假定我们不可能中途停住。
示例：“动物实验有损对生命的尊重。如果不尊重生命，即可能越来越容忍诸如战争及杀人等等暴力行为。那么，社会将很快就会沦为战场，人人都会时刻担忧自己的生命。这将是文明的末日。为了防止出现这种可怕结果，应当立即宣布动物实验为非法。”由于动物实验为合法存在已有相当时日，而文明也并未面临末日，因此似乎非常明显：这一系列事件未必就会件件发生。即使认同动物实验的确危及对生命的尊重，而不尊重生命可以导致容忍暴力，但到此也可能就到了滑坡事件的终止点：我们也许不会非得一路滑跌到文明的末日。因此，并没有充分的理由让我们接受“必须宣布动物实验为非法”的结论。
与“事后归因”类似，“滑坡”谬论迷惑性也很强，难以识别，因为有时的确可以预知某事之后的一系列连锁反应。请看一个似乎并无谬误的例子：“如果我英语101课程不合格，就不能毕业。如果不能毕业，可能就找不到好工作，因此明年我也就很可能只能打零工或者做汉堡了。”
建议：检查论证中属于“如果甲，那么乙，而如果乙，则丙”之类的连锁引申，确保事件系列关联合理。
类比失当
定义：很多论证需要类比两种或更多事物、观点或情形。如果所比较的两件事物就所讨论的问题而言实际上并非真正类似，这种类比就不恰当，基于其上的论证也就存在“类比失当”的逻辑谬误。
示例：“枪支和铁锤一样，都是具有金属构件的可以杀人的工具，但是限制购买铁锤是很荒唐的，因此限制购买枪支也同样荒唐。”的确，枪支和铁锤都有一些共同特性，但是这些特性（均有金属构件、同为工具、都可能用于暴力）在决定是否需要限制购买枪支时并非属于要害问题。实际上，限制枪支乃是因为枪支很容易用于远距离大规模杀人。这一特性铁锤则没有，用铁锤击杀一大群人，恐怕很不容易。因此，这样的类比是不恰当的，而据此给出的论证也同样不恰当。
只要动动脑筋，就可以在世间几乎任何两个事物之间进行某种类比：“我的作文像个泥潭，遇雨都会膨胀扩大（闷在屋里出不去时学习的时候就更多），都让人郁闷。”因此，单凭类比了两个事物这一点，其本身证明不了多少问题。
在关于堕胎问题的争论之中，就常常借助类比：胎儿经常与成人相提并论，然后便争论说：某一医学手段如果侵犯了成人权利，那么同样也侵犯了胎儿的权利。诸如此类的观点是否成立，要看相关类比是否有力：成人和胎儿具有那些赋予成人权利的相同的属性吗？如果关键属性是拥有人类基因密码或者拥有度过丰富人生的可能，那么成人和胎儿的确都拥有这样的属性，上述观点及其类比也将是有力的；但如果关键属性是自我意识、理性或者能够自立生存，那么成人和胎儿就不是都具有这种属性，而相关类比也就是不恰当的。
建议：明确对于命题来说哪些才是重要的属性，再看所比较的两个事物是否都具备这些关键属性。
诉诸权威
定义：为使论证更为有力，我们往往举出享有盛誉或权威的资料作为参照，介绍他们对所探讨问题的立场观点。然而，如果只是举出某个人物的大名，或者诉诸某一号称权威但却难说是名副其实的专家，就期望能够影响读者，那就犯下了“诉诸权威”的逻辑谬误。
示例：“我们应当废弃死刑。众多的知名人士，例如影星Guy Handsome都曾公开表示反对死刑。”虽说Guy Handsome在演艺方面也许是个权威人士，却没有什么特别的理由认为他在政治方面的见解也应能打动人心——也许，就死刑问题而言，他若可称为权威，那么文章作者同样可以称为权威。
建议：要避免“诉诸权威”的谬误，有两个很简单的办法：第一，确保所援引的权威在所讨论的话题方面确实是专家；第二，不要只是说“Authority博士认同某某观点，因此我们也应该认同”之类，还应该充分阐述该权威所依据的理由或者论据。这样的话，读者获得的就不仅仅只是某人的声名。此外，还可以选择那些公认比较中立或理性的权威，而非那些可能会被指立场偏颇的权威。
诉诸公众
定义：这一谬误的拉丁文意为“挟众”。“诉诸公众”有若干形式，但共同之处是利用期望受到欢迎以及符合主流的这种大多数人的心理，从而赢得受众的支持。其最为常见的一种形式是“从众谬误”，即试图说服受众采取某一行动或接受某一说法，只因为（据说）大家都这样。
示例：“同性恋婚姻很不道德。70%的美国人都这样认为！”虽然在决定需要对什么问题立法方面，大多数美国的意见是有关系的，但是他们的意见却无疑不能决定某事是否符合道义。曾几何时，相当数量的美国人都赞成种族隔离，但是这些人的意见却不能成为种族隔离符合道义的论据。这样的论证，是在诉诸我们期望与其他美国人协调一致的心理，从而认可其结论。
建议：确保没有要求读者因为其他所有人都相信、所有冷静的人都相信、相信就会受到人们的欢迎，如此等等，所以他也应该相信你的结论。谨记：流行的观点并非总是正确！
人身攻击及“你也一样”
定义：“人身攻击”及“你也一样”类似“诉诸权威”，也是将人们的注意引向人身而非论证或者论据。这两类谬误的结论通常都是“不应相信某某人的看法”之类，而不能相信的原因则要么是因为某某乃是一个坏人（人身攻击）或者伪君子（你也一样）。涉嫌人身攻击的论证，针对的乃是对手本人而非对手的论证。
示例：“Andrea Dworkin出版了若干专著，认为色情会伤害女性。但是，Dworkin是个丑陋的怨妇，因此不值得相信。”这里对Dworkin的长相和性格的刻薄描述，与她的论证是否有力毫无关系，因此，依此为论据乃是逻辑谬误。
“你也一样”谬误是指责对手实际上也做了他自己所反对的事情，因此他的论证不听也罢。试举一例：假定父母对你讲明了不能吸烟的道理，列举了很多非常好的理由，例如有损健康、费钱等等，而你则回答说：“你说的我不能接受，因为你在我这个年龄也抽烟。你也是这样！”但是，即使父母在过去也做过他们现在反对的事情，这一事实也无损于他们的论据（吸烟有害健康而且费钱），因此，你的回答存在逻辑谬误。
建议：务必专注于对方的论证，而非对方的品格。（当然，如果的确是在讨论某人的品格，则应例外：如果你认为“克林顿总统不值得信任”，那么证明他不值得信任的言行的那些论据就是与论题相关的，不属于逻辑谬误。）
诉诸怜悯
定义：“诉诸怜悯”是指激起人们的恻隐之心，从而接受自己的结论。
示例：“我知道考试是依据答题结果给分的，但是我应该得A，因为我的小猫病了，我的车抛了锚，我又得了感冒，所以我的学习是很艰难的！”这里的结论是“我应该得A”，但是“得A”的标准衡量的是课程学习情况以及知识运用情况。这里要我们接受的理由（挺过了艰难一星期的人应该得A）显然是不可接受的。所给出的那些情况也许让人觉得相关，甚至觉得其结论值得考虑；但是，这些情况在逻辑上确实并不相关，因此，其论证也是存在谬误的。再举一例：“向企业征税是不对的，别忘了他们的慈善捐款以及他们已经付出的生产成本有多少！”
建议：确保没有仅仅只是激起人们的恻隐之心，从而认同自己的观点。
诉诸无知
定义：这一谬误基本上等于说：“瞧，手头尚无这一问题的结论性证据，因此，应该接受我对这一问题的结论。”
示例：“若干世纪以来，人们都一直在努力证明上帝的存在，但迄今尚无人能够证明。因此，上帝并不存在。”与此观点相反但存在同样逻辑谬误的说法则是：“多年来，人们一直试图证明上帝并不存在，但迄今尚无人能够证明。因此，上帝确实存在。”这两种论证都是利用证据不足来支撑结论的正确性。不过，在有一种情形中这样做，却不属于逻辑谬误：如果训练有素的科研人员通过精心设计的方法长期探求某一目标而最终未能发现，而按理又应该能够发现，那么，“实际未能发现”这一事实就可以构成“该事物并不存在”的某种证据。
建议：仔细检查是否存在已经指出证据不足而又依据证据不足得出了某种结论的那些部分。
草人谬误
定义：增加论辩力量的方法之一是预测可能的驳论然后先发制人。而“草人”谬误则是先把对方的观点变成一个容易推翻的版本，然后将其驳倒而得分；但是，这如同一拳击倒一个草人或者吓唬小鸟的假人，影响实在有限，因此，击败对手已遭扭曲的观点，影响也同样十分有限。
示例：“女权主义要求全面禁止色情作品，而且，凡阅读色情作品者都应受到处罚！但是，这样严厉的举措无疑是不稳妥的。因此，女权主义是错误的，不应干涉色情作品及其读者。”这里，女权主义的观点经夸大而遭削弱——实际上，多数女权主义者并不主张完全“禁止”色情作品或者惩罚仅仅阅读的人；女权主义主张的更多乃是限制诸如儿童色情这些东西，或者是主张受到色情作品伤害的人士为索赔而起诉相关出版商或者出品人，并非起诉读者。因此，这种论证并不能真正得分，其中存在逻辑谬误。
建议：面对对手要宽容厚道。陈述对方的论辩时应该尽量做到同样有力、准确，并且能够将心比心。如果能够驳倒对方观点最为有力的某种辩驳，则属真正有所成就。
转移论题
定义：这是指在论证过程中偏离正题，转向某一次要问题，从而转移人们对要害问题的注意力；之后的论证往往再不会回归原题。
示例：“考试分析采用曲线形式最为公平。无论如何，师生关系融洽的话，教学效果就更好。”不妨将这里的论据及其结论各自单列出来，看看其中存在什么问题：
论据：师生关系融洽的话，教学效果就更好。
结论：考试分析采用曲线形式最为公平。
这样就看得很清楚：论证偏离了正题——能够使人融洽相处，并不一定就更公平；公平与正义有时需要我们做出某些将会导致矛盾冲突的事情。但是，人们可能觉得，诸如师生关系融洽之类的问题确实很重要，因而就容易忽视这一点：上述论证并未给出能够支撑“曲线更公平”的任何证据。
建议：借用提纲形式，把论据与相应结论分列出来，看看论证中提出了多少问题，各个论据能否支撑相应结论。
虚假两分
定义：“虚假两分”指设法让人认为只有两种选择，然后排除其中之一，于是便剩下似乎是唯一的答案，亦即论证想要的首选答案。但是，实际上往往存在很多其他选择，而不仅只有两个；如果我们一一都考查一下，也许就不会那么快就选择论证所想要的那一个了。
示例：“Caldwell Hall大楼状况很遭，我们要么把它拆掉重建，要么就只能继续拿学生的人身安全冒险。显然，不应拿任何人的安全冒险，那么就必须把它拆掉。”这里的问题在于没有指出还有这种可能：我们可以维修该建筑，或者采取某种措施保证学生安全；例如，如果只有若干教室破损，那么我们也许可以不给这些教室排课。
建议：注意检查在声称只能二者选一的时候，是否真的如此，是否还有其他选择没有提到；如果有，就不应略去不谈，而应该也给出排除的理由。同理，让人认为只存在三个、四个等等选择这样的谬误，虽然还没有正式的术语以名之，但是如果实际存在的选择不止那些，则与“虚假两分”性质相同，也应避免。
窃取论题
定义：这一谬误颇为复杂，表现为多种形式，较之此前讨论过的谬误更难判别。但基本可以这样认为：“窃取论题”就是要求读者径直接受结论而并不给出任何真正的证据；论证要么是建立在与结论内容相同的论据之上（即所谓“循环论证”），要么就是略去了论证以为依据的某一重要（然而可疑的）假定。有时，人们会把“窃取论题”用作一种常见的评语，泛指结论缺乏充分理据的那些情形，但这不是我们这里所说的含义。
示例：“主动型安乐死在道义上是可以接受的，协助他人籍由死亡而摆脱痛苦乃是正当的，符合伦理的。”试将这一论证的论据及结论分列如下：
论据：协助他人籍由死亡而摆脱痛苦乃是正当的，符合伦理的。
结论：主动型安乐死在道义上是可以接受的。
把这里的论据稍加变换，就能看出同一个意思重复了两遍：“正当的，符合伦理的”很大程度上也就是“在道义上是可以接受的”，而“协助他人籍由死亡而摆脱痛苦”意思就是“主动型安乐死”。因此，该论据实际上在说：“主动型安乐死在道义上是可以接受的”，而这与相应结论完全一样！对于为什么安乐死可以接受，该论证并未给出任何真正的理由，反而难免让人质疑：“呃，那么，你为什么认为安乐死可以接受呢？”。这里的论证“窃取”（亦即规避）了真正的问题。
再举一例，其中论证所需但却含糊的论据完全被略去：“杀人在道义上是错误的。因此，主动型安乐死在道义上是错误的。”这里略去了论据“主动型安乐死是杀人”，而这一论据又是大可争议的——这样的论证同样是只字不提相关论据从而“窃取”了或者“规避”了主动型安乐死是否属于杀人这一问题，意在让人们只注意到没有争议的论据“杀人在道义上是错误的”，而看不到相关的假定。
建议：避免“窃取论题”的方法之一是把自己的论据及其结论按提纲形式一一列举出来，检查其中是否缺少某些环节、步骤，从而造成论据之间或者论据与结论之间的脱节；补上缺失的那些命题。如果这些命题存在争议，而又被含混敷衍过去，那就可能发生了“窃取论题”谬误。之后，检查一下论据所表达实质内容（但说法不同）是否与结论相同；如果是，那么就属于“窃取论题”。要旨在于：务必不要把试图论证的问题用作假定或者无可争议的证据。
词语歧义
定义：即在论证中某一重要词语的两个或更多含义之间游移不定。
示例：“向慈善组织捐助是正确的，因此，慈善组织向我们要钱也是正确的。”这里“正确”（right）一词有歧义，既可以指某事是对的或者好的（例如“这次测验我的答案都正确”），也可以指某人对某物拥有权利（例如“人人都有生命权”）。有些论证会有意偷换词语含义，常见的此类词语有“自由”、“正义”、“正确”等等；有些词语歧义问题则属于差错或者误解。但是，无论属于哪种情况，重要的是应保证在论证中所使用的主要词语的含义始终一致。
建议：找出论证中的重点词语，检查它们是否可能造成歧义，如是，则应确保没有同时使用其不同含义。
谬误的排查
以下是查找论证中逻辑谬误的几个方法：
站在反对相关结论的角度，反观论证中哪些部分比较可疑，哪些部分最为薄弱，然后重点加强这些部分。
列出论证的各个要点，在其下方分别列出相应的证据，这样也许就会发现某个提法并无过硬证据，或者可以更为严格地审查所采用的证据。
了解自己特别容易发生哪些逻辑谬误，检查自己的文章中是否出现这些谬误。有些人频频“诉诸权威”，有些人则更容易“类比失当”或者发生“草人”谬误。重读一下自己以前的文章，看看是否存在今后需要注意防范的某种逻辑谬误。
注意宽泛的说法，它们较之有所限制的说法需要更多的证据。说法中若有涵盖全部的词语如“所有”、“非”、“无”、“每个”、“向来”、“从不”、“无人”、“人人”之类，虽然有时是合适的，但较之不那么绝对的词语如“有些”、“很多”、“不多”、“有时”、“通常”等等，需要给出多得多的证据。
检查并复核涉及他人品格的用语，尤其是涉及对手品格的用语，确保这些用语准确、得当。
配套练习
点击本链接，即可看到一篇存在很多逻辑谬误的论证示例（之后再点相关链接即可看到对应解说）。此外，还提供了关于同一主题但谋篇布局更为精心合理的另一篇文章。
参考书目
本文初稿参考了下列论著，但这里并未全部列出。希望读者能够自己检索，查找最新相关论著。不要以这里的格式为标准书写自己文章的“参考文献”，因为这里的格式未必符合你所需要的格式。关于引用标注格式，请参考北科罗拉多大学图书馆（UNC Libraries）的文献引用标注指南。
Hurley, Patrick J. A Concise Introduction to Logic. Thornson Learning, 2000
Lunsford, Andrea and John Ruszkiewicz. Everything's an Argument. Bedford Books, 1998.
Copi, Irving M. and Carl Cohen. Introduction to Logic. Prentice Hall, 1998.
英文原文如下：
————————————————————————————————————
Fallacies
What this handout is about
This handout is on common logical fallacies that you may encounter in your own writing or the writing of others. The handout provides definitions, examples, and tips on avoiding these fallacies.
Arguments
Most academic writing tasks require you to make an argument—that is, to present reasons for a particular claim or interpretation you are putting forward. You may have been told that you need to make your arguments more logical or stronger. And you may have worried that you simply aren't a logical person or wondered what it means for an argument to be strong. Learning to make the best arguments you can is an ongoing process, but it isn't impossible: "Being logical" is something anyone can do, with practice! Each argument you make is composed of premises (this is a term for statements that express your reasons or evidence) that are arranged in the right way to support your conclusion (the main claim or interpretation you are offering). You can make your arguments stronger by
1. using good premises (ones you have good reason to believe are both true and relevant to the issue at hand),
2. making sure your premises provide good support for your conclusion (and not some other conclusion, or no conclusion at all),
3. checking that you have addressed the most important or relevant aspects of the issue (that is, that your premises and conclusion focus on what is really important to the issue you're arguing about), and
4. not making claims that are so strong or sweeping that you can't really support them.
You also need to be sure that you present all of your ideas in an orderly fashion that readers can follow. See our handouts on argument and organization for some tips that will improve your arguments.
This handout describes some ways in which arguments often fail to do the things listed above; these failings are called fallacies. If you're having trouble developing your argument, check to see if a fallacy is part of the problem!
It is particularly easy to slip up and commit a fallacy when you have strong feelings about your topic—if a conclusion seems obvious to you, you're more likely to just assume that it is true and to be careless with your evidence. To help you see how people commonly make this mistake, this handout uses a number of controversial political examples—arguments about subjects like abortion, gun control, the death penalty, gay marriage, euthanasia, and pornography. The purpose of this handout, though, is not to argue for any particular position on any of these issues; rather, it is to illustrate weak reasoning, which can happen in pretty much any kind of argument! Please be aware that the claims in these examples are just made-up illustrations—they haven't been researched, so you shouldn't use them as evidence in your own writing.
What are fallacies?
Fallacies are defects that weaken arguments. By learning to look for them in your own and others' writing, you can strengthen your ability to evaluate the arguments you make, read, and hear. It is important to realize two things about fallacies: First, fallacious arguments are very, very common and can be quite persuasive, at least to the causal reader or listener. You can find dozens of examples of fallacious reasoning in newspapers, advertisements, and other sources. Second, it is sometimes hard to evaluate whether an argument is fallacious. An argument might be very weak, somewhat weak, somewhat strong, or very strong. An argument that has several stages or parts might have some strong sections and some weak ones. The goal of this handout, then, is not to teach you how to label arguments as fallacious or fallacy-free, but to help you look critically at your own arguments and move them away from the "weak" and toward the "strong" end of the continuum.
So what do fallacies look like?
For each fallacy listed, there is a definition or explanation, an example, and a tip on how to avoid committing the fallacy in your own arguments.
Hasty generalization
Definition: Making assumptions about a whole group or range of cases based on a sample that is inadequate (usually because it is atypical or just too small). Stereotypes about people ("frat boys are drunkards," "grad students are nerdy," etc.) are a common example of the principle underlying hasty generalization.
Example: "My roommate said her philosophy class was hard, and the one I'm in is hard, too. All philosophy classes must be hard!" Two people's experiences are, in this case, not enough on which to base a conclusion.
Tip: Ask yourself what kind of "sample" you're using: Are you relying on the opinions or experiences of just a few people, or your own experience in just a few situations? If so, consider whether you need more evidence, or perhaps a less sweeping conclusion. (Notice that in the example, the more modest conclusion "Some philosophy classes are hard for some students" would not be a hasty generalization.)
Missing the point
Definition: The premises of an argument do support a particular conclusion—but not the conclusion that the arguer actually draws.
Example: "The seriousness of a punishment should match the seriousness of the crime. Right now, the punishment for drunk driving may simply be a fine. But drunk driving is a very serious crime that can kill innocent people. So the death penalty should be the punishment for drunk driving." The argument actually supports several conclusions—"The punishment for drunk driving should be very serious," in particular—but it doesn't support the claim that the death penalty, specifically, is warranted.
Tip: Separate your premises from your conclusion. Looking at the premises, ask yourself what conclusion an objective person would reach after reading them. Looking at your conclusion, ask yourself what kind of evidence would be required to support such a conclusion, and then see if you've actually given that evidence. Missing the point often occurs when a sweeping or extreme conclusion is being drawn, so be especially careful if you know you're claiming something big.
Post hoc (also called false cause)
This fallacy gets its name from the Latin phrase "post hoc, ergo propter hoc," which translates as "after this, therefore because of this."
Definition: Assuming that because B comes after A, A caused B. Of course, sometimes one event really does cause another one that comes later—for example, if I register for a class, and my name later appears on the roll, it's true that the first event caused the one that came later. But sometimes two events that seem related in time aren't really related as cause and event. That is, correlation isn't the same thing as causation.
Examples: "President Jones raised taxes, and then the rate of violent crime went up. Jones is responsible for the rise in crime." The increase in taxes might or might not be one factor in the rising crime rates, but the argument hasn't shown us that one caused the other.
Tip: To avoid the post hoc fallacy, the arguer would need to give us some explanation of the process by which the tax increase is supposed to have produced higher crime rates. And that's what you should do to avoid committing this fallacy: If you say that A causes B, you should have something more to say about how A caused B than just that A came first and B came later!
Slippery slope
Definition: The arguer claims that a sort of chain reaction, usually ending in some dire consequence, will take place, but there's really not enough evidence for that assumption. The arguer asserts that if we take even one step onto the "slippery slope," we will end up sliding all the way to the bottom; he or she assumes we can't stop halfway down the hill.
Example: "Animal experimentation reduces our respect for life. If we don't respect life, we are likely to be more and more tolerant of violent acts like war and murder. Soon our society will become a battlefield in which everyone constantly fears for their lives. It will be the end of civilization. To prevent this terrible consequence, we should make animal experimentation illegal right now." Since animal experimentation has been legal for some time and civilization has not yet ended, it seems particularly clear that this chain of events won't necessarily take place. Even if we believe that experimenting on animals reduces respect for life, and loss of respect for life makes us more tolerant of violence, that may be the spot on the hillside at which things stop—we may not slide all the way down to the end of civilization. And so we have not yet been given sufficient reason to accept the arguer's conclusion that we must make animal experimentation illegal right now.
Like post hoc, slippery slope can be a tricky fallacy to identify, since sometimes a chain of events really can be predicted to follow from a certain action. Here's an example that doesn't seem fallacious: "If I fail English 101, I won't be able to graduate. If I don't graduate, I probably won't be able to get a good job, and I may very well end up doing temp work or flipping burgers for the next year."
Tip: Check your argument for chains of consequences, where you say "if A, then B, and if B, then C," and so forth. Make sure these chains are reasonable.
Weak analogy
Definition: Many arguments rely on an analogy between two or more objects, ideas, or situations. If the two things that are being compared aren't really alike in the relevant respects, the analogy is a weak one, and the argument that relies on it commits the fallacy of weak analogy.
Example: "Guns are like hammers—they're both tools with metal parts that could be used to kill someone. And yet it would be ridiculous to restrict the purchase of hammers—so restrictions on purchasing guns are equally ridiculous." While guns and hammers do share certain features, these features (having metal parts, being tools, and being potentially useful for violence) are not the ones at stake in deciding whether to restrict guns. Rather, we restrict guns because they can easily be used to kill large numbers of people at a distance. This is a feature hammers do not share—it'd be hard to kill a crowd with a hammer. Thus, the analogy is weak, and so is the argument based on it.
If you think about it, you can make an analogy of some kind between almost any two things in the world: "My paper is like a mud puddle because they both get bigger when it rains (I work more when I'm stuck inside) and they're both kind of murky." So the mere fact that you draw an analogy between two things doesn't prove much, by itself.
Arguments by analogy are often used in discussing abortion—arguers frequently compare fetuses with adult human beings, and then argue that treatment that would violate the rights of an adult human being also violates the rights of fetuses. Whether these arguments are good or not depends on the strength of the analogy: do adult humans and fetuses share the property that gives adult humans rights? If the property that matters is having a human genetic code or the potential for a life full of human experiences, adult humans and fetuses do share that property, so the argument and the analogy are strong; if the property is being self-aware, rational, or able to survive on one's own, adult humans and fetuses don't share it, and the analogy is weak.
Tip: Identify what properties are important to the claim you're making, and see whether the two things you're comparing both share those properties.
Appeal to authority
Definition: Often we add strength to our arguments by referring to respected sources or authorities and explaining their positions on the issues we're discussing. If, however, we try to get readers to agree with us simply by impressing them with a famous name or by appealing to a supposed authority who really isn't much of an expert, we commit the fallacy of appeal to authority.
Example: "We should abolish the death penalty. Many respected people, such as actor Guy Handsome, have publicly stated their opposition to it." While Guy Handsome may be an authority on matters having to do with acting, there's no particular reason why anyone should be moved by his political opinions—he is probably no more of an authority on the death penalty than the person writing the paper.
Tip: There are two easy ways to avoid committing appeal to authority: First, make sure that the authorities you cite are experts on the subject you're discussing. Second, rather than just saying "Dr. Authority believes x, so we should believe it, too," try to explain the reasoning or evidence that the authority used to arrive at his or her opinion. That way, your readers have more to go on than a person's reputation. It also helps to choose authorities who are perceived as fairly neutral or reasonable, rather than people who will be perceived as biased.
Ad populum
Definition: The Latin name of this fallacy means "to the people." There are several versions of the ad populum fallacy, but what they all have in common is that in them, the arguer takes advantage of the desire most people have to be liked and to fit in with others and uses that desire to try to get the audience to accept his or her argument. One of the most common versions is the bandwagon fallacy, in which the arguer tries to convince the audience to do or believe something because everyone else (supposedly) does.
Example: "Gay marriages are just immoral. 70% of Americans think so!" While the opinion of most Americans might be relevant in determining what laws we should have, it certainly doesn't determine what is moral or immoral: There was a time where a substantial number of Americans were in favor of segregation, but their opinion was not evidence that segregation was moral. The arguer is trying to get us to agree with the conclusion by appealing to our desire to fit in with other Americans.
Tip: Make sure that you aren't recommending that your audience believe your conclusion because everyone else believes it, all the cool people believe it, people will like you better if you believe it, and so forth. Keep in mind that the popular opinion is not always the right one!
Ad hominem and tu quoque
Definitions: Like the appeal to authority and ad populum fallacies, the ad hominem ("against the person") and tu quoque ("you, too!") fallacies focus our attention on people rather than on arguments or evidence. In both of these arguments, the conclusion is usually "You shouldn't believe So-and-So's argument." The reason for not believing So-and-So is that So-and-So is either a bad person (ad hominem) or a hypocrite (tu quoque). In an ad hominem argument, the arguer attacks his or her opponent instead of the opponent's argument.
Examples: "Andrea Dworkin has written several books arguing that pornography harms women. But Dworkin is an ugly, bitter person, so you shouldn't listen to her." Dworkin's appearance and character, which the arguer has characterized so ungenerously, have nothing to do with the strength of her argument, so using them as evidence is fallacious.
In a tu quoque argument, the arguer points out that the opponent has actually done the thing he or she is arguing against, and so the opponent's argument shouldn't be listened to. Here's an example: Imagine that your parents have explained to you why you shouldn't smoke, and they've given a lot of good reasons—the damage to your health, the cost, and so forth. You reply, "I won't accept your argument, because you used to smoke when you were my age. You did it, too!" The fact that your parents have done the thing they are condemning has no bearing on the premises they put forward in their argument (smoking harms your health and is very expensive), so your response is fallacious.
Tip: Be sure to stay focused on your opponents' reasoning, rather than on their personal character. (The exception to this is, of course, if you are making an argument about someone's character—if your conclusion is "President Clinton is an untrustworthy person," premises about his untrustworthy acts are relevant, not fallacious.)
Appeal to pity
Definition: The appeal to pity takes place when an arguer tries to get people to accept a conclusion by making them feel sorry for someone.
Examples: "I know the exam is graded based on performance, but you should give me an A. My cat has been sick, my car broke down, and I've had a cold, so it was really hard for me to study!" The conclusion here is "You should give me an A." But the criteria for getting an A have to do with learning and applying the material from the course; the principle the arguer wants us to accept (people who have a hard week deserve A's) is clearly unacceptable. The information the arguer has given might feel relevant and might even get the audience to consider the conclusion—but the information isn't logically relevant, and so the argument is fallacious. Here's another example: "It's wrong to tax corporations—think of all the money they give to charity, and of the costs they already pay to run their businesses!"
Tip: Make sure that you aren't simply trying to get your audience to agree with you by making them feel sorry for someone.
Appeal to ignorance
Definition: In the appeal to ignorance, the arguer basically says, "Look, there's no conclusive evidence on the issue at hand. Therefore, you should accept my conclusion on this issue."
Example: "People have been trying for centuries to prove that God exists. But no one has yet been able to prove it. Therefore, God does not exist." Here's an opposing argument that commits the same fallacy: "People have been trying for years to prove that God does not exist. But no one has yet been able to prove it. Therefore, God exists." In each case, the arguer tries to use the lack of evidence as support for a positive claim about the truth of a conclusion. There is one situation in which doing this is not fallacious: If qualified researchers have used well-thought-out methods to search for something for a long time, they haven't found it, and it's the kind of thing people ought to be able to find, then the fact that they haven't found it constitutes some evidence that it doesn't exist.
Tip: Look closely at arguments where you point out a lack of evidence and then draw a conclusion from that lack of evidence.
Straw man
Definition: One way of making our own arguments stronger is to anticipate and respond in advance to the arguments that an opponent might make. In the straw man fallacy, the arguer sets up a wimpy version of the opponent's position and tries to score points by knocking it down. But just as being able to knock down a straw man, or a scarecrow, isn't very impressive, defeating a watered-down version of your opponents' argument isn't very impressive either.
Example: "Feminists want to ban all pornography and punish everyone who reads it! But such harsh measures are surely inappropriate, so the feminists are wrong: porn and its readers should be left in peace." The feminist argument is made weak by being overstated—in fact, most feminists do not propose an outright "ban" on porn or any punishment for those who merely read it; often, they propose some restrictions on things like child porn, or propose to allow people who are hurt by porn to sue publishers and producers, not readers, for damages. So the arguer hasn't really scored any points; he or she has just committed a fallacy.
Tip: Be charitable to your opponents. State their arguments as strongly, accurately, and sympathetically as possible. If you can knock down even the best version of an opponent's argument, then you've really accomplished something.
Red herring
Definition: Partway through an argument, the arguer goes off on a tangent, raising a side issue that distracts the audience from what's really at stake. Often, the arguer never returns to the original issue.
Example: "Grading this exam on a curve would be the most fair thing to do. After all, classes go more smoothly when the students and the professor are getting along well." Let's try our premise-conclusion outlining to see what's wrong with this argument:
Premise: Classes go more smoothly when the students and the professor are getting along well.
Conclusion: Grading this exam on a curve would be the most fair thing to do.
When we lay it out this way, it's pretty obvious that the arguer went off on a tangent—the fact that something helps people get along doesn't necessarily make it more fair; fairness and justice sometimes require us to do things that cause conflict. But the audience may feel like the issue of teachers and students agreeing is important and be distracted from the fact that the arguer has not given any evidence as to why a curve would be fair.
Tip: Try laying your premises and conclusion out in an outline-like form. How many issues do you see being raised in your argument? Can you explain how each premise supports the conclusion?
False dichotomy
Definition: In false dichotomy, the arguer sets up the situation so it looks like there are only two choices. The arguer then eliminates one of the choices, so it seems that we are left with only one option: the one the arguer wanted us to pick in the first place. But often there are really many different options, not just two—and if we thought about them all, we might not be so quick to pick the one the arguer recommends!
Example: "Caldwell Hall is in bad shape. Either we tear it down and put up a new building, or we continue to risk students' safety. Obviously we shouldn't risk anyone's safety, so we must tear the building down." The argument neglects to mention the possibility that we might repair the building or find some way to protect students from the risks in question—for example, if only a few rooms are in bad shape, perhaps we shouldn't hold classes in those rooms.
Tip: Examine your own arguments: If you're saying that we have to choose between just two options, is that really so? Or are there other alternatives you haven't mentioned? If there are other alternatives, don't just ignore them—explain why they, too, should be ruled out. Although there's no formal name for it, assuming that there are only three options, four options, etc. when really there are more is similar to false dichotomy and should also be avoided.
Begging the question
Definition: A complicated fallacy; it comes in several forms and can be harder to detect than many of the other fallacies we've discussed. Basically, an argument that begs the question asks the reader to simply accept the conclusion without providing real evidence; the argument either relies on a premise that says the same thing as the conclusion (which you might hear referred to as "being circular" or "circular reasoning"), or simply ignores an important (but questionable) assumption that the argument rests on. Sometimes people use the phrase "beg the question" as a sort of general criticism of arguments, to mean that an arguer hasn't given very good reasons for a conclusion, but that's not the meaning we're going to discuss here.
Examples: "Active euthanasia is morally acceptable. It is a decent, ethical thing to help another human being escape suffering through death." Let's lay this out in premise-conclusion form:
Premise: It is a decent, ethical thing to help another human being escape suffering through death.
Conclusion: Active euthanasia is morally acceptable.
If we "translate" the premise, we'll see that the arguer has really just said the same thing twice: "decent, ethical" means pretty much the same thing as "morally acceptable," and "help another human being escape suffering through death" means "active euthanasia." So the premise basically says, "active euthanasia is morally acceptable," just like the conclusion does! The arguer hasn't yet given us any real reasons why euthanasia is acceptable; instead, she has left us asking "well, really, why do you think active euthanasia is acceptable?" Her argument "begs" (that is, evades) the real question.
Here's a second example of begging the question, in which a dubious premise which is needed to make the argument valid is completely ignored: "Murder is morally wrong. So active euthanasia is morally wrong." The premise that gets left out is "active euthanasia is murder." And that is a debatable premise—again, the argument "begs" or evades the question of whether active euthanasia is murder by simply not stating the premise. The arguer is hoping we'll just focus on the uncontroversial premise, "Murder is morally wrong," and not notice what is being assumed.
Tip: One way to try to avoid begging the question is to write out your premises and conclusion in a short, outline-like form. See if you notice any gaps, any steps that are required to move from one premise to the next or from the premises to the conclusion. Write down the statements that would fill those gaps. If the statements are controversial and you've just glossed over them, you might be begging the question. Next, check to see whether any of your premises basically says the same thing as the conclusion (but in other words). If so, you're begging the question. The moral of the story: You can't just assume or use as uncontroversial evidence the very thing you're trying to prove.
Equivocation
Definition: Equivocation is sliding between two or more different meanings of a single word or phrase that is important to the argument.
Example: "Giving money to charity is the right thing to do. So charities have a right to our money." The equivocation here is on the word "right": "right" can mean both something that is correct or good (as in "I got the right answers on the test") and something to which someone has a claim (as in "everyone has a right to life"). Sometimes an arguer will deliberately, sneakily equivocate, often on words like "freedom," "justice," "rights," and so forth; other times, the equivocation is a mistake or misunderstanding. Either way, it's important that you use the main terms of your argument consistently.
Tip: Identify the most important words and phrases in your argument and ask yourself whether they could have more than one meaning. If they could, be sure you aren't slipping and sliding between those meanings.
So how do I find fallacies in my own writing?
Here are some general tips for finding fallacies in your own arguments:
* Pretend you disagree with the conclusion you're defending. What parts of the argument would now seem fishy to you? What parts would seem easiest to attack? Give special attention to strengthening those parts.
* List your main points; under each one, list the evidence you have for it. Seeing your claims and evidence laid out this way may make you realize that you have no good evidence for a particular claim, or it may help you look more critically at the evidence you're using.
* Learn which types of fallacies you're especially prone to, and be careful to check for them in your work. Some writers make lots of appeals to authority; others are more likely to rely on weak analogies or set up straw men. Read over some of your old papers to see if there's a particular kind of fallacy you need to watch out for.
* Be aware that broad claims need more proof than narrow ones. Claims that use sweeping words like "all," "no," "none," "every," "always," "never," "no one," and "everyone" are sometimes appropriate—but they require a lot more proof than less-sweeping claims that use words like "some," "many," "few," "sometimes," "usually," and so forth.
* Double check your characterizations of others, especially your opponents, to be sure they are accurate and fair.
Can I get some practice with this?
Yes, you can. Follow this link to see a sample argument that's full of fallacies (and then you can follow another link to get an explanation of each one). Then there's a more well-constructed argument on the same topic.
Works Consulted
We consulted these works while writing the original version of this handout. This is not a comprehensive list of resources on the handout's topic, and we encourage you to do your own research to find the latest publications on this topic. Please do not use this list as a model for the format of your own reference list, as it may not match the citation style you are using. For guidance on formatting citations, please see the UNC Libraries citation tutorial.
Hurley, Patrick J. A Concise Introduction to Logic. Thornson Learning, 2000
Lunsford, Andrea and John Ruszkiewicz. Everything's an Argument. Bedford Books, 1998.
Copi, Irving M. and Carl Cohen. Introduction to Logic. Prentice Hall, 1998.
文章链接中推举的实例：
————————————————————————————————————
Sample Arguments with Fallacies
What this handout is about
This is an adjunct to our fallacies handout. It presents a sample argument with many fallacies and another, less fallacious, argument.
Sample argument
Try to spot the fallacies in the following passage. To see a breakdown of this argument with comments on the various fallacies, click here.
The feminist argument that pornography is harmful has no merit and should not be discussed in college courses. I read "Playboy" magazine, and I don't see how it could be harmful. Feminists might criticize me for looking at porn, but they shouldn't talk; they obviously look at it, too, or they couldn't criticize it. Many important people, including the Presidents, writers, and entertainers who have been interviewed by the magazine and the women who pose in it, apparently agree. Scientific studies so far have not proved that pornography is harmful, so it must not be harmful. Besides, to be harmful, pornography would either have to harm the men who read it or the women who pose in it, and since they both choose these activities, they must not be harmful. Feminists should take a lesson from my parents—they don't like loud music and won't have it in their houses, but they don't go around saying it's harmful to everyone or trying to prevent others from listening to it. Ever since feminists began attacking our popular culture, the moral foundation of our society has been weakened; the divorce rate, for example, continues to rise. If feminists would just cease their hysterical opposition to sex, perhaps relationships in our society would improve. If feminists insist, instead, on banning porn, men will have no freedom and no pleasure left, and large numbers of women will be jobless and will have to work as prostitutes to support themselves. In light of these consequences, feminists shouldn't be surprised if their protests are met with violence. Truly, the feminist argument is baseless.
A less fallacious argument for roughly the same claim
How might we make a stronger argument for the claim that "the feminist argument that pornography is harmful has no merit"? Let's try to construct an argument that avoids the fallacies above. Please note that much of the "evidence" here will be made up to serve as a model—don't use this page as a source for any actual research on pornography! We'll see what a good argument could look like, even if the evidence needed to make that argument doesn't currently exist.
The feminist argument that pornography is harmful lacks adequate support. First, the feminist argument typically alleges that pornography increases men's willingness to rape women, or at least to think of them only as sex objects. But this argument ignores the fact that the print pornography industry alone earns more money each year than the entire "legitimate" bookselling industry. For that to be true, there must be many, many men and women who read pornography regularly. And yet crime statistics suggest that not many men rape women. Furthermore, most men today believe in women's equality, as a study by Dr. Knowitall and her research group, of the Institute on the Status of Women, demonstrates. Feminists acknowledge that scientific studies have failed to show that porn harms women. If there had been only a few such studies, or if we had reason to believe they were unreliable, we should conclude that nothing has yet been shown about whether porn harms women. But I think that when reliable studies have repeatedly failed to show a relationship, that fact constitutes some evidence that the relationship doesn't exist. So it seems unlikely that porn is harming women in the way the feminist argument alleges.
In the absence of positive evidence from studies, we have to rely on common sense. Can people distinguish between the sometimes-degrading scenarios they see in porn and real life? I believe they can. I think pornography is a lot like television and movies—it presents images that, while they certainly do have some impact on us, we all realize are nothing more than fiction. Young children may have difficulty distinguishing between fantasy and reality, but they are not often exposed to pornography. Men and women who look at porn should know better than to think that it gives a realistic picture of sexual relationships between men and women. If porn cannot be shown to harm women as a class by making them more vulnerable to sexual violence or causing men to think of them as inferior, how else might it be harmful? Feminists have often argued that the porn industry is harmful to the women who work within it—that many of them are abused and exploited. I agree with them that if an industry is mistreating people, it needs to be reformed, and they are doing a public service by pointing out such abuses. But what sort of reform are feminists proposing?
One suggestion I know about has been made by Catharine MacKinnon and Andrea Dworkin, who argue that there should be a civil rights statute that allows anyone who has been harmed by porn to seek civil damages from pornographers. My concern about this proposal is that although it will not legally be censorship, since the law would not empower the government to stop anyone from producing material based on the ideas it contains, the civil rights statute will have the same effect as censorship. Pornographers may be so afraid of facing lawsuits that many of them will stop producing porn—and a situation where people are afraid to put forward certain kinds of writing or pictures because they will face legal consequences seems to violate the spirit, if not the letter, of the first amendment. Porn, like books, may express certain ideas about men, women, and sex, and those ideas may have political ramifications—but just as controversial books are protected, porn should be. It may even do more good than harm by provoking thoughtful discussion and debate about men, women, and sexuality.
To see an explanation of this argument, click here.
